PDA

View Full Version : Maintain current



ridelong
08-14-2008, 07:22 PM
All,

If the problem is runaway heat/current, why not just use a PWM with current feedback?

I can design one if it is worth while.

Will post design for free. Just want to screw the screwers.

Any opinions?

sp1r0
08-14-2008, 07:42 PM
Sure, I'm up for a good design. Let's mention some spec's first.

1. Needs to cover a frequency from 1Hz-500kHz(maybe 200kHz if higher freq's are a problem with current demands)
2. Duty cycle 10%-90%
3. At theoretical 100%, the power MOSFET needs to conduct 200Amps DC at 12volts, 24 volts max. I know this would mean big heat sinks and possibly fans... I've seen good power MOSFET's capable of dissipating 500W for $35 each.
4. Money not an option. This should be the "best of the best".
5. What have I missed

Thanks for that offer, are you going to build it?
I could help in testing and design.

sp1r0
08-14-2008, 07:51 PM
I would be willing to build a circuit around this if you design it:

http://www.st.com/stonline/books/pdf/docs/8220.pdf

sp1r0
08-15-2008, 09:48 PM
Yeah I thought so, gotta do it myself...

ridelong
08-16-2008, 08:06 PM
sp1r0,

Doodled the first schematic last night.

Went down to the basement to prototype it this afternoon, and am missing some stuff. Found what I need at Digi-key. Will order tomorrow if they are open.

Read your suggestions. First, I want cheap. Was going to design this thing to work only to 10 amp (for my gen). Decided to extend to 30 amp, because that seems like what the high amp guys run.

Frequency will be determined by the feedback loop. In this circuit, the fet will turn on and off due to the current.

Haven't figured out how to transfer a schematic.

c02cutter
08-16-2008, 08:25 PM
If designing for use that people will build, design for maximum and let the rest figure it out for settings. Any electrical component needs to be designed on over kill in this environment to allow for someone that does not know to use it. If designing at lower amps, you'll get a lot of people building it and blowing it immediately. So go for the over kill when it comes to components and such. Get the schematic as a working unit, and let them figure it out from there. Most do not really understand an electrical schematic, other wise there would be a bunch of zero's pwm's available. Remember heat is an issue with this kind of device.

overtaker
08-16-2008, 09:36 PM
Good advice co2cutter!

sp1r0
08-17-2008, 11:15 AM
Hey ridelong, try mouser.com for electronics, I think you will find they are far superior. That FET I recommended a few posts ago is abit pricey at ~$34, and may need a pricey heat sink as well, but as an alternative, could you parallel a few IRFP064's? It's in a TO-220 package and handle's 70A max. What do you think of this shcematic? How would you modify it to run multiple FET's in parallel?

alt.nrg.org: It's in the Plans & Schematic section.

BIGGUN
08-17-2008, 01:39 PM
Damn it.....I hate being stupid or at least not knowledgable.

A pwm is my next undertaking and I guess I have to blindly build from threads that I can find on the web. What is going to be different on this design than the ones that are currently offered? Most of the build is over my head but I can understand using a pwm and how it helps in production.

Thanks,
Kevin

ridelong
08-17-2008, 02:36 PM
All,
I am going to start with a IRL1404ZPBF-ND From Digi-Key. It is 40 volt/ 75 amp on, but the on resistance is only 0.0031 ohms. At 30 amps, that will waste 900 X 0.0031= 2.8 watts in the device . No massive heatsink. Plus, it is only $3.67.

ridelong
08-17-2008, 05:59 PM
BIGGUN,

What is different from others is I want to maintain a set current under varying conditions. I don't care about the frequency or pulse width, the feedback loop will adjust the pwm to stay at the set current.

Jaxom
08-18-2008, 03:57 PM
This is a great idea. It's fairly well established that HHO production is a function of current flow. Steady current flow regardless of temperature/electrolyte/etc. will do a LOT to help isolate the effects of the other factors that affect a cell's output.

ridelong
08-18-2008, 05:22 PM
I just had another idea. Instead of using current feedback, which is good, why not use the generator temperature to set the pwm. Temperature feedback. This would drive the gen as hard as possible until the gen comes up to set temperature. I'm thinking an LM335 would be a great temperature feedback element. I have used them a few times, and they work great. Haven't figured out how to couple the LM335 with the electrolyte solution to get good temp data.

I want to ger the current feedback done first, then play with temp feedback.

Ordered the stuff to build the current feedback today. Should be done in about a week.

sp1r0
08-19-2008, 10:05 AM
Temp is a function of current, probably wouldn't see much difference in overall performance between the two systems. Just my 2c...

ridelong
08-20-2008, 06:08 PM
Just finished prototyping (kinda) the current feedback pwm. Don't have the fet or the current sense resistor yet (probably get tomorrow), so I faked out some feedback stuff. Using the scope, looks like it works. Will post again when it really works.

Atechguy
08-20-2008, 06:52 PM
HI! there , are you guys making these as kit or to be sold if they are quality protype, i have the skills to put one together just no good at designing a stable product.

GOplayer
08-20-2008, 10:21 PM
Common Sense Questions
1: What would the name be of a device that sends DC to multiple targets but one at a time and alternating its delivery point at a high speed? (Would the result be the same at the HOD Generator as if a PWM is used? (In old cars we had a high voltage power distributor rotating to power the sparkplugs. Along this line but using modern electrical circuitry without mechanical components, what kind of switch can do this type of work?)
2: What is the list of attributes we know thus far that positively affects the electrolysis when it comes to applied electricity to the HOD Generator? (What is the most effective electrolysis process documented and what type of an electrical current controllers they use?)
Once we know the answers to these questions, a better focused research can be applied. I say this because so many solution providers are out there who put together PWMs and people are buying them but use them effectively…I do not find definitive reports thus far. Perhaps you can direct my research to such site where PWM are conclusively showing a significant increase in HHO production? Also I’d like to know how is a multi HOD Gen. setup is being powered up and is there a PMW or some other kind of switch that is used and how?
Z

ridelong
08-21-2008, 06:50 AM
Atechguy,

I thought a current feedback pwm would be a popular thing, and offered to design one.

When I finish it, I am going to make the schematic available. I don't know how to post a picture yet.

I will post a parts list also.

Atechguy
08-21-2008, 08:36 AM
Atechguy,

I thought a current feedback pwm would be a popular thing, and offered to design one.

When I finish it, I am going to make the schematic available. I don't know how to post a picture yet.

I will post a parts list also.

Thats MARVELOUS!! , i need more control of amperage.Thanks Ridelong.:)

ridelong
08-21-2008, 06:11 PM
Got my parts today, minus the current sense resistors.

Used parts from my junk drawers to make a sense resistor close to what I ordered.

Fired the prototype current feedback pwm up about an hour ago.

Sucker works GREAT. Current stays dead on. Been running for about !/2 hour in the basement, will check it after this post.

I am using a 1 x 2 inch heatsink on the TO220 fet. Running at 6.5 amps, after 20 minutes, fet is barely warm. The jumper clips are hotter than the fet.

Will have to do some high current runs, but I don't have my hho stuff set up to run much more than 8-10 amps.

sp1r0
08-21-2008, 08:40 PM
Sounds good, keep that FET cool!

ridelong
08-22-2008, 06:29 AM
The pwm works great.

Measuring actual current will be interesting. The frequency is low, so the current measuring has to be integrated.

The voltage must be resistor divided to be proportional to current. I'm thinking 200 mv= 200 amps so it works with the cheap digital readout modules.

Will work on this over the weekend, unless I go diving.

Atechguy
08-22-2008, 01:02 PM
The pwm works great.

Measuring actual current will be interesting. The frequency is low, so the current measuring has to be integrated.

The voltage must be resistor divided to be proportional to current. I'm thinking 200 mv= 200 amps so it works with the cheap digital readout modules.

