DC to AC Square wave question

[HC]

Hey, all, I thought this would be easy but it's not been so far. What
I want to do is take a DC power supply and output an AC voltage. The
power supply I'm using puts out 6 volts DC and is marked that it can
handle 1.66 amps output. All I want to do is take that and turn it
into square-wave AC; not sinusoidal AC, but strictly square-wave AC
and at the same 6 volts as the input. The mechanism needs to be able
to handle the maximum of 1.66 amps and the 6 volts. The load is NOT
inductive and, to the best of my knowledge is NOT capacitive; I
believe the load is STRICTLY resistive. Many Bothans died to bring us
this information.

What I've done is search for: buck converters, buck-boost converters,
dc ac converters, dc ac inverters, dc chopper, dc square wave ac, and
similar but most of what I find falls under either some patent site,
or dc to ac inverters for taking 12 volts dc and making 120 volts ac
sine-wave for driving appliances, or taking dc voltage up or down to
DC.

All I want is to take my DC source and alternate the polarity like AC
but I want it to be square-wave and I'd like to be able to vary the
frequency. The oscillator like I'm using now (555-based) should
handle the frequency variations (but I'm always open to suggestions,
of course, or I wouldn't be posting here); what I'm hoping for is a
"chopper" circuit to turn this output into square-wave AC.

What I've done is create an oscillating circuit using a 555 timer and
a pot that allows me to vary the frequency of output from about 160 Hz
to about 10.86 kHz at roughly 50% duty cycle. What I would like to do
is use that to drive a circuit that can alternate the polarity of my
DC power supply at the same frequency. It would seem that I need a
MOSFET to control the power (several, in fact). I have used this
timer circuit to drive a MOSFET (a BUZ11) and it works; I get variable
frequency, pulsed DC power. That is, use the 555 output to drive the
gate of the BUZ11, the power supply positive at the Drain, the power
supply negative at the Source, and what I get is pulsed DC. But what
I want, again, is square-wave AC.

I have found that in DC motor control (particularly for robotics)
people use H-Bridge's to control electric motor direction which would
seem to do what I need to have done; the schematics I've seen use 4
MOSFETs (two P channel, two N channel) to apply source voltage in one
polarity or another. I could use the output of the 555 timer (or
another oscillator) with an inverter of some kind (I'm NOT an expert
at this stuff by any means, so some technicality will be lacking in
anything I say but I try to be accurate) to take the output and make
one MOSFET gate low (N channel) and one MOSFET gate high (P channel)
when one polarity is required and reverse it for the other.

That leaves me using one 555 timer to control oscillations, 4 MOSFETs
to control power (two of each channel), some number of logic
comparators (I'm not educated enough in electronics to know what the
component would be called) to make the output high for one MOSFET and
low for another (for the gates), and maybe some transistors if the 555
output couldn't handle the brief current of charging the gates.

I have an old froo-froo LED flasher kit I put together years ago that
uses two transistors and some caps and resistors; this makes me think
there has to be a simple way to do what I'm trying here using just two
MOSFETs but I cannot find it and, quite honestly, I think I may be too
much of an amateur to know what the right terms are to search for it.

Anyway, sorry for the length but I wanted to explain what I was after
and that I did try to find the answer before posting. Of course, if
it goes like it does when I ask where the restroom is in a business
then someone will point right over my shoulder behind me and say,
"There.".

Thanks in advance.

--HC

 

[redbelly]

On Jan 3, 10:04 pm, HC <hboo...@gte.net> wrote:

Hey, all, I thought this would be easy but it's not been so far. What
I want to do is take a DC power supply and output an AC voltage. The
power supply I'm using puts out 6 volts DC and is marked that it can
handle 1.66 amps output. All I want to do is take that and turn it
into square-wave AC; not sinusoidal AC, but strictly square-wave AC
and at the same 6 volts as the input. The mechanism needs to be able
to handle the maximum of 1.66 amps and the 6 volts. The load is NOT
inductive and, to the best of my knowledge is NOT capacitive; I
believe the load is STRICTLY resistive. Many Bothans died to bring us
this information.

{ Snip }

Thanks in advance.

--HC

While I'm not the one to answer this question, those people who can
would likely want more details about the load. What, specifically,
are you trying to do? There might be an easier and/or better way to
do whatever it is you are trying to do.

