525 build thread

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Three more things to confirm in my troubleshooting of an inherited build:

1) I jumpered sw5B sweeper to r64, and confirmed that the unit otherwise appears to behave properly.  This kills the giant thumps since it effectively bypasses the switch section, so something is up in this switch deck.  (I did put in the spacers from another switch, and it tightened it up mechanically, but not sonically)  Specific question: With sw5B jumpered out, switching from '0' to '2' the unit mutes entirely, then comes back up.  Is this normal behavior for sw5A, or evidence of other issues? 

2) following dmp's advice on meter trim, I discovered relatively similar steps in meter/GR correspondence.  What I don't see addressed anywhere, is what the attack control does to meter behavior.  In this unit, a slower attack causes the meter to show greater gain reduction than it does at the stock (fastest) setting.  As expected, the actual result of slower attack is that more transients pass through, and output level rises as expected.  A 1M attack control does work, but does the meter misbehave because of it, or is this unit also funky in this way?

3) I aligned the GR meter following dmp's clues, with the ceiling control set to position 10.  I then checked the measurements with it set on 2 and 20, and found that I got slightly less GR per LED on 2, and even less on 20.  At 7 LED's in particular I measured -4.6dB GR on 20, -5.8dB GR on 2, and -6.3dB GR on 10. 

Thanks for clues and observations. 
 
1) I jumpered sw5B sweeper to r64, and confirmed that the unit otherwise appears to behave properly.  This kills the giant thumps since it effectively bypasses the switch section, so something is up in this switch deck.  (I did put in the spacers from another switch, and it tightened it up mechanically, but not sonically)  Specific question: With sw5B jumpered out, switching from '0' to '2' the unit mutes entirely, then comes back up.  Is this normal behavior for sw5A, or evidence of other issues? 
I haven't seen this behavior. The 'A' side of the switch is going to the sidechain input. At '0' the input is grounded, at '2' it is being fed to the sidechain. It sounds like there might be a big DC voltage from the output of the A1 doa? This would cause a sudden spike to go into the sidechain and trigger it to compress the signal. A DC voltage on the input was the first thing I thought of for the thumps on your output 'B' side switch as well, since the resistors connect each switch position, and effectively make the 'B' side of the switch a voltage divider. You can check the switch pretty easily - disconnect the input and wiper from the 'B' side, attach R64, and measure the resistance of the wiper to ground as you turn the switch. Verify that it matches the schematic.
I'd try a different doa and check the components around it for proper value / connection. Check that the in/out pots are correct and grounded correctly. Since both sides of the switch are showing some funny behavior I'd look at the doa output.
 
2) following dmp's advice on meter trim, I discovered relatively similar steps in meter/GR correspondence.  What I don't see addressed anywhere, is what the attack control does to meter behavior.  In this unit, a slower attack causes the meter to show greater gain reduction than it does at the stock (fastest) setting.  As expected, the actual result of slower attack is that more transients pass through, and output level rises as expected.  A 1M attack control does work, but does the meter misbehave because of it, or is this unit also funky in this way?

3) I aligned the GR meter following dmp's clues, with the ceiling control set to position 10.  I then checked the measurements with it set on 2 and 20, and found that I got slightly less GR per LED on 2, and even less on 20.  At 7 LED's in particular I measured -4.6dB GR on 20, -5.8dB GR on 2, and -6.3dB GR on 10.

I'll preface this with... I spent HOURS trying to figure out how the sidechain of this compressor worked, including simulations, which just crashed - I eventually gave up.  But I can confirm the observation that the meter shows much different gain reduction as the attack is changed. I can't recall if it is the same as you describe (slower attack, more LEDs).
As to #3, the meter circuit in this compressor does not really capture the compression behavior of the FET and gain reduction structure of the first DOA. I wouldn't be terribly surprised to learn of odd behavior like this. In order for the meter to really track GR correctly, it would need a matching VCR2N to capture the functional form of the attenuation behavior. Since it is a feedback compressor it is possible I suppose that the input to the sidechain might change the relationship between the meter and actual gain reduction. 
 
