PWM [or ANY OTHER] chip that can drive an H bridge?

[Ignoramus30105]

Gentlemen, I greatly appreciate your help. You helped me with so many
decisions already. I would not be where I am without it. However, I am
somewhat confused by something.

I am not really asking for very much. I cannot believe that I have to
go to great lengths or create esoteric solutions. I suspect (and hope)
that we are missing something simple.

Let me make it very clear as to what I want.

I want a IC chip that has the following characteristics:

1. Generates pulsed, square wave, preferably at 5V above ground
alternating with ground.

2. Can have its frequency controlled by a separate pot and able to
function around 30-1000 Hz

3. Can have its "duty cycle" (percentage of time it spends in the HIGH
section) controlled by another, separate pot.

4. Preferably use about 18VDC power.

5. Preferably is a DIP.

It may be called PWM or something else. Let's not get stuck on
designation here. What matters is that it does what I want.

Is it really true that such chips are nonexistent?

MAX038 does almost what I want, but it is a little expensive, requires
a different power supply compared to the gate driver and also
amplification of its output. Otherwise it is great.


i

 

[Ignoramus30105]

Gentlemen, I greatly appreciate your help. You helped me with so many
decisions already. I would not be where I am without it. However, I am
somewhat confused by something.

I am not really asking for very much. I cannot believe that I have to
go to great lengths or create esoteric solutions. I suspect (and hope)
that we are missing something simple.

Let me make it very clear as to what I want.

I want a IC chip that has the following characteristics:

1. Generates pulsed, square wave, preferably at 5V above ground
alternating with ground.

2. Can have its frequency controlled by a separate pot and able to
function around 30-1000 Hz

3. Can have its "duty cycle" (percentage of time it spends in the HIGH
section) controlled by another, separate pot.

4. Preferably use about 18VDC power.

5. Preferably is a DIP.

It may be called PWM or something else. Let's not get stuck on
designation here. What matters is that it does what I want.

Is it really true that such chips are nonexistent?

MAX038 does almost what I want, but it is a little expensive, requires
a different power supply compared to the gate driver and also
amplification of its output. Otherwise it is great.

I think that I finally found what I am looking for.

XR2206

http://www.exar.com/product.php?ProdNumber=XR2206&areaID=7

$3.59 at Jameco. It fits the bill perfectly.

i

 

[Jim Thompson]

I think that I finally found what I am looking for.

XR2206

http://www.exar.com/product.php?ProdNumber=XR2206&areaID=7

$3.59 at Jameco. It fits the bill perfectly.

i

I was going to suggest an LM339 circuit but, if you're happy, I'm
happy ;-)

...Jim Thompson

 

[John Woodgate]

I read in sci.electronics.design that Ignoramus30105

I want a IC chip that has the following characteristics:

If that's REALLY what you want, a 555 will do it. So I guess that ISN'T
what you really want.

 

[[email protected]]

I was going to suggest a 555, 50c for one at digikey. But if Jim is
happy, then I'm happy too.

Jeff Stout

 

[Ignoramus30105]

I was going to suggest an LM339 circuit but, if you're happy, I'm
happy ;-)

Thanks Jim... That IC looks reasonable to you, right?

i

 

[Winfield Hill]

Jim Thompson wrote...

I was going to suggest an LM339 circuit but, if you're
happy, I'm happy ;-)

Don't tell me you designed the xr2206 too?

 

[Joerg]

Is it really true that such chips are nonexistent?

You could make one out of a Schmitt trigger. 74HC14 or CD40106. Look at
the app notes for "oscilltor" that are often part of the data sheets.
Place another resistor across the one from input to output but with a
diode in series. Then you can control the duty cycle. That would come to
about 10 cents plus some discretes. Oh, and the potmeters.

Regards, Joerg

 

[Ignoramus30105]

I am happy too. Like I said, I will first make a circuit without the
timer chip and will use my wavetek for timing. I will test it without
using heavy current or high voltage on the power line.

Once (or if) that works, I will add XR2206 and proper pots and caps,
and after testing will try it for real to weld something.

If all works, then I will try to integrate this contraption with my
welding machine, by placing it inside the welder and connecting it
properly before the high frequency arc start circuit. That way I could
use arc start and DC inverter without IGBTs being damaged by high
voltage.

My plan for that is to modify that DC electrode negative/DC electrode
positive switch to be DCEN/AC/DCEP. If it is in the AC mode, the DC
inversion circuit that I built would be powered up.

I would, then, place the adjustment pots on the welder's front
panel.

There is plenty of space inside the welding machine.

http://igor.chudov.com/projects/Welding/00-Hobart-CyberTig-Welder/

I am hoping that if all works, I will convert my $9.99 welding machine
(what I paid for it) into a $2,000 welding machine. It would be a
perfectly well settable AC/DC welder with digital controls, quite
modern in fact.

The only minus of this welding machine would be lack of portability
and 3 phase power requirement. I can live with that.

Again, I am doing this for myself, I have no plans to sell this
welder.

i

I was going to suggest a 555, 50c for one at digikey. But if Jim is
happy, then I'm happy too.

Jeff Stout

--

 

[Tim Wescott]

Gentlemen, I greatly appreciate your help. You helped me with so many
decisions already. I would not be where I am without it. However, I am
somewhat confused by something.

I am not really asking for very much. I cannot believe that I have to
go to great lengths or create esoteric solutions. I suspect (and hope)
that we are missing something simple.

Let me make it very clear as to what I want.