Will work on this over the weekend, unless I go diving.
Glad to hear , i need a stable pwm , just when i think i have it under control , it seems to get hot enough to blow a 30 amp fuse , i will be changing from 8 gauge to 10. to put a liitle resistance , i believe the my cell only needs a couple of grains of NOah or just distilled water ,has 50 small plates .:eek:

ridelong
08-23-2008, 10:41 AM
Atechguy,

I don't think the current feedback pwm I'm building will help you. I made it very slow to minimize the amount of switching, to minimize fet heating.

For example, if it is set to 5 amps, it might draw 10 amps for 1/2 second, then be off for 1/2 second, thereby maintaining 5 amp average.

In pwm, the fet is full on or full off, that's why they are so efficient.

My circuit will regulate average current very well, but by being a pwm cannot regulate current at any given time.

Atechguy
08-23-2008, 01:42 PM
Atechguy,

I don't think the current feedback pwm I'm building will help you. I made it very slow to minimize the amount of switching, to minimize fet heating.

For example, if it is set to 5 amps, it might draw 10 amps for 1/2 second, then be off for 1/2 second, thereby maintaining 5 amp average.

In pwm, the fet is full on or full off, that's why they are so efficient.

My circuit will regulate average current very well, but by being a pwm cannot regulate current at any given time.

Will this not help to control current rising out control especially after the cell gets warm from under hood heat and current flow, when i start cool it draws about 6 amps but after hour or less it can be 22 amps. ??I know the series design can be a problem but i didn't want to add another cell and i can't easily open up this cell and reconfig. plates, do you recommend a more fully controlable PWM?? Thanks

ridelong
08-23-2008, 06:42 PM
Atechguy,

I played with my current feedback pwm today. I can speed it up (x1000) and still run a cool fet. If you use a slow blow fuse, I think this circuit might work for you. Make sure you don't have something intermittently shorting out.

Do you know how to post a link or a picture? I'll post schematic, but I still need to work out how to adjust current and read it.

The prototype is strung all over my workbench. I am in the process of making a neat one in a box, but still don't have the current sense resistor, so i'm stuck until it arrives.

Atechguy
08-23-2008, 07:27 PM
Atechguy,

I played with my current feedback pwm today. I can speed it up (x1000) and still run a cool fet. If you use a slow blow fuse, I think this circuit might work for you. Make sure you don't have something intermittently shorting out.

Do you know how to post a link or a picture? I'll post schematic, but I still need to work out how to adjust current and read it.

The prototype is strung all over my workbench. I am in the process of making a neat one in a box, but still don't have the current sense resistor, so i'm stuck until it arrives. updates i now suspect i may have a short in a cell ,now when i start cold it is about 21 amps when previosly it was 6 amps. I sorry i didn't make my own .
No i haven't posted pictures before, i have reduced my wiring down to 10 gauge hope it helps abit, what does that mean speed adjusted to x1000 ??pusles?? I can get a slow blow fuse, i pretty sure i don't have short. Thanks

ridelong
08-24-2008, 04:13 PM
Atechguy,

What is 1000x? A pulse width modulator works by changing the on and off times. The circuit parameters determine the total on-off cycle times. I first made my pwm very slow, about a second for 1 on-off cycle. I did it slow so the fet would stay cool. After reading your post, and realizing that this won't help you, I tried speeding up the circuit. It worked fine and the fet is cool. Now, 1 on-off cycle ocurrs in 0.004 seconds. BTW it is only 250x original.

If your circuit is somehow drawing high current the short on time of the pwm should limit it enough using a slow blow fuse to not blow. If it is a dead short, it will blow with or without a pwm.

Atechguy
08-24-2008, 04:35 PM
Atechguy,

What is 1000x? A pulse width modulator works by changing the on and off times. The circuit parameters determine the total on-off cycle times. I first made my pwm very slow, about a second for 1 on-off cycle. I did it slow so the fet would stay cool. After reading your post, and realizing that this won't help you, I tried speeding up the circuit. It worked fine and the fet is cool. Now, 1 on-off cycle ocurrs in 0.004 seconds. BTW it is only 250x original.

If your circuit is somehow drawing high current the short on time of the pwm should limit it enough using a slow blow fuse to not blow. If it is a dead short, it will blow with or without a pwm.

I just cleaned out the cell to and put straight distilled water in , iwas amazed at the production i was getting at 6.25 amps, i also emailed the company i bought the cell off of and there willing to help me out, i'm not sure if there is a short or not but was suspecting that miniature metal flakes could build up and cause a short, i am still interested in your pwm . To be able to have some control with your pwm for changing conditions like winter time , to add more amps for more production when it is cold out.Thanks

Jaxom
08-29-2008, 01:15 PM
Any progress on this? I had two thoughts:

First, use an inductive-capacitive filter on the input side to quell the spikes in current draw when the FET switches. This should stabilize the current through the fuse so that normal (as compared to slow-blow) fuses can be used.

Second, can this circuit be modified to change the output current based on an external input? I.e. use a 0-5v input to proportionally vary the output between 0 and 40A? I would be VERY interested in such a design, but I'm too rusty on my electronics to design it myself.

Good work so far! Keep it up!!

ridelong
08-29-2008, 04:40 PM
Jaxom ,

I have a potentiometer on the circuit that changes the setpoint current from 3 to 30 amps.

Then the feedback circuit changes the pwm to keep it at the current you set no matter what the load does. If the load doubles, the pwm decreases the "on" time 50%, keeping the current the same.

I'm running it on a generator in my basement at 7 amps, works great.

I don't know how to post pictures, can anyone tell me how? If so, I can scan my pencil drawing of the schematic and post it.

ridelong
08-29-2008, 04:51 PM
Thinking through what is needed, the current thing is OK, but temperature feedback is better.

1. The pwm would run full "on" until the generator got to the temperature
you set it to.

2. This would give maximum generation and get the gen up to temperature quicker.

3. The same circuit could be set up so if the car is off, the circuit could monitor the temp of the electrolyte, and turn on an electrolyte heating device if the electrolyte got below, say 35 degrees f.

I'm pretty excited about this one. Got about 90% of the stuff to make one.

Will report when finished.

Jaxom
08-29-2008, 04:59 PM
If the pics are small enough, youu can use the "Manage Attachments" button below the text box on the reply screen to attach them. If they're too big for the size limit, you'll need to scale them down or have them hosted elsewhere and link to them.

I'd love to see the schematic for this circuit. I really think it's a great way to prevent the cell from overheating and resolve the problem of the electrolyte becoming more concentrated as the water in the solution is used. We can stop measuring our electrolyte by the grain...just use a little more than needed and let the "RideLong Regulator" control the current.

Temperature feedback will make a better safety device to be sure, but I like current feedback as an actual power supply controller. Maybe both?

ridelong
08-29-2008, 06:43 PM
Jaxom ,

Cool, I'm going to redraw and post it. Thanks

Ran some tests last night, the circuit actually reduces current a little bit as load increases. Some nonlinearity, but not bad, about 5 %.

But I still think temp feedback is it.

Thanks again.

ridelong
08-30-2008, 03:58 PM
Here is a link to the schematic of the current feedback pwm.

NOTE:
I REMOVED THE LINK. THERE IS A MISTAKE ON THE SCHEMATIC. WILL FIX AND PUT LINK BACK. 9/30/08

I POSTED THE CORRECTED SCHEMATIC. 10/02/08


www.mikecramer.com/images/ifeedpwm.jpg

sp1r0
08-31-2008, 07:06 AM
Couple observations. Are you using +12V AND -12V? Also I don't see the PWM section, frequency adjust and duty cycle adjust?

Jaxom
08-31-2008, 11:06 AM
What he has labeled as -12v is meant as the (-) side of the 12v supply...this is the ground connection. It looks like the frequency is not variable (there's no need for it to be,) and the duty cycle is controlled by the current feedback loop.

