More features suggestions for PWM product

Philldapill

A while back, someone did a post asking how to make a power PWM circuit for hydrogen production. I put in my two cents, and figured I'd go on ebay and see if anyone sells a simple thing. Wow. They do.

http://shop.ebay.com/?_from=R40&_trksid=m38&_nkw=PWM&_sacat=See-All-Categories

I'm just amazed at how much these things sell and the price! So, I figure I could make one better with all the bells and whistles. I've got some MOSFETs with a 0.0035mOhm Rds rated for 75 amps, although I know I won't get anywhere close. Most of these controllers have a fixed frequency, and adjustable duty and that's it. I'm making a list of features that I could add fairly easily and for hardly any money.

1. Adjustable Duty Cycle(obviously)
2. Adjustable Frequency(maybe for other inductive loads?)
3. An LCD screen that displays Duty and Freq.
4. A couple ADC's that measure voltage and current and display these as power, voltage, current on LCD
5. Four buttons to increase/decrease Freq. and Duty
6. Possibly use the ADC's to allow the user to set the desired current(Prop adjusts duty to output desired current)
7. Circuitry and all housed in a nice pretty enclosure($10)
8. ???

I'm sure the propeller can do all this stuff and more(very easily too!). Suggestions very welcome!

By the way, I've been thinking about how to maximize my MOSFETs. In particular, I'm using a driver chip for them, but I want to go a little further...·The initial transition from off to on dissipates almost half the power. In a steady full-on state, the mosfet barely gets warm at 20A. I figure that if I add an inductor right before the FET, and a diode in antiparallel with the indcutor, I can reduce the power dissipated in the mosfet by allowing a sort of delayed current flow as it turns on. However, when it turns off, the normal power would be dissipated, and the diode in antiparallel with the inductor would soak up most of the energy in it. Here's a schematic.

Anyway, I'm just looking to make a buck and the propeller seems perfect for all this. What do you think?

jazzed

It seems that the "ZeroFossilFuel" oscillator design would be much more efficient than a Propeller PWM because of feedback design. However, it does look like one could use a Propeller or other micro-controller to monitor with an ADC and adjust the oscillator properties with digital potentiometers. I still don't see the advantage of using up to 30A to produce 2 liters per minute of Hydrogen though in an automotive application. How much power can be derived from an alternator? For comparison, doesn't a starter take like 40A to turn an engine? Emperically, at 40A one lead-acid battery discharges pretty fast.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

Philldapill

jazzed, I totally agree. Producing hydrogen for an automotive application on board is just silly to me. I'm not a believer in the whole thing, but I do see a market for a power controller. I want to make one with all the bells and whistles so that people that ARE into the hydrogen thing can feel like they are doing some thoroughly scientific and groundbreaking work.

As for the oscillator... The resonant frequency of a circuit in those designs is up in the MHz range, yet so many of these garage expirements that claim to "have done it" use some arbitrary 20kHz(static) frequency. I'm not even suggesting that these people try to hit that sweet spot. The PWM is simply to provide a means of controlling current. If they want to do some expirements with pulsing·different frequencies through the water, hey, that's there thing. I just want to supply them with the means.

TJHJ

Hello all, Sorry to knock the idea here (If this is not allowed to be placed on the forums, please remove it or tell me and I will. ) One of the largest challenges you will run into is dealing with the heat, and keeping the mosfet's alive.
but we are about two weeks away from releasing our PWM board kit (high current load), Some of the features you talk about are included. Were running reliability tests right now, I have one sitting on a table switching 50 amps that has been running so far for 44 hours straight, with out a problem yet.( Woot....)

Adjustable PWM and Frequency (100hz- 100khz.) and duty (0-100%)
Pulse on Pulse configuration.
Fixed Duty cycle or closed loop current control.
50+ amps of switching power. With proper heat sinking and current limiting.
0-5v adjustable limit on off switch.
Oscilloscope display for configuration to a VGA monitor. ( With true triggering and frequency measurement)
Digitally stored values for constancy.
Full menu based configuration.

The cost currently is looking to be right at $200. If there is interest here as soon as I can get the online store up and running I can post the link, or Send me a PM for more information.

TJ

*edit* The price is a bit higher than ebay but its built to actually last, we have spent a lot of time developing a reliable product that wont nuke its self with in 5 min of use. We purchased a few of these "ebay" kits of varying designs, its quite funny to watch them melt when you try and use them anywhere near their limit.

Phil what duty are you running the mosfets at? To have them not generate much heat at 20 amps you got skills I need, Using twin 75 amp fet's anything over 30% we saw a rapid increase in temperature and associated failure.


Disclaimer.
I or any employee associated with this project can not be held responsible for the actions, damages or other harm from the use of this product.

Post Edited (TJHJ) : 8/6/2008 7:20:40 AM GMT

Philldapill

Wow, TJ, sounds pretty nice! You are right about keeping the mosfets cool. I'm thinking about adding a temperature sensor right on or near the FETs, so that if they get too hot, the controlling can stop operation. I know the propeller can handle that...

I've been thinking this over, and it seems that as you increase the frequency, there will be more switching cycles, leading to more heat being dissipated. It only seems logical that there should be some kind of derating of current vs. frequency. It seems the relationship SHOULD be linear since the relationship between switching cycles vs. energy dissipated is linear. I guess there is another feature... automatic frequency vs. current limitation calculations?

Is your project being done soley with the propeller, TJ?

Philldapill

This is at 100%. My FET driver seems to look a little screwy on the scope under the recommended conditions - i.e. with capacitor bypassing. The signal looks fine when it is high, but when it goes low, the signal oscillates at VERY high freq between high and low, which gets the IC and FET hot. The fet's do the best under 100% duty because there is no switching. I think the gate capacitance is something like 4.9nF so it's kinda how and takes a while to charge, hence a bit of dissipation during that time. I admit, they do get a little warm to the touch when under 20A continuous, but nothing that will burn you too bad. @ 20A, 1.4W isn't THAT hot...

A little off topic - I bought these FETS in a box of 800 for $50 on ebay. I sold about 700 of them for about a buck a piece and man I wish I hadn't... These things are absolutely amazing and haven't found anything that matches it since. Not even the IRF3205(110A, 55V). What FETs are you using TJHJ? And what driver? Mine is a sample from TI(like nearly every component I have for tinkering LOL)

TJHJ

1. Yes it is being built solely on the propeller. What else can pull of a semi real time scope effect [noparse]:)[/noparse]

IM using a logic level mosfet, but at 20 amps im having to dissipate alot more heat than you a bout 7 watts, not really that much but we saw a lot of early failures when there should have been no reason, after figuring out how to keep them cool they have maintained a nice reliable state. I ve got to find some of these E-bay fet's, lol.

The switching time for the fet I just did with trial and error until I got the relationship right, now all I hope is the next batch I order maintains that same spec.

Best of luck

Philldapill

Futurlec.com sells IRF3205 fets for fairly good pricing... you get a $0.10 discount when buying 25 or more...

1+ - $0.95
25+ - $0.85
100+ - $0.75

That pricing isn't bad... and the Rds(ON) is like 8mOhms.

hinv

Hi Phil & TJ,

I didn't hear how your projects went. Do you have them completed?