Can anyone identify the source of this distortion artifact? in mic preamp

[lassoharp]

In the included link there is a sample of two tom hits that compare two different mic pres. Each sample is identical except for the mic pre (each was made with the same tom, same mic (Senn 421), same cable, same DAW).

The first sample uses a recently restored Altec 9470 and the second one uses the onboard pres of an RME interface.

There is apparently an issue with the 9470. It can be heard in the tail of the tom hit as a fizzy/buzzy echo sound that at first listen seems to sound like typical drum annoyances (lug, spring rattle etc). It may be, but they are absent on the RME sample and this test has been repeated with four different mic pres under the same conditions , including another identical 9470 module from the same era and they all produce the same results that you hear in the sample. It was also tested with a different drum and same results. So there is clearly something going on with the one 9470 module and I'm not sure what else to check. The module had both electrolytics replaced (C7, C8), also C1 and C2 replaced and all the resistors replaced with carbon film. It passed a clean sine wave on the scope and the headroom measured exactly to the factory spec for high power strapping (+ 27dbm) and all voltages measured within spec. The transients look clean. Any ideas on what might be causing this?

*The bottom end has been rolled off in the samples so that the buzzy echo can be heard more clearly. You'll need to listen on decent speakers or headphones. Be sure and hit the repeat button on the SoundCloud player as it may take a few listens to adjust your ears to it and it's a very short sample.

 

[radardoug]

If you look at the waveforms close up, theres more distortion on the first one, plus a step in gain towards the end of the sample. Fixed gain preamp, maybe its overloading?

 

[lassoharp]

If you look at the waveforms close up, theres more distortion on the first one, plus a step in gain towards the end of the sample. Fixed gain preamp, maybe its overloading?

Yes it's fixed gain. He had the 20db pad engaged. Its probably not quite as clean as the RME but the transient still looks ok. It's the pronounced buzzy thing on the tail that's puzzling. It also has a level control on the output so that the interface doesn't get overloaded and those levels were watched carefully and the issue was still there. The person who owns this is a drummer who records lots of drum only tracks that are very tom heavy and depend on lots of resonance so it really sticks out.

 

[radardoug]

So you never did actual distortion measurements?

 

[lassoharp]

So you never did actual distortion measurements?

If you mean using a distortion analyzer, no. Only what I mentioned in the post plus looking for obvious clipping in the waveform in a DAW.

When you said you see distortion in the waveform, where are you seeing it? In the initial transient or the trail?

 

[scott2000]

Weird...Almost sounds like wireless router /cell phone noise ....

 

[emrr]

20dB pad means it would have to be overloading at something like 0dBu from the mic. Possible. But the others don’t - impossible.

 

[freshtapescent]

Check the matching on R5/6, R7/8, and R9/190, and as always with 9470s, triple check all necessary grounds are tied and no ground traces are cracked, etc. As the transient isn't lopped off, we didn't just bang the rails...it sounds like the pushing and the pulling is not happening as complimentarily as it should be. When you clip 9470s it sounds bad and you'll 100% know it.

 

[emrr]

Yeah getting transistor matches in these is a bitch.

 

[radardoug]

If you mean using a distortion analyzer, no. Only what I mentioned in the post plus looking for obvious clipping in the waveform in a DAW.

When you said you see distortion in the waveform, where are you seeing it? In the initial transient or the trail?

In the tail. Plus there is a very obvious gain step.

 

[lassoharp]

Check the matching on R5/6, R7/8, and R9/190, and as always with 9470s, triple check all necessary grounds are tied and no ground traces are cracked, etc. As the transient isn't lopped off, we didn't just bang the rails...it sounds like the pushing and the pulling is not happening as complimentarily as it should be. When you clip 9470s it sounds bad and you'll 100% know it.

All resistors were replaced with matched sets so should be good but faulty connections due to broken traces and lands have already been a problem in this particular module. Good point about the push pull balance being off. Hopefully its not the transistors drifting out of spec. Have you by chance ever tried the rebuild John Hall recommended some years ago using 2n2484s and 2n2219s?

 

[Dan Elleson]

It's not crossover distortion. It sounds more like it's starved for power. Is your power supply adequate? supply voltages high enough? and able to deliver enough current fast enough? Probe the supply rails in the pre amp and watch what happens to it when you hit it with some loud input

 

[lassoharp]

It's not crossover distortion. It sounds more like it's starved for power. Is your power supply adequate? supply voltages high enough? and able to deliver enough current fast enough? Probe the supply rails in the pre amp and watch what happens to it when you hit it with some loud input

PSU is Int'l Power 24VDC @ 1.2A for 500-525 ma load (4 amp modules). Should be fine

 

[abbey road d enfer]

I would look for an oxidized contact.
BTW have you loaded the output as recommanded?

 

[emrr]

I would look for an oxidized contact.
BTW have you loaded the output as recommanded?

These don’t require loading. I’ve documented it somewhere around here.

I am also suspicious of a contact, or even an intermittent transformer connection.

 

[abbey road d enfer]

These don’t require loading. I’ve documented it somewhere around here.

Yes, I know, it's in the datasheet. However, with time some characteristics drift and can impair stability. Loading the output may reduce this instability, and thus indicate what could be done to fix the problem.
As a diagnosis tool.

 

[moamps]

@lassoharp
I would like to ask are the new resistors carbon film or carbon composite because this distortion reminds me of the distortion of a carbon microphone. I would first check resistors R15 and CR2.

 

[lassoharp]

Yes, I know, it's in the datasheet. However, with time some characteristics drift and can impair stability. Loading the output may reduce this instability, and thus indicate what could be done to fix the problem.
As a diagnosis tool.

The output is loaded with a 1 K pot

 

[abbey road d enfer]

The output is loaded with a 1 K pot

So it's a dead end...

 

[lassoharp]

@lassoharp
I would like to ask are the new resistors carbon film or carbon composite because this distortion reminds me of the distortion of a carbon microphone. I would first check resistors R15 and CR2.

The amp transistors (R1-R8) were replaced with carbon films as were the dropping resistors (R11, 12 and 13). R15 was replaced with a metal film. The NFB resistors (R9 and R10) were left stock (carbon comp). btw that is one thing that makes the module in question differ from the other three 9470s. In the other 3, only R15, R13 and the diode pairs were replaced. They ended up being replaced in the one amp because I was tracking down a noise issue early on that had continued after reflowing all the joints and that sounded much like a noisy resistor.