PDA

View Full Version : Retrofit PWM to Throttle



sjrobinson
01-20-2013, 04:30 PM
Im not sure if I touched on this before. While I'm rather inactive with my HHO project I'm still wondering about this subject. Has anyone been able to rig a pwm to the throttle body that way HHO is only produced in direct correlation to the amount of air flowing into the engine?
Granted the production would not be as fast to respond as the engines air intake but when cruising with a relatively slow change in throttle position this could prove beneficial as to not waste HHO and to provide the ideal amount. Any thoughts?

Stevo
01-23-2013, 11:32 PM
Yep, better yet it is MAP sensor based (0||5v). I will post details in a new thread once I get the time to complete the install. Not enough hobby time lately :(

sjrobinson
02-18-2013, 01:00 AM
So you would reroute the map sensor signal wire to the PWM to cause the amps to the cell to increase or decrease with the voltage of the MAP sensor readings?
That is an interesting approach.

My idea is more mechanical I suppose. And then HHO would not be wasted when the throttle is closed and you are still moving in gear. itd be a simple retrofit to the throttle body. or for a signal based like the map sensor, fitting it to the TPS signal wire.

myoldyourgold
02-18-2013, 10:55 AM
Stevo, I have found on some vehicles the MAP signal changes very quickly almost erratically depending on both the vehicle and the driver. So much so, that the reactor can not react that fast and in one case was on seemingly most of the time except with large changes over extended periods of time and in another was off for long periods of time. I also found that in some cases it worked somewhat satisfactory but still not as good as I wanted. To make things more difficult the vehicles were commercial vehicles that had very little down time so had to do most of the testing while they were working, which was good for some testing but not for others. I am still working on a solution. I first thought it was a faulty MAP and changed it out but that made no difference on all the vehicles that were having a problem. The other problem is that there is a number of vehicles that use a frequency signal instead of a 0-5 volt signal. I am sure there is a solution and will find it soon because I am going to have some good equipment to test on here shortly and some expert hands on help to speed things up.

sjrobinson
02-18-2013, 04:55 PM
So if there are issues with doing it electrically, why not mechanically?
It wouldnt work with diesel engines, but gas engines I see no problems during highway driving.

myoldyourgold
02-19-2013, 10:57 AM
There are issues doing it no matter how you do it. They are all solvable though just takes time and money. LOL The problem is the delivery of on demand HHO is not as controlled as the delivery of fuel and air. Because of the difficulty to maintain the ratio of air/fuel/HHO over a wide range of engine demand, there is going to be problems, trade offs and plenty of room for improvement. Using the MAP/MAF is one of the methods that is used by the current fuel systems and is very accurate. I am sure a solution will be forth coming to control the HHO using the MAP/MAF and other sensors.

sjrobinson
02-19-2013, 12:46 PM
Yeah I see. Id try the tb retrofit myself I see difficulties in fabricating it myself.

Stevo
03-25-2013, 12:13 AM
Man, I keep forgetting to manually subscribe to these threads I post in. Sorry guys!

Anyway, myold I am working on an Arduino (open source micro controller) based solution and am currently writing the software to allow use of another very exciting open source project that plugs into any OBDII CAN (or older) bus. With this, we can read way more than just MAP, but also just about every sensor (which will come in standardized OBD data format). The cost of the system is approximately $120 USD if you source the parts yourself. So when mentioned MAP, I also meant any other readable sensor such as MAF or even TPS. I fully plan to open source the controller code on GitHub once ready for alpha. Right now my tests are utilizing a China made 50A (peak) robot motor driver module which is controlled via Arduino PWM. This system will not be like anything I have seen and the logic will be 100% tweakable granted you have the understanding. I could build in some potentiometers for min/max and gain control (to name a few), but I would have to charge for that system as it takes time to develop and I am in the red concerning time these days.

