I'm just wondering if you couldn't control the duty cycle of a PWM with rpm's to create a hydrogen on demand system.

Russ

alpha-dog Wrote:I'm just wondering if you couldn't control the duty cycle of a PWM with rpm's to create a hydrogen on demand system.

Russ

Talk to this guy.

http://www.fuel-saver.org/showthread.php?tid=2126

HI ALPHA DOG:
Been playing with those ideas for a few years myself. The tach pulse is good place to tap into something that is relative to RPM.. another is to place a few magnets around the drive train and a magnetic latch switch connected to a switching transistor ..thats a pain in the ass, but it does give you and on and off proportionate to speed which is an direct indicator of how many rpms..... I like using the pulse to switch a scr that controls the input current to the generator.. as the speed increases so does the pulses ..henceforth the scr switches more often. allowing a more on current state than off.. as you slow down the pulses decrease henceforth the opposite occurs more demand ..more current through swicthing.. less demand . less current because of switching speed. Hope this helps you some. Mike of MOBILEHYDRICITY htp:/www.freewebs.com/mobilehydricity

Our little research group and I have discussed this before too. Good suggestions there by MobileH - I thought that the TPS may be used in some way too, or a mechanical linkage connected to a pot from the throttle.

(03-29-2009 10:13 AM)mobilehydricity Wrote:  HI ALPHA DOG:
Been playing with those ideas for a few years myself. The tach pulse is good place to tap into something that is relative to RPM.. another is to place a few magnets around the drive train and a magnetic latch switch connected to a switching transistor ..thats a pain in the ass, but it does give you and on and off proportionate to speed which is an direct indicator of how many rpms..... I like using the pulse to switch a scr that controls the input current to the generator.. as the speed increases so does the pulses ..henceforth the scr switches more often. allowing a more on current state than off.. as you slow down the pulses decrease henceforth the opposite occurs more demand ..more current through swicthing.. less demand . less current because of switching speed. Hope this helps you some. Mike of MOBILEHYDRICITY htp:/www.freewebs.com/mobilehydricity

Main problem with using a direct pulse train in the manner you describe is that, while you are varying frequency, the main method of controlling current flow is normally done by varying the duty cycle, and not frequency.

Unless you also vary the duty cycle, varying pulse frequency will have little effect on the average current you supply to your generator.

e.g. 50% duty cycle at 1000Hz provides the same average current as a 50% duty cycle at 2000Hz.

OK. So the above is not 100% correct. If triggering a fixed width pulse, then you will have control over the average current, but this still limits the frequency range over which you have control. above a certain frequency and your output from control signal will be more or less continuously on, meaning that your generator will be at maximum output above this frequency..

Have a look at triggering a 555 ic, with fixed width output pulse, from momentary input pulse. (Switch de-bounce circuit). Should give some idea of what I am trying to describe.

My own preference would be to use a microcontroller to read input signal, and generate varying output duty cycle using PWM.
Bear in mind that efficient switching using fets can be frequency dependent.

(04-01-2009 04:16 AM)benny Wrote:  

(03-29-2009 10:13 AM)mobilehydricity Wrote:  HI ALPHA DOG:
Been playing with those ideas for a few years myself. The tach pulse is good place to tap into something that is relative to RPM.. another is to place a few magnets around the drive train and a magnetic latch switch connected to a switching transistor ..thats a pain in the ass, but it does give you and on and off proportionate to speed which is an direct indicator of how many rpms..... I like using the pulse to switch a scr that controls the input current to the generator.. as the speed increases so does the pulses ..henceforth the scr switches more often. allowing a more on current state than off.. as you slow down the pulses decrease henceforth the opposite occurs more demand ..more current through swicthing.. less demand . less current because of switching speed. Hope this helps you some. Mike of MOBILEHYDRICITY htp:/www.freewebs.com/mobilehydricity

Main problem with using a direct pulse train in the manner you describe is that, while you are varying frequency, the main method of controlling current flow is normally done by varying the duty cycle, and not frequency.

