Wrong Application for Bipolar capacitors?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

rjb5191

Well-known member
Joined
Oct 26, 2017
Messages
162
Location
Montana, USA
I'm working on "modding" channel 8 on an allen and heath Zed r16 and have ran into some issues after I made some changes.  I'm experiencing large pops when engaging the EQ switch and high pass filter (maybe some significant DC?) switch and experience some weird distortion when the fader is at near max.  I'm working out eliminating other causes but one thing I'm thinking about is that I replaced most of the polar electrolytic coupling caps with bipolar electrolytics and I'm wondering if this is a potential cause.  I'm wondering if there is something I don't know about bipolar electrolytics in an application like this and somehow they could be allowing DC to go where it shouldn't.  I've attached a schematic.
 

Attachments

  • ZED_R16_003-817_monoip.pdf
    208.9 KB · Views: 33
Not exactly. Bipolar should be better. It's probably a difference in leakage maybe. Look up the datasheets for the old and new caps and see if there's a difference in leakage. Or just skip to the part where you put drain resistors at the right places. For example, put a 47K across pins 1 and 3 of SW2. That should remove the popping of the HPF.
 
I see some floating nodes that may cause the problems.
1. HP filter add 100K to +side of C4 to ground. It is floating.
2. Add 100K from (+C9, +C11, CN3-Ring  node) to ground.
Duke
 
rjb5191 said:
I'm working on "modding" channel 8 on an allen and heath Zed r16 and have ran into some issues after I made some changes.  I'm experiencing large pops when engaging the EQ switch and high pass filter (maybe some significant DC?) switch
Yes clicks, pops, thumps are generally evidence of DC. That should be easy enough to detect with a VOM.
and experience some weird distortion when the fader is at near max.
Sometimes evidence of path saturation. An incorrect DC operating point for some interior circuit could result in premature saturation.
I'm working out eliminating other causes but one thing I'm thinking about is that I replaced most of the polar electrolytic coupling caps with bipolar electrolytics and I'm wondering if this is a potential cause.
In general that should not cause DC issues. Bipolar electrolytic capacitors are typically just two polar caps back to back, so not a game plan for ideal performance, perhaps better than polar caps exposed to a negative DC bias.
I'm wondering if there is something I don't know about bipolar electrolytics in an application like this and somehow they could be allowing DC to go where it shouldn't.  I've attached a schematic.
No bipolar caps do not pass DC, but the fact that you replaced them all suggests rechecking your work.

From a quick glance of the schematic, using TL07x op amps do not provide predictable DC offsets, while modest DC voltages are generally innocuous.

If you are suspicious of DC are any of the potentiometers scratchy when turned (evidence of DC)?

I really don't enjoy second guessing other engineer's console designs but one perhaps opportunity for an upgrade is the number of "disk" (ceramic disc?)  capacitors in EQ circuits.  NPO /COG ceramic caps are as good as film however if they are using common ceramic discs consider mylar (polyester), or better dielectric. Mid band EQ circuits are one place where inferior dielectrics can actually be audible.

JR
 
squarewave said:
Not exactly. Bipolar should be better. It's probably a difference in leakage maybe. Look up the datasheets for the old and new caps and see if there's a difference in leakage. Or just skip to the part where you put drain resistors at the right places. For example, put a 47K across pins 1 and 3 of SW2. That should remove the popping of the HPF.

Thank you.  Is there something similar to this that can be done on the fader?
 
Audio1Man said:
I see some floating nodes that may cause the problems.
1. HP filter add 100K to +side of C4 to ground. It is floating.
2. Add 100K from (+C9, +C11, CN3-Ring  node) to ground.
Duke

Thank you. It looks like +c9 and +c11 are on the same node so just one node to ground should suffice, correct? Why not for c10?
 
