Topic: Hum problems when using the ADI2 PRO FS R BE as a tester

I am using an ADI2 PRO FS R to test various phono stage designs.
Unfortunately, there appears to be a hum problem which limits its usability, or at least makes it less convenient.
I use the DAC output in single ended mode and take the output from the preamp in balanced or single ended mode.
The hum is independent of the output gain setting but can be proportionally reduced by placing a resistive attenuator (3k in the signal, plus 3k in the ground, with 10 ohms across the output for example) between the DAC output and the phono stage input.
I use a laptop to control the measurements using REW, and it improves if the laptop is connected to mains ground.
A USB isolator also makes no difference, except to cause intermittent connectivity failures.
The units are not connected to any other piece of gear.
Is there a way to improve this situation, as although I have a functional work around, it's not particularly elegant.

2 (edited by KaiS 2022-07-23 01:21:06)

Re: Hum problems when using the ADI2 PRO FS R BE as a tester

You can try the following (listed in no specific order):

• Place the resistive attenuator at the input of the phono stage, not somewhere in-between the cabling.

• Lay the send and return cables from and to ADI-2 as close together as possible to reduce the ground loop’s area.

• Locate where the hum is generated, disconnect the DAC to phono stage cable and ground phono stage’s input.

• Check the hum level through the configuration without the phono stage (bypassed), but all cables stay in place.

• Ground ADI-2’s chassis on one of it’s back screws.

• Instead of single-ended, try an impedance-balanced connection from the phono-stage to ADI-2:
Use a balanced cable, cold (XLR pin 3) is grounded at the phono stage’s output side (e.g. inside the RCA plug) through a resistor equal to phono stage’s output impedance.

• Try connecting screen (XLR pin 1) through a ca. 100 Ohm resistor, to remove a ground loop between ADI-2’s in and out.

• Use an as high as possible ADI-2 output level (Reference Level +24 dB) and dimension the attenuator accordingly.

3

Re: Hum problems when using the ADI2 PRO FS R BE as a tester

KaiS list misses only the power supply - try running the ADI on battery. This might be leakage current buzzing, not typical hum.

Regards
Matthias Carstens
RME

4 (edited by KaiS 2022-07-23 07:50:51)

Re: Hum problems when using the ADI2 PRO FS R BE as a tester

MC wrote:

KaiS list misses only the power supply - try running the ADI on battery. This might be leakage current buzzing, not typical hum.

The laptop likely contributes to this type of stray noise too.

Both PSU’s stray currents should be largely eliminated, or at least reduced, by grounding ADI-2 Pro’s chassis.

Battery is still worth a try.

5 (edited by wynpalmer 2022-07-23 20:20:46)

Re: Hum problems when using the ADI2 PRO FS R BE as a tester

I have a battery arriving today.
1. the resistive attenuator is placed at the input of the phono stage. I need to have c. 60dB of attenuation to remove the hum and the noise from the ADI-2 pro so that the phono stage can be successfully measured without separating the distortion and noise floor measurements.
2. The phono stage noise is quite low- at about -87dBA relative to 1mv input and with 70dB of gain.
3. I need to use coherent averaging to measure the distortion, but it's barely different from the loop back distortion of the ADI-2 pro.
4. I've dressed the cables in various ways, including far apart and essentially twisted with minimal effect.
5. The phono stage is designed to force signal current to flow from the input ports to the output ports. Having the input port ground shorted to signal ground will cause ground current to flow backwards through the preamp and generate noise if the ADI-2 is attached at the input.
6. No hum pickup is measured if the input ports are shorted.
7. There is no evident hum in loop back mode- i.e. the phono stage is shorted.
8. The phono stage has XLR and phono inputs and outputs. The XLR outputs are balanced, the inputs are not. The + and - pins of the input XLR are connected to the + and gnd connection of the phono, just as the phono input is. The XLR chassis ground is connected to the phono chassis ground, which is not connected to the signal ground anywhere, unless chosen to be.
I can short the various chassis grounds to various points in the signal ground "space" using wires. There is negligible effect.
All of these things were done before I posed this question.
The phono stage output impedance is from an opamp with c. 130dB of feedback at 60 Hz. The output impedance is essentially whatever the cable resistance is.
I guess that I was looking for some solution that I had not considered.
Thank you for your inputs, much appreciated.


