[TomS]

OK, my cheapo LG scope crapped out at the wrong time here, so I'm trying to do really something basic by hand - measure the relative voltage gain of various power amps.  I was doing this to develop a simpler procedure and options for matching dissimilar LF and HF amps in my biamped Emerald Physics CS2 setup.  Once you know the upper/lower amp difference, the output board on my Didden DCX XO can deal with about 15db difference (relative level cuts) easily, without having to boost.  Ideally an approach where it would not require a signal generator, scope, non-inductive speaker dummy loads, etc.  Yes, I do know I could do it acoustically with an SPL meter, with constant test tones to the woofers, bypassing the XO

Since I had no scope I downloaded a quick and dirty PC based o'scope from http://www.virtins.com/.  For audio signal out and in, I am using the M-Audio MobilePreUSB external sound card that I normally use for Room EQ Wizard measurements.  Virtins seems to work fine for basic stuff and includes signal generator, spectrum analyzer, multi-meter, freq, etc. and can convert Vpp, Vrms, DBv, DBu, etc. on the fly.

Since I don't have their nice mini-jack to o'scope probe converter, I decided to wing it and just used the line in via an RCA plug.  I set up a 100mv p-p 1khz sine output, measured/set the gain of the MobilePre to unity and then tried to measure some gain stages.

Burson Buffer = ~ +12dBv (mine has extra gain courtesy of Wayne)
Odyssey Candela preamp = ~+13dBv (about what Alex said it was)

All seems well for that, where it's sort of measuring voltage across the high input impedance of the pre.

Here is the part where I feel pretty dumb as I never really worried about measuring amps into a load much.  I have some really nice 250w non-inductive 8R Dale resistors to use, so I hooked them up to a few amps (obviously no speakers).  I just used a twisted pair wired direct from the load to RCA, into the MobilePre, taking care to set the generator low enough that the amplified signal level wouldn't overload the input of the MobilePre coming back into the PC (it clips long before the amps).

So I try to measure a couple of amps:

- Samson S1000 500wpc sub amp advertised at 35.2dB gain.
   Using the DBv scale, with 100mv p-p in, I measure input of -30dBv and output of -13dBv, which sounds like only +17dB (voltage right?).  I check the peak voltage levels on the scope and it actually sounds about rigth.

- Bella Extreme 3205 40wpc tube amp.  I don't know the gain spec, but it's probably something in the high 20 db's.  Same deal, except -30dBv to -17dBv for about +13dB.

Be gentle, but what obvious thing am I missing here - measuring on 8R load direct into the pre, calculation problem w/dbu's, something going on with the Mobilepre, ???

Tom

[JoshK]

What the most voltage input that the mobile pre can handle? 

Can you explain a bit clearer how you have the amp > load > mobile pre setup?  It sounds like you have it the logical way, but I am not sure.  By the way, I would be the blind leading the blind here, as I have not done any real measuring of amps before.  I am just helping you rule out some obvious things.

[TomS]

I'll have to find the manual for the MobilePre, but I can see it clip fairly easily, plus it has a clip led to warn you.  I'd say 1v max by looking at the scope.  It has a gain control for each input.

I have a mono 1/4" to RCA adapter from the MobilePre 1/L channel output to RCA to the DUT, such as power amp input RCA jack.  I have a pair of wires coming from the load resistor/amp output back to an RCA plug, then into the 1/L channel input on the MobilePre via another RCA to 1/4" mono adapter.

I loop the MobilePre output to the input via 1/4" to 1/4" to check/set the levels/gain first, then hook it back to the DUT.

Seems simple enough to me, but I'm missing something obvious.

[mgalusha]

I just use a signal generator and an RMS AC Voltmeter (my old HP 334A distortion analyzer).

I set the signal generator for .1V RMS out @ 1KHz as measured with the meter then feed this to the amp (with a 8R load on the output) and measure the voltage there.

Vout/Vin = Voltage Gain
20*Log10(voltage gain) = db gain.

so a voltage gain of 20 is 26dB of gain.

Most DVM's don't do well with high frequencies but are OK at 1KHz. Even a non RMS voltmeter should be OK since you are measuring relative levels, not absolute voltage.

Perhaps you can use the mobile pre as the signal generator and just measure the output directly without using Virtin. Not sure if it will work but it seems like any preamp should be able to output 100mV without clipping. You can download a PC signal generator from Marchand, it's basic but free and works good enough for this.

BTW, even though the meter will read in dB directly, I prefer to measure the actual voltage and then do the math. Seems a tad bit more accurate as the meter has better resolution on the voltage scales as they are not logarithmic.

Hope this helps.

mike