525 build thread

[dmp]

Can you  measure for the control voltage?
If you feed a audio wave to the unit and measure the DC voltage at the attack pot, you should see the control voltage go up as you increase the signal level feeding the unit.

 

[duantro]

I swapped out a fresh vcr2n, and no difference. DMP are you saying measure the dc voltage at the attack pot, while increasing signal level?

 

[dmp]

Yes - I would start out measuring on the attack pot (the side that faces the VCR2N) for the control voltage.  You probably will not have any voltage there. If you don't, then work your way back through the sidechain to see where things are getting messed up.
The basic operation is an audio signal goes to the ceiling control, which sends a signal to the sidechain (the ceiling changes a dual voltage divider to adjust the threshold and output together) . The signal to the sidechain  is amplified and rectified to a DC control voltage. That voltage changes the resistance of the FET (VCR2N) in the first 2520 amp circuit, to modify the gain of the amp.

 

[duantro]

Thanks so much DMP, for the detail explanation. I was sick over the weekend and now am backed up with work, but will check this asap!

 

[duantro]

I finally got around to looking at this. Spent about 5 hours searching for components and retracing wiring, etc. Still the same results. I'm not getting any dc voltage at the att connection on the meter board. I am getting between -2vdc to -12vdc on the top lead of r50 depending on how hard I send input audio. I double checked the wiring and solder joints going to the release time board. Input pot changes the meter reduction, but does nothing to the sound. Output pot is working, but when it is ccw there is still like 30 % audio coming through. Sw5 just makes out put gain hotter as turned cw. So frustrating...

 

[dmp]

You have a negative control voltage on R50 but not on the attack pot???
Which side of R50: CR5 or CR12/CR11?

 

[duantro]

You have a negative control voltage on R50 but not on the attack pot???
Which side of R50: CR5 or CR12/CR11?

On the CR11/CR12 side. I'm wondering if I should just jumper a wire from R50 to the inside att/c3 junction on the meter board, as it seems to die en route somewhere. The previous owner attempted a few repairs, perhaps a solder trace got too hot.

 

[dmp]

From the schematic the R50/CR11/CR12 node should connect to the Attack pot / C3  - so yeah something seems funny
I don't remember what the routing looks like and I don't have one of mine with me to look at now, so can't be much help.

 

[duantro]

That fixed the side chain! Compression at last! ....However, I'm still getting around 20-30% signal passing through, with the output pot turned off. Time to search for that and calibrate.

 

[dmp]

Nice!

The ceiling also attenuates the signal

 

[duantro]

Nice!

The ceiling also attenuates the signal

Right. I'm talking about the output pot itself. Regardless of the ceiling control position, I still get signal with output off. I think it might be time to double check the calibration now. Thinking of making sure R7 isn't way off.
Also, i need to do the mounting screw mod to disconnect gold finger 5 and 13 from each other. Does this image look correct to cut the red line?
Edit: i cut that and the same section from pin 13.

 

[duantro]

I finally got some free time to check these out, and I'm now getting low output. Signal drops down after R1. Q5 checks out E-B @.6VDC. 2520 voltages are 15.6VDC. Did I cut the wrong traces at gold finger 5 &13? DMP did you cut the trace on the top or bottom? I cut the bottom. I'm still getting continuity to chassis from pin 5 and 13, so thinking this is the issue.
EDIT: this cut on pin 5 is incorrect!

 

[dmp]

I went back to try to figure out what you were asking and thinking it is this mod, 10 years ago so I don't remember details. But the 5&13 had to do with the ground plane - should be pretty easy to see what is connecting to the ground plane.

From the 500 series specs, pin 5 is audio ground and pin 13 is power ground.

I think I found the proverbial smoking gun with these 525s.
The ones I built both had hum and a non-musical distortion (same as discussed on  pg 14 by Mich and Peter). The distortion was not caused by the FET - it was caused by the meter circuit. After turning R55 all the way off (so the meter would not light at all), the distortion disappears. There's a ground problem when the meter circuit shares the same ground as the in/out pot grounds. I think the current from the meter circuit was causing a ground loop between in/out pots and A3. Easy fix - I cut the trace leading from the neg side of CM1 to the GND on the meter board, and ran a wire from Neg CM1 to audio ground at the hole by gold finger 5. With this change, both of my 525s are quiet & distortion free. Even at heavy compression they sound smooth - a huge improvement. And the hum is gone
A second mod I'd recommend - after tightening down the mounting nuts a few times (or a zillion in my case), the silkscreen over the ground plane developed shorts, which connects audio ground to the chassis, not desirable. Easy mod was to cut the traces by gold fingers 5 & 13 to disconnect the pcb ground plane from audio ground. Then the short to the chassis through the mounting bolts makes the pcb ground plane a chassis ground, which is better for noise rejection.
With these mods the 525s are fantastic. I just played through them with guitar and singing and they are smooth and clean.

 

[duantro]

I went back to try to figure out what you were asking and thinking it is this mod, 10 years ago so I don't remember details. But the 5&13 had to do with the ground plane - should be pretty easy to see what is connecting to the ground plane.

From the 500 series specs, pin 5 is audio ground and pin 13 is power ground.

Crazy it’s been that long ago!
Yeah I understand everything about the mod, except deciphering where the trace cut goes. Do you remember if the ground plane that pin 5 &13 need to be  disconnected from are on the top, bottom, or both side?

 

[dmp]

do the nuts on the top side sit on a ground plane?  should be able to trace to the connections with a multimeter

 

[duantro]

Yes they do. I've deducted that the trace I cut on pin 5 underneath wasn't correct. I've soldered it back and cut the top trace to the left of pin 5. Audio and power ground now are unconnected from the chassis ground plane. I'm still not getting any gain reduction now  and output is low and intermittent. If I push on the output pot it comes back in, however it is still quiet. I'm thinking I may have heated up the solder pins so many times that the pots might be dead. I'll order some new pots and see if that will save this thing from the trash can!

 

[dmp]

The calibration of trimmer R7 adjusts the amp gain. If you are getting low signal with the output all the way up, you need to adjust R7

Regardless of the ceiling control position, I still get signal with output off.

Either your output pot was dying, or it was not connected to ground.  Make sure once you get the output pot replaced, that it is seeing audio ground (pin 5)

 

[duantro]

Finally got the new input and output pots installed, and the signal bleed is now gone, along with other erratic behavior. The amp gain is still lower than the other units, and r7 isn't doing anything. I'll order another r7 trimmer and see if that fixes it.

 

[duantro]

Finally getting around to more debugging. Compression works, but the output( and threshold sensitivity?) is 30db lower than working units. I've swapped 2520's, and a new r7 trimmer. I'm sending .666 VAC signal into units. If i probe on working unit at audio ground and out of A1 I get ..662VAC on working unit and .006 on faulty one. Something is shorting out. I've changed the input and output pots. The input pot seems to function but the gain reduction meter shows less sensitivity. can anyone think of places to check?

 

[dmp]

Trace back in the circuit until you find where the signal is dropping so much. Like Input to A2, output of A1, input to A1
Try pulling A3 also to turn off compression while you get the signal working.