Regards,

Mark

 

[Anthony Fremont]

HC wrote:

Hey, all, I thought this would be easy but it's not been so far. What
I want to do is take a DC power supply and output an AC voltage. The
power supply I'm using puts out 6 volts DC and is marked that it can
handle 1.66 amps output. All I want to do is take that and turn it
into square-wave AC; not sinusoidal AC, but strictly square-wave AC
and at the same 6 volts as the input. The mechanism needs to be able
to handle the maximum of 1.66 amps and the 6 volts. The load is NOT
inductive and, to the best of my knowledge is NOT capacitive; I
believe the load is STRICTLY resistive. Many Bothans died to bring us
this information.

Use an h-bridge. An LMD18200T can handle 3A continuously. It's $14 from
digikey. You can get a pc-board to break out the weird pin spacing from
www.sparkfun.com for ~$2. You just feed the signal from the 555 to the
"direction" input of the h-bridge. The outputs will flip-flop accordingly.
There are probably other parts, but I happen to be using one of these right
now to do precisely what you are describing.

 

[ehsjr]

HC wrote:

Hey, all, I thought this would be easy but it's not been so far. What
I want to do is take a DC power supply and output an AC voltage. The
power supply I'm using puts out 6 volts DC and is marked that it can
handle 1.66 amps output. All I want to do is take that and turn it
into square-wave AC; not sinusoidal AC, but strictly square-wave AC
and at the same 6 volts as the input. The mechanism needs to be able
to handle the maximum of 1.66 amps and the 6 volts. The load is NOT
inductive and, to the best of my knowledge is NOT capacitive; I
believe the load is STRICTLY resistive. Many Bothans died to bring us
this information.

What I've done is search for: buck converters, buck-boost converters,
dc ac converters, dc ac inverters, dc chopper, dc square wave ac, and
similar but most of what I find falls under either some patent site,
or dc to ac inverters for taking 12 volts dc and making 120 volts ac
sine-wave for driving appliances, or taking dc voltage up or down to
DC.

All I want is to take my DC source and alternate the polarity like AC
but I want it to be square-wave and I'd like to be able to vary the
frequency. The oscillator like I'm using now (555-based) should
handle the frequency variations (but I'm always open to suggestions,
of course, or I wouldn't be posting here); what I'm hoping for is a
"chopper" circuit to turn this output into square-wave AC.

What I've done is create an oscillating circuit using a 555 timer and
a pot that allows me to vary the frequency of output from about 160 Hz
to about 10.86 kHz at roughly 50% duty cycle. What I would like to do
is use that to drive a circuit that can alternate the polarity of my
DC power supply at the same frequency. It would seem that I need a
MOSFET to control the power (several, in fact). I have used this
timer circuit to drive a MOSFET (a BUZ11) and it works; I get variable
frequency, pulsed DC power. That is, use the 555 output to drive the
gate of the BUZ11, the power supply positive at the Drain, the power
supply negative at the Source, and what I get is pulsed DC. But what
I want, again, is square-wave AC.

I have found that in DC motor control (particularly for robotics)
people use H-Bridge's to control electric motor direction which would
seem to do what I need to have done; the schematics I've seen use 4
MOSFETs (two P channel, two N channel) to apply source voltage in one
polarity or another. I could use the output of the 555 timer (or
another oscillator) with an inverter of some kind (I'm NOT an expert
at this stuff by any means, so some technicality will be lacking in
anything I say but I try to be accurate) to take the output and make
one MOSFET gate low (N channel) and one MOSFET gate high (P channel)
when one polarity is required and reverse it for the other.

That leaves me using one 555 timer to control oscillations, 4 MOSFETs
to control power (two of each channel), some number of logic
comparators (I'm not educated enough in electronics to know what the
component would be called) to make the output high for one MOSFET and
low for another (for the gates), and maybe some transistors if the 555
output couldn't handle the brief current of charging the gates.

I have an old froo-froo LED flasher kit I put together years ago that
uses two transistors and some caps and resistors; this makes me think
there has to be a simple way to do what I'm trying here using just two
MOSFETs but I cannot find it and, quite honestly, I think I may be too
much of an amateur to know what the right terms are to search for it.

