My Gaiclone "Pops" :(

[xcortes]

After months of reading the threads in this forum and others and ordering BrianGT's boards and components (Thanks Brian!) I finished my first GC a couple of days ago. It uses a 225VA toroid (24V dual secondaries) and a 20k series attenuator (made with a shorting 11 position switch from electroswitch and Vishay Dale resistors- at around $30 its a great bang for the buck if you can leave with 6dB leaps in volume level!-).

On the first trials I heard a couple of "pops" coming out of the speakers. Not annoying but clearly there. The best way to describe them is like the pops that you get from a damaged vinyl record. Then nothing.

This morning the amp was on with no signal (the cd I played last night had finished hours before!) and I heard the pops again. Interestingly I noticed they happened when my wife turned on or off the bathroom lights. These lights are powered by small transformers. I tried repeating the experience and voilá, each time you turn the lights on or off the pop appears. I tried with different lights and it happens with those that are powered via transformers, apparently not with those powered directly from the ac lines (but have to verify this).

My only idea is that when the lights are turned on they draw so much current that momentarily "turn the amp off". I thus tried to replicate the sound by turning the amp on and off rapidly and it happens (I get some pops although not sure they're the same).

Supposedly the LM3875 does not produce on-off transients. Could this be oscillation? Could this be related with very small input caps that really doesn't store much energy and from a large energy draw from the lights? Could this be only some kind of noise in the lines induced by the lights?

Doesn't seem like they will damage the speakers or something (I use full range speakers so no tweaters!).

I will keep experimenting but any ideas will be mostly appreciated as I'm by no means an expert.

Thanks,

xavier

[jeff mai]

Originally posted by xcortes
This morning the amp was on with no signal (the cd I played last night had finished hours before!) and I heard the pops again. Interestingly I noticed they happened when my wife turned on or off the bathroom lights.

Mine does the same thing. Except it does it when virtually any thing in the house is switched on or off and possibly also the neighbors house.

Not particularly loud pops, but annoying anyway. My valve amp doesn't let any of this through.

After receiving advice and trying a few things, I believe the reason is a lack of PSU filtering. The valve amp has much more series impedance in the supply (valve rectifier and choke) which is probably more effective at squashing spikes on the AC line.

You can try an encapsulated mains filter (the sort combined with an IEC socket.) I'd suggest getting a fairly good one rated for lots of current - 5 amps. This still may not work.

Other tries might be regulated supplies, a battery power supply, or a choke input supply. There is a way to suppress the offending spark at it's source which may be an option if only the bathroom light causes the problem, but you should have an electrician do this work.

[xcortes]

did you try different grounding configurations? in mine all ground connections (including the attenuator ones) go to the same point. i'm thinking of connecting the psu and mains wire ground to one point and the signal grounds to another.

[jackinnj]

High gain, high current -- susceptibility to all sorts of nasties.

1) Check to make sure that all of the units are grounded commonly.

2) You might want to put an LC filter on the mains. A bifiliar wound choke on each line, a pair of 10nF caps (could be 100nF-- horses for courses) rated for the appropriate line voltage, etc.

3) Not entirely germaine -- but the reservoir caps should have 100nF bypass caps mounted as closely as possible to the chip.

[netgeek]

Instead of trying to "roll your own" (and playing with mains voltages while doing it) - look for filtered power input modules (e.g. "Corcom") and perhaps supplement these with commonly available "surge protectors" (offered for use with typical PC's, monitors, etc.).

Winding your own inductors and then tacking on capacitors to create home made "filters" and other experiments is just asking for trouble if you're not VERY familiar with the techniques, the voltages, and the risks... I'd advise that you look for things which are commercially available and experiment with different combinations of those.

Just my $0.02..........

[peranders]

Your problem may come from the output also. If you don't have any output filter, the noise may be picked up by the speaker cables (long good antennas!)

How about your protective earth and signal ground?

[jackinnj]

Originally posted by netgeek

Winding your own inductors and then tacking on capacitors to create home made "filters" and other experiments is just asking for trouble if you're not VERY familiar with the techniques, the voltages, and the risks... I'd advise that you look for things which are commercially available and experiment with different combinations of those.

Just my $0.02..........

naah, in this case it's pretty simple and the technique is described in virtually every ham radio handbook (and these are available at libraries everywhere in the civilised world) ... of course in the case of ham radio the problem is one of noise migrating FROM the equipment to the power line.

the other problem which I neglected to mention could be poor relay contactors in household appliances -- could be a design flaw or just dirt and old age -- the EU has very tough EMI standards but if an applicance is made in China, anything can happen. if there is a bum contact the electrical noise will propagate for several hundred feet from the source.

[roibm]

mine does the same too....
any switch on/off, pop or white noise if the switch is not really good(like the heater in one of my aquariums).
I knew there must be others having the same problem so I waited. Looks like I was right.

note: I have an EMI filter on the mains(the plug-in on the amp is a computer plug I got from an old unused PSU). I also have two ferrites on the mains cable.

please can someone describe the proper way of grounding? like grounding everything(inputs, amp boards, psu, mains) in star?

[xcortes]

It also happens with my recently acquired phono stage (Lehmann black cube)!!!!!!!!!!!!!!

The good news is that the problem doesn't seem to be directly related with my gainclone

The bad news is that I now have to solve the problem for two systems!!

The appartment is not old but I've noticed many low quality construction and finishing materials in it.

Does anyone think that a filter such as the popular Jon Risch one will eliminate the problem?

What about the Auricap tweak installing two .47 uF 600V Auricaps in parallel with the red lead to the hot screw and the black lead to the neutral screw?

In the meantime I'll do two things:

1. Try a PC surge suppressor (I know It will negatively affect sound but it's a free experiment).
2. Measure if there are any AC voltage variations (peaks?) correlated with the pops.

Thanks all

[jeff mai]

I have no pops when I replace the transformer and rectifiers with a battery power supply - none. Therefore I strongly suspect that the problem is coming directly from the mains.

That Xavier's problem occurs in two systems with differing components, also seems to indicate a direct mains issue. I don't think improved grounding, etc. will improve the situation, not meaning to discount its importance, of course.

Note that while the PSSR of LM3875 is quite good at 120Hz, it is only 50-60db at 10kHz. Of course, at this frequency the PSU caps should be more effective. Is this true? Should the caps absorb a spike of this nature?