I'm going to order the parts for this thing next week and build one. I've been through electronics school and can handle the construction but it's been a long time and I'm a bit rusty on the design aspect. I have two more quesions:
1. Did you go with the original low freq. or the high freq. FET switching?
2. How hard would it be to use a 0-5v input in place of the pot?

ridelong
08-31-2008, 11:36 AM
Jaxom,

Good answer.

The pot is part of a voltage divider.

The divider keeps the fet on until the negative feedback voltage of the .01 ohm resistor exceeds the divider voltage, then the fet turns off.

In full ccw position (pot shorted out), the voltage applied to the 49.9k to the opamp input is 0.03 volts (3 amp at .01 ohms).

At full cw pot, the voltage applied to the 49.9k is 0.3 volts (30 amp at .01 ohms).

---------------------------------------------------------------------------------------------------
To make the circuit work at 5 volt = 30 amp, and 0 volt = 0 amps, remove the 20k, 500 ohm pot and the 20 ohm.

Now connect the 5 volt source to a 4.7k resistor, and connect the other end of the 4.7k to the 49.9k (where the pot used to be connected to).

Connect a 300 ohm resistor to the junction of the 4.7k and 49.9k resistor. Connect the other end to ground.

Of course the 5 volt source must share ground with the circuit.
----------------------------------------------------------------------------------------------------

Should be good to go.

I went with the higher frequency, easier to meter that way. Frequency is determined by the .01 uf capacitor. I started with a 10 uf.

Q-Hack!
08-31-2008, 11:45 AM
This doesn't look like a Pulse Width Modulator, or PWM, in the true since of the word. It looks like it is just a amperage control device. It is straight DC throughout the circuit. There is no Frequency being generated any place. A good design, but not technically a PWM.

ridelong
08-31-2008, 12:00 PM
Q-Hack!

It is a pulse width modulator in the purest sense of the term.

The comparator generates a pulse train with the duty cycle modulated by the feeback strength of the voltage signal developed across the current sense resistor (.01 ohm).

Jaxom
08-31-2008, 12:09 PM
Q-Hack: Not to be rude, but do you know what a pulse-width modulator is? The very definition is: a circuit which switches power on and off, which uses a variable on-time (a.k.a. duty cycle) to control power flow through the circuit.

What most of the HHO guys call a PWM is a variable-frequency pulse-width modulator. It's a little more complex in that you can adjust the frequency using a potentiometer, but it's still basically the same device.

And yes, it's an amperage control device. That's the whole point. Did you read the thread?


RideLong: Thanks for the info on the voltage input. I'll put it to use. The only thing is, I'll probably up the resistors to 47K and 3K to raise the input impedance. The reason is: I'm using one of the factory EFI sensors for the 5v input, and this device will be piggybacked onto the PCM's input channel. I want the least possible impact on the factory circuit so that I don't compromise the PCM's functions. Sound reasonable?

Q-Hack!
08-31-2008, 12:13 PM
Err... I am not convinced... can you hook it up to a O-scope and show me the wave form?

In my mind that LM339 is only a voltage comparator and varies its output by what you put in. You have a positive DC voltage going into pin 5. The output (pin 2) is forward biasing your FET. The negative input (pin 4) will vary the voltage into the LM339 depending on current flowing through the FET. The LM339 doesn't generate frequencies, it just compares voltages.

I suppose one could say that the LM339 turns on and off the FET creating a really slopped square wave, but to call this a PWM is a bit of a stretch.

Jaxom
08-31-2008, 12:22 PM
In my mind that LM339 is only a voltage comparator and varies its output by what you put in. You have a positive DC voltage going into pin 5. The output (pin 2) is forward biasing your FET. The negative input (pin 4) will vary the voltage into the LM339 depending on current flowing through the FET. The LM339 doesn't generate frequencies, it just compares voltages.

Exactly. The input to the LM339 varies based on current through the FET. As it varies, the LM339 switches it's output on and off. The .01uF cap sets the switching frequency as it charges and discharges. The LM339 output switches the FET on and off. Put it all together and the LM339 switches the FET on and off, with freq. based on the size of the (.01uF) capacitor, and duty cycle based on current through the FET.

ridelong
08-31-2008, 12:28 PM
Jaxom,

Upping the resistors should work fine.

BTW Q-hack, my square wave isn't slopped at all.

The evidence for this is a cold fet at 7 amps.

Please wiki the definition of comparator.

Q-Hack!
08-31-2008, 12:29 PM
Q-Hack: Not to be rude, but do you know what a pulse-width modulator is? The very definition is: a circuit which switches power on and off, which uses a variable on-time (a.k.a. duty cycle) to control power flow through the circuit.

What most of the HHO guys call a PWM is a variable-frequency pulse-width modulator. It's a little more complex in that you can adjust the frequency using a potentiometer, but it's still basically the same device.

And yes, it's an amperage control device. That's the whole point. Did you read the thread?


RideLong: Thanks for the info on the voltage input. I'll put it to use. The only thing is, I'll probably up the resistors to 47K and 3K to raise the input impedance. The reason is: I'm using one of the factory EFI sensors for the 5v input, and this device will be piggybacked onto the PCM's input channel. I want the least possible impact on the factory circuit so that I don't compromise the PCM's functions. Sound reasonable?

You may be right, and I apologize if I came off rude there... I am just trying to understand what this circuit is doing. (I am not an EE) ;) I was always under the assumption that a PWM was something that you could adjust the width of a square wave (be it by variable or fixed components) to me this looks like it is just varying the current through the FET, creating something more like a sinewave or sawtooth shaped waveform.

If it is a PWM in the true since of the word and I am just a fool, then I apologies and will back down off my semantics soap box. :rolleyes:

Q-Hack!
08-31-2008, 12:34 PM
Exactly. The input to the LM339 varies based on current through the FET. As it varies, the LM339 switches it's output on and off. The .01uF cap sets the switching frequency as it charges and discharges. The LM339 output switches the FET on and off. Put it all together and the LM339 switches the FET on and off, with freq. based on the size of the (.01uF) capacitor, and duty cycle based on current through the FET.

You know what... I didn't even look at the capacitor at the bottom of the circuit as being something that would charge and discharge with the variation in current. I was thinking it was some sort of filter device... I really need to brush up on my circuit design. Thanks for the clarification.

Jaxom
08-31-2008, 12:38 PM
Glad we got it straightened out. :D

For future reference, a comparator output is either full-on or full-off, depending on which inpt has higher voltage. That's where the squarewave aspect comes into play.

ridelong
08-31-2008, 04:25 PM
Jaxom,

Are you going to control HHO production with a 0-5 volt signal relative to RPM? If so, I don't think it will help. The present generators barely put out enough HHO to idle a lawn mower, you need every liter you can get.

Tell me if I'm incorrect in my assumptions.

Just ran the circuit at 17 amps. That fet is the best device I've ever seen. At 17 amps (80% duty cycle) and the heatsink in the schematic, the fet was barely warm to my finger. Watch out for the current feedback resistor though, it gets hot.

Jaxom
08-31-2008, 05:25 PM
You're on the right track, but it's a bit more complex than that. First off, I'm using the MAP sensor to generate my 5v signal. It's a .5-4.5v output that is proportional to engine load. Secondly, I'm actually going to use 2 of these drivers. I'm modifiying my radial cell (http://www.hhoforums.com/showthread.php?t=571) to have 2 positive connections and two ground connections (instead of one of each.) The generator has enough surface area to theoretically handle 90 amps, and bench testing shows that it generates ~2lpm at 20A. I'm hoping that the production will increase proportional to current flow, so that I can get 6lpm at 60A with both regulators at full DC (although I haven't run it that hard yet.)

I know 60A seems like a lot for an automotive electrical system, but consider this: the engine is rarely at full load. Most of the time, the MAP voltage will run around 1.5-2.5v, except under full throttle acceleration and steep uphill grades. The battery should be able to supply enough current to keep up with the cell demands for these short periods, then when the load is reduced (at idle or when coasting) the battery can recharge off the alternator since the generator current will be significantly lowered. I may need to change to a high-capacity marine battery. Generator temperature should run in a similar fashion, with the cell heating up under heavy loads and then cooling off again at cruise/idle. I intend to use a circulation pump and a SS tube coiled in front of the radiator to help keep the cell cool.