Anyway, sorry for the length but I wanted to explain what I was after
and that I did try to find the answer before posting. Of course, if
it goes like it does when I ask where the restroom is in a business
then someone will point right over my shoulder behind me and say,
"There.".

Thanks in advance.

--HC

By "square wave AC" do you mean you want a square wave
output that goes 6 volts peak to peak or do you want
it to go positive 6 volts, then negative 6 volts, for
12 volts peak to peak?

Ed

 

[HC]

On Jan 4, 1:30 am, ehsjr <eh...@bellatlantic.net> wrote:

HC wrote:
Hey, all, I thought this would be easy but it's not been so far. What
I want to do is take a DC power supply and output an AC voltage. The
power supply I'm using puts out 6 volts DC and is marked that it can
handle 1.66 amps output. All I want to do is take that and turn it
into square-wave AC; not sinusoidal AC, but strictly square-wave AC
and at the same 6 volts as the input. The mechanism needs to be able
to handle the maximum of 1.66 amps and the 6 volts. The load is NOT
inductive and, to the best of my knowledge is NOT capacitive; I
believe the load is STRICTLY resistive. Many Bothans died to bring us
this information.

What I've done is search for: buck converters, buck-boost converters,
dc ac converters, dc ac inverters, dc chopper, dc square wave ac, and
similar but most of what I find falls under either some patent site,
or dc to ac inverters for taking 12 volts dc and making 120 volts ac
sine-wave for driving appliances, or taking dc voltage up or down to
DC.

All I want is to take my DC source and alternate the polarity like AC
but I want it to be square-wave and I'd like to be able to vary the
frequency. The oscillator like I'm using now (555-based) should
handle the frequency variations (but I'm always open to suggestions,
of course, or I wouldn't be posting here); what I'm hoping for is a
"chopper" circuit to turn this output into square-wave AC.

What I've done is create an oscillating circuit using a 555 timer and
a pot that allows me to vary the frequency of output from about 160 Hz
to about 10.86 kHz at roughly 50% duty cycle. What I would like to do
is use that to drive a circuit that can alternate the polarity of my
DC power supply at the same frequency. It would seem that I need a
MOSFET to control the power (several, in fact). I have used this
timer circuit to drive a MOSFET (a BUZ11) and it works; I get variable
frequency, pulsed DC power. That is, use the 555 output to drive the
gate of the BUZ11, the power supply positive at the Drain, the power
supply negative at the Source, and what I get is pulsed DC. But what
I want, again, is square-wave AC.

I have found that in DC motor control (particularly for robotics)
people use H-Bridge's to control electric motor direction which would
seem to do what I need to have done; the schematics I've seen use 4
MOSFETs (two P channel, two N channel) to apply source voltage in one
polarity or another. I could use the output of the 555 timer (or
another oscillator) with an inverter of some kind (I'm NOT an expert
at this stuff by any means, so some technicality will be lacking in
anything I say but I try to be accurate) to take the output and make
one MOSFET gate low (N channel) and one MOSFET gate high (P channel)
when one polarity is required and reverse it for the other.

That leaves me using one 555 timer to control oscillations, 4 MOSFETs
to control power (two of each channel), some number of logic
comparators (I'm not educated enough in electronics to know what the
component would be called) to make the output high for one MOSFET and
low for another (for the gates), and maybe some transistors if the 555
output couldn't handle the brief current of charging the gates.

I have an old froo-froo LED flasher kit I put together years ago that
uses two transistors and some caps and resistors; this makes me think
there has to be a simple way to do what I'm trying here using just two
MOSFETs but I cannot find it and, quite honestly, I think I may be too
much of an amateur to know what the right terms are to search for it.

Anyway, sorry for the length but I wanted to explain what I was after
and that I did try to find the answer before posting. Of course, if
it goes like it does when I ask where the restroom is in a business
then someone will point right over my shoulder behind me and say,
"There.".

Thanks in advance.

--HC

By "square wave AC" do you mean you want a square wave
output that goes 6 volts peak to peak or do you want
it to go positive 6 volts, then negative 6 volts, for
12 volts peak to peak?

