Hi All
This is my first post so please be gentle
The project I am trying to build is an amp tester, I have a 1000w resistive load bank and I will feed a 40hz sine wave into the amp and read the output. That part is fine, but I wanted to drive the amp until clipping. The idea I had is to use a voltage divider (the same one I am using to read voltage output of the amp) and read the sign wave. The reading should go up and down and should only be the same voltage read twice at the very top or very bottom of the wave. When the output is being clipped, the top and bottom of the wave should be flat, so I would receive more than two readings of the same voltage. would the Arduino (UNO or Pro mini) be precise enough to measure this? With no other commands for the arduino, I thought it should get over 150 reads per sine. Which should show clipping quite quickly when it occurs.
The other problem is with the code (I am still learning), I can make it set off the clipping led when it has two of the same numbers (using if oldvalue == new value), but I dont know how to trigger it with three common numbers.
Any help is greatly appreciated and if you need anymore info, let me know.
Hi jremington thanks for the reply
I did read the "how to use this forum" before posting, I havent written any code yet, which is why I put it in the project guidance part. I was looking at the feasibility of reading voltages quickly and accurately. I have done some research and found that using two 100k resistors for a voltage divider will give a 2.5v reading from 5v and if connected to the amp output with a 10uf capacitor, it should oscillate around a 2.5v centre.
The Arduino is fast enough to detect that. However, the ADC is only 10 bits and hence has an accuracy of around 5mV. So two consecutive readings can easily result in the same reading; you need to add an additional check (e.g. reading >= 1020) to detect the clipping.
Note: no experience with this, just my thoughts for when I would design something like what you want to build.
Pottomas:
When the output is being clipped, the top and bottom of the wave should be flat, so I would receive more than two readings of the same voltage
It depends on the quality you want to measure. Clipping does not necessarily produce a flat cut of the wave, and distortions will increase already before such a hard clip.
Is the real (specified) load really resistive, or is it a complex impedance?
You better measure the peak voltages, both positive and negative, and compare these values against the expected unclipped values. Peak values are easily obtained from a rectifier (diode) and capacitor, with almost no time constraints on the sample rate. For automated measurement you'll have to measure the input peak voltages as well, or you'll have to control the input amplitude by other means, in order to tell the cut off amplitude.
@mrburnette, I think you might be right, I was hoping to have it fully automated, but I think I will have to increase the gain myself while I am monitoring the Oscope
@DrDiettrich, The load is a 4 ohm wire wound resistor. The hope for the project was to hook up a car amplifier to the load bank and run it until clipping and produce a max watts (car amplifiers are sometimes a little overated :o )
Rookie question, as I am putting a constant tone through the amp would the max voltage output change from before clipping to after clipping. As the reason the signal is being clipped is the amp cant provide the power to maintain the full sine.
As the reason the signal is being clipped is the amp cant provide the power to maintain the full sine.
Ummm...
If I remember that audio class from many, many years ago, the Z of the speaker (4 Ohm in this case) is calculated as a complex resistance at 1KHz. Therefore, a 4 Ohm resistor would match the power level at 1K Ohm - now, because 4 Ohm is a very low value, the percent of change in the non-complex impedance is small and maybe it can be neglected at 40Hz ... maybe not. It's going to take someone with more current knowledge of the digital amplifiers to put that question to bed.
Inductive loads (loudspeakers) result in a (frequency dependent) phase shift between voltage and current. But IMO this will typically result in a higher clipping level, because the maximum current occurs at a lower voltage, not at the hard clipping level determined by the amplifier supply voltage. OTOH the resistance of a loudspeaker is much lower than its nominal impedance at 1kHz, so that at lower frequencies the current increases and can cause earlier clipping.
I'm still not sure what really shall be measured. A clipping level at 40Hz may be appropriate for a bass (low frequency) booster, but not for audio amplifiers in general.
The first amp I will be testing will be a bass amp, I can always repeat with a 1kHz tone for testing for other amps. I am still waiting for my scope to be delivered, so I can do some testing.
Thanks TomGeorge, I have had a look into these, but I would need to calibrate it for each amp I test. I am trying to devise a way to measure any amp (up to 1000w) and signal when it is clipping. I thought of testing distortion instead of clipping, but I thought testing for clipping would be the easier route.