I also had another idea about the regulator circuit: Use a potentiometer in place of the 3K resistor on the 5v input channel so that the "gain" on the current output can be adjusted relative to the 5v input. I.E. instead of [0-5v=0-30A], it could be adjusted to [0-5v=0-10A] or any level in between. I'll need to series the pot with something small to set a minimum current flow...maybe a 500ohm and a 2.5K pot? That should yield 0-5A min. and 0-30A max. if the feedback loop's response is linear.

sp1r0
08-31-2008, 09:19 PM
Sounds like a circuit of a friend of mine built way back. A pre-stage to a digital audio amp. Just because you use a comparator and not omp amps. I guess other designs I've seen use opamps that generate controllable frequency sine waves, cut them off at a certain voltage level and obtain the square wave that way... anyway all that aside. Ridelong brings to the table some real work IMHO, and also a very cheap and workable circuit. PWM in THE purest sense based on current control. My question is what cap or resistor do we change to change the frequency? AND also If some of our cells draw more current is the driver circuit sufficient to driver multiple parallel FET's. Awesome effort dude! Much appreciated so far.

ridelong
09-01-2008, 12:29 PM
Jaxom,

The pot will work like you want. It will be interesting to see your results, good luck.

SP1R0,

Change the .01 capacitor to change frequency. Lower capacitance=higher frequency, higher capacitance=lower frequency.

The IRL1404zpbf fet should be good to 30 amps, but I only tested my circuit at 17 amps. The fet was barely warm at 17 amps.

The circuit should drive multiple fets if more current is needed.

DigitalMocking
09-01-2008, 05:18 PM
Jaxom,

The pot will work like you want. It will be interesting to see your results, good luck.

SP1R0,

Change the .01 capacitor to change frequency. Lower capacitance=higher frequency, higher capacitance=lower frequency.

The IRL1404zpbf fet should be good to 30 amps, but I only tested my circuit at 17 amps. The fet was barely warm at 17 amps.

The circuit should drive multiple fets if more current is needed.

I did a mockup of your pencil drawing if anyone wants to copy the design.

http://www.geekforever.com/images/misc/pwmcircuit.jpg

DigitalMocking
09-01-2008, 05:21 PM
One of the things I'm interested in is varying the output of my cell using this kind of circuit, I want to tie my TPS (throttle position sensor) to the where the Pot is now so when you hit the pedal you get more HHO production.

I'm beginning to believe we're producing too much HHO in some of the cells we've got. What I'm building should do about 1lpm @13 amps, and I don't know that I need that much production for puttering around town in my little 2 liter engine.

ridelong
09-01-2008, 06:20 PM
Thanks for the cleaned up drawing.

I could of done it in PADS, but was lazy.

Read the posts about the MAF sensor on this thread, pretty good.

ridelong
09-01-2008, 07:06 PM
Jaxom,

If you have any trouble getting parts let me know.

JojoJaro
09-03-2008, 10:16 AM
What FET can we use that's good up to 100-125A.

I have a lot of spare capacity in my dual Alternator setup and I would like to push my gen up to 125A and see what that gives me.

Jaxom
09-03-2008, 02:00 PM
My big issue is going to be digging through Digi-Key's online catalog to get the order together. I always have a hard time finding stuff there for some reason....takes me forever and a day. Then again, this will only be my third time dealing with them.

ridelong
09-03-2008, 07:00 PM
JOJOJARO,

You might try IRF1324S-7PPBF-ND, from Digi Key. It is 1 milliohm at 160 amps. Better sink it good though. The good news is,they are only 5 bucks.

sp1r0
09-03-2008, 08:15 PM
Hey ridelong, if we wanted to parallel 2 or 3 of those FET's you use, would it be too much for the LM339 to drive them?

ridelong
09-03-2008, 08:41 PM
sp1r0,

Should drive 3 or 4 fine, no problem.

h2ocommuter
09-04-2008, 01:31 AM
I did a mockup of your pencil drawing if anyone wants to copy the design.

http://www.geekforever.com/images/misc/pwmcircuit.jpg

I apreaciate this pic. I am bad enough with the scamatic in front of me, the pencil pic. was omg.

I'll get the parts and give it a try.

Thank you too ridealong..:cool:

Keith's Garage
09-04-2008, 07:39 PM
First off, Thank you so much for coming up with this circuit. My only issue is that the ground wire and the current sense resistor are getting really hot. I am testing this circuit on a breadboard and these two wires are getting so hot that they are melting the breadboard. Any ideas?
Thanks

sp1r0
09-04-2008, 08:20 PM
Breadboard's are good for prototyping but are good for MAYBE 2 amps. Sounds like you should go on to building this thing with high guage wire with short distance between connections.

ridelong
09-05-2008, 06:17 AM
Keith's Garage,

You built it COOL!

The power resistor is going to get hot, but not hot enough to melt solder, it's the nature of the beast.

The power stuff (HHO wires, fet, current sense resistor, etc) should be 10 ga. wire. Solder them solidly for good connections. Look at my original post with the link to the schematic, it has some additional info on it.

Jaxom
09-05-2008, 11:02 AM
RideLong....PM'd you about part numbers.

h2ocommuter
09-05-2008, 06:07 PM
Kieth, thanks so much for helping could you send a pic. so I could see how to lay it out on a viro board or whatever kind of board it needs.

I am just learning and it would help if I had an example to see.

failure is not an option for me to learn what I am doing.

I am the only dog trying to eat this elephant.

ridelong
09-05-2008, 07:31 PM
Here is a parts list for the current feedback pwm, and where to get.

PART DIGI-KEY P/N COST
-------------------------------------------------------
12 VOLT ZENER 1N5349BTPMSCT-ND .40
500 OHM POT CT2201-ND 2.71
LM339 LM339NFS-ND .47
FET IRL1404ZPBF 3.67
20 OHM 20.0XBK-ND .50
20 K OHM 20.0KXBK-ND .50
49.9K OHM CMF49.9KQFTR-ND .30
(2) 24.9K OHM CMF34.9KQFTR-ND (.60) .30
1 K OHM CMF1.00KQFTR-ND .30
133 OHM CMF133QFCT-ND .30
.01 UF 1460PH-ND .17
10 MEG OHM 10MQBK-ND .27


PART MOUSER ELECTRONICS P/N COST
------------------------------------------------------
.01 OHM 25 WATT 71-RH25-0.01 3.76

hydrotinkerer
09-05-2008, 08:23 PM
I apreaciate this pic. I am bad enough with the scamatic in front of me, the pencil pic. was omg.

I'll get the parts and give it a try.

Thank you too ridealong..:cool:


Didn't someone forget on the redraw (left of page 12volt zener 1/2 watt) or did I miss something?

Jaxom
09-06-2008, 08:36 AM
Thanks for the parts list...that'll make things much easier.

h2ocommuter
09-09-2008, 12:54 AM
Can someone help me out on this one.

I ordered the parts to build three of these today and realized I didn't have the knowledge of where the temp probe is comming from in the circuit.
Next rereading everything here two more times I cannot figure out if the hydroxy cell gen. is fed equal 12v from the battery and that same 12v from the battery feeds the pwm circuit?
or if the lower wire (the D wire from the FET) comming from the gen is actually the temp probe for the feedback or the ground for the gen.

I am a vergin

Q-Hack!
09-09-2008, 02:03 AM
I don't think he has posted his temp control circuit as of yet. Just the current feedback PWM.

ridelong
09-09-2008, 12:22 PM
h2ocommuter,

I Haven't posted the temperature feedback pwm schematic.

The schematic on this thread is current feedback, and will hold a set current regardless of what the load does. The potentiometer in the schematic will allow a setpoint from 3 to 30 amps to be dialed in.