myoldyourgold
03-25-2013, 01:43 AM
Anyway, myold I am working on an Arduino (open source micro controller) based solution and am currently writing the software to allow use of another very exciting open source project that plugs into any OBDII CAN (or older) bus. With this, we can read way more than just MAP, but also just about every sensor (which will come in standardized OBD data format). The cost of the system is approximately $120 USD if you source the parts yourself. So when mentioned MAP, I also meant any other readable sensor such as MAF or even TPS. I fully plan to open source the controller code on GitHub once ready for alpha. Right now my tests are utilizing a China made 50A (peak) robot motor driver module which is controlled via Arduino PWM. This system will not be like anything I have seen and the logic will be 100% tweakable granted you have the understanding. I could build in some potentiometers for min/max and gain control (to name a few), but I would have to charge for that system as it takes time to develop and I am in the red concerning time these days.

Stevo, that sounds really good. It is what the small vehicle needs. I only use the small vehicles to test on because the 30 plus liter engines/vehicles would not fit in my shop. LOL The other problem is they are all 24 volt. I am also looking for simple foolproof solutions to get the right amount of HHO for the right demand. The exact ratio is not known because it depends on so many things. On the big engines dividing the demand into 3 or 4 sectors yields better results in all of my testing of these big diesel engines. More HHO at higher demand and lesser amounts at lower demand. Each engine has to be tuned in those 3 to 4 zones to see the difference. Once you have it right you can enjoy some good gains. Even if you are off a little you are still smiling because theses burn over 50 gallons an hour at the higher demand level and they work up there a lot. Saving even 10% is worth it over a year of use. More of course is better!! The high amp draw on the big engines is like a fly hitting the wind shield. There torque is so much that 200 to 300 amps does not seam to change the fuel consumption by any meaningful amount when adding HHO. 20 to 30 LPM takes some amps no matter how efficient you are. I have a 12 volt controller that can handle 2 zones and off at idle which works on a 0-5 volt sensor. I will be doing some serious dyno testing in the next couple of weeks or so on a 7 liter diesel 12 volt system. This will give some very much needed information. I can increase its zones by using PWM's and multiple reactors. The use of multiple reactors is always more efficient compared to on large one. The results should be interesting. I am not convinced that the MAP/MAF is the best single. These up coming tests will sort all of that out. With a consistent volume of HHO it shows gains in all ranges of demand which puzzled me at first on prior tests. I then figured out that it is possible but there is much more to be had with the right ratio based on demand. At one point you are getting the max gain for that point or demand and all other points/demand, just minimal gains but gains just the same. This is all measured with a very sophisticated flow meter and if in a vehicle, a GPS etc. It is then plotted in a graph so you can see the difference when HHO is on and when it is off going over the exact same route with the same load if in a vehicle. It plots the route also to make sure it is the same. Expensive equipment but necessary. You have to be much more precise on a small car where an extra suitcase or passenger robs you of mileage when it comes to how much HHO is need for a given demand. Will be looking for more of your results.

sjrobinson
03-25-2013, 09:03 AM
What makes what youre describing more than just an OBD2 reading program like torque? Or did I read that wrong

What id suggest- I think Ive brought this up before, is to reprogram a standalone ECU for HHO. Enabling you to reprogram the fuel trim as well as response from sensor input. I'm sure if you got a digital PWM you could incorporate it into the standalone ECU. If that makes sense.

Stevo
03-26-2013, 12:19 AM
Well, for me, I just use an OpenPort 2.0 to put a custom tune on. Pretty easy since its a Subaru.. As far as ECU tech is concerned now is a great time to get open source hardware such as Raspberry Pi and others to take care of everything else such as sensor tweaking and PWM. You would want to have your ECU chipped for most makes. My previous Subaru yielded reasonable gains and better in-town drivability with what I would consider a sub-par HOD system mounted in the trunk WITHOUT modifying the OTS Stage 2 maps. What worked best? Resetting the ECU after the install and following ECU quick learning procedure after the car was at normal operating temps >=180 F. And no, there is no standard procedure across the makes. Each make has its own procedure and I imagine each procedure could very well yield different results.