JohnRoberts said:
Yes clicks, pops, thumps are generally evidence of DC. That should be easy enough to detect with a VOM. Sometimes evidence of path saturation. An incorrect DC operating point for some interior circuit could result in premature saturation.In general that should not cause DC issues. Bipolar electrolytic capacitors are typically just two polar caps back to back, so not a game plan for ideal performance, perhaps better than polar caps exposed to a negative DC bias. No bipolar caps do not pass DC, but the fact that you replaced them all suggests rechecking your work.

From a quick glance of the schematic, using TL07x op amps do not provide predictable DC offsets, while modest DC voltages are generally innocuous.

Thank you.  It's too tight in there to get in with a VOM with the channel cards plugged in.  I'd need to find or build some kind of ribbon extender to test the channel card outside of the chassis.

If you are suspicious of DC are any of the potentiometers scratchy when turned (evidence of DC)? 

Just the fader, none of the pots.  Also, weirdly enough there is some motorboat oscillation that occurs on one of the digital routing switches when the fader is above 0.  Also, all of these problems seem to occur only when the fader is above 0 which makes me suspect something in the fader buffer circuit.  I replaced the 4580 with an AD8599 and added decoupling capacitors.  It's a SIP8 SMD chip and it was my first time removing one and I lifted two of the pads so I had to reconnect with point to point wires but I already measure resistances and continuity and everything looked OK but I will check again.

I really don't enjoy second guessing other engineer's console designs but one perhaps opportunity for an upgrade is the number of "disk" (ceramic disc?)  capacitors in EQ circuits.  NPO /COG ceramic caps are as good as film however if they are using common ceramic discs consider mylar (polyester), or better dielectric. Mid band EQ circuits are one place where inferior dielectrics can actually be audible.

Aren't most of the ceramic disks in the opamp feedback loops.  Its looks like most of the filter caps are actually mylar film.  This is a pretty good sounding board and I bought it cheap with issues to get some experience with working on this type of gear.

JR

Thanks again for all the help!!!
 
rjb5191 said:
Thank you. It looks like +c9 and +c11 are on the same node so just one node to ground should suffice, correct? Why not for c10?

No need as C10 has a return path through R53. +C4 has a floating path when SW2 is in #6 position and need a return path like 100k.
Duke
 
Are all the channels behaving exactly the same?

The disks(?) that caught my attention was 680pf in parallel with 220pf...

The 15pf caps are pretty high impedance down at audible frequencies so probably not an issue.

You may be able to measure some internal DC circuit nodes at insert jack...

JR
 
JohnRoberts said:
Are all the channels behaving exactly the same?

The disks(?) that caught my attention was 680pf in parallel with 220pf...

The 15pf caps are pretty high impedance down at audible frequencies so probably not an issue.

You may be able to measure some internal DC circuit nodes at insert jack...

JR

Thank you John.  I actually noticed that all channels are not behaving the same.  Channels 1-7 are working fine.  Channel 8 has the problems I described previously.  Channels 9-16 are pretty much not working at all now and I can only hear low volume distorted sound when I hook up a source to any of those inputs.  Those channels are "downstream" of channel 8 on the ribbon cable going to the summing section on the LR output card.  That was with sending those channels to the mixbus but I need to test those channels with the PFL button engaged to see if the same problem presents itself.  I'll need to check my work on channel 8 again more closely; maybe a pad lifted and is wreaking all sorts of havoc.  Unfortunately, these chinese boards are quite delicate. I haven't been able to get to this work yet, but hopefully later today or tomorrow.  I can't thank the knowledagble people here enough for there help and I'm striving to one day become as competent so I can help others as well.
 
I figured out the issue!  There was a typo in some instructions to switch the first two aux sends from pre to post fader.  There ended up being two jumpers connect to the pre and post feeds and a feedback loop was created.  Doh!  Onward with the mods!
 
rjb5191 said:
I figured out the issue!  There was a typo in some instructions to switch the first two aux sends from pre to post fader.  There ended up being two jumpers connect to the pre and post feeds and a feedback loop was created.  Doh!  Onward with the mods!
Good job...

JR
 

Latest posts

Back
Top