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

Help Support GroupDIY:

emrr

Well-known member
Joined
Apr 12, 2006
Messages
7,915
Location
NC, USA
An intermittent transformer winding sounds a lot like a noisy resistor or discharging cap, or tube flicker noise, etc.
 

emrr

Well-known member
Joined
Apr 12, 2006
Messages
7,915
Location
NC, USA
But it usually sounds like it somewhat continuously, if current is involved. I've never seen a bad 9470 transformer but there's always a first. I'd still lean more toward transistor problems, though what I've always experienced has been lack of low end, with the output stage being imbalanced. If the output was severely out of balance or running on one side for some time in the past, there might be a transformer problem. I'd metal film the shit out of that thing and F some carbon.
 

lassoharp

Well-known member
Joined
Jan 3, 2009
Messages
2,067
Location
USA
But it usually sounds like it somewhat continuously, if current is involved. I've never seen a bad 9470 transformer but there's always a first. I'd still lean more toward transistor problems, though what I've always experienced has been lack of low end, with the output stage being imbalanced. If the output was severely out of balance or running on one side for some time in the past, there might be a transformer problem. I'd metal film the shit out of that thing and F some carbon.
For transistor issues, do you think it would show the same balance related symptoms if it were one of the low level pairs vs the output?
 

abbey road d enfer

Well-known member
Moderator
Joined
Jan 22, 2008
Messages
13,456
Location
Marcelland
For transistor issues, do you think it would show the same balance related symptoms if it were one of the low level pairs vs the output?
Yes. However is would be quite strange since transistors generally don't drift significantly. Unless they have been replaced?
There is no easy way for assessing matching of any pairs. Before trying, you must first measure if the currents in the output xfmr are balanced. This not too difficult to do. If it confirms there's an unbalance, I think there's no other way than unsoldering the transistors pairs and measuring them out-of-circuit. It's not an easy task; the manual warns about not overheating the transistors when replacing them.
 

lassoharp

Well-known member
Joined
Jan 3, 2009
Messages
2,067
Location
USA
Update: The preamp was returned to me and I tested it on a floor tom using a couple different mics and repeated this test at another local studio. What I found was that the fizz was not consistent in either presence (being there or not being there) or amplitude (no measurement, just how much could audibly be heard and how long it lasted). In some cases there was virtually no detectable fizz. In other cases there was an extremely low level (much lower than what is present on initial samples I posted). On other samples it was present and loud but in very short, burpy bursts. Then on some it was present just as it was in the original samples - very loud with a long tail. The loudest, longest instances were produced using a D112 under the tom. The other samples used an SM57 or Senn 421 on the top head. The D112 under the tom had much more low end content and resonated longer and the extended ringing seemed to be what aggravated the preamp the most. I found I could control it on one of the top head samples by damping the head with my finger. This seems consistent with the fact that all of the samples had relatively clean sounding transients, but then followed by the weird fizzy distortion on the tails.

I ran a 1K and 80Hz sine wave and checked the overload on a scope and top of the wave was bending slightly before the bottom in both tests suggesting the biases were off a bit. There's no adjustment so I swapped sides with the output pair and from that point on there has been no more fizz. (daily checks for 7 days worth). The sine wave peaks bend closer but now the bottom bends just slightly ahead of the top. So they're aren't perfect but the obnoxious distortion is gone.

I could think of no easy way to measure current as abbey road suggested. The DCR on the OT pri measures exact to two decimal places . . . just the same the sections on the input do (you don't see this very often) .

So my question now is whether the distortion was due imbalance based only on drifts in component values or whether it was possibly due to a connection issue that got remedied after the swap. I had already had a connection issue due to a faulty joint in the emitter circuit of Q5 partially losing connection and resulting in a string of different Q5 voltages over several power up cycles along with intermittent noise bursts.

I'm hesitant to do anything else to it as long as the distortion is gone because:

1) if it is the transistors, replacing the Q1-Q4 is a no fly as they are obsolete and unobtanium

2) replacing the 3053s (Q5-6) is no problem, but yet is a problem bc of what Emrr has spoken about numerous times - the circuit is fussy about beta matching and most new 3053s won't fall in that range . .. . . so how much time and money does one allot for this chore?

3) reflowing and desoldering as always is risky on the older pc boards and if the pad/land completely goes then you to have look at hardwiring connections. I've already had to do it once on this same amp but only for a very short trace run in the PSU section. To do it for long traces you'd probably want to cut a section out of the old trace to prevent parallel currents. I have no problem doing this when a pad has 100% failed, but I don't like doing it as preventative move just in case something fails

4) possible damage to the existing transistors

There is also the option of doing a full rebuild via the John Hall suggested transistor replacements, but again I'm hesitant to do it unless absolutely forced to and as of now the distortion has gone.

I'm curious to hear others thoughts as to the best way to proceed from here.
 

groovyloop

New member
Joined
Nov 7, 2021
Messages
2
Location
Russia, Kaliningrad
for such defects it is a good idea to put in sinus wave from any program and set the level just to hear the distortion. then trying to touch by fingers all components & move signal wires and listen - if the distortion change. If you find the point that change/remove distortion then it can be analized. it may be an oscilation.
 

groovyloop

New member
Joined
Nov 7, 2021
Messages
2
Location
Russia, Kaliningrad
if it is an oscilation just use your fingers while listening the result. fingers change the capacitance between ground and schematics also change the resistance if you press them more. It is the best instrument for finding the oscilation point.
 

FIX

Paul Wolff
Joined
May 5, 2021
Messages
102
Location
Nashville
Remember that a drum has a very low frequency component that if you follow the distortion curve could very well be heard. If you look at the insides of the old 808 drum machine, they have this flapping sound that they add to the drum sound. That LF flapping, which is natural to a drum, can cause distortion that doesn't seem in sync with he tom tone.
 

abbey road d enfer

Well-known member
Moderator
Joined
Jan 22, 2008
Messages
13,456
Location
Marcelland
if it is an oscilation just use your fingers while listening the result. fingers change the capacitance between ground and schematics also change the resistance if you press them more. It is the best instrument for finding the oscilation point.
Don't do that with a tube power amp, though... :)
 

lassoharp

Well-known member
Joined
Jan 3, 2009
Messages
2,067
Location
USA
Remember that a drum has a very low frequency component that if you follow the distortion curve could very well be heard. If you look at the insides of the old 808 drum machine, they have this flapping sound that they add to the drum sound. That LF flapping, which is natural to a drum, can cause distortion that doesn't seem in sync with he tom tone.
Are you talking about distortion that sounds like what I posted in the link?
 
Top