Need help with an unusual problem

[Larpy]

Hello all,

Last fall I was having trouble with vinyl playback.  I started a thread that went 4 pages before I realized I had all the good advice I needed and it was time for me to troubleshoot on my own.  No need to rehash that thread (https://forum.bottlehead.com/index.php?topic=15152.0, but in the course of trying to track down the problem, I created all sorts of new problems.  As best I can tell, the original problem was a bad transistor (or a bad solder connection to that transistor) in my Eros.  I now feel confident that I have fixed the original issue (and all of the collateral damage I created in trying to find that problem) except for one that I simply cannot figure out.

There is something "off" in my vinyl playback and the only non-objective evidence I have for it consists of asymmetrical waveforms I get when I record LPs.  It is a weird problem whose fix eludes me.

Here is what I'm experiencing in a nutshell.  The audio signal I get from playing an LP through my Eros to my preamp (routed through) the "preamp out" to a DAC to my laptop = asymmetrical waveform.  See first attachment.

The signal I get recording a CD from my CD player to my preamp (routed through "preamp out") to a DAC to my laptop = symmetrical waveform.  See second attachment.

Doesn't matter which LP I play or which input my analog signal is plugged into at the preamp.

OK, so you’d think there’s something amiss in my analog chain, but get this:

The signal I get from my turntable and Eros directly to a DAC to laptop = symmetrical waveform(!).  See third attachment.


So there’s something wrong with my preamp?  Apparently not.  I swapped it out with my Moreplay and I got the same results as above.

What on earth could cause this?

 

[Paul Birkeland]

It's been a while since I looked at your initial thread, can you remind me of the cartridge being used and any step-up devices? 

It is technically possible to overdrive the grid of the 6922 in the Eros, but you'd have to try really hard to do that, and it wouldn't disappear in your third measurement. 

It is also possible to overdrive a Moreplay and the BeePre 2, but again you'd have to have an exceptionally hot amount of output coming out of the phono preamp.  If this is all that's happening, then there are plenty of inexpensive and easy solutions to that problem. 

 

[Larpy]

The cartridge is a Hana ML, which puts out .4mV.  I'm using Sowter 1990 SUTs wired in the 1:10 configuration, so the signal to the Eros is 4V, at the bottom of the Eros' recommended input range.  So that shouldn't be an issue.

The thing that baffles me is that the signal from the Eros directly to my Apogee Duet DAC interface is fine, but when that same cable is routed to either of my two preamps, where it gets routed from the selector switch to line level output jacks, the signal is asymmetrical.  I've tried substituting a different cable from the preamp line level out to the Apogee, but I still get the same asymmetrical waveforms.

 

[Paul Birkeland]

It might help to get a scope and a generator and feed some small signals into the Eros/preamp combo to see what magnitude of voltage leaving the Eros has this clipping behavior.  A cheapo USB scope and an app on a PC/phone/tablet with an 1/8" to RCA cable would be enough to get this done.

 

[Larpy]

I've never used an oscilloscope but I'm game to try.  Does anyone have a recommendation for a cheap one?  (I looked on Amazon and ebay and was bewildered by the variety.)

 

[Deluk]

If you are playing an LP do you need to run it through a DAC? If you were saying an ADC then I might think well OK. A CD being already digital doesn't need a DAC after it comes from the player either. I have a Behringer UCA202 for recording LP's. very cheap and there are different versions. See some of the workflows here. Most revolve round Audacity.
https://forums.stevehoffman.tv/threads/ripping-vinyl-to-laptop-computer-what-is-really-the-best-way.984126/

 

[Larpy]

I was a little sloppy with my wording.  It's an Apogee Duet, so both a ADC and a DAC.  It's the interface I use to digitize LPs.  I mentioned it because it's the device that allowed me to see how my analog signal is, when routed through the line out jacks of both my preamps, distorted/asymmetrical.  I use Vinyl Studio for recording, and that's where my waveform screen shots came from.

 

[sl-15]

I am wondering if you have listened to the recorded files? They obviously look like that something is wrong but how do they sound?

 

[Larpy]

Yes, I have listened to them.  Whether the waveforms are symmetrical or asymmetrical, they sound a little harsh: no obvious distortion, but not the sound I'm used to.

Among the many odd things here is that, last fall, in a fit of frustration, I ordered another Eros kit and built an entirely new one (gave my other one to a young audio friend), and all of its voltages check out, so presumably it's not the problem.  And just a few weeks ago, I schlepped my Linn LP12 to a Linn dealer (7 hours away) and had it serviced, and it sounded spectacular after he installed a few upgrades (never visit a Linn dealer with an LP12 if you're not prepared to be talked into an upgrade ).  So I know for sure it's not the turntable.  The cartridge is brand new and I tried another brand new one as well to rule it out as the problem.  In hindsight, I should have brought my Eros with me to the Linn dealer, but I didn't think of it.

I'm willing to learn how to use an oscilloscope, but I'm hoping someone will recommend one that's cheap but useful.  I know absolutely nothing about oscilloscopes and every time I look for one online I get overwhelmed.  I searched for "cheap oscilloscopes" on the diyaudio website, and the threads it yielded were themselves overwhelming, with many of the recommended oscilloscopes no longer available.

 

[Doc B]

I have not used one for low level.audio stuff, but the little $40ish digital scopes on Amazon have worked well for me for tuning oscillators and such.

 

[Larpy]

OK, I am now the proud owner of my very own pocket oscilloscope.  I barely know how to use it, so please be patient with me, but I hooked up the oscilloscope the output from the Eros (RCA jacks) and looked at the voltage while playing an LP.  The highest peak to peak voltage I saw was a little over 16 volts AC.  Is that too high?  Is that the source of my problem (bad sound and asymmtrical waveforms when I digitize LPs)?