There is not an off the shelf hardware/software combo for sale that I know of. I can assure you that the hardware and programming languages to build such a unit indeed is. Takes time, dedication and know-how.. We are all tweakers here when it comes to building and exploring this technology.

sjrobinson
03-26-2013, 12:04 PM
I imagined you could do something similar with a chipped obd1 ecu from honda. Like how I can use different free programs to reprogram the fuel map for a p28 ecu that will work with my obd2 f22b1 engine. Thats more used for performance but can be tweaked for hho.

Free software and $100 to buy an already chipped ecu and writable chip.

Stevo
03-26-2013, 02:07 PM
Hondata? I looked into that mod for my Civic a few years ago. Definitely helpful for re-mapping.

You would need pin outs to be able to send analog signals to the microcontroller unit in order to control the PWM duty cycle for HOD though. Not sure if that is an upgrade option.. I know that AEM offers that, but their ECU's are über expensive. :eek:

sjrobinson
03-26-2013, 07:22 PM
Yeah. AEM ecus run about $1500. But I'm not quite talking about hondata as you specify to hondata and they rewrite the fuel map for you. Or reflash the ecu in the k series honda engines.

What you can do with hondas is take a stock ECU from their obdI cars and use it with a compatible OBDII car (if you need it for OBDII), or just use the ECU in the OBDI honda. My 97 accord can use an integra/civic p28 ecu. They sell such ECUs altered with removable chips that can be written on using fuel mapping software.

You can also pick up the stock ecu at a junkyard, solder a chipping kit on and do it all yourself.

With this method, I think 90s Hondas would at least be the easiest to modify for hho anyways.

daleh007
04-05-2013, 10:14 PM
Would using simple RPM measurement make any sense? At least under more constant throttle settings you would know how much volume per unit time is needed to fill the cylinders and could try metering an appropriate amount of HHO to match? I've also been thinking about this but have an older carburated motor.

Stevo
04-06-2013, 02:16 PM
Yep, you could use the analog voltage signal as an analog input to drive the PWM input.

sjrobinson
04-07-2013, 07:41 PM
On the accords there is a point in the harness where you can hook a tachometer up to. That could used. But that doesnt accommodate for engine load. It is a start though.

I guess it depends on the most efficient mix of HHO.
Is it best as a ratio to fuel intake? Air intake? Engine speed? Or even the amount of fuel burned measured by o2 sensors?

Stevo
04-07-2013, 08:15 PM
I think RPM would be best unless you have a really crazy fast hydrogen generator. Most generators won't respond fast enough to provide a really accurate ratio. It is, however, a lot better than running the reactor 100% the entire time. At idle the reactor should be producing nothing. My complicated setup will have the benefit of integrated LCD as well as logging so I don't mind the extra wait time and preparation. If you don't care about those things then you can avoid the hassle.

You would just need a micro controller to accept the 12V analog signal from the tach port and output digital PWM 0/5V signal to the power module. Any Arduino can do this very easily.

myoldyourgold
04-07-2013, 10:09 PM
Most generators won't respond fast enough to provide a really accurate ratio.

Stevo is 100% correct about not being fast enough to get the ratio 100% accurate over 100% range of the engine. This is the problem using the MAP or MAF signal which are all over the place and change to fast. They are still the most accurate and with the proper electronics can be used. I think in small engines you need to look very seriously at the torque curve and have max HHO at max torque. Off at idle, closed throttle, or just above idle. Divide the remaining up into 2 or 3 zones and you will see great results if you have the ratio right in each. This can all be done very simply using an adjustable vacuum switch or pressure switch if turbo charged. Cheap and easy. You will need to use solenoid valves to keep the pressure in the various zones when off which helps limit the lag time when they come on and off in the different zones. The pressure is very little more than is there all the time so is safe. This requires multiple reactors and/or pwm's so one reactor can be run in two separate zones. The smaller amount off the pwm and the larger amount off electrolyte concentration based on vacuum or pressure. This will cut down the number of reactors. This has proven to be the most cost effective method for me. I am still working on a controller to achieve this in one device though. Some serious dyno time is going to be required.

