How to use LM556 PWM for Freq & Duty cycle?

If you want to design and build your own driver for the learning experience, I heartily suggest reading Laszlo Balough's article.

You can also purchase driver chips.

Thanks, some interesting reading for later (Bookmarked).

Bill_Marsden said:

I'm not sure if I'm using the FET correctly, and there is absolutely no short circuit protections, but I believe this will do variable frequency and variable duty cycle independent of each other nicely....

Click to expand...

Bill, in your schematic you're using an N-ch MOSFET with the source and drain reversed; ie: it's upside-down.

Also, with a typical power MOSFET, you need to get Vgs to 0 in order to fully turn it OFF, and Vgs to 10v to fully turn it ON. The driver device output therefore needs to get down to the same potential as the drain. This pretty much means that the driver has to have rail-to-rail outputs, and a standard BJT op amp can't get there. The LF353 wouldn't work either, as it's not rail-rail. A LM6152 might for a low-power MOSFET, as it has rail to rail I/O and a 75MHz bandwidth.

However, as the power handling capability of a MOSFET goes up, so does the gate charge, Miller charge, etc. - meaning the gate requires more current to switch the thing on or off. In order to keep the switching time as short as possible to avoid operation in the heat-generating linear region, you need a driver that can source or sink a lot of current in a short amount of time to overcome the capacitance of the gate.

Use something like a LM353 for the op amp, which is a drop in replacement for 1458's, but faster.

Click to expand...

I believe you meant LF353 here. Maybe LM358?

Could be, I'm still digesting the article ThingMaker3 pointed out. If I read it correctly you don't need push pull transistors to the FET, one should work, so why do they show two?

Yes, I meant the LF353. My old roots are showing.

I really need to bone up on these gadgets, they didn't have them when I was in school. I really like what I see. I'm going to pick this discussion up on another thread in a bit, maybe someone could be inspired to write an article for the ebook. It would also be a good subject for the experiments section.

Thanks again guys. You got me up late nite reading...
What about BUZ100 mosfet as its 50v 60a 250 watts and only $1.39 ea.
Or, IRFZ1010Z 55v 75a 140w
And IRF3710 100v 57a 200w (2 for $1.00) compared to IRFZ34N 55v 60a 65w...

If I was using straight current (29a), I be in the 400 watts area. So going to have to double/triple up on mosfets.

OK, I decided to keep it simple and use a 555 to drive the MOSFET, that and corrected the polarity. I really should have stuck with something I know and used a BJT. Schematic check please?



The 2 555s could be merged into a 556.

At the risk of being an ex-sailor at odds with an ex-marine, the N-channel MOSFET was right-side up the first time. Source goes towards negative, drain goes towards positive. They're good ol' voltage-controlled electron-flow-model gizmos!

Whether a 555 can drive one or not will depend on which specific MOSFET is being used.

I figured 200ma would be sufficent, capacitive load nonwithstanding. MOSFETs need more? About how much more do you figure? I was aiming for the rail to rail characteristics, and simplicity.

I have the source and drain labeled correctly at least?

I'll do another redraw in a bit.

Some power MOSFETs need an amp or more to turn on and off at usable rates. Depends on the Miller charge (which will be listed in the datasheet), and how fast one wants to switch the switch.

I never made it into the military thanks to my big bro - they didn't want another troublemaker in, not that I am
They purposely scrambled sections of a final test after a bunch of other tests days & weeks before, and I came up 2 answers short of 80%.
A 30 min test and I spent 5 minutes looking for the guy who gave me the test. Never found him. If I just stayed, I would of been in the naval air reserves - electronics area. Guess it was not ment to be...

Why would I need the LF353 then as one side of the 556 could do the freq and the other side the duty?
Would not the 1st pic I posted on page 1 at the top work, or did I wire it wrong?

As is, you have the 555/6 doing the frequency now with the LF353 doing the duty (and other side of 556 to drive mosfet).
If anything, I would think I would need a better chip for stable freq generating and let the 555 do the duty, or will it not pass higher frequency modulation through it?.
As is, 555/6 limited in max freq. as I would like to experiment up to 6.7mhz.
But thats another day in the far future...

Don't think I would need much in mosfet protection as little or no inductance loading for what I am doing... Wire lenght to plates would be apx 2'

When I get a chance I'll take another look at them, which will be about 16 or hours from now. I might even redraw them for clarity, I tend to be one of the local draftsmen around here.

The op amp in my diagram makes a nice triangle wave, which is easy to convert into a PWM signal, and linear to boot. It is also the basis of a Class D amp, which is a different subject. A straight 555 oscillator makes a sawtooth wave, which can also be used, but isn't as linear in the adjustment.

Elad said:

Would not the 1st pic I posted on page 1 at the top work, or did I wire it wrong?

Click to expand...

It is wired wrong. Pin 5 will be above 13V or will be ground, depending on where you are in the cycle of the first half of the chip. To control duty cycle of the second half of the chip, you need pin 11 to vary, under your control, between 6 and 11 Volts.