Unless you also vary the duty cycle, varying pulse frequency will have little effect on the average current you supply to your generator.

e.g. 50% duty cycle at 1000Hz provides the same average current as a 50% duty cycle at 2000Hz.

OK. So the above is not 100% correct. If triggering a fixed width pulse, then you will have control over the average current, but this still limits the frequency range over which you have control. above a certain frequency and your output from control signal will be more or less continuously on, meaning that your generator will be at maximum output above this frequency..

Have a look at triggering a 555 ic, with fixed width output pulse, from momentary input pulse. (Switch de-bounce circuit). Should give some idea of what I am trying to describe.

My own preference would be to use a microcontroller to read input signal, and generate varying output duty cycle using PWM.
Bear in mind that efficient switching using fets can be frequency dependent.

Couldn't agree with you more. In the thread , I was refering to a simple and basic concept I tinkered with from many years past ... And your right .. there is a point where as the pulses are so fast that the switching device appears to saturate ..being on most of the time. I prefer sampling and comparators to an a to d
then process the info like it was a ramp with a given known for a set of input variables. But for beginers , the magnetic latch is cool to experiment with... For the most part I have ussually seen square waves out of the ecu tach which can make a pretty a good ramp to start with for beginers . Microprocessors are the bong .. Pics and even basic stamps are all that is needed. My focus lately has been more on fine tuning the timing of events . I haven't found it to do much good to fire everthing up at the start of engine ignition. Most ecus are in a closed loop mode until the engine temp warms up... I recently posted a circuit using the old efie model with a temp sensing circuit that has an adjustable threshold. I personally , found that the incidence of the ECU staying in closed loop went down and furthermore , overall I showed a somewhat better performace , turning the HHO on after the engine has reached operating temp. Any rate ..gotta run for now.

This appears to have been done, and it is for sale on ebay. They tie the PWM output to the map or maf.
http://cgi.ebay.com/ebaymotors/ws/eBayIS...MEWNX%3AIT

Why buy it when you can have fun doing it yourself for twice the cost

I have a Parallax basic stamp bs2 with pwmpal and I almost have enough info to build my own pwm (thanks to Benny).. The programming part is the least of my worries. I have managed to read the RPM signal from my ECU as well as my MAF signal which is just a frequency based 5V square pulse. You can then make up any sort of formula (even include an offset from a temperature probe) to create an output pulse of varying frequency or duty cycle.. however I'm not sure how to connect it up to these mosfet thingies..

This is what Ive got so far.
http://www.fuel-saver.org/Thread-How-to-make-a-PWM

P.S. Just bought myself a USB plug and Propeller Proto Board for about half the price of the stamp setup and it can supposedly do 8 processes simultaneously.. It would be perfect for running a PWM, O2 Effie, MAF frequency modifier and more..
http://www.parallax.com/Store/Microcontr...roductName

I totally agree on the point of doing it yourself. It may take some time, but the lessons learned through success and failure will bring much understanding to what you are doing.


I've seen a PWM or two for sale that can be controlled from a microprocessor. Your Stamp should do this just fine. Just read the RPM/MAF or whatever with the Stamp and hook up an output pin to the PWM. No MOSFETs required in between. I'll see if I can find that specific PWM again....


Personally I think the Stamp is a bit pricey, but very easy to program.

I've used the SX chip as it is about 1000 times faster than the Stamp and costs about $4.00 each. The programming language is SX/B which is about 98% the same as the BASIC used for the Stamp. I did start with the Stamp (the intro kit for sale at Radio Shack). Then I found out that the SX was much faster and switched to it. Only problem is the SX just had it End of Life (EOL)announcement. Parallax will place a huge order to ensure supply for about 5 years to come. I think its always possible a large order may be placed that could wipe out than inventory. Now I'm going to learn the Propeller, it sounds like a great processor.

Edit - 11-23-09 - correction