Very confused about HHO

[ridelong]

Philldpapill,

The circuit I posted is electrically a "P", not a "PID". Essentially, a comparator that runs full blast until the temp signal from the LM324 reaches the pot voltage "setpoint", then the comparator turns off the power fet. When the temp signal falls below the setpoint, it turns the power fet on again. This action cycles the power fet on and off according to temperature.

I used a stainless tube submerged in the electrolyte with a LM34 for a probe. The ss acts like an integral loop component to stabilize the circuit. Therefore the circuit acts like a "PI" loop. It is very stable.

The fet is a Digi-Key P/N IRL1404ZPBF-ND. The circuit is designed to control 30 amps continuous.

Russ

[ridelong]

canpower,

I would pick the HRF3205. The .008 on resistance would allow you to draw 27 amps at a 6 watt dissipation in the device. This would allow you to run fanless with a decent heatsink.

Russ

[canpower]

canpower, can you do your circuit using surface mount components? Can you make your own circuit boards? I have a design that uses some MOSFETs from ST Microelectronics. You can sign up for an account, and get them for free. They are 55Vds, 0.0085(nominal)Rds_on.

They main thing to look for, is a mosfet with a very low Rds_on, and a low gate capacitance. The UCC27322 driver is an excellent driver from TI. In fact, I use it in nearly all my power designs. However, it does have a tendancy to explode - for totally unknown reasoning to me.

I can send you my EAGLE cad files which has the circuit schematics and board layout in it, if you'd like.

Hi Philldpapill,

Thank you so much for your generocity and kindness, I really appreciate the helping hand. And also thank you ridelong for your very useful info & suggestions, very helpful, thank you

Unfortunately, at this point in my life, I am in between housing moves, and am working out of boxes, and only have a small area to work with (VERY small...lol) and I do not currently have the equipment on hand to fabricate SMT prototypes. I am actually PACE certified (PACE SMT solder equipment courses) and have worked in the electronic assembly and manufacturing field doing electronic assembly, running wave solder machines, pick & place machines, ect... for many years (not now, in the past, thank God! LOL)

I also used to etch my own double sided copper clad boards at one time, using the UV photoresist technique, but gave that up years ago. Too messy!

When you drive your mosfet via the UCC27322, do you put a V-gate resistor between the driver and the mosfet gate, or just drive it direct out of the 2 outputs of the UCC27322? Do you connect BOTH of the UCC27322's outputs to the gate of the mosfet, or only one, and leave the other floating?

I'm assuming you at least put a 1uF electrolytic and a 0.1uF ceramic cap close to the UCC27322's output as the datasheet suggest, from the output to ground, as close as possible it says, don't you?

Do you use a DC bypass cap between the output of the UCC27322 and the mosfet's gate? I've seen this in a few schematics, and have wondered why it is there, and I also wondered how you are generally interfacing the UCC27322 to the mosfet, as you say you regularly use them in your designs.

What do you use to drive the input of the UCC27322? Just a 555 timer circuit configured to your needs?

I also wondered a bit about the UCC27322's datasheet where it states the input to the UCC27322, it is a "bit" obscure in it's description of the required input voltages.

It states the UCC27322 input voltage to input pin 2 as "" -5 to 6V "" OR VDD +0.3V, and then further confuses me with the statement "WHICH EVER IS LARGER".

Does this mean you have to either power the input with a 5V logic pulse for example, (falls within the -5 to 6V recommendation) OR lets say if VDD is 12V, then you can also power it optionally with a 12.3V input signal??

Usually I'm pretty good about reading these things, been doing it all my life, but it's always been hammered into my head that rail voltages are NOT to exceed supply voltages, so I'm a bit confused. I hope you can clarify this for me. I hope I'm not just asking a totally OBVIOUS question, and look foolish in doing so, I think I'm slipping in my old age of 42....LOL.

I would REALLY appreciate your Eagle cad files, but I guess I'll have to install Eagle again, I've been thinking about doing so anyhow, but keep toiling with whether or not to install Protel instead, as it's what I've always used in the past.

By the way, I've asked this question and not had an answer, what frequencies should the PWM be running at, and why? Is there a pulsed duty cycle also, like a certain on/off waveform of a certain frequency? Such as a pulsed 44Khz waveform for example. If so, what is the duty cycle, what are the frequencies??

Thanks again guys, I DO appreciate all your help!

[biggy boy]

not sure Ide go into it near the pcv valve. the closer to the throttle body the better, and if you can get upstream of the MAF sensonr that wouldl be good as well

Helz Don't you mean after the MAF, down stream? If you put it before the MAF and the little glowing heater used by the MAF it may go BOOM

[Helz_McFugly]

details details, im sure he knew what I meant. ill post my "Helzish to English" decoder sheet later.

[ridelong]

I did a lot of testing of PWM freq's, duty cycle. The freq (100hz-100khz) makes no difference as far as I could tell.

I personally like to run low frequencys (100hz) because the electronics are easier to design, and the FET switching losses are reduced.

