Amplifier output testing

[edouble101 314]

I have about 30 1990's amplifiers. Mostly Precision Power. I want to test the amplifiers both for functionality and for max unclipped output. I will record the data on a spreadsheet. I will also record by video.

Method :

Power supply - 12v car battery

Source - Ipod

Load - (2) 200w 8ohm non-inductive resistors

Oscilloscope - DSO Nano

panel meter for measuring power supply voltage

Amplifier output, resistor, DSO Nano, and volt meter connections

"Switched" turn-on

Positive and negative blocks

Wattage will be calculated using P=E^2/R. Resistance of the dummy load will be measured before and after the test.

Are there any faults in my testing method?

First amp tested is a PPI Ax400. Max unclipped output across all four channels is at 15.16v a/c @ 1000hz. Dummy load resistance was 4.2 ohms starting and after the test. This would make max unclipped power at 54.72 wrms. This was at 12.54v d/c.

Video of LR channel being tested on a PPI Ax400

[edouble101 314]

The point of this particular thread is to see if you guys think that my testing method is correct.

[shizzzon 461]

It is my understanding that you either need to measure using TRUE RMS everything or measure voltagecurrent AND the phase of each.

[95Honda 486]

As long as your meter actually reads RMS, this will work. You could also take the Pk-Pk measurement with your scope and do the math...

Having a battery as a power source is tough because it is unregulated. Try and do the test quick before the battery voltage drops too much.

Also try and do it quick before those dummy loads heat up. If they do (they look like the Partsxpress ones) mount them to a metal plate or heatsink. For high power tests I usually had them sitting in a 5 gallon bucket of cold water. (clean water isn't really conductive at all)...

Again, like you stated, check the resistance of the load. The tolerance may not be too tight on those...

And yes, follow the tips above and you are doing it right...

Have fun....

[Shogen 325]

cool idea .... you have too much time on your hands... soon as money starts rolling for me I'm gonna have to take away some of that free time for help with my build....

[shizzzon 461]

Yea, De-ionized water is non-conductive enough.

Mineral Oil is non-conductive as well.

[95Honda 486]

I just use bottled water... The cheap 1 gallon purifed stuff without Flouride...

If you don't have city water with a ton of crap in it, you can use your tap also...

[Sencheezy 689]

nice, looking good so far.

[edouble101 314]

As long as your meter actually reads RMS, this will work. You could also take the Pk-Pk measurement with your scope and do the math...

Having a battery as a power source is tough because it is unregulated. Try and do the test quick before the battery voltage drops too much.

Also try and do it quick before those dummy loads heat up. If they do (they look like the Partsxpress ones) mount them to a metal plate or heatsink. For high power tests I usually had them sitting in a 5 gallon bucket of cold water. (clean water isn't really conductive at all)...

Again, like you stated, check the resistance of the load. The tolerance may not be too tight on those...

And yes, follow the tips above and you are doing it right...

Have fun....

My meter does not read RMS. I am measuring a/c voltage then doing the math. Am I still doing this right?

[edouble101 314]

cool idea .... you have too much time on your hands... soon as money starts rolling for me I'm gonna have to take away some of that free time for help with my build....

Actually I do not have much free time. This is a rainy/snowy day project

I will make time to help you though

[ricksi30 923]

Actually I do not have much free time. This is a rainy/snowy day project

I will make time to help you though

That's very nice!

[95Honda 486]

As long as you do the math right you should be OK.

[///M5 2,833]

First amp tested is a PPI Ax400. Max unclipped output across all four channels is at 15.16v a/c @ 1000hz. Dummy load resistance was 4.2 ohms starting and after the test. This would make max unclipped power at 54.72 wrms. This was at 12.54v d/c.

2 somewhat major flaws in your testing

1) Significant digits. Reporting 54.72 wrms is not valid. You need to take into account the accuracy of the devices you are testing and reporting beyond their accuracy isn't appropriate IMO. The easiest and most appropriate is to state the uncertainty. ie, 54.72 +/- 5wrms or whatever the overall electronic uncertainty is.