The temperature feeedback pwm isn't done yet. I have the temperature to the gen controller part done, but I still have to do the low temp heater circuit.

When I finish it, I'll post it on the temperature feedback pwm thread.

HHO King
09-09-2008, 08:07 PM
Take a look at this one. I think it runs at a fixed frequency but for most it would work perfectly. The link is here: http://www.extremehho.com

I built one similar with an led gauge instead of the analog, but now i'm questioning that decision. It fluctuates too much.

What is the optimum frequency to split water?

Q-Hack!
09-09-2008, 11:39 PM
ZeroFossilFuel has an interesting theory on the subject...

http://www.youtube.com/watch?v=vKjUzsNj8NM

h2ocommuter
09-11-2008, 12:15 PM
ridelong,
Throughout my searching I have finally found kevin west and he has duplicated Stanley Meyers System; he has many pionts that we could all learn what will produce huge amounts of gas at very low amperage.

he has rewired the alternator winding in such away as to get 600 wats of energy and pulses it with a truely simple pwm and then squeses it down with some microwave oven capacitors, with result of a Tidal Wave sign form and puts it through the electrolizer with plain distilled water and bang nearly excactly like (if not) like Stanley Meyers.

This guy has produced the jewel.

http://blog.waterforfuel.com/

I won't dis this guy because he is a doer but, I think he is going to run into the heat controle factor and then will have to revamp things if he is going to actually duplicate SM work. Other factors are in play that he is ignoring. So it looks like he is on the right track but he may be going to cleveland befor he gets to seattle. The guy has some real nice ideas.

ridelong
09-11-2008, 05:07 PM
All,
The schematic I posted has a mistake.

I built one for somebody and started testing about an hour ago. It started to get hot, and the scope showed screwy waveforms, not clean squarewaves like previously.

The mistake is the 10 meg resistor. It should connect to pin 2, and the other side to pin 5.

Sorry for the screwup.

Jaxom
09-11-2008, 06:06 PM
Directly to pin 5, instead of through the 49.9K?

ridelong
09-11-2008, 06:18 PM
Jaxom,

Exactly. Right to the comparator pin 5.

The 10 meg acts as hysteresis, squaring the output. With it on the
49.9k, it didn't supply any hysteresis, and the waveform was all over the place, making the fet run in linear mode, and FRYING HOT.

Bet that's why Kevins garage melted the solder on his prototype.

Sorry again for any problems I caused.

h2ocommuter
09-12-2008, 02:59 PM
HHO King,
When we are splitting H and O with electrolisys the covolent bond is being overcome by the magnitism of the water molicule sticking to the electrodes and that sticking is not using a frequency value in my opinion.
When we use a frequency to dissaciate H and O using frequency we are using a catistofic break down of the covelent bond of the molicule not touching the electrodes.

http://blog.waterforfuel.com/ has a lot of experience at these processes but I am not sure where I read this from some patent somewhere maybe.

Jaxom
09-13-2008, 11:38 AM
Digi-Key kills me. I finally got a chance to get my order together for the parts for this thing, and most of the resistors can't be bought in quantities of less than a thousand. I just don't need $561 worth of 1K, 24.9K, and 49.9K resistors. You also can't get fewer than 10 of the zeners or caps.

Since I'm setting the circuit up with a 5v input in place of the voltage divider network, can I drop the zener and 135ohm from the circuit? Or are they still needed for another reason (like to stabilize the voltage to the IC?)

ridelong
09-13-2008, 04:13 PM
Jaxom,

You can remove the zener and 135 ohm. They setup a stable reference voltage for the divider.

h2ocommuter
09-15-2008, 08:41 PM
Jaxom,
I found the same thing, so I called and they told me to change a couple of letters around and they were there at a 5 to 10 pack quantity. I ordered enough parts for a min, of three ( Shunt potentiometer ) Or pwm and where I had to get more, I won't have to order them next time.

1 10 10 0 0 1N5349BTPMSCT-ND 0.39000 $3.90
2 3 3 0 0 CT2201-ND 2.71000 $8.13
3 3 3 0 0 LM339NFS-ND 0.47000 $1.41
4 3 3 0 0 IRL1404ZPBF-ND 3.67000 $11.01
5 5 5 0 0 20.0XBK-ND 0.09800 $0.49
6 5 5 0 0 20.0KXBK-ND 0.09800 $0.49
7 3 3 0 0 CMF133QFCT-ND 0.29000 $0.87
8 10 10 0 0 1460PH-ND 0.16600 $1.66
9 5 5 0 0 10MQBK-ND 0.05400 $0.27
10 10 10 0 0 CMF49.9KQFCT-ND 0.25700 $2.57
11 3 3 0 0 CMF1.00KQFCT-ND 0.29000 $0.87
12 6 6 0 0 CMF24.9KQFCT-ND 0.29000 $1.74

You will be able to figure out where the changes are.

h2ocommuter
09-15-2008, 09:47 PM
ridelong,
If you could express what to do in this circumstance: I intend on using 6 volt big truck batteries to power my electrolizers in many applications for sale. so I wonder if I need to do somthing different besides changing the 20 meg to eclude the 49.9 k so it is only touching the 2 and 5 pins? Hope I said that right.
I don't know how many amps I willl be pulling, I am just now building my first "whoper"
I hope to produce 2 gal min gas continually. I want to keep my temp to ngt 135 deg.
So I will have to let you know the progress.

Jaxom
09-16-2008, 11:33 AM
Good deal...thanks for the info. The FETs and other stuff have already been shipped (should be here tomorrow) so I'll probably just source the resistors locally for the sake of time. I'll keep this info handy for the next time around though.

Also, 2gal/min is a lot of HHO...you'll have to really push some current to get that from a 6v supply. I'd expect a current draw of at least 150-180 amps.

ridelong
09-16-2008, 07:25 PM
h2ocommuter,

The circuit that I published on this thread will only handle 30 amps. Also it was designed to work on 12 volts. It won't work on 6 volts. It would probably fry the fet on 6 volts because there would not be enough gate drive.

To do this circuit for 6 volts would require a total redesign.

I would need the maximum current, operating voltage spread (for example 12 to 14.4 volts).

With that data, I could find what will work for you

h2ocommuter
09-21-2008, 12:47 AM
Thanks ridelong

Thats ok,

Normal big trucks have 4, 6V batteries. so i thought of using just one of them.
No prob, I don't need to use just one it was just a thought. no biggie...

Thanks for the offer.

donnylynn
09-24-2008, 10:50 AM
Here is a link to the schematic of the current feedback pwm.

http://www.mikecramer.com/images/ebay/ifeedpwm.jpg

In your circuit you use a resistor to sense current. What if you put a series ammeter there instead? Is it possible to use it as a sense resistor? or does the voltage across it change with respect to current? I know that somehow the ammeter divides off a percentage of the current to create a voltage to drive the meter mechanism, but dont know the specifics of how it works. Just thinking of performing 2 functions with one part. I was considering this ammeter from harbor freight.

http://www.harborfreight.com/cpi/ctaf/Displayitem.taf?itemnumber=95778

ridelong
09-24-2008, 11:36 AM
donnylynn,
The ammeter will act as a resistor in the circuit.

The problem will be scaling. The resistor in the schematic will develop 10 mv/amp across it. The pot voltages are set to match the resistor voltage. The current can be determined by the voltage across the resistor (100 mv=10 amps).

If the ammeter is not close to .01 ohm, the pot scaling will not be known.

Jaxom
09-25-2008, 06:31 PM
Well I finally got this thing assembled. Digi-Key messed up my order and sent me surface-mount ICs so I had to wait for the right parts to come in. It's not tested yet, but it's built and modified for progressive control based on a 0-5v input with the output range adjustable between 0-10 and 0-60 amps (0v=0A, 5v=[5-30A adjustable] on each of two output channels.) The whole unit measures 4"x4"x1", not counting the adjustment knob, and it's a simple 5-wire hookup. I still need to find a box to put it in. :cool:

ridelong
09-26-2008, 06:48 AM
Jaxom,

Glad to hear you finished it.