How about this simple circuit then (except 13.8v)?

Bill, I do like your circuit though

Well, I came here for advice and will go with Bill's circuit. Still need to order the parts yet (am a very busy guy ).
Anyway, need more information on Mosfets as not too sure what am looking at. Should I be looking at watts or max amps? (why am looking at IRFP260)
Am looking at these and for your input as to what to use. Like to have enough rating as not to worry about killing em...
Like I mention earlier, I do not intend on going over 30a max - but never know if water gets to hot and wife driving not paying attention, rather be safe so was looking at the top one...

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=IRFP260NPBF-ND 200v 50a 300w 20Vgs
or this looks good too (next line)?
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=IRL2203NPBF-ND 30v 116a 180w 16Vgs (5VDC on the gate will drive the device to saturation)(from: http://technology.niagarac.on.ca/people/mcsele/i2k.htm -again, hope not to be using any transformers for my app. so no inductance feedback)

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=497-5204-5-ND 600v 60a 200w... Vce(on) (Max)@Vge, Ic 2.5v@15v, 20a

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=IRFZ48RPBF-ND 60V 50a 190w 20Vgs

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=IRF1010ZPBF-ND 55v 75a 140w 20Gvs

Or, you have something else in mind for my application?


2nd) Also wondering why most are using DC and not AC instead.
It would double the shaking of the water over just a dc pulse. Don't think the water will care
But I do not see hi current P mosfets. How would I go about making that without transformers and then have to start worrying about inductance feedback from them?
Any circuit ideas on AC PWM/driving?

Look at both Watts and max Amps. Watts rating must be higher than your anticipated switching losses plus I^2*(on resistance)*(duty cycle). I like to go for 2x anticipated maximums, but I'm gun shy from blowing up too many MOSFETs when I first started playing with them.

For an AC PWM, try a search on "h-bridge." An h-bridge can be built with only N-channel MOSFETs if one uses a "charge pump." (That would be another handy search term, as I'm sure you noticed.)

pntrbl said:

I've got a buddy with a homemade HHO device on his Toyota pickup. It's only a 4 cyl. with a 4 speed but he did get a 31.5 mpg run out of it. No baseline to compare to tho so the juries still out .....

I don't think he wants to know.

What's fun is when he takes the output tube and sticks it in a bottle of soapy water.The resulting bubbles blow up like a firecracker! The strange thing is you don't see or feel anything like an explosion. No heat and with a cigarette lighter for ignition you're thumb's right in the bubble. No light to speak of altho it was daytime. A big bubble will snap so hard it'll make your ears ring.

Fun stuff.

SP

Click to expand...

31miles = 49.879km
1gallon = 3.78liter
13.195km for 1 liter?!
without using the HHO generator my car also can get the same mileage

btw, we use SI unit in malaysia....

linchiek said:

31miles = 49.879km
1gallon = 3.78liter
13.195km for 1 liter?!
without using the HHO generator my car also can get the same mileage

btw, we use SI unit in malaysia....

Click to expand...

hey linchiek, if your car is a late model one with injection and computer controlled systems and such, it will probably have a oxygen sensor on the exhaust system, detecting whether to increase the fuel mix or to lean it off. Because HHO actually introduces more O than the system is set for, that extra O---oxygen---for those that need to be told...will make the computer think it need to enriched the fuel mix going to the engine. If that is the case, which I think it is, you will need to add a EFIE..to bluff the Oxygen sensor circuit into letting the engine run as normal and not on an enriched fuel mixture, then you will see and improvement in your fuel mileage.

The basic essentials for a successful HHO system are 316L stainless plates, a PWM to control the amount of power being delivered hydrogen generator and a EFIE...electronic fuel injection enhancer, or otherwise called a oxygen sensor simulator...then you will have success...

let this be a short note to other doubters out there....HHO is here to stay and it does work.......ok..

daveinsingapore said:

let this be a short note to other doubters out there....HHO is here to stay and it does work.......ok..

Click to expand...

Please define the word "work" as used in the context of your claim. Some valid numbers would be most enlightening. Thank you.

Above is a circuit to trick your computer and lean out the engine, but requires each one for each O2 sensor.
My 97 Toyota Avalon XLS has 3 O2's
Instructions for this circuit is here: http://www.free-energy-info.com/D17.pdf
(Note that the schemetic above is a newer version than in the pdf file...)
For me, am only going to try this circuit below 1st and remove/ignor the check engine lite
Another guy up here has installed his Smacksbooster and is only doing this circuit shown below to lean out his oldsmobile to the point where is starts to run bad, switches on his HHO (apx 25 amp draw when warm) and engine smooths right out. So its doing something.
He's yet to run a tank of fuel as only lives a few miles from work.
I though will install an exhaust gas temp meter so I won't allow it to burn up my valves by being too lean....

We would all be interesed in the actual before and after mileage, along with an emissions test result printout.

BTW, the circuit I drew did have a problem, but I'm working on it. If you want the info it is at http://forum.allaboutcircuits.com/showthread.php?t=12405&page=2 starting with post #19.