Ed

Hello, Ed. I want the latter; I want it to be +6 volts on one lead at
one moment with ground/negative on the other lead at that moment, then
swap them quickly (hence my use of the term square-wave) to reverse
the polarity taking the first lead and making it ground/negative and
the second lead +6 volts. The +6 volts comes from a wall-wart that
converts 120 VAC to 6 VDC. I want the device I am looking for to just
switch the positive output from the wall-wart with the negative output
of the wall-wart so at one moment one lead from the device is
connected to the positive 6 volts and the other lead is connected to
the ground/negative of the wall-wart, and at the next moment swap
them.

--HC

 

[BobG]

Take wall wart apart and bypass the diode thats converting the 6VAC to
DC.

 

[Rich Grise]

On Thu, 03 Jan 2008 19:04:11 -0800, HC wrote:

Hey, all, I thought this would be easy but it's not been so far. What I
want to do is take a DC power supply and output an AC voltage. The
power supply I'm using puts out 6 volts DC and is marked that it can
handle 1.66 amps output. All I want to do is take that and turn it into
square-wave AC; not sinusoidal AC, but strictly square-wave AC and at
the same 6 volts as the input. The mechanism needs to be able to handle
the maximum of 1.66 amps and the 6 volts. The load is NOT inductive
and, to the best of my knowledge is NOT capacitive; I believe the load
is STRICTLY resistive.

If by "the 6 volts", you mean from 0v at the bottom to 6V at the top,
all you need is a switch.

To normalize it, i.e., make it zero-center, you'd just need a MONGO
capacitor in series. The bigger the cap, the less the "signal" will
droop between cycles.

I'm sure you can calculate the cap value needed to give an acceptable
droop - just use that time constant formula; I'd also use a pretty
hefty transistor - maybe even a complementary switch so you get
a good solid pullup and pulldown.

Good Luck!
Rich

 

[HC]

On Jan 3, 11:05 pm, "Anthony Fremont" <nob...@noplace.net> wrote:

HC wrote:
Hey, all, I thought this would be easy but it's not been so far. What
I want to do is take a DC power supply and output an AC voltage. The
power supply I'm using puts out 6 volts DC and is marked that it can
handle 1.66 amps output. All I want to do is take that and turn it
into square-wave AC; not sinusoidal AC, but strictly square-wave AC
and at the same 6 volts as the input. The mechanism needs to be able
to handle the maximum of 1.66 amps and the 6 volts. The load is NOT
inductive and, to the best of my knowledge is NOT capacitive; I
believe the load is STRICTLY resistive. Many Bothans died to bring us
this information.

Use an h-bridge. An LMD18200T can handle 3A continuously. It's $14 from
digikey. You can get a pc-board to break out the weird pin spacing fromwww.sparkfun.comfor ~$2. You just feed the signal from the 555 to the
"direction" input of the h-bridge. The outputs will flip-flop accordingly.
There are probably other parts, but I happen to be using one of these right
now to do precisely what you are describing.

Thanks, Anthony. That is probably what I'll wind up doing.

--HC

 

[HC]

On Jan 4, 1:45 pm, BobG <bobgard...@aol.com> wrote:

Take wall wart apart and bypass the diode thats converting the 6VAC to
DC.

I like that idea; it's simple and would probably work great. My only
reservation in this application is that I want to be able to vary the
frequency. But that's a good enough idea to maybe try to see if going
to AC instead of DC will do what I want to do. Thank you.

--HC

 

[HC]

On Jan 4, 5:05 pm, Rich Grise <r...@example.net> wrote:

On Thu, 03 Jan 2008 19:04:11 -0800, HC wrote:
Hey, all, I thought this would be easy but it's not been so far. What I
want to do is take a DC power supply and output an AC voltage. The
power supply I'm using puts out 6 volts DC and is marked that it can
handle 1.66 amps output. All I want to do is take that and turn it into
square-wave AC; not sinusoidal AC, but strictly square-wave AC and at
the same 6 volts as the input. The mechanism needs to be able to handle
the maximum of 1.66 amps and the 6 volts. The load is NOT inductive
and, to the best of my knowledge is NOT capacitive; I believe the load
is STRICTLY resistive.

If by "the 6 volts", you mean from 0v at the bottom to 6V at the top,
all you need is a switch.

To normalize it, i.e., make it zero-center, you'd just need a MONGO
capacitor in series. The bigger the cap, the less the "signal" will
droop between cycles.

I'm sure you can calculate the cap value needed to give an acceptable
droop - just use that time constant formula; I'd also use a pretty
hefty transistor - maybe even a complementary switch so you get
a good solid pullup and pulldown.

