challenge: record your amplifier voltage and do DeltaWave comparison

[pjug]

I bought a lowly Behringer UCA202 to see if I could do some voltage recording of amplifiers driving speakers at high levels. This is not a very good ADC and limited to 48KHz but I thought I'd give it a try, then upgrade if I gmake any headway.

I am using the track Tricycle by Flim and the BB's available here for $1.29: https://us.7digital.com/artist/flim-and-the-bbs/release/tricycle-11282941
This because of this track's extreme crest factor that allows sending large voltage peaks into real speakers while not blowing you out of the room.

Here is what I get recording voltage out of my pre-amp compared to the downloaded track, not terrible for the cheapo ADC, and what I would hope to approach with my power amplifier measurement:

To measure voltage, I am using a 220Ohm/3.3Ohm resistor divider, no buffer or anything [if you do this with a bridged amp please record with a floating system, e.g. laptop on battery power!]. Here is a pretty unsuccessful measurement of my Outlaw M2200 driving an Usher V601 speaker (this a shorter clip of the same track, to save my ears. So different spectrum from above):

I think I need to tidy up my setup (first go is just with clip leads). Also move my speaker farther away since it was just a couple feet away and I imagine the speaker EM emissions may be interfering.

Anyway, I wonder how wgood the results might be for folks with better equipment. Anyone want to try this? I'd love to see what kind of match is possible with AHB2, Ncore, purifi driving real speakers.

 

[Blumlein 88]

Might be good to do a loop back just with Berhinger as a base line. That is going to be best result possible for the recording interface.

I'm also wondering why there is 0 ppm drift at the speaker, but 22 ppm for the preamp.

 

[pjug]

Might be good to do a loop back just with Berhinger as a base line. That is going to be best result possible for the recording interface.

I'm also wondering why there is 0 ppm drift at the speaker, but 22 ppm for the preamp.

Might be good to do a loop back just with Berhinger as a base line. That is going to be best result possible for the recording interface.

I'm also wondering why there is 0 ppm drift at the speaker, but 22 ppm for the preamp.

On loopback, do you mean with 1KHz tone using REW? I would only use the Berhinger ADC in the loop, and use a better DAC going in. Does that sound right? Do you know how to translate results from a loopback test into expectation of what should be possible in the DeltaWave comparison?

On the drift, maybe I had the drift correction on in DeltaWave on one comparison and not the other. I was fiddling with settings trying to get good results.

 

[Blumlein 88]

On loopback, do you mean with 1KHz tone using REW? I would only use the Berhinger ADC in the loop, and use a better DAC going in. Does that sound right? Do you know how to translate results from a loopback test into expectation of what should be possible in the DeltaWave comparison?

On the drift, maybe I had the drift correction on in DeltaWave on one comparison and not the other. I was fiddling with settings trying to get good results.

No play music thru the Berhinger feeding the output back into the input. Analyze in DW. Original file vs the recorded one.

Sounds like you are feeding a preamp/DAC to your system and recording the result at the speaker terminals with the Behringer adc. Which is also a good way to do it. Is that how you have it configured?

 

[pjug]

No play music thru the Berhinger feeding the output back into the input. Analyze in DW. Original file vs the recorded one.

Sounds like you are feeding a preamp/DAC to your system and recording the result at the speaker terminals with the Behringer adc. Which is also a good way to do it. Is that how you have it configured?

Yes the top comparison is loopback from DAC output back through the Behringer ADC. But the DAC is pro-ject pre box ds2 DAC instead of the Behringer DAC. The bottom comparison is measuring speaker voltage through a divider (connection tapped at power amplifier terminals).

I can do the loopback as you describe but I think the Pro-ject DAC is better than the Behringer I'll try to do it tonight if I can get to it.

 

[Blumlein 88]

Yes the top comparison is loopback from DAC output back through the Behringer ADC. But the DAC is pro-ject pre box ds2 DAC instead of the Behringer DAC. The bottom comparison is measuring speaker voltage through a divider (connection tapped at power amplifier terminals).

I can do the loopback as you describe but I think the Pro-ject DAC is better than the Behringer I'll try to do it tonight if I can get to it.

You don't really need to do that extra comparison. You are doing it correctly, using your playback DAC and the Behringer. I simply wanted to know how things were connected and such.

Good idea btw, I wish more would do this. Of course I've been telling myself that and haven't done it myself for years. BTW, if you are worried about EMI, you can twist your clip leads around each other. This will reduce the pickup of such noise. The twisting need not be that tight for audio frequencies.

 

[pjug]

You don't really need to do that extra comparison. You are doing it correctly, using your playback DAC and the Behringer. I simply wanted to know how things were connected and such.

Good idea btw, I wish more would do this. Of course I've been telling myself that and haven't done it myself for years. BTW, if you are worried about EMI, you can twist your clip leads around each other. This will reduce the pickup of such noise. The twisting need not be that tight for audio frequencies.

Thanks for the tip on twisting the leads. When you did this in the past did you use an un-buffered divider like I am doing? I wonder if I would get better results using an op-amp.

 

[Blumlein 88]

Thanks for the tip on twisting the leads. When you did this in the past did you use an un-buffered divider like I am doing? I wonder if I would get better results using an op-amp.

When I've done this in the past I've done what you are doing.

One could use a buffer, but I don't know you'd get better results. It would be safer to not blow something up. Of course when I last did this Deltawave was available yet.

 

[pjug]

When I've done this in the past I've done what you are doing.