EDIT: The battery arrived and was installed. It made no difference.

6 (edited by KaiS 2022-07-23 23:35:53)

Re: Hum problems when using the ADI2 PRO FS R BE as a tester

wynpalmer wrote:

1. the resistive attenuator is placed at the input of the phono stage. I need to have c. 60dB of attenuation to remove the hum and the noise from the ADI-2 pro so that the phono stage can be successfully measured without separating the distortion and noise floor measurements.
2. The phono stage noise is quite low- at about -87dBA relative to 1mv input and with 70dB of gain.

• ADI-2’s nominal single ended output level is +19 dBu, 6.9 V.
If you’re heading for 1 mV input your passive attenuator should have 76 dB attenuation.

This would improve hum by 16 dB.

wynpalmer wrote:

3. I need to use coherent averaging to measure the distortion, but it's barely different from the loop back distortion of the ADI-2 pro.

8. The phono stage has XLR and phono inputs and outputs. The XLR outputs are balanced, the inputs are not. The + and - pins of the input XLR are connected to the + and gnd connection of the phono...

• If you are using ADI-2 Pro‘s XLR Out in single ended mode, pin 3 should NOT be grounded, but left open.
If you short a high signal level to ground, this can degrade signal quality.
ADI-2 Pro‘s XLR Out is NOT servo-balanced.

ADI-2 Pro’s THD specs can be found in the manual, that’s about what you can expect in loopback mode, without the use of a Double-T-Filter for the fundamental, on ADI-2’s input.

wynpalmer wrote:

7. There is no evident hum in loop back mode- i.e. the phono stage is shorted.

Clear evidence ADI-2 Pro’s in/out is clean, hum origins from external sources.

wynpalmer wrote:

6. No hum pickup is measured if the input ports are shorted.

So, in the interconnect ADI-2 to Phono the hum source is located.
Try as short as possible cables, add an extra interconnect ground wire of decent gauge.
Screen the (improved, see above) attenuator.

Remember, your signal level is just 1 mV in this stage, EVERY ever so little stray noise current spoils the result there.

Locating the passive attenuator INSIDE the phone stage (from RCA to circuit board) would exclude some interconnect resistances.


wynpalmer wrote:

The phono stage output impedance is from an opamp with c. 130dB of feedback at 60 Hz. The output impedance is essentially whatever the cable resistance is.

You’re missing an inline output resistor.
Few OP-Amps (mostly some lesser ones) can happily drive an almost pure capacitive load, like a cable, without getting instable.

7 (edited by wynpalmer 2022-07-24 14:01:31)

Re: Hum problems when using the ADI2 PRO FS R BE as a tester

KaiS wrote:
wynpalmer wrote:

1. the resistive attenuator is placed at the input of the phono stage. I need to have c. 60dB of attenuation to remove the hum and the noise from the ADI-2 pro so that the phono stage can be successfully measured without separating the distortion and noise floor measurements.
2. The phono stage noise is quite low- at about -87dBA relative to 1mv input and with 70dB of gain.

• ADI-2’s nominal single ended output level is +19 dBu, 6.9 V.
If you’re heading for 1 mV input your passive attenuator should have 76 dB attenuation.

This would improve hum by 16 dB.

wynpalmer wrote:

3. I need to use coherent averaging to measure the distortion, but it's barely different from the loop back distortion of the ADI-2 pro.

8. The phono stage has XLR and phono inputs and outputs. The XLR outputs are balanced, the inputs are not. The + and - pins of the input XLR are connected to the + and gnd connection of the phono...