WHEN YOU USE SOLENOID VALUES THE REACTOR MUST BE WIRED SO IT CAN NOT COME ON WHEN CLOSED!!!

Stevo
04-07-2013, 10:23 PM
Solenoids are a great idea. Really easy to activate via the controller too. Torque curve on my turbo boxer motor can be read from the OBDII data so that should be easy. I will probably experiment with the algorithm A bit after the first install. Still waiting on electronics to arrive from China..

sjrobinson
04-09-2013, 11:37 AM
My original thought was the reaction time to produce HHO. But the system would be for cruising where the throttle doesnt vary greatly.

What about a butterfly valve with a storage container that opens in direct relation to the throttle. There would be a reservoir the HHO would go into that would store excess gas that is ready to be sucked into the engine. That way you dont have to worry about the on-demand need as the HHO would already be produced ahead of time.

Of course storing the HHO would be dangerous but if left at a low pressure it would be fine.

So you have the HHO produced in intervals by the cell and contained. You would have to figure out how much HHO is being used based on how much the engine is taking in and create a formula used so that the system knows when to produce more HHO at what amperage.

But Im making it complicated... Could work though.

freetv6969
04-10-2013, 07:42 PM
i cam with a simple idea
using pressostato, it´s a valve that activate/deactivate one electric contact with air pressure.
you put a T in the gas output and the car vacuum activate/desativate the generator.
can´t be more simple

sjrobinson
04-12-2013, 05:34 PM
That is a great idea. Simple, I like it. But that doesnt give the spectrum of amps to HHO production needed, just on and off. Which is why I stuck with the PWM retrofit to something.

Stevo
05-14-2013, 11:07 PM
Got my OBD-II to Arduino setup working as desired with proper prototype shield, custom box, 3.3v OLED mounted in a cheap gauge pod and proper Molex connectors to sum it up. I have easy access to most all OBD-II PIDs including throttle position (which I have considered).

I have been considering taking a normalized average of these PIDs to determine the final signal duty cycle to the motor module:


Throttle position
Air Pressure
Calculated Engine load


I think I might still give throttle position approximately 50% overall influence in order to better handle city driving, but I might need to factor in things like vehicle speed and RPM in order to ensure optimal system output on the highway.

I still need to complete the custom motor module box design and print it as well as find a decent ammeter if the cheap Chinese motor module doesn't properly report amp draw. We shall see.. more to come..

myoldyourgold
05-15-2013, 04:44 AM
Sounding better and better Stevo! Will follow it closely.

Stevo
06-23-2013, 04:08 PM
This is my latest power mosfet prototype. What this box does is simply control the current to the drain (in this case it's the reactor) based on a microcontroller 5V PWM signal. Since a microcontroller can listen to the car's computer via standard OBD-II interface, it can respond to various sensor readings such as throttle position by adjusting the duty cycle of the signal and thus controlling the current supplied. One thing I really like about this design is that it provides a high level of modularity and this device can be used in just about any application where power needs to be controlled variably via PWM signal. So you could use one of these to drive a water injection system as well. :)

http://www.hhoforums.com/attachment.php?attachmentid=2300&stc=1&d=1372017757

Quebecker
06-25-2013, 08:52 AM
This is my latest power mosfet prototype. What this box does is simply control the current to the drain (in this case it's the reactor) based on a microcontroller 5V PWM signal. Since a microcontroller can listen to the car's computer via standard OBD-II interface, it can respond to various sensor readings such as throttle position by adjusting the duty cycle of the signal and thus controlling the current supplied. One thing I really like about this design is that it provides a high level of modularity and this device can be used in just about any application where power needs to be controlled variably via PWM signal. So you could use one of these to drive a water injection system as well. :)

http://www.hhoforums.com/attachment.php?attachmentid=2300&stc=1&d=1372017757

Good Joob Stevo

I think you're on the right track !!!

sjrobinson
06-25-2013, 05:32 PM
So that basically alters the amperage to the cell based on the data?

