AUDIX 35102 module weird ramp up?

[mkiijam]

I have this racked Audix module that appears to slowly ramp up when powered on. The power supply seems fine, as I have a steady +24V at the power pin on the module, but the audio appears plenty loud, then dies off by about 90% then dies off to 100% then slowly comes back up, first with lots of harmonic distortion (headroom issue I'm guessing) then finally it works normal. This all happens in about 60 seconds. It is almost like it is some slow ramp on feature.

 

[mjrippe]

Sounds like a cap dying on the power rails.

 

[mkiijam]

Is this not a ramp up? Seems to me that the DC bias won't come up until C11 charges up?

 

[mjrippe]

It is not supposed to ramp up, so probably a cap that is failing.

 

[mkiijam]

C11 doesn't take time to charge up? When I watch the DC offset on PIN 3 of the Ne5534 it slowly comes up. I have two modules that are doing the exact same thing

 

[ruffrecords]

C11 doesn't take time to charge up?

The time constant (30K x 100uF) is 3 seconds.

Cheers

Ian

 

[Matt Syson]

When I tested those modules in 1982 in the factory they did not take a significant time to 'ramp up' but like all electyrolytics even when new they took about a minute for the internals to re form to a stable 'leakage' current value so switches would click for the first few minutes but then no one EVER starts using these within a minute of switch on. The IBA specifications said that gear had to reach 'normal' operating conditions within a few minutes and then remain essentially constant 'forever' (until after next power up:down cycle anyway. Electrolytic capacitors function as an electro chemical process and a small 'leakage current is entirely normal. This is WHY there are usually 2 capacitors in series with the centre point taken to ground to bleed off the leakage of the first capacitor. Typical leakage was IIRC less than 3uA or a percentage of the capacitance value. Thus DOUBLE capacitors where a signal path is switched as the click (change in DC level) had to be minimised. A 'click' being defined as a disturbance of more than a couple of dB above a steady 0dBu reading on a PPM meter where the gain was raised such that NOISE would read 0dBu. A test that would challenge a lot of the gear that is produced these days but then PROPER tests are not carried out because people focus on the pretty lights and whizzy things. A 'stability test' is a fun one where you turn ALL channels acc gain controls and all EQ to max and listen to the resultant output (around +4dBU or more of noise which is hopefully 'fine grass'(white or a bit coloured noise).