AC supply

[Tarzan]

Hi, I have a Variac but like to automate a certain test on power amps.
Therefore I need to change the voltage (going up to 240V AC) and with adjustable current.
In fact a power supply from "0" to 240V AC and 5Amps.
I thought of using a LF power amp and make the DC supply voltage of the power stage current wise adjustable.
Feed it with a variable small ac signal and voila...
But, and there I'm stuck.
Who can push me further in this jungle. All ideas are welcome.
Thanks

[Nigel Goodwin]

Originally posted by Tarzan
Hi, I have a Variac but like to automate a certain test on power amps.
Therefore I need to change the voltage (going up to 240V AC) and with adjustable current.
In fact a power supply from "0" to 240V AC and 5Amps.
I thought of using a LF power amp and make the DC supply voltage of the power stage current wise adjustable.
Feed it with a variable small ac signal and voila...
But, and there I'm stuck.
Who can push me further in this jungle. All ideas are welcome.
Thanks

You talking huge amounts of expense to do it like that, a far simpler way (and probably more effective?), would be to use your existing variac - but control it via a motor, a stepper would be good (if you could get one strong enough?), or use a geared motor with a servo loop around it. The feedback for the servo loop could be the AC output of the variac.

Just a thought!.

[Tarzan]

Building a LF power amp with an APEX powerchip is feasable.
Feeding it with a variable 50Hz source is no problem at all.
Controlling the current is in fact the main problem.
I want to control the current (or keep an eye on it) during the ac voltage rise that feeds the DUT (device under test).
If the current reaches a preset current level at a certain ac voltage, the rise of the voltage has to deminisch to avoid problems in the DUT
As this is to happen automatic without human intervention (appart from the settings like current and voltage presets) a variac even with servo controlled motor drive is no option.
It takes me about an hour to do a test (manually).
An automatic system would do the trick in 5 minutes.
Thanks for the ideas.

[clem_o]

Put a small-valued series resistor on the return from the DUT to ground, monitor the voltage on that resistor, which will be proportional to the current through the DUT.

With a bit of filtering and a window comparator, you should be able to get something workable...

Cheers!

[Tarzan]

Hi,
Why do I allways think to complicated?
Maybe because I worked for a big Multinational?
The idea of the small value resistor and comparator looks great.
Current issue is "solved"
Now the voltage adjustment.
I do not fancy the idea of using a motor driven variac.
If nothing else shows up I have to go that way.
Thanks

[clem_o]

Hi,

I thought you were going with the LF amplifier? It would certainly do the work..

A second-hand industrial DC servo amp will do the job nicely and save you having to build anything. Most of the present-day units are switchmode though, so some output filtering will be needed. Drive it with an AC signal at 50 or 60Hz, and guess what - these servo amps have built-in variable current limit, usually set by some multi-turn trimpot.

Cheers!

[Tarzan]

I'm not so familiar with these servo amps.
Can you tell me a bit more?
Manufacturers, technical sites etc...

I still have that audio amp in my head but your suggestion on that serial resistor is a good starting point.

Thanks

[clem_o]

Hi -

Servo amplifiers are industrial drives usually used for motors. The rated BW of most servo amps is about 2KHz or so, which is more than enough to do a 50Hz or 60Hz sine-wave output. The feedback system of these servo amps can usually be configured to be voltage, current or position-derived (using an external position to voltage converter).

These amps should be useable to make something like a high-powered AC signal - check this link as an example:

http://www.a-m-c.com/content/prods/d...alogbrush.html

(look at the specs for example of the 100A25 model... !!)

And on the same website check "Motion 101" to learn a bit more...

Hope this helps!

Cheers

[Tarzan]

Well as a week went by and I've been thinking (again) I will not go for the servo system with the HV supply.

If I use a psu with eg. +/- 65 V, 5A and feed a power amp consisting of 2 units of eg. 500 Watt and connect them in bridge mode.
As load we use a normal 1 KW power trafo: 120V to 220V.
Feed both power amps with a 50 Hz signal coming from a low voltage trafo that feeds the control circuitry.
One amp is in phase, the other 180°.
These buffers are preceeded with an DC controlled opamp (VCA).
The opamp is fed with the 50Hz signal and with the control voltage coming from the secundairy of the power transformer (via a small power trafo eg 220 in, 12V out).

The control voltage comes from an opamp that can be tuned by a potmeter and also set the output voltage.

Another opamp can be set in such a way that when the current, measured by a current transducer (cheap at Conrad DE), reaches a predefined level will shut the VCA of so no signal is fed to the amps and no secundary output is available.

I think this is the way to go.
Easy to build, technology is available, parts to.
And above all not so expensive and everything is galvanic seperated.

Comments are very much appreciated.

Zilog

[clem_o]

Hi Zilog,

Sounds to me that's pretty much what's inside the servo amp, save for galvanic isolation at the outputs (easy to accomplish with a 1:1 isolation transformer at the power input). But hey, your proposal seems like it should work, and better to be comfortable with the design / approach since you're the one who will use it!

Cheers!

Clem