It will be interesting to see what the setup does for mpg.

HHO King
09-26-2008, 06:42 PM
Extreme HHO has adjustable frequency pwm's completely assembled with the gauge and fan.

The link is http://www.extremehho.com

paulm39083
10-04-2008, 06:19 AM
I too have trouble with " The Digi - Key so I took your parts list and plugged it into where you can list the part number and add the list to your order on their website

all the parts came up as wrong number or " error" I put the in exactly as in your list except I did not include the 1-12 numbering

I copied and pasted so I wouldn't screw it up and did anyway.

Any advice

ridelong
10-04-2008, 11:20 AM
I always call in my orders 1-800-344-4539.

The list is three parts, description, part number, and price.

Only the part number should be used. For example, the first part of the list would be 1N5349BTPMSCT-ND. Sorry if I wasn't clear.

Here is a parts list for the current feedback pwm, and where to get, AND THE SAME LIST WITH THE P/N ONLY.

PART DIGI-KEY P/N COST
-------------------------------------------------------
12 VOLT ZENER 1N5349BTPMSCT-ND .40
500 OHM POT CT2201-ND 2.71
LM339 LM339NFS-ND .47
FET IRL1404ZPBF 3.67
20 OHM 20.0XBK-ND .50
20 K OHM 20.0KXBK-ND .50
49.9K OHM CMF49.9KQFTR-ND .30
(2) 24.9K OHM CMF24.9KQFTR-ND (.60) .30
1 K OHM CMF1.00KQFTR-ND .30
133 OHM CMF133QFCT-ND .30
.01 UF 1460PH-ND .17
10 MEG OHM 10MQBK-ND .27


PART MOUSER ELECTRONICS P/N COST
------------------------------------------------------
.01 OHM 25 WATT 71-RH25-0.01 3.76


--------------------------------------------------------------------------------

Here is a parts list WITH PART NUMBERS ONLY for the current feedback pwm.

PART DIGI-KEY P/N COST
-------------------------------------------------------
1N5349BTPMSCT-ND
CT2201-ND
LM339NFS-ND
IRL1404ZPBF
20.0XBK-ND
20.0KXBK-ND
CMF49.9KQFTR-ND
CMF24.9KQFTR-ND
CMF1.00KQFTR-ND
CMF133QFCT-ND
1460PH-ND
10MQBK-ND


PART MOUSER ELECTRONICS P/N COST
------------------------------------------------------
71-RH25-0.01

galkaen
10-13-2008, 08:42 AM
Hi Ridelong, thank you for your input, it is of great value. I have made a PCB layout of your PWM Circuit and will like to know if it is correct. This is my first experience with PCB design and also my electronics skills are minimal. I will post the layout if you agree to check it out. Thank you again for all your time.

ridelong
10-13-2008, 06:32 PM
galkaen,

If you post it, I will check it.

I think this circuit is simple enough for perf board. That is what I built mine with.

stevo11
10-16-2008, 04:33 PM
Heya all,
great thread so far. I placed parts order should have it in 3 days. I have a little background in electonics, nothing much to speak of but I can at least read resistors (with a chart) and solder. Will I be needing a scope to toon this in?



Stevo

Jaxom
10-16-2008, 05:46 PM
Nope, no scope necessary unless you're picky and want to check the output waveform to verify that it's working properly.

H2OPWR
10-16-2008, 06:38 PM
Heya all,
great thread so far. I placed parts order should have it in 3 days. I have a little background in electonics, nothing much to speak of but I can at least read resistors (with a chart) and solder. Will I be needing a scope to toon this in?



Stevo

Many multi-meters have the ability to read duty cycle as well as frequency. Mine does and only cost $125.00. With a pwm that is all you really care about

stevo11
10-17-2008, 11:32 AM
Heya all,
am i using a pcb with all the holes in it or is it better to draw it out and etche it? I also ran across PCB123, but i think you need to be a rocket scientist to use it.



Stevo

Jaxom
10-17-2008, 02:40 PM
I built mine on perf board. Etching a PCB is probably overkill...this is a pretty simple circuit.

Never heard of PCB123...care to elaborate?

stevo11
10-17-2008, 08:24 PM
PCB123 (http://www.pcb123.com/) is a free download software for layout.



Stevo

galkaen
10-17-2008, 09:15 PM
Hi all!
How can I upload a PCB file of ridelong's pwm so you all can use it? I tried attaching it but it didn't worked.

h2ocommuter
10-18-2008, 01:35 AM
Hi all!
How can I upload a PCB file of ridelong's pwm so you all can use it? I tried attaching it but it didn't worked.


galkaen,
if you could email it to me I may be able to understand how to assemble this thing.

I don't know where the 49.9 K goes to if the 10 meg goes to the # 5 and #2 pins.

besides that i am just guessing as to the assembly.

dzndlideas@aol.com

galkaen
10-18-2008, 11:04 AM
h2ocommuter,

You will need express PCB to open it but i'm working on a .bmp version of the schematic too, and as soon as i learn how to post it; i will. I will send you the file also. Remember; I have not tried this yet! If you find it to work please post it and let us know.

This is the file:

http://www.hhoforums.com/attachment.php?attachmentid=905&stc=1&d=1224341956

stevo11
10-18-2008, 02:33 PM
Ty for the schematic Galkaen. I can see where the +&- (cell+ and cell-)are going to the cell, I also see where the - is coming into the board (-12vdc). My question is where is the posative coming in?



Stevo

ridelong
10-18-2008, 04:20 PM
galkaen,

Nice layout. A few comments.

1 Add "+" to the 12vdc label to make it clear.

2 The 20k near the 12 volt zener is 20 ohm.

3 The .01 ohm 25w is way larger and gets hot so it should be elevated off the board.

4 The power section should be beefed up to handle 30 amps. Place overlapping .300 pads so the power components tie together in a big solder puddle. Do this for +12vdc and Cell+, Cell- and d of the fet, s of the fet and .10 25w, .01 25w and -12vdc

5 If you use the heatsink in the bill of materials, you might add some mounting holes for it.

galkaen
10-19-2008, 11:47 AM
Hi ridelong!

Redid the whole thing! Thanks for the advice. Here's the new layout. As for the heatsink holes i will put them in if you give me the specs. I was planning on bending the fet and using a top mounted heatsink or an aluminum case. Thank you again!

Also, stevo11: This is not my schematic. It is Ridelong's work. I only made the PCB Layout. The credit goes to him.

Thanks everybody for making this a great forum!

Galkaen

resago
10-19-2008, 11:52 AM
how about something hi-res? at least 4x as big and clear.

ridelong
10-19-2008, 01:17 PM
galkaen,

The power section looks nice.

The current sense resistor is .550 inch thick, and the soldering holes are
.080 inch diameter, spaced 1.680 inch.

The heatsink is an AAVID THERMALLOY p/n 531302B02500G. It is Digi Key p/n HS386-nd.

Atm0spher
10-19-2008, 02:13 PM
so let me know if you accidently build an extra one. i would buy it from ya you know to help you out. good work guys.... i think. i see all the squiggly and stuff but where do all the components go? just kidding.......... so how do they work?

Jaxom
10-20-2008, 02:41 PM
Resago if you decide to build on of these let me know. I have most of the components handy.

resago
10-20-2008, 02:44 PM
cool. When I get the cell tuned I'll let you know.

galkaen
10-21-2008, 09:08 AM
how about something hi-res? at least 4x as big and clear.

Hi Resago;

Display your email and i will send you the .pcb file which is scalable and has more detail.

Galkaen.

stevo11
10-21-2008, 03:58 PM
Heya all,
I am just waiting on PCB board before i do this. My local radio shack did not have any if you can believe that. After searching for hours for the parts to put this together I thought i found some good sites for some of the items. here goes cooling fan w/ heatsink .79 cents (http://www.goldmine-elec-products.com/prodinfo.asp?number=G13970) I am gonna connect that the fet. I also got some DC amp and voltage meters (http://www.skycraftsurplus.com/). you can get some cheeper at Harbor freight but I kinda like these better for a couple of bucks extra. Hope this will help anyone looking for these parts.