Just remembered about the attack pot / GR meter - the attack pot is AFTER the meter in the sidechain circuit. In a feedback compressor, the sidechain reacts to the audio, generates a voltage which then has to go through the attack resistance and charge the attack capacitor BEFORE the fet attenuates the signal. But the GR meter indicates the sidechain voltage BEFORE the attack RC circuit, not the voltage on the base of the FET, which is what actually correlates with GR.
Because it is a feedback compressor, the sidechain voltage actually increases as you slow down the attack (before the attack RC), because more audio signal actually slips by the GR doa on a slow attack. So, long confusing explanation aside, it is another error due to a poorly designed meter circuit.
 
 
Thanks, confirms many things.  I think you're right; need to try a different A1 and also pull SW5 enough to measure it fully.  SC voltage increase with attack R change is making sense.  Input level to SC affecting metering is probably also correct, and would be fudged in the original with it's tiny low resolution meter movement.  We may see more here, and it may be less satisfying to know the limitations. 


----------edit--------

Switch is correct, and wired in properly.

DC at output of any A1 2520 I try is around 10mV on average, largest seen during switching is about 30mV.

A3 removed, sw5 thumps and chirps at every position.
A3 removed, sw5B jumpered sweeper to r64, 0-2 thumps, gain higher on 0.
A3 removed, sw5A jumpered sweeper to ground, less chirps, lower thumps.

Obviously something more to yet discover....
 
Hey Guys,
I am searching A Rotary Switch [1X 2 POLE 11 WAY MBB ROTARY] and found it here --> http://www.banzaimusic.com/Rotary-Switch-2-level-2x1x12.html
Can I use it? and I didn't find any info which resistor I'll use for the Rotary Switch ?
Thanks for any feedback...
Cheers,
 
2 levels, 2x1 pole 12 positions, without stop end position. Switching: Make-before-break.
Looks ok, but not sure what they mean above.
 
Thanks Pete and Ben !

I already have  used this one for Pete's Tr**dnt project  --->http://fr.mouser.com/ProductDetail/Alpha-Taiwan/SR2921F-0112-19R0B-E9-S-W/?qs=sGAEpiMZZMuLS1mv1rPY1mZCAvtEStLP
and works fine... if you say it's ok I am buying it  :-*
Second thing is what resistor I will solder on this resistor? if you have any idea or document and if you share it I will be very happy  :)
Thank you gentlemans!

Cheers,
 
I used http://uk.rs-online.com
352-187 & 352-244, but looks like the wafers are out of stock for the moment.

Second thing is what resistor I will solder on this resistor?
On the switch? 1/4w metal film are just fine.
 
peter purpose said:
Second thing is what resistor I will solder on this resistor?
On the switch? 1/4w metal film are just fine.

Thanks Peter,
I meant what value of resistors I neeed to solder? I searching and already searched on all pages but I just found this picture :-X :-[
wafermark.jpg


Thanks again!

 
jandoste said:
peter purpose said:
Second thing is what resistor I will solder on this resistor?
On the switch? 1/4w metal film are just fine.

Thanks Peter,
I meant what value of resistors I neeed to solder? I searching and already searched on all pages but I just found this picture :-X :-[
wafermark.jpg


Thanks again!

Thanks again!

Hi Peter,
I really can't read [ because of my eyes ]  the schematic :eek:
I read like this [ I may be wrong but my wife helped me :]
2= 47k    ->8k2
4= 16k    ->6k8
6= 5k1    ->5k1
8= 5k6    ->4k3
10=3k8  ->3k8
12=2k7  ->2k7
12=1k8  ->2k2
14=1k5  ->1k6
16=1k1  ->1k3
              ->1k1
am I write?
Thanks
 
jandoste said:
Hi again  ::)
I'm sorry for my stupid question but I can't find on mouser 9k1 resistor so can I use 10k instead of 9k1?
Thanks for any advance.
cheers,
http://www.mouser.com/Search/Refine.aspx?Keyword=271-9.1k-rc
 

Latest posts

Back
Top