Circuit to amplify square wave

[[email protected]]

I need to build a circuit which will amplify a square wave input from a
function generator. I have with me an 8V DC power supply that can
source over a hundred amperes. I had conceived doing the amplification
using a solid state relay with the square wave driving the input side
of the SSR. This technique will work if the square wave goes from 0V to
some finite voltage, say, 5V, as in a TTL signal. But my input is a
square wave going from -5V to +5V. How do I get my desired
amplification with just the +8V power supply? I don't have a -8V
source.

I have been told that I need to use an opamp for this. This seems like
a simple task, but I don't know enough circuit design to do this. Any
help appreciated.

-SD

8V----SSR output----load---ground

 

[Ge0rge Marutz]

Without getting into too much detail, look at a switching power supply
with a dual primary/secondary transformer. Any car audio amp over 50W
or so will have one you can use as an example. Just cut and paste the
sections you need.

There are a couple very good books at your local Borders or Barnes and
Noble bookstore that cover this very topic if you need to design
something from scratch.

Ge0rge

 

[Jim Thompson]

I need to build a circuit which will amplify a square wave input from a
function generator. I have with me an 8V DC power supply that can
source over a hundred amperes. I had conceived doing the amplification
using a solid state relay with the square wave driving the input side
of the SSR. This technique will work if the square wave goes from 0V to
some finite voltage, say, 5V, as in a TTL signal. But my input is a
square wave going from -5V to +5V. How do I get my desired
amplification with just the +8V power supply? I don't have a -8V
source.

I have been told that I need to use an opamp for this. This seems like
a simple task, but I don't know enough circuit design to do this. Any
help appreciated.

-SD

8V----SSR output----load---ground

If you have only a single +8V supply, your output is constrained to be
bounded by 0V and +8V.

"OpAmp" implies relatively low power output, why "...over a hundred
amperes"?

What does "amplify" mean? -5V to +5V input produces what range of
outputs? Centered at +4V? Minimum always at 0V?

?????

...Jim Thompson

 

[[email protected]]

I do need build the circuit from scratch. Do you of a specific book
which addresses this problem? I looked in "The Art of Electronics" by
Horowitz - couldn't find what I'm looking for.

-SD

 

[John Popelish]

I do need build the circuit from scratch. Do you of a specific book
which addresses this problem? I looked in "The Art of Electronics" by
Horowitz - couldn't find what I'm looking for.

Does this need to be a linear amplifier (output amplitude proportional
to input amplitude) or must it output a constant, high amplitude, in
spite of changes in input amplitude?

 

[[email protected]]

opamp is not my idea - it's just hear-say. I don't even know whether an
opamp needs to be used in the first place.

I need to pass high currents through my load (here, an electromagnetic
actuator). The desired current is not steady, but periodically
switching between +A amperes and -A amperes at frequencies up to 100
Hz. The actual value of the current A is variable - I will gradually be
increasing it in steps until my load burns out. The value of the
current through the load is of importance to me, not the voltage drop
across it. I used the word amplify to signify that a high power circuit
is being driven by a function generator, a low power device. I guess
'amplify' is not the right word in this situation; my bad - a square
wave pulse on the input side of a solid state relay acts as a gate
only, and its actual amplitude does not influence the output current or
voltage.

I am pretty sure that all I need is the 8V DC power supply box. We used
to have a box made by a company (which probably doesn't exist now!)
which did exactly what I want, and using the exact same power supply
that I have. It's just that it doesn't work and probably parts of it
are stolen/missing.

If someone could point me to a book, say, which has this explained, or
sketch an quick and dirty circuit diagram, that'll be invaluable.

-SD

 

[[email protected]]

O.. and by the way, I plan to regulate the current magnitude A by
either toning down the voltage from the power supply to something below
8V, or alternatively, keeping the 8V fixed and connecting a
potentiometer in series with my load.

 

[Tim Wescott]

opamp is not my idea - it's just hear-say. I don't even know whether an
opamp needs to be used in the first place.

I need to pass high currents through my load (here, an electromagnetic
actuator). The desired current is not steady, but periodically
switching between +A amperes and -A amperes at frequencies up to 100
Hz. The actual value of the current A is variable - I will gradually be
increasing it in steps until my load burns out. The value of the
current through the load is of importance to me, not the voltage drop
across it.

Your idea of using a solid-state relay (or a pair of them) is valid
then. Many general-purpose function generators have a TTL-level "sync"
output that could be used; if you are stuck with +/-5V and the SSR can't
handle the input you can always use a blocking diode:


------>|---o-------
From | To SSR
signal .-.
generator | |
| |
'-'
|
|
===
GND
created by Andy´s ASCII-Circuit v1.24.140803 Beta www.tech-chat.de

Things to check for on your SSR are the response speed (100Hz is 10ms,
the relay should be capable of a fraction of that), current handling
ability, and ON resistance. If you can't find a DPDT solid state relay
(do such things exist) to reverse your power then you'll have to make an
'H' bridge out of four individual ones. This will create a definite
danger of turning on both relays in one leg of the 'H', causing what
Zsa-zsa Gabor would call "shoosht through" and the rest of us would call
"smoke".

I used the word amplify to signify that a high power circuit
is being driven by a function generator, a low power device. I guess
'amplify' is not the right word in this situation; my bad - a square
wave pulse on the input side of a solid state relay acts as a gate
only, and its actual amplitude does not influence the output current or
voltage.