Stevo

stevo11
10-23-2008, 11:53 PM
Heya All,
I have a few quick questions about the layout of this board.

1. .01F (upper left side) I believe is 1460PH-ND it is like a ceramic disk. which side is the + or -? If I look at the disk side with the letters is it the right side + ?

2. on the lower right side above 12vdc is 135. does part # cmf133qfct-nd go there?



Stevo

ridelong
10-24-2008, 11:01 AM
stevo11,

The .01 ceramic capacitor has no polarity, it works either way.

Yes, the cmf133qfct resistor goes in the lower right marked 135.

h2ocommuter
10-25-2008, 12:12 AM
ridelong,
I am building my first pwm and I am double guesing what I see on the pcb layout. It seems to contradict what you had written on your original drawing and the one #2 the program design; where both of them call for the #2 pin to be connected to the middle post on the mosfet. is this somthing that is insignificant of an error I must ignor to get this built right?

stevo11
10-25-2008, 12:27 AM
Pwm is complete. I had one sheet of photographic printing paper and wouldnt you know I forgot to mirror the image. I triead a bunch of times to print the mirrored image on what was left of the sheet but my printer would not take it (paper Jam). I assembled it anyway having to flip the board upside down backwards to align parts to the proper diagram. I posted pics of what I used to etche and my pwm.

For those who want to etche a board but can not find the ferric chloride like me here is a link for an alternative. Ferric Chloride alternative (http://www.instructables.com/id/EFVWDTSLWYEV2Z8QB1/)



Stevo

galkaen
10-25-2008, 09:44 AM
Hi stevo

Did it worked fine? Im looking at the pics and i think is not gonna work like that because the pins don't match.
Let me know. Im working to make an updated version and will like to know if it overheats, so i can put thicker traces.

stevo11
10-25-2008, 11:40 AM
Hope this helps. when I printed the board i forgot to mirror it. so I had to assemble it backwards. here is a pic of it from the back. does the backside of the fet (metal part) need to face CELL- or 24.9K?



Stevo

galkaen
10-25-2008, 02:40 PM
Hi stevo

Ok i see what you did!

The pcb is ok, but you soldered the components on the wrong side of it. The traces go on the bottom of the pcb. if you have a chem-wick you can still save all the components and even the board.

ridelong
10-26-2008, 06:32 PM
h2ocommuter,
The LM339 number 2 pin connects to the gate pin of the FET. The pinout of the FET is GDS, 123. The pcb layout corresponds to the correct connection.

stevo11
10-26-2008, 10:52 PM
BLAHHH hooked it up and it did not work. so i made another board and switched all the parts. still did not work. I think i am going to try and test all the parts and go from there.



Stevo

stevo11
10-27-2008, 01:03 AM
i tested everything but lm339, .01F capacitor and .01 OHM 25 WATT 71-RH25-0.01 3.76 only cause i dont know how to test them with a dmm. everything else seams to be fine. the only thing i can remember happening is the first time i plugged in the pwm something went "pop" not really loud. I thought it was an alligator clip on edge and straightend out. not really sure. I did do a visual inspection of circut and did not see anything that went pop. I kinda figured i would see something if it made that noise.

If anyone has some thoughts or wants to elaborate, feel free.



Stevo

sp1r0
10-28-2008, 06:48 AM
you probably blew out the lm339 by giving it reverse voltage. If you heard a "pop" you may have blew out a cap too. The worse thing to have is a bunch of parts lying around you are unsure of. It will make your future tinkering hell, believe me I've been through it. Throw all the parts you have away except the board, and start over.

stevo11
10-30-2008, 09:21 AM
this time i want to take all precautions. are there any components that need to be in line a specific way other than the diode 12v 1/2w on top, the mosfet, and lm339?

The only precaution i took for static was touching the frame of my puter before touching the fet and lm339. I am not even sure if that works, my brother told me to do it years ago while assembling a puter. Are there any type of antistatic gizmo recomended that i can make?



Stevo

HALS-GUNSMITHING
10-30-2008, 07:04 PM
Their is a braclet with an alagotor clip to keep you grounded to whatever you are working on.

Keith's Garage
10-30-2008, 07:54 PM
Ok, I need some help
I built this thing, everything was working perfect. I was able to adjust it to 20 amps, then the amperage just started droping for no reason. The max I can get now is about 16 amps. I checked the generator and it is still capable of pulling 40 amps.
What happened?

stevo11
10-30-2008, 08:33 PM
I took a computer cord and cut and taped all the wires but the ground and turned that into a bracelet. should work.

another question. do i just attach the 12vdc cooling fan to the +- terminals of the pwm. or is it going to burn the fan out, the fan is rated for 1.68 watts



Stevo

HALS-GUNSMITHING
10-30-2008, 09:24 PM
If you attach it to the pwm terminals attach it to the power in not the power out.

stevo11
10-30-2008, 09:45 PM
ty for the info

ridelong
10-31-2008, 06:36 PM
Ok, I need some help
I built this thing, everything was working perfect. I was able to adjust it to 20 amps, then the amperage just started droping for no reason. The max I can get now is about 16 amps. I checked the generator and it is still capable of pulling 40 amps.
What happened?

Keith,

What power source are you using? Could the voltage be dropping?

Measure the source voltage and the voltage delivered to the gen. If they are very close to the same, the pwm is running full on , and the pwm can't regulate. This could be because the gen impedence has increased for some reason.

I noticed that when the gen impedence decreases (electrolyte heats up), the circuit will actually back the current down a little, but not drastically

stevo11
11-01-2008, 01:27 AM
third attemp with failure... I hook the pwm to a battery charger at 12vdc with a car light attached in place of a cell. at 10 amps the lamp is barely orange not even glowing, when i turn the pot nothing happens. at 50 amps the lamp is only a little bit more orange but still not glowing, when i turn the pot nothing happens. I will post some detail pitcture of both sides of the pwm tomorrow. maybe someone will see the problem.



Stevo

Keith's Garage
11-01-2008, 04:20 PM
Ok, now I have a new problem. When I finally got the PWM installed on a vehicle, the heat from the FET got so high that the solder started to melt. And now the PWM wont regulate the amperage.
First, how are you guys keeping the heat down
Second, what do you think burnt up?

donnylynn
11-01-2008, 06:40 PM
I assume you mean the solder on the FET. If you got hot enough to melt the solder, then you probably burned up the FET. Was it mounted to a heat sink? What is the part number of the FET?

ridelong
11-01-2008, 07:43 PM
Keith's Garage,

Make sure you have the correct schematic!

My first pencil drawing had a mistake on it. I have since fixed the pencil drawing, but the guy who drew it with a layout program never fixed his.

The mistake WILL MAKE THE FET OVERHEAT and probably toasted the fet. The fet will get only mildly warm at 14 amps with the listed heatsink and no fan if the circuit is operating correctly.

Here is the correct schematic:

www.mikecramer.com/images/ifeedpwm.jpg

stevo11
11-01-2008, 08:08 PM
I dont know if it matters but my pwm regulatesthe amps at 3.3 wether its on 10 amp or 50 amp. i just can not adjust it up or down.

Can anyone post detail pics of there pwm both sides?

Stevo

stevo11
11-01-2008, 09:53 PM
I am posting how the layout of the pcb board with the components placed in specific areas. I am not an EE so i just named most of the stuff R. one of my questions is are R9 and R4 in the proper positions. I also conected all the unused pins in LM339 and brought them to a ground. Does the pot need to be wired in any way in particular? TY for all the support for a nub.