Good Luck!
Rich

Hey, Rich, yes all I want to do is switch "poles" from the wall-wart.
I was hoping to use a couple of MOSFETs for the switches (like what
you are saying about complementary switches) but when I search for DC
AC inverters I keep getting inundated with results about commercial
rigs for use in driving small appliances in your car, boat, etc. The
next most common thing I find is when I search for MOSFETs and power
switching I get H-bridges which are a bit more complicated than what I
need (not by much but I like things to be as simple as possible). I
read somewhere that a person can use two transistors to chop DC into
AC but I do not know how to do it and have yet to find a design that
seems that it would do it. I feel there should be an elegant and
simple solution to do what I'm doing. Anthony posted about using a
single package H-bridge controller which would work, it seems, no
sweat, but I'd kinda like to make this device from scratch.
Ultimately I need the solution so if I need to use a commercially
produced H-bridge (like the one Anthony suggested) then I'll do it,
but before I break down and do that I'm going to hold out for
something just silly simple.

Thank you.

--HC

 

[whit3rd]

On Jan 3, 7:04 pm, HC <hboo...@gte.net> wrote:

Hey, all, I thought this would be easy but it's not been so far.  What
I want to do is take a DC power supply and output an AC voltage.

... I want it to be square-wave and I'd like to be able to vary the
frequency.
 
What I've done is create an oscillating circuit using a 555 timer and
a pot that allows me to vary the frequency of output from about 160 Hz
to about 10.86 kHz at roughly 50% duty cycle.

As others have said, the output stage is going to be a bridge (two
P MOSFETs and two N MOSFETs). Easiest drive would require
true and inverse square waves, and your timer only gives one
of those, so I'd add a CD4013 flip-flop.

The flip-flop will divide the clock by 2 (you'll want to change the
'555
range to 320 -21700 Hz), and has Q and not-Q outputs so there are
no missing drive signals. It also makes a precise 50% duty cycle,
which is important for several effects:
(1) electrolytic corrosion
(2) inductor saturation
(3) speaker distortion

You should know that this kind of bridge output CAN get troublesome
if the transition of the transistors takes significant time (because
it
shorts the power supply during the transition). There are tricks
to prevent this effect, but your wallwart supply will just droop
for a microsecond... if you filter the power into the '555 and flip-
flop,
it shouldn't hurt anything else.

 

[HC]

On Jan 4, 9:22 pm, whit3rd <whit...@gmail.com> wrote:

On Jan 3, 7:04 pm, HC <hboo...@gte.net> wrote:



Hey, all, I thought this would be easy but it's not been so far. What
I want to do is take a DC power supply and output an AC voltage.
... I want it to be square-wave and I'd like to be able to vary the
frequency.

What I've done is create an oscillating circuit using a 555 timer and
a pot that allows me to vary the frequency of output from about 160 Hz
to about 10.86 kHz at roughly 50% duty cycle.

As others have said, the output stage is going to be a bridge (two
P MOSFETs and two N MOSFETs). Easiest drive would require
true and inverse square waves, and your timer only gives one
of those, so I'd add a CD4013 flip-flop.

The flip-flop will divide the clock by 2 (you'll want to change the
'555
range to 320 -21700 Hz), and has Q and not-Q outputs so there are
no missing drive signals. It also makes a precise 50% duty cycle,
which is important for several effects:
(1) electrolytic corrosion
(2) inductor saturation
(3) speaker distortion

You should know that this kind of bridge output CAN get troublesome
if the transition of the transistors takes significant time (because
it
shorts the power supply during the transition). There are tricks
to prevent this effect, but your wallwart supply will just droop
for a microsecond... if you filter the power into the '555 and flip-
flop,
it shouldn't hurt anything else.

Hey, Whit3rd, thank you for your reply. Okay, an H-bridge it is. Is
it important which MOSFET I use for the P channel and the N channel in
regards to whether or not they are "complementary" (a term I've
seen)? I have BUZ11's to use for the N-channel (and some others like
an IRF520, IIRC), and I have some IRF9540's for P channel. All of
them, as I recall, will easily handle my 1.66 amp load, but I've heard
the term "complementary" in a variety of locations. Any worries
there?