One could use a buffer, but I don't know you'd get better results. It would be safer to not blow something up. Of course when I last did this Deltawave was available yet.

Yes I was thinking, and still hoping, that a buffer isn't needed. I wouldn't think so because the resistor divider values are couple orders of magnitude less than the Behringer ADC input impedance (Behringer spec is ~27Kohms).

I hope some others will try this to get towards knowing how good of a match is possible. It doesn't take too much time or (I hope) expensive measuring equipment. Just take care not to short anything out!

 

[b4nt]

You choosed 220Ohm/3.3Ohm resistor divider. No need of a buffer for that, this is aprox 3.3Ohm output impedance.

Addind a buffer (extra cost, noise, DHT...) would be required if you would have much more than 3.3Ohms, and needing to compensate for ADC input impedances.

To be on the safe side with a brided amp, one coud use 110Ohm/3.3Ohm/110Ohm. Attenuation will be the same, and each amp output will be protected by a 110Ohm resistor in case of error/mistake. And connecting to the 3.3 resistor (laptop for ADC running on battery only).

I'm not sure you will learn a lot so, unless one is driving the same speakers, at same power level, for the same audio track. You would then get the diff for the amps and precise speakers.

 

[pjug]

I'm not sure you will learn a lot so, unless one is driving the same speakers, at same power level, for the same audio track. You would then get the diff for the amps and precise speakers.

What I am hoping to find, or with group effort at least get close to, is how close amplifier voltage can track a digital file input when driving real speakers (whatever the amp or speakers may be). I think I may eventually find that the cheap ACD I'm using won't be useful to get there. But so far it seems something else in my methodology is not as good as my ADC.

 

[b4nt]

Your cheap ADC adds noise at -85dB as far as I've read about it (embeds a poor op amp buffer?). It should also add minimal amount of distortion, and limitations in trebles/bass. But the very basic DACs chips (like in our PCs and phones) are correct, you should get a good picture.

Amp driving capabilities aren't the same, and different speakers should have up to a big impact on the signals.

Compare either same speakers and different amps, or same amps and different speakers. Ask a friend to bring his amp in your home? By the way, you could listen and record.

 

[pjug]

Compare either same speakers and different amps, or same amps and different speakers. Ask a friend to bring his amp in your home? By the way, you could listen and record.

I can do this too, but first I want to figure out how to make better measurement/recording. Next thing I try will probably be 8-ohm resistor load instead of speaker to see if the DeltaWave match is better.

 

[b4nt]

I can do this too, but first I want to figure out how to make better measurement/recording. Next thing I try will probably be 8-ohm resistor load instead of speaker to see if the DeltaWave match is better.

Compare the original audio track shapes and your records using any app like Audacity. See if the shapes around attacks and so are similar. If they are close to, you probably won't need a more expensive sound card.

Compare the track shapes, the shapes of capture after DAC, the shapes of captures after the amp.

 

[pjug]

Compare the original audio track shapes and your records using any app like Audacity. See if the shapes around attacks and so are similar. If they are close to, you probably won't need a more expensive sound card.

Compare the track shapes, the shapes of capture after DAC, the shapes of captures after the amp.

I did this and they seem to track well eyeballing it. Probably noise level or interference causing problems with the DW figures of merit looking bad. It might be some wireless gear too close. Just not as easy to do this as I thought it would be.

 

[Helicopter]

What are we doing here?

I understand the recording you made, and I am set up to make similar recordings, but I am puzzled by the charts and I am not familiar with DW.

Is the recording mono or stereo?

Is the chart showing the whole song with no time axis?

Is the chart showimg the recording you made or the null of that recording and the original?

What is the point? To evaluate amp performance under speaker load?

If so, how is this useful?

 

[pjug]

What are we doing here?

I understand the recording you made, and I am set up to make similar recordings, but I am puzzled by the charts and I am not familiar with DW.

Is the recording mono or stereo?

Is the chart showing the whole song with no time axis?

Is the chart showimg the recording you made or the null of that recording and the original?

What is the point? To evaluate amp performance under speaker load?

If so, how is this useful?

-Just one channel (remember I'm just using clip leads on a resistor voltage divider)
-Chart just shows spectra comparison but the numbers below are what I'm really looking at.
-Comparison is the voltage recording vs original dowloaded file used as input to the system.
-Yes point is to look at amplifier performance driving speakers.

As to what's the point: other threads have gone at this with microphone recordings or just listening to compare different amplifiers driving same speakers. To me recording voltage seems like a better approach to see if there are differences. So I tried this and it is trickier than I thought. So good, lets figure it out.

 

[b4nt]

I beleive he is willing to see some representative shapes of its amp outputs, driving his speakers.

Comparing captured track shapes and spectrum shapes using DW is a solution?

What is missing, is how audio renders at speakers side, for the listener.

 

[pjug]

I beleive he is willing to see some representative shapes of its amp outputs, driving his speakers.

Comparing captured track shapes and spectrum shapes using DW is a solution?

What is missing, is how audio renders at speakers side, for the listener.

I'm not caring what comes out of the speaker, only what effect its load might present to the amplifier. In my case my speakers are easy to drive but still I'm not getting good results.

 

[b4nt]

I'm not caring what comes out of the speaker, only what effect its load might present to the amplifier. In my case my speakers are easy to drive but still I'm not getting good results.

Not good results: not so good listening experience. Which is also room dependent.

You won't get your answers by asking such measures, from unknown setups and environments. But by asking a friend to bring an amp in your home, for listening tests... or maybe by changing your environment, or your speakers.