QA403 Input Attenuator interaction with phono stage

[GlasgowGrip]

I thought I’d start my own thread rather than clutter up the very interesting Stereo Phono Input testing thread that’s been going on, with a tangential question. Hopefully the answer is simple but I’m not sure what it is I’m seeing…

Background here is that I’ve been testing some vintage phono stages in Marantz receivers and integrated amps, both restored and unrestored. I’ve made a lot of progress and really gotten a much better understanding of what to do/not do and once again how darn versatile the QA403 is. And many thanks to @VAR and @restorer-john, by the way!

Okay, so one thing I’ve been investigating is a common Marantz phono stage (same board was used in a number of different models) that can be made to “motorboat” - exhibit a loud low frequency oscillation - if certain modifications are made and certain conditions presented to it. Namely, the mod is replacing certain vintage transistors that tend to become noisy with modern low noise substitutes. The performance is fine but there are stability issues that are seemingly related to differences in Cob of the transistors - but that’s not really the point here…

Anyway, my question relates to an interaction I am finding in some cases between the phono stage and the Input of the QA403. I’m weak on this theoretical stuff so any enlightenment would be welcome! My crude assumption is that the input impedance of the QA403 must be changing as I change the “Full Scale Input (dBV)” setting in the software, and if it gets too low it can set off oscillation in the phono stage.

I get that there is a relay involved and a physical resistance is switched in and out when going between the 18 and 24 settings. But what is happening between the other settings? And specifically, can anyone tell me what is the impedance seen by the phono stage output when the QA403 input is switched between 24, 18, 12, 6 and 0? And my apologies if I missed where this is spelled out in the FAQ’s or Wiki - I’d gladly go read that, too.

The one thing I found is a statement by @matt in another thread Input Transformer Measurement - #6 by matt that reads “the input impedance appears as follows: The BNC input goes to a 4.7uF cap in series with a 470 ohm, followed by a shunt of 100k ohms. Overall, we consider it a 100k series input.”

But clearly it’s not constant and must be changing as I switch between Full Scale Input settings… since I am able to trigger oscillations in the phono stage output if I go too low… yes? Maybe?

If I understand what the QA403 input is doing, then I can better solve the problem caused by the phono stage modifications.

Thanks in advance! -Steve

[restorer-john]

I just took some measurements of the QA-403’s output impedance (generator) across the range from 50Hz up to 20kHz and it came to around 111ohms (100 is the rated), per leg.

So, you can use the generator to test your input impedance at various input levels! Get a 100k resistor, put it in series and sweep (FR) the inputs at various attenuator settings like I did.

Across the entire set of ranges, I got a -6.01dB (6dB) reduction (dead flat line from 20-20k), so that means the input impedance is pretty much a constant 100k for audio frequencies.

I’d be looking at the front ends of the phono stages for potential RF getting IN and causing the problem you’ve encountered.

The QA403 seems to ouput significant “hash” on my CRO when turned on via USB, whether the software is running or not- just powered up. My 200kHz bandwidth distortion meter shows 150uV constantly with a small fluctuation around half a second which corresponds with the moving pattern on the scope. The same pattern of “hash” appears whether running on my main PC or a laptop. Grounding does not affect it.

The direction of moving fluctuation in the CRO display changes direction for + and - outputs on the QA.

I threw a 0.0068uF greencap in parallel with a gen output and the hash went down to ~80uV. There may be a sweetspot for a fixed value RC or C that doesn’t affect the highest sampling rates for ADC and DAC and kills the out of band noise I’m seeing.

Maybe @matt can comment, but I’d love to see who else can see what I’m seeing.

-John

[restorer-john]

@VAR Can you connect your QA403 direct to a proper CRO at 1 or 5mV per div/ 0.2mS and tell me what you see? Just power up the QA- no software running. Compare the outputs and report back.

[VAR]

I think @restorer-john is probably correct about noise getting into the phono stages, though I don’t think it would be RF causing the noise that I typically see. Most likely it is power supply noise in my case that probably could use some cap replacements. But the phono stages work good enough for the folks whose units I get, so I can’t say going in and replacing caps and maybe transistors would improve things. After all, and please don’t shoot me, it is still vinyl. While I regularly listen to vinyl and usually enjoy it, I prefer streaming HD music or my own music ripped as .wav files that are on my “cloud”. While I doubt there is a market for having a “phono test mode” built into a future QA40x, maybe we will be able to come up with some impedance matching network that can be easily connected. I am going to post the thd/snr plot for a $5 phono preamp from ali express that I measured for a friend, a PP500. As far as THD/SNR it looking really clean considering it us a cheap 12vdc wallwart. The FR is another matter, it did not look so good. But it looks really clean and is using the same setup that I use to measure other phono stages (MM), so that may go to the idea that the older vintage gear needs a little “help” ? Unless it is owned by @restorer-john

[MarkT]

The attenuation network something like this inverse-RIAA network?

Where R8 and R7||C4 represent the actual phono preamp input stage rather than being part of the network, leading to these responses:

[GlasgowGrip]

[quote=“restorer-john, post:3, topic:1300, full:true”]
Can you connect your QA403 direct to a proper CRO at 1 or 5mV per div/ 0.2mS and tell me what you see? Just power up the QA- no software running. Compare the outputs and report back.[/quote]

Thanks for your replies John and Scott, just fascinating stuff, I am learning so much from you guys. I had a hectic day and no time at the bench today, but I should be able to do some playing around with this tomorrow afternoon. John, I don’t have a proper CRO but do have a pretty decent Keysight DSO so I’ll take a shot at your request then.

[restorer-john]

Have a look here: QA403 generator noise. IDLE vs RUN