"Amplifier" is more or less correct here -- what folks were fishing for
was whether you needed to take +/-5V and turn it into +/-100V at 1mA, or
if you wanted 0V, 8V at 100A -- which appears to be what you want.

I am pretty sure that all I need is the 8V DC power supply box. We used
to have a box made by a company (which probably doesn't exist now!)
which did exactly what I want, and using the exact same power supply
that I have. It's just that it doesn't work and probably parts of it
are stolen/missing.

If someone could point me to a book, say, which has this explained, or
sketch an quick and dirty circuit diagram, that'll be invaluable.

-SD

You are looking for a heavy-duty, low voltage, low speed (thankfully)
'H' bridge. 'H' bridges are dirt common anywhere electrical watts are
being turned into mechanical watts, so there's lots of choices out there.

 

[Tim Wescott]

O.. and by the way, I plan to regulate the current magnitude A by
either toning down the voltage from the power supply to something below
8V, or alternatively, keeping the 8V fixed and connecting a
potentiometer in series with my load.

(8V)(100A) = 800 Watts. That's quite a big pot. If you can adjust the
power supply that's probably the way to go.

 

[Jim Thompson]

opamp is not my idea - it's just hear-say. I don't even know whether an
opamp needs to be used in the first place.

I need to pass high currents through my load (here, an electromagnetic
actuator). The desired current is not steady, but periodically
switching between +A amperes and -A amperes at frequencies up to 100
Hz. The actual value of the current A is variable - I will gradually be
increasing it in steps until my load burns out. The value of the
current through the load is of importance to me, not the voltage drop
across it.

[snip]

Your description is clear as mud ;-)

But it sounds like an H-bridge on the output of your power supply
might do what you want.

Can you post a block diagram with performance expectations either on
newsgroup alt.binaries.schematics.electronic or on a URL?

...Jim Thompson

 

[Ge0rge Marutz]

"Simplified Design of Switching Power Supplies" By John D. Lenk

This is my favorite. It helped me design my first 3KW switcher for an
automotive amp project I had going. Does a great job at the basics and
gives plenty of examples.

The only problem I see you having is with the magnetics. You would be
well served to consult with a transfomer company to get what you need.

Many other books exist and are on the shelves at Barnes and Noble or
Borders. Just do what I did and do a search for "switching power
supply" on either companies web site.

Ge0rge

 

[Pooh Bear]

I do need build the circuit from scratch. Do you of a specific book
which addresses this problem? I looked in "The Art of Electronics" by
Horowitz - couldn't find what I'm looking for.

To be honest you clearly need a better grasp of the basics before you
should consider switching 100 Amps or so.

What electronics experience *do* you have ?

Graham

 

[Rich Grise]

I need to build a circuit which will amplify a square wave input from a
function generator. I have with me an 8V DC power supply that can
source over a hundred amperes. I had conceived doing the amplification
using a solid state relay with the square wave driving the input side
of the SSR. This technique will work if the square wave goes from 0V to
some finite voltage, say, 5V, as in a TTL signal. But my input is a
square wave going from -5V to +5V. How do I get my desired
amplification with just the +8V power supply? I don't have a -8V
source.

I have been told that I need to use an opamp for this. This seems like
a simple task, but I don't know enough circuit design to do this. Any
help appreciated.

-SD

8V----SSR output----load---ground

Just limit the negative excursion:

Signal Generator +5 ~ -5 ----[470R]---+--- SSR Input
|
---K
/ \ 1N914, 1N4148, etc
---
|
Signal Generator Ground --------------+--- SSR Input return

Good Luck!
Rich

 

[Ken Smith]

opamp is not my idea - it's just hear-say. I don't even know whether an
opamp needs to be used in the first place.

I need to pass high currents through my load (here, an electromagnetic
actuator). The desired current is not steady, but periodically
switching between +A amperes and -A amperes at frequencies up to 100
Hz.

Can you get at both of the leads of the load? Can you seperate the
frequency source and the amplitude?

Consider this:

From
variable -------+------------------------
supply ! !
O S1A O S2
/ /
! !
! !
+--------[ LOAD ]--------+
! !
O S2B O S1B
/ /
! !
------------+------------------------+

The switches can be arrays of MOSFETs.

 

[Ge0rge Marutz]

Ahh yes, an H-bridge!!!

Ge0rge

 

[[email protected]]

Thanks to all those who replied. I think Ken Smith's circuit best does
what I am trying to do. Ken, I know how to use solid state relays as
switches in locations S1A, S2A, S1B and S2B in your circuit diagram. I
have been told that MOSFETs do the same job (as the SSR's) but more
effectively. I am not too sure how to connect the MOSFETs. An SSR has
four terminals for low power input and return (function generator in
this case), and high power input and return (here, the 8V power
supply). So this makes sense. A MOSFET has only three, right?

high power input
/
signal generator -----| MOSFET
\
return

Where does sig generator return go? Another way of asking the same
question is, if the two leads of the signal generator get connected
across MOSFET gate and X, what is X? On the high power side, I assume
that the two ends of the wire on which the MOSFET is 'mounted' get
connected across MOSFET source and drain.

-SD

 

[[email protected]]

No sir. This seems to chop off one half of my input signal. I can
easily produce the output of your circuit by halving the amplitude and
adding an offset in my signal generator. I NEED a zero offset square
wave (going from +a to -a volts) on the high power (output) side of the
SSR (or any circuitry used). Thanks.