Stevo

sp1r0
11-02-2008, 07:38 AM
I am posting how the layout of the pcb board with the components placed in specific areas. I am not an EE so i just named most of the stuff R. one of my questions is are R9 and R4 in the proper positions. I also conected all the unused pins in LM339 and brought them to a ground. Does the pot need to be wired in any way in particular? TY for all the support for a nub.



Stevo

not bad for a n00b, not bad at all. I double and triple checked your layout, everything is correct, except you have no power going to your lm339. Here's what it should look like...

Just to be clear here, your PCB layout is on top of the pcb, that is, you should mount and solder the parts right on the etched (top) part of the board according to your pcb layout. If you wanted to mount the parts on top of the board and solder underneath, like most hobbyists do, you will have to mirror the image.

sp1r0
11-02-2008, 07:48 AM
Actually if you want to get technical, ridelong says to ground all unused inputs of the lm339, just good practice, creates less noise in the circuit. Good luck. I hope it works for you!

Note: Do not connect the outputs to ground or you may damage the lm339!
P.S., I've not built this circuit yet, just going from ridelong's latest schematic and stevo11's pcb layout.

stevo11
11-02-2008, 01:38 PM
Its ALIVE:eek:!!! Its ALIVE:eek:!!!! (My Gene Wilder impresion, young frankenstien) I started to type this, sat down and ate lunch. Came back from lunch to test it again and when i turn the amps up it will not come back down. I am also measuring amps with an DM set to 10a and at its lowest pot setting runs at 2.48 - 2.60. when i turn pot all the way up it goes to 3.45. I am not sure but when i turn on the pwm is it running 24.8 amps and when maxxed 34.5 amps?

TY for all the help, yall been great:).



Stevo

stevo11
11-02-2008, 01:45 PM
Ok, i just searched for the breakdown and found it on Ridelong schematic, 10mv=1amp, i new i saw it somewhere.



Stevo

ridelong
11-02-2008, 02:16 PM
Actually, the 339 is an open collector output, so it is fine to ground all the unused pins. I just put inputs only out of habit.

ridelong
11-02-2008, 02:42 PM
stevo11,

You can measure across the 25 watt, .01 ohm resistor with a voltmeter set to 200 mv or 2000 mv (2 volt).

On 200 mv scale, if it reads 143.5 mv, it means 14.35 amps.

On the 2000 mv scale, if it reads 144 mv, it means 14.4 amps.

Hope this helps.

BTW, the lowest pot setting should be betweem 3 and 5 amps. The highest setting is around 30 amps.

stevo11
11-02-2008, 10:37 PM
I have attempted to build this pwm 6 times, one i had going for a little while and the others basically failed.

I have enough parts left for 1 more attempt and I dont want to waste them. I am looking for someone to build it for me please. Any takers out there?



Stevo

galkaen
11-04-2008, 02:41 PM
Ridelong

Would you be kind and post a picture of your finished unit, please. Even if it is made on a perf board it would be of great help to see the actual connections.

Thanks in advance:
Galkaen

ridelong
11-04-2008, 02:49 PM
Galkaen,

Here it is.

It's drawing 10.64 amps.

http://www.mikecramer.com/images/iproto.jpg

h2ocommuter
11-05-2008, 07:45 PM
I have attempted to build this pwm 6 times, one i had going for a little while and the others basically failed.

I have enough parts left for 1 more attempt and I dont want to waste them. I am looking for someone to build it for me please. Any takers out there?



Stevo

Stevo,
I believe I should get my next one right but I am going over the last one untill I find the problem; My mistake I am sure.

I have an issue I am going to scrutinize untill I decern the answer.
I like the example of ridelongs and I will be able to get a better concept of what it could look like.

I am courious as to the what looks like a heat sink but I don't see the mosfet. thats ok I will have this one done in a couple of days the last one took 3 to 4 days. ( I had no clue) LOL

I am having fun but strugling to get it right.

ridelong
11-06-2008, 10:01 AM
h2ocommuter,

The black thing is the heatsink for the fet.

The fet is mounted horizontally on the heatsink with the legs pointing toward the terminal strip.

The gate lead is the closest pin to the perf board.

Hope this helps.

h2ocommuter
11-06-2008, 05:01 PM
Stevo,
I believe I should get my next one right but I am going over the last one untill I find the problem; My mistake I am sure.

I have an issue I am going to scrutinize untill I decern the answer.
I like the example of ridelongs and I will be able to get a better concept of what it could look like.

I am courious as to the what looks like a heat sink but I don't see the mosfet. thats ok I will have this one done in a couple of days the last one took 3 to 4 days. ( I had no clue) LOL

I am having fun but strugling to get it right.

I had put the 20 ohm and the20k in the reverse places, at that piont I was getting a consistant voltage to the cell but no adj. were poossile, So at that piont I put a jumper wire across the sweep "between the 49,and the 20k" and that was not the answer and I had no out put to the cell at all, so I shorted the ground directly to the ammeter on the cell and adj the sweep and I poofed the 339 I think. I tore it down and found the 20 ohm and the 20k mixed up and my heat sink was rather big and I had atached the .01 to it with big wires and I pulled the mosfet apart. I replaced it with a scavengered one but I could not redeem the unit. I am going to build another.


Happy just to learn this stuf. And thanks to all.....

h2ocommuter
11-07-2008, 09:14 PM
not bad for a n00b, not bad at all. I double and triple checked your layout, everything is correct, except you have no power going to your lm339. Here's what it should look like...

Just to be clear here, your PCB layout is on top of the pcb, that is, you should mount and solder the parts right on the etched (top) part of the board according to your pcb layout. If you wanted to mount the parts on top of the board and solder underneath, like most hobbyists do, you will have to mirror the image.

I missed the Power on the 399 also I missed it on the original ridelong drawing.
I apriciate it.
I made a pcb layout visualizing it like ridelongs pic. show's.
I would like someone to check it out for me. e-mail.
I am building it using spades and I turned the pot upright too.
I figured it would go into a closed box when finished. set it and forget it.

sp1r0
11-10-2008, 04:32 PM
So you got it working or what? We expect progress, heh.

galkaen
11-13-2008, 06:14 PM
Hi Guys!

This is my latest layout of Ridelong's PWM Schematic.
If you are interested in the .PCB file email me (galkaen@gmail.com).
Ill be glad to share it.

scratch1676
12-09-2008, 02:20 AM
h20commuter did you get your pwm working? I read the post and you made me laugh will all the trouble you had. I went thru the same stuff and it was living heck. you reminded me of myself.

jeev_alexander
02-22-2009, 02:26 PM
Keith's Garage,

Make sure you have the correct schematic!

My first pencil drawing had a mistake on it. I have since fixed the pencil drawing, but the guy who drew it with a layout program never fixed his.

The mistake WILL MAKE THE FET OVERHEAT and probably toasted the fet. The fet will get only mildly warm at 14 amps with the listed heatsink and no fan if the circuit is operating correctly.

Here is the correct schematic:

www.mikecramer.com/images/ifeedpwm.jpg
Hi ridealong, Thanks 4 ur circuit,i made this circuit but there is no current limiting is happening...can you advice what could e the reason? The FET is getting warm mildly ,no pulse is getting generated from LM339,only deviation from ur spec is the resistor instead of .01 i put five .05 in parallel so the effective resisitance is .0125,guess its ok.

galkaen
03-10-2009, 11:47 AM
Hi Ridelong,

Finally have the parts but couldn't get the right fet. Will it work with (ISL9N302AP3)?

TKS

Give me your email and i will send you my new pcb file.

galkaen
03-26-2009, 09:35 AM
It would be nice if you return to this forum thread and help us build this thing.
I have tried 5 times to put it together but it just doesn't work. Please HELP!

I've been succesfull with ZFF PWM, which is a lot more complicated. But i'm willing to build your circuit because of its simplicity.

Again PLEASE HELP!

ridelong
03-26-2009, 05:20 PM
Galkaen,

Sorry haven't been on this thread for a while.

The ISL9N302AP3 should work with a good sink.

I installed the temperature feedback unit designed September 2008 and have been using it since.

Russ