Can I see the electrical schematics for it or is this a patent pending product?

Stevo
06-25-2013, 11:05 PM
It's an N-channel MOSFET box that includes the voltage regulator for the gates. Pretty simple really. The difficulty is putting it all together in such a small form factor. This one is good for 25A continuous current. Change the fan out and you can go 24V if you wanted to.

The micro controller is a separate unit. I guess my original wording makes it sound like the mc is included. Oops.

Google: "Arduino N-channel MOSFET"

sjrobinson
06-28-2013, 09:49 AM
Ok. Thanks.

Stevo
06-30-2013, 03:27 PM
I've got a rev 1 PCB designed now and am working on the case. This will be a pretty darn rigid one and I made sure to include some pretty large wire pads (up to 10 AWG) and did a partial override to the thermal reliefs on the source and drain traces in order to achieve less resistance. Boards should be here soon and I will go over details.

Here is the main thread I am starting over at HODInfo:

http://hodinfo.com/forum/topics/smart-hod-system-for-obd-ii-vehicles

Stevo
08-03-2013, 05:32 PM
I completed the bench test against the reactor @12A. Unit ran at room temperature and very smoothly at 173Hz from 0-100% duty cycle. I need to finish my stainless bubbler then it's install time.

Vr4-NIGEL
08-04-2013, 04:01 PM
Good work guys you've helped me no end as it was a problem i have as i know i need some way of controling hho input over the rev range as not to over lean mixture i know i have to use evoscan to tune fuel maps but could not see how i could contol hho but will be intrested in seeing how this pans out :cool:
stevo would you be willing to build one of these of others namely me lol ?

Stevo
08-04-2013, 08:20 PM
would you be willing to build one of these of others namely me lol ?

I will be posting this as a kit some time within the next 8 - 12 weeks (depending on my work schedule). The current model is getting a big upgrade with current sensing and ability to handle 60A without getting any larger. Stay tuned..:D

Vr4-NIGEL
08-05-2013, 01:03 PM
Cool will keep an eye on here then i think what you have done will help me control the hho that my engine receives at anytime giving me more peace of mind

Stevo
08-05-2013, 01:06 PM
Should also help provide better average results as running a reactor at 10 or more amp draw at idle (when there is little or no load on the engine) is not going to save you anything.

Stevo
08-28-2013, 10:20 PM
Added a current carrier for detecting amp draw, wrote a simple test to limit current to 12.5 amps and threw in a cheap Radioshack project box.

http://hodinfo.com/forum/topics/smart-hod-system-for-obd-ii-vehicles?commentId=2068750%3AComment%3A138045&xg_source=activity

http://api.ning.com/files/QuP-U7H6v7Y8mf634iyzi6yeXCvc71lA-zEnEcQ-7jwi-ZXqmKYw2Xb1oe2cliIHHwkobdpNS-WuUYZ-Yr6BIq81U9EGfPM6/DSC_0066.JPG

Stevo
11-21-2013, 11:44 PM
I found some time to get this completed (minus the display). I have successfully retrofitted the PWM to the throttle signal plus mapped RPM and engine load. Lots of other things I could map it to as well. The best thing about this setup is it turns off 100% reliably at a stop while idling. I'm in the middle of a mileage tainted tank right now (lots of idling with the system on for different tests), but will begin with an ECU reset and full tank of gas. I expect 20-30% range increase while commuting similar to what I realized in my previous car of the same earlier model.

Sorry about linking to the thread instead of posting here. This site doesn't provide the writing features I need yet.

http://hodinfo.com/forum/topics/smart-hod-system-for-obd-ii-vehicles?commentId=2068750%3AComment%3A138987

Stevo
12-16-2013, 03:14 PM
Sorry about linking to the thread instead of posting here. This site doesn't provide the writing features I need yet.

So, I'm just going to port the blog entry over to this account to help spread the idea/word. Here is the link on these forums: http://www.hhoforums.com/entry.php?4-Arduino-Controlled-PWM