I want a IC chip that has the following characteristics:

1. Generates pulsed, square wave, preferably at 5V above ground
alternating with ground.

2. Can have its frequency controlled by a separate pot and able to
function around 30-1000 Hz

3. Can have its "duty cycle" (percentage of time it spends in the HIGH
section) controlled by another, separate pot.

4. Preferably use about 18VDC power.

5. Preferably is a DIP.

It may be called PWM or something else. Let's not get stuck on
designation here. What matters is that it does what I want.

Is it really true that such chips are nonexistent?

MAX038 does almost what I want, but it is a little expensive, requires
a different power supply compared to the gate driver and also
amplification of its output. Otherwise it is great.


i

What the #@$% is your problem?

I told you, with at least two options.

Sphero told you, also with at least two options.

Both of us gave you solutions that would be programmable with two pots.

Both of us gave you solutions that will work from 5V to 18V (or more)

Both of us gave you solutions with chips that are in my 1982 National
Linear databook, before surface mount was found in anything but weird
military systems.

Others chimed in with addenda and cautions &c.

If you can read you already have your answer. If you can't read go back
to school.

 

[Jim Thompson]

I was going to suggest a 555, 50c for one at digikey. But if Jim is
happy, then I'm happy too.

Jeff Stout

One package of LM339 has 4 comparators... can do both your oscillation
AND duty cycle requirements.

...Jim Thompson

 

[Jim Thompson]

Thanks Jim... That IC looks reasonable to you, right?

i

Sure.

...Jim Thompson

 

[Joerg]

Hello Jim,

One package of LM339 has 4 comparators... can do both your oscillation
AND duty cycle requirements.

But the LM339 is 10-12 cents while a 74HC14 is 6-7 cents. And it's a
six-pack from which you'd only need one.

Regards, Joerg

 

[Jim Thompson]

Hello Jim,

But the LM339 is 10-12 cents while a 74HC14 is 6-7 cents. And it's a
six-pack from which you'd only need one.

Regards, Joerg

I got you beat on adjustability ;-)

...Jim Thompson

 

[Jim Thompson]

Jim Thompson wrote...

Don't tell me you designed the xr2206 too?

No. But I've used it as the core for a super-quality function
generator by adding self-tweaking elements around it.

...Jim Thompson

 

[Joerg]

Hello Jim,

I got you beat on adjustability ;-)

Not if I am allowed to use more than one of the Schmitt inverters. Even
with just one you can adjust but then it would be an adjustment of low
phase and high phase, not the frequency.

Regards, Joerg

 

[Rich Grise]

Gentlemen, I greatly appreciate your help. You helped me with so many
decisions already. I would not be where I am without it. However, I am
somewhat confused by something.

I am not really asking for very much. I cannot believe that I have to
go to great lengths or create esoteric solutions. I suspect (and hope)
that we are missing something simple.

Yes, we're missing, "What's the goal?" i.e., "What's the application"?
"What's the end result you're trying to accomplish?"

You're detailing us to death here - just tell us what you have, what
you want it to do, and let the experts figure out how to get from point
A to point B.

Thanks,
Rich

 

[Rich Grise]

I was going to suggest a 555, 50c for one at digikey. But if Jim is
happy, then I'm happy too.

Jeff Stout

I did a PWM with a 555 once, and it died first time out. Finally ended
up with a pair of MPSU06s in an astable multivibrator.

But this was for a spool gun, that got its power from the welder itself,
through the cable. =:-O Very nasty electrical environment!

Good Luck!
Rich

 

[Ignoramus30105]

Yes, we're missing, "What's the goal?" i.e., "What's the application"?
"What's the end result you're trying to accomplish?"

You're detailing us to death here - just tell us what you have, what
you want it to do, and let the experts figure out how to get from point
A to point B.

I will be happy to oblige. I want to thank you for attention and hope
that you would be able to provide some expert guidance.

I have a 200A DC Tig welder. Here it is:

http://igor.chudov.com/projects/Welding/00-Hobart-CyberTig-Welder/

I want to design a device to convert it to a AC/DC welder by means of
a H bridge based inverter, to produce square wave AC.

I am going to use Toshiba 200 A IGBTs MG200Q2YS40.

I will drive them using half bridge drivers such as IR21094 chips.

What I need is a generator of pulses to drive the input of
IR21094. This generator would send signals to tell IR21094 when to
open the high side and when to open the low side.

To be an optimal tig welder, I would like the frequency of pulses to
be settable with a potentiometer, from 20 to 400 Hz or so, and to have
the width of pulse, as a percentage of pulsing period, to also be
settable by a separate pot. That percentage is called a "duty cycle",
and I would like it to be relatively wide ranging, say between 20% and
80%.

I apologize if my explanation is unclear and will be glad to clarify
it and answer any questions.

As to what I am looking for right now: I am relatively clear as to
what to do around IGBTs and gate drivers. I am not as clear as to what
is the easiest way to build a generator of pulses that would satisfy
the above requirements.

I am looking more for an easy way than for a very cheap way, within
reason of course (I did not buy a brand new tig welder, after all). For
instance, I would rather buy a $40 "module" that would do what I want,
rather than buy a few 50 cent components and spend a week figuring it
out.

That said, if figuring out is necessary, I would do it.

So, if you have suggestion for a simple pulse generator, preferably
giving 5V pulses and with about 18V power supply, I would be happy to
hear from you.

i

 

[Robert Latest]

I was going to suggest an LM339 circuit but, if you're happy, I'm
happy ;-)

I already did that (LM393 actually) in quite some detail. Can't
beat a 8-cent 8-pin DIP part.

robert