What the CD4013 does sounds like exactly what I've been missing for
driving both an N channel and P channel MOSFET; the ability to take a
high output pulse and drive both kinds. That helps a lot. I'm going
to grab the datasheet on that and see if I might have one somewhere in
my odds-n-ends box.

Electrolytic corrosion? Like in electrolysis of water like we did in
school? That takes me back. I thought we did that with just DC.
It's off topic, sorry, but what impact does 50% duty cycle square-wave
AC have on electrolysis (if that's what you're talking about with
'electrolytic corrosion')?

Thanks again.

--HC

 

[whit3rd]

On Jan 4, 8:20 pm, HC <hboo...@gte.net> wrote:

I've heard
the term "complementary" in a variety of locations.  Any worries
there?

Probably not in this application; if there were linear-amplifier
application
of the bridge amp, there'd be distortion if the transconductance of
the
output transistors didn't match. For switching, that won't
be a problem.

Electrolytic corrosion?  Like in electrolysis of water like we did in
school?  That takes me back.  I thought we did that with just DC.
It's off topic, sorry, but what impact does 50% duty cycle square-wave
AC have on electrolysis

If you ran outdoor lights with wet wires and DC, the (+) wire would
corrode away to nothing very fast. Same wet wires and AC, you should
expect less corrosion (and symmetric corrosion of both wires).
The use of shock-safe voltages and AC is common in wet locations,
because that's a backup plan if waterproofing and/or insulation
fail.

 

[HC]

On Jan 5, 1:37 am, whit3rd <whit...@gmail.com> wrote:

On Jan 4, 8:20 pm, HC <hboo...@gte.net> wrote:

[on the subject of MOSFET selection ]I've heard
the term "complementary" in a variety of locations. Any worries
there?

Probably not in this application; if there were linear-amplifier
application
of the bridge amp, there'd be distortion if the transconductance of
the
output transistors didn't match. For switching, that won't
be a problem.



Electrolytic corrosion? Like in electrolysis of water like we did in
school? That takes me back. I thought we did that with just DC.
It's off topic, sorry, but what impact does 50% duty cycle square-wave
AC have on electrolysis

If you ran outdoor lights with wet wires and DC, the (+) wire would
corrode away to nothing very fast. Same wet wires and AC, you should
expect less corrosion (and symmetric corrosion of both wires).
The use of shock-safe voltages and AC is common in wet locations,
because that's a backup plan if waterproofing and/or insulation
fail.

Hey, Whit3rd, thanks for the answer. Sorry for the delay in response.

That makes a lot of sense about the wire degrading. I remember
rusting out a few alligator clips when they got submerged in solution
for electrolysis.

I did build an H-bridge (finally). It took a while for me to get the
parts together (I buy all of my stuff about a hundred miles from here
so I'm not there as often as I'd like). I got it to work after a
little bit of a problem: apparently, while the rest of the electronics
universe uses the convention for schematics of "top side is positive
voltage" and "bottom side is ground" the world of MOSFET documentation
deviates. The "drain" on the top of the schematic diagram for the p-
channel MOSFET in the datasheet (http://www.ortodoxism.ro/datasheets/
fairchild/IRF9540.pdf) goes to ground while the source goes to
positive. In retrospect I can see how to understand this from the
names; SOURCE, um, where the positive power comes from, DRAIN, where
it's going. But it would have been easier if the diagrams followed
what I believe to be standard (it's a given here, by the way, that
Fairchild knows a lot more about this stuff than I do so when there is
a discrepancy between what I think should be done and what they do,
they're probably right and I'm most certainly wrong; i.e. I'm not
saying they're wrong by any means). It also would have helped if I
had looked at the test-circuit schematics at the end of the document
more closely (which is what ultimately straightened me out) and saw
that the battery was depicted with negative at the top and positive at
the bottom.

In any case, nobody's fault but my own and my lack of experience with
these things. I put that little bit of info in here in case anybody
reads this later on and has similar confusion to my own. There is a
little blue cloud still hanging around in my electronics room from my
frustration at why what should have been so simple just bloody
wouldn't work and why the little suckers were getting so hot.
Live and learn. At least I didn't blow up any of the MOSFETs, I
caught them in time before they got too hot.

When I got it working it worked very well, thank you. I ran it up to
about 10kHz with no problems. The CD4013 was what I needed.

Thank you, and all who helped, again.

--HC