2) You unclipped measurement will vary by a fair amount over what is clipped and what is not, visualization on a cheap scope is not very clear as to when the signal is no longer a pure sine tone. Personally I'd measure THD and test at a consistent level of distortion. There is a reason power is related to THD in other tests although a standardization of the true measurement of distortion would be helpful.

[///M5 2,833]

I'd do a repeatability measurement as well. ie, measure power, measure load, measure power, measure load, measure power. That way you'll see how much variance is imparted from the load itself. The more variance the more cycles I'd use to determine the uncertainty.

[NB50 206]

As long as your meter actually reads RMS, this will work. You could also take the Pk-Pk measurement with your scope and do the math...

Having a battery as a power source is tough because it is unregulated. Try and do the test quick before the battery voltage drops too much.

Also try and do it quick before those dummy loads heat up. If they do (they look like the Partsxpress ones) mount them to a metal plate or heatsink. For high power tests I usually had them sitting in a 5 gallon bucket of cold water. (clean water isn't really conductive at all)...

Again, like you stated, check the resistance of the load. The tolerance may not be too tight on those...

And yes, follow the tips above and you are doing it right...

Have fun....

My meter does not read RMS. I am measuring a/c voltage then doing the math. Am I still doing this right?

No, your meter does read RMS and you are calculating from an already RMS voltage.

[edouble101 314]

@M5 - THD testing is another route I would like to look into.

@bromo - Thanks for clearing that up. I will need to educate myself further on the tools I am using!

[Impious 1,476]

M5 and 95Honda both hit on my main points. Accuracy of the equipment and load, basing the measurement on a visualization of a waveform on a O-scope of probably not great resolution, along with maintaining proper supply voltage throughout all of the tests are all issues. A long with that, clipping isn't a THD measurement so trying to compare these numbers to rated power or even necessarily usable constant power isn't going to be possible.

[95Honda 486]

M5's point was good about tolerance.

Also, a word on visible clipping. I could not visually discern anything that less than 5% THD (as measured with an H/P distortion analyzer) when looking at the waveform on a high-resolution O-scope.

Laymans terms, you can't really get a good idea of THD looking at a sine wave on a scope....

[edouble101 314]

I am in the process of updating my computer's audio (currently RealTek onboard audio). I think I am going with a M-Audio Audiophile 192 sound card. After I get that calibrated and get my mic calibrated I will post up my THD testing method. I will be using TrueRTA and a mic.

Note: Keep in mind I am not doing these tests for anything more than my entertainment. I enjoy this hobby. I am comfortable with a certain level of inaccuracy in my testing methods. Basically, I don't want to invest goobs of money in test equipment, less than $300. I do appreciate everyone's input.

[///M5 2,833]

You need to test the amplifier's THD as a signal NOT from a mic.

I wouldn't buy anything until you really determine the steps.

I'd also be really surprised if you could even see 5% on that scope.

[edouble101 314]

You need to test the amplifier's THD as a signal NOT from a mic.

I wouldn't buy anything until you really determine the steps.

I'd also be really surprised if you could even see 5% on that scope.

This is a point where my testing inaccuracy will be defined and accepted. I am not going to buy a better scope unless it is under $300. The only other method that I know of is with software and a mic, I have both. Your thoughts?

[///M5 2,833]

You need enough voltage input to test the amplifier, not going to get that with a soundcard. Microphone is useless as you will be measuring the THD of the combined amplifier and speaker.

Measurement equipment is not cheap. The mic I use is $1500, the analyzer close to $50k, power supply and load are also in the >$30k range and so on. Obviously anything cheaper CAN be used, but expect the uncertainties to climb to nearly pointless levels. Using that scope is there. If you are within 10% of THD with a variance of 5-10% in which level you see the power will be all over the board.

It may be possible to use an analog filter to take out the fundamental and then use the sound card for the harmonic analysis. It's another variable to quantify, but not so easy.