Intermittent popping and crackle in LM3875

[jcon2]

Perhaps someone might assist me in troubleshooting this odd problem. Recently boxed up Peter Daniel's premium 3875 kit in a dual mono arrangement. Used a an old BackUPS Chassis, 160 VA toroids from Group buy with Uriah, silver plated wire, Cardas jacks, etc.
Rectifier output for each channel is 35-0-35 Vdc.
DC offset with input shorted was 17mV for Right and 11mV for left.
Will play music and sounds quite good.
Initially I had some RF that is very faint and improved with better shielded interconnects.
Inputs and outputs are isolated from chassis as using acrylic back panel.
Had AC hum that went away with using Peter Daniel Star Ground Ala Patek
in Commercial Gainclone post #77.
Problem is I have a audible popping/ crackle every 3-6 sec on both channels.
Doesn't really vary with volume of ipod or iphone.
So as it is in both channels thought it might be on ac input.
Eliminated switch and was still there.
Tried different ac outlet and still there. Tried 2 different inputs and still there.
I doubt cold solder joints as is new circuit board, Hakko iron, wonder solder and visual inspection after deflux looks good.
I did add R1 Caddock 220 ohm due to initial hum issue.
Removed led and no change.
So any ideas what could be problem?
Only thing I can think of now is Rectifiers and Caps.
It just seems odd that it is in both channels with dual mono. Fear that I am missing something obvious.
Thanks
Jack C

[Atilla]

Some more elimination could be useful. Does it happen if you only use 1 channel and disconnect the second one entirely?

If you have another supply availavle, even something with lower voltage/current specs, try using that with 1 channel and see if it's the PSU causing that. Bad caps or potentiometer is a good guess.

Oh, 1 more thing. Sometimes you can get pops and clicks when certain equipment turns off, like lights with dimmer switches, refrigerators, whatnot. Maybe it is your AC, but something else is causing that. If you try it in another location and it doesn't happen, then your build is OK.

[jcon2]

Attilla,
Thanks for the thoughts.
Have tried the following.
Uncased it. And started with one channel at a time.
Hooked up the board with a Plitron Transformer and a snubberized rectifier from a BrianGt 3886 build. Had an extra as didn't go dual mono on that one.
DC voltage was 24V-0-24. and DC offset was 11mv.
Has significant hum with no input connection only an interconnect to RCA jack without ipod attached.
With Ipod attached will again play music just fine. With ipod muted still get intermittent low level audible pop and or crackle every 3-6 sec.
As the ipod is muted there is not really a pot to act up. Also does the same behavior with iphone.
Also am testing through a variac so should have some ac filtering.
I am thinking either something is amiss with the 3875 board Or my AC is bad in my room. No other electronic equipment it exhibiting this behavior though.
Head phone amp, Cheap receiver etc.
There just aren't that many parts to trouble shoot.
And a few shots of test setup. Thanks
Jack C

[gootee]

If there are no RF low-pass filters on the inputs, then that is most-likely the problem. Even the pops and cracks from things like switches and appliances are often or usually due to high frequencies (i.e. "RF"). That's because something that is narrow in the time domain is wide in the frequency domain. e.g. Switching that occurs over a very short time interval will generate a very broad frequency spectrum, reaching well up into the RF frequencies. And the P-N semiconductor junctions in all diodes, transistors, and IC chips will rectify the RF, causing "unpredictable" things to happen inside IC chips, often generating pops and cracks at their outputs when RF bursts go into their inputs (or power pins, or outputs; For RF, everything is an input!).

I am not familiar with that particular amp's design. But if I recall correctly Peter is of the "minimalist" religion and might not have included any input RF filter.

If your amps don't have RF input filters, see my post at F5 - how to reduce bandwidth response and install a low-pass filter with a -3dB frequency of 200 kHz or higher, as close to each chip's signal input pin as is practical. If the chipamps are in an inverting configuration, it will be a little trickier. In that case, see the note near the end of that post, and also make sure that the gain won't be changed.

If you put up a link to the schematic of your amps, we could get more specific.

Alternatively, often you can just add a small capacitor directly across the + and - signal input pins of an opamp or chipamp, instead of using a regular in-line low-pass filter. A typical value would be 220 pF, usually either a film cap or an NPO or C0G ceramic.

Cheers,

Tom

[sandyK]

Have you tried without the input cables, but with someting like 1Kohm resistors soldered across the input jacks, and then listened ?
Also, how long are your speaker leads ? RF can be picked up via speaker leads and finds its way in via the feedback network. For example, it is not that uncommon for low level hum from strong local TV stations to get in this way. The hum then goes up and down in sympathy with the brightness of the scene from the TV station.The cure in that case is to loop a couple of turns of the speaker leads through a large toroid as used in big RF baluns.

SandyK

[abraxalito]

Do you have a star earth? From the look of your pics, you've got more than one earth - for example you have a PSU pcb and a power amp PCB but only one wire between them (the one going through the red croc lead?). If your star point is on the amp board, then you'll need more than one ground running from the PSU - to the caps, and to the transformer centre tap.

[jcon2]

Thanks for all you input. I finally took Attilla's and my wife's advice and moved the amp to another room in the house. Hooked it up to my Brick Wall AC line filter and pop and crackle were gone. Then plugged it in to another AC outlet in the same room and again dead quiet. Tried the other channel and the original rectifier and also quiet. Back to my work room and the original AC outlet and pops return.
I live in Seattle and generally RF is a problem.
Tom,
Thanks for the suggestions and will try that next up if the problem persists.
Will now try and solve the problem of where the clicks, pops, etc. originate in this house, or elsewhere. We have the usual assortment of dimmers, motors, and electronics.
Sandy,
In this case the speaker leads were only 3-4 feet for debugging. But you make a good point. Have used the toroid choke for interconnects in the past with some success.
Abraxalito,
I agree. The last photo was a bit of funky set up in an attempt to trouble shoot. The star arrangement is clearer on the 2nd photo of the original post.
So it looks like I have a solution/ cause. Thanks for all the thoughtful comments.
Jack C.

[Peter Daniel]

I would be giving the same advice as Tom: Commercial Gainclone kit- building instructions

BTW, really nicely done although it looks like you waited quite a while before boxing that kit; I switched to a blue solder mask at least 2 years ago

[Atilla]

If mains interference is causing that, then you could try a simple line filter as well. It should be able to do the exact same thing.

[AndrewT]

if you do/can identify the offending switch/appliance that is interfering with your music system, it is always better to attenuate the interference at the source.

There are interference suppressors sold for exactly that purpose. They attenuate the spark across the opening contacts of a switch.

But, in any case fit RF filters to the inputs of all your audio equipment.

I set my Power Amps filters to 680us.
If I can, I set any source equipment input filtering to half this ~330us.