The voltages inside the Eros all measure correctly.  The cartridge's output is 0.4 mV and I'm using Sowter 1990 SUTs wired in a 1:10 configuration.

 

[Larpy]

To clarify: I'm still getting the same results as before:  the signal from the Eros directly into the Apogee Duet interface (~16 volts AC p-to-p) is a symmetrical waveform.

The signal from the Eros to the Moreplay's tape out jacks to the Apogee Duet (~25 volts p-to-p) is asymmetrical.  But the signal from my CD player to the Moreplay's tape out jacks to the Apogee Duet (also ~25 volts p-to-p) are symmetrical.

So if 16 volts from the Eros is enough to cause the asymmetrical waveform (clipping?), then why does the waveform look symmetrical when I digitize an LP straight from the Eros (bypassing the Moreplay and its tape out jacks)?

 

[Larpy]

It might help to get a scope and a generator and feed some small signals into the Eros/preamp combo to see what magnitude of voltage leaving the Eros has this clipping behavior.  A cheapo USB scope and an app on a PC/phone/tablet with an 1/8" to RCA cable would be enough to get this done.

Here is what I tried:  I played a frequency sweep through my phone's headphone out jack into the Eros input and recorded the Eros output through my Apogee Duet.  I hooked one of the Eros output jacks to the oscilloscope and saw that (by adjusting the volume of the headphone out signal) I could modulate the amplitude of the frequency sweep's amplitude from around 1 vAC p-to-p on up.

The signal on the oscilloscope looked OK to me, but I admit I have no idea what I'm looking for--the same kind of asymmetrical waveform?  If so, I didn't see that.

The waveform I recorded through the Apogee Duet interface looked symmetrical, but because it was just one (changing) frequency. I decided to play an a piece of music and repeat the process, recording the waveform as I turned the amplitude of the music track from 1 vAC out of the Eros up to about 16 vAC.  Resulting waveform was symmetrical.  See attachment below.

Not sure what I should try next.

 

[Paul Birkeland]

The highest peak to peak voltage I saw was a little over 16 volts AC.  Is that too high? 

That is, for lack of a better term, damn high! That would have to mean you have about 15mV going into the Eros, which would be a 0.4mV cartridge feeding the 1990 wired as a 1:20...  When you listen to a recording, does it sound like the EQ is working properly? (that there is present bass and the treble isn't overwhelming?)

I would triple check the SUT wiring and the cartridge specs.  If all else fails, then do your best to track down an old cell phone or tablet with an 1/8" TRS headphone jack and get an 1/8" to RCA cable, then we can send a very low level signal into the SUT and see what comes out. 

You definitely have tracked down the source of the problem though, as you're overdriving the second stage of the Eros. 

-PB

 

[Larpy]

Attached is a photo of the Sowter 1990 SUT wiring.  It appears to be correct: according to Sowter, "for 1:10 operation connect the primary windings in series (Orange to Red)."  That's exactly what I've done.

The cartridge is a Hana ML.  Hana website identifies its output as 0.4mV.

I have an old laptop with a headphone out jack.  I'm using a stereo 1/8" TRS jack to RCA jacks cable to feed the Eros.  So I'm ready for your "if all else fails" option: what kind of very low level signal should I feed the SUTs?

 

[Paul Birkeland]

You will want to turn the volume all the way down on the laptop, then up the minimum amount off zero volume that you can.  Plug that into your Eros (powered off is fine) and measure the AC voltage with your scope across the RCA jacks and then across the 47K resistors on the secondary side of the input transformer.  This will allow you to verify the ratio and those transformers ought to take a fair bit more signal than the typical MC cartridge without a ton of issues. 

If you see a 1:10 step-up, then let me know what the minimum voltage you have that comes out of the laptop and perhaps we can use it and your scope to check the EQ. 

You may also find that this all works a lot better if the laptop is running off its battery.

 

[Larpy]

I'm back to troubleshooting after a few days of kitchen renovating.  Lots going on right now.

The lowest waveform I could get on the oscilloscope from my laptop to the input of the Eros wasn't as low as I thought it should be:  99-114mV pk-pk (-3mV mean).  At the other end of the Sowter SUTs I measure 1v pk-pk (mean bounced around from -4v to -15v).

This would seem to indicate a proper 1:10 step up of the voltage.

By the way, at the output of the Eros, I measured 6v pk-pk.

The next step?

 

[Paul Birkeland]

OK, 6V P-2-P is about 2V RMS, which is plenty low for the Eros. 

I think it's worthwhile to to put low level tones into the input of the Eros again and measure the output.  Try a 3.8kHz, a 1kHz, and a 300Hz tone.  All the output voltages should be different.

 

[Larpy]

Got the sine waves from YouTube.

300hz at input 213-293mv pk-pk    at output 13-14v pk-pk
1Khz at input 217-245mv pk-pk        at output 7.12-7.32v pk-pk
4khz at input 126-174mv pk-pk    at output 10.8-11.4v pk-pk

Way above the other day's 6v at output.

 

[Paul Birkeland]

300hz at input 213-293mv pk-pk    at output 13-14v pk-pk
1Khz at input 217-245mv pk-pk        at output 7.12-7.32v pk-pk

This is what I would expect to see.  I chose the two frequencies that will be about half/double 1kHz.

4khz at input 126-174mv pk-pk    at output 10.8-11.4v pk-pk

Why did the input level go down?  That may be something that has to be looked at on its own.  Still, you have a bit less than half of the input voltage that you had a 1kHz and you have more output than expected, which indicates that your EQ is not properly working.