• If you are using ADI-2 Pro‘s XLR Out in single ended mode, pin 3 should NOT be grounded, but left open.
If you short a high signal level to ground, this can degrade signal quality.
ADI-2 Pro‘s XLR Out is NOT servo-balanced.

ADI-2 Pro’s THD specs can be found in the manual, that’s about what you can expect in loopback mode, without the use of a Double-T-Filter for the fundamental, on ADI-2’s input.

wynpalmer wrote:

7. There is no evident hum in loop back mode- i.e. the phono stage is shorted.

Clear evidence ADI-2 Pro’s in/out is clean, hum origins from external sources.

wynpalmer wrote:

6. No hum pickup is measured if the input ports are shorted.

So, in the interconnect ADI-2 to Phono the hum source is located.
Try as short as possible cables, add an extra interconnect ground wire of decent gauge.
Screen the (improved, see above) attenuator.

Remember, your signal level is just 1 mV in this stage, EVERY ever so little stray noise current spoils the result there.


wynpalmer wrote:

The phono stage output impedance is from an opamp with c. 130dB of feedback at 60 Hz. The output impedance is essentially whatever the cable resistance is.

You’re missing an inline output resistor.
Few OP-Amps (mostly some lesser ones) can happily drive an almost pure capacitive load, like a cable, without getting instable.

Replies to various elements of your response.

The output section of the phono stage normally drives a balanced buffer that has an effective noise gain of two for both output phases. It has a 47 ohm series resistance in line with both output phases.
However, in this particular test mode the output is taken directly from the second stage of the phono equalizer. It does not have a series R, but it has a noise gain of c. 11 and is quite stable into the 3' cable that I am driving. Both outputs have been simulated and tested into appropriate capacitive loads, and stability is not an issue.
(Incidentally, I'm a retired ADI Senior Design Fellow, and I used to design high performance opamps, amongst many, many other things).
76dB attenuation is too high as I am using the same attenuator for MM and MC with a difference of 20dB in gain and I'm measuring various things, including RIAA accuracy with its 40dB gain range. I'd prefer to have the lowest attenuator value that I can get away with.
I have the shortest feasible cabling- it's about 50cms.
I am quite aware of the ADI-2 PRO FS R specs, and the difference between the balanced and single ended stages.
I also am aware of the distortion results. Amongst other things, I have characterized the total loopback distortion versus output level and frequency as a baseline for the preamp performance. This characterization includes several three-tone metrics and a number of many tone cases. In most cases, but not all, the measured preamp results are indistinguishable from the loopback performance of the ADI-2 box.
The REW and Virtins pro tools that I use, have the input scale set to reflect the fact that the output is +19dBU in single ended mode. I actually set the output gain to -1dB in recognition that the scaling is not from 24-18dBu, but from 24 to 19dBu, and set the level accordingly. If this seems convoluted, it's because it's easier for me to mentally keep track of the gain ratios with a clean 6dB difference. If I gave the impression that I drive the XLR inputs directly, I'm sorry.
I do not drive the phono input with the ADI-2 XLR outputs, just the single ended. The XLR input is intended for use with a phono cartridge where the -ve pin can be safely connected to phono input ground as cartridges have no inherent ground reference.
In loopback mode the gain is essentially unity. Surely any potential for hum loops is suppressed by the lack of gain. Adding 70dBs of gain changes that dynamic.
In any case, it has been an interesting exercise and the system works adequately with the crude attenuator that I have constructed. Now I need to make it more robust...

Re: Hum problems when using the ADI2 PRO FS R BE as a tester

To overcome ADI-2’s THD limitation, try a two-tone generator setup where ADI-2’s left channel plays one, and right channel the other tone.

Combine the tones with passive (resistive) summation, best done in the attenuator circuit.
Now both tones contain their harmonic distortions of course, but: zero IMD products!

Place two (active bootstrapped or passive) Double-T-Filters at the DUT’s output to reduce the fundamentals.

This way you can reveal distortions of your DUT that are way below ADI-2’s.