The only thing a straight PWM is good for is reducing current in a HHO gen.

Russ

[canpower]

I did a lot of testing of PWM freq's, duty cycle. The freq (100hz-100khz) makes no difference as far as I could tell.

I personally like to run low frequencys (100hz) because the electronics are easier to design, and the FET switching losses are reduced.

The only thing a straight PWM is good for is reducing current in a HHO gen.

Russ

So the frequency is meaningless really? I find that hard to believe. I thought resonant frequency was the reason "why" this whole thing worked properly in the first place. Could it be that the reason why none of the frequencies you used made any difference is because none of them were the "right" frequencies, or in other words some kind of "resonant frequency" to something.

Just wondering.

Philldpapill, are you going to reply to my questions I've asked you? I would really appreciate your input as you've already built this using the UCC27322.

I'm ready to build something, but just sitting here doing nothing, because I have TOO MUCH info now, and I just can't decide on what circuit is worth building to start out with. And NOBODY will disclose the PWM and EFIE circuits they are using.

[ridelong]

I did hours of testing, and there is no resonant frequency below 500 Khz.

I did read that water is resonant near the ghz freq's that microwave ovens operate, but that would make the water boil.

I've read about the resonance that makes super generation of HHO, but I couldn't find it. I think if there was a resonant frequency, all of the work on this site would of found it by now.

The PWM units sold today are for current limiting. Also, PWMs won't stop the current increase as the electrolyte gets warmer without current or temperature feedback.

As for a current feedback pwm, I also designed this: www.mikecramer.com/images/ifeedpwm.jpg
It will maintain the setpoint current as the electrolyte warms up.

Russ

[canpower]

Hi Philldpapill,

Thank you so much for your generocity and kindness, I really appreciate the helping hand. And also thank you ridelong for your very useful info & suggestions, very helpful, thank you

Unfortunately, at this point in my life, I am in between housing moves, and am working out of boxes, and only have a small area to work with (VERY small...lol) and I do not currently have the equipment on hand to fabricate SMT prototypes. I am actually PACE certified (PACE SMT solder equipment courses) and have worked in the electronic assembly and manufacturing field doing electronic assembly, running wave solder machines, pick & place machines, ect... for many years (not now, in the past, thank God! LOL)

I also used to etch my own double sided copper clad boards at one time, using the UV photoresist technique, but gave that up years ago. Too messy!

When you drive your mosfet via the UCC27322, do you put a V-gate resistor between the driver and the mosfet gate, or just drive it direct out of the 2 outputs of the UCC27322? Do you connect BOTH of the UCC27322's outputs to the gate of the mosfet, or only one, and leave the other floating?

I'm assuming you at least put a 1uF electrolytic and a 0.1uF ceramic cap close to the UCC27322's output as the datasheet suggest, from the output to ground, as close as possible it says, don't you?

Do you use a DC bypass cap between the output of the UCC27322 and the mosfet's gate? I've seen this in a few schematics, and have wondered why it is there, and I also wondered how you are generally interfacing the UCC27322 to the mosfet, as you say you regularly use them in your designs.

What do you use to drive the input of the UCC27322? Just a 555 timer circuit configured to your needs?

I also wondered a bit about the UCC27322's datasheet where it states the input to the UCC27322, it is a "bit" obscure in it's description of the required input voltages.

It states the UCC27322 input voltage to input pin 2 as "" -5 to 6V "" OR VDD +0.3V, and then further confuses me with the statement "WHICH EVER IS LARGER".

Does this mean you have to either power the input with a 5V logic pulse for example, (falls within the -5 to 6V recommendation) OR lets say if VDD is 12V, then you can also power it optionally with a 12.3V input signal??

Usually I'm pretty good about reading these things, been doing it all my life, but it's always been hammered into my head that rail voltages are NOT to exceed supply voltages, so I'm a bit confused. I hope you can clarify this for me. I hope I'm not just asking a totally OBVIOUS question, and look foolish in doing so, I think I'm slipping in my old age of 42....LOL.

I would REALLY appreciate your Eagle cad files, but I guess I'll have to install Eagle again, I've been thinking about doing so anyhow, but keep toiling with whether or not to install Protel instead, as it's what I've always used in the past.

By the way, I've asked this question and not had an answer, what frequencies should the PWM be running at, and why? Is there a pulsed duty cycle also, like a certain on/off waveform of a certain frequency? Such as a pulsed 44Khz waveform for example. If so, what is the duty cycle, what are the frequencies??

Thanks again guys, I DO appreciate all your help!

Philldpapill, I don't mean to sound like a nag, but I would REALLY appreciate an answer to the questions I've asked you about the UCC27322 usage in your circuit in the post above that you did not address with a reply as of yet.

Your experience with the IC is greatly respected, and I'm waiting to hear back from you about how you used it "before" I try building a prototype circuit using the UCC27322 IC.

I hope you can take the time to answer the questions, I would truly greatly appreciate it!
Thanks

Chris

[canpower]

Hmmmm.....
looks like I hit another brick wall.