Alternative method load for power testing amps?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have a powerful amp that needs to have it's power output tested and it's way beyond any power resistors I have at a potential 600 W RMS.
So I was thinking about an alternative load and the following occurred to me.
What I do have on the bench is an electronic load capable of dissipating upto 750 watts on DC. So, if I rigged up a large bridge rectifier on the amps output and fed the DC from that to the electronic load then in theory it should work ?
I have diodes capable of handling the power.

Can anyone see any problems with that arrangement ?
 
Administrator
Joined 2007
Paid Member
I can see where your coming from on this but I suspect there may be problems in how the amp sees the load. If the AC from the amp is rectified (and is it smoothed as well ?) then the loading on the amp could be destructive as the bridge will only conduct to maintain the peak voltage on the caps. The conduction angle will be small and the peak currents very high, just like with a tranny and bridge and cap.

The load has to appear as essentially resistive for this to work. Even if there are no caps and the load is say a FET (s) hung off the bridge then I'm still not sure tbh. It depends how the load reacts to the changing voltage and whether its reasonably linear (like a resistor).
 
Those issues occurred to me too. There would be no caps, I can't see a need and it would make the load reactive. The electronic load can be set in constant current or constant resistance mode, effectively in constant current it would reduce the resistance to maintain current so constant resistance would be the correct mode.
 
Administrator
Joined 2007
Paid Member
Without experimenting and trying I wouldn't like to say one way or the other really.

Resistors are linear, can the same be said for the electronic load as voltage and frequency (and not to mention the voltage being full wave pulses) vary.

Dunno is the honest answer.

An old electric shower heating cylinder makes a good dummy load :)
 
Member
Joined 2010
Paid Member
Cheap dummy loads

When it comes to 230V appliances, the resistance is usually too high for practical audio applications and of course,
they vary as the temperature rises too, so a directly water cooled element like an electric kettle is preferable and
some early types with open, spiral wound elements can easily be altered to any resistance you like.
A single heating element rated at 1.2 kW can be folded and trimmed to 8R with about 1 kW rating.
 
I have close neighbours so testing on speakers isn't an option.
However I found out if I lay the speaker on its face on the floor most of the noise is soaked up by the carpet.

The best thing to test an amp with is speakers as they are the correct load.

What about placing the baffles of the speakers against eachother (or as close as possible) and connecting one speaker out of phase? Theoretically, this would create an acoustic short-circuit, and the sound escaping this setup would be largely reduced. No?
 
I use 45off 20r 5W wired in series parallel to create a 225W 4r0 dummy load.
I have 4 of these and can test all my amplifiers to full power. My highest was a Krell Klone @>400W into 2r0. But it is cheap and easy for me to build up more if needed.

BUT,
my testing method is probably different from many.
I have described it a few times.

I arrange a 0dB/-20dB switch on the input signal to the amplifier.
Pass in a test signal at the test frequency and monitor the output signal on the scope. Adjust the scope with test switch set at -20dB to get roughly 6 to 8 vert divs for the Vpp output.
Change the scope three settings less sensitive. The signal may now drop out of lock. Flick the test switch to 0dB. The scope shows if the output is clipped or not. Flick back to -20dB.
Determine how far to adjust the test signal so that next time the output is very near clipping flick and read. Flick back. Each time the scope is read the amp is on full power for maybe 1 second at most. No overheating effects for the amplifier, nor for the test load.

Once you have the test signal set to exactly the value that is just short of clipping, you are now ready to measure the output voltage. Use your best DVM. Flick that switch to 0dB and leave it there just long enough for your DVM to stabilise it's voltage reading. This may take 1 second or 2 seconds at the very most. Flick back as soon as the reading is obtained.

You should find that your test load is still cold, or at worst only just perceptibly warm.
Similarly the amplifier heatsinks should be little changed from their -20dB power temperatures. For a 600W ClassAB amplifier, the -20dB output will be 6W. The amp temp should be near enough the same as quiescent conditions.

Now, think about longer period high power testing.
What is one achieving?
I think you are testing the heatsink aluminium. The output stage gets hotter and hotter as the main heatsink gradually approaches the final full power dissipation temperatures. This can be easily modeled by hand calculation. I would know what that is. I would also know by hand calculation the Tc of the output devices was for this long period test. This would give me NO NEW INFORMATION. I don't need to test the heatsink.

I want to test the semiconductors. A 10ms test (if I had the skills and equipment) or a 1second test tells me all I need to know about how the amplifier copes with worst case signals and worst case loads.

This short period testing also shows the effect of PSU voltage droop. You can set the test signal slightly high. The first half cycle is not clipped. The second is very slightly clipped, the third and so on, by the time the scope is showing the 10th half wave you have probably seen the PSU at full droop. You can actually see the clipping develop in real time. You can measure the difference between the no clip and the progressive clip, to see how far the power output rises/falls with changes in PSU voltage under load.

Thermal,
UK post is not too expensive. Would you want to buy a few dozen 20r 5W radial wire wound resistors? I bought 1000 from the USA via Ebay. They get used for soft starting and dummv loads and discharging capacitors.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.