aleph3 built, but humming...

[cowanrg]

so, i finally have an aleph3 put together fully, but its got a hum to it. its pretty slight, but you can still hear it.

ive checked the DC offset, its at 6mv or something really low, and it plays sound just fine, and everything else appears to be working.

Magura suggested that its because i dont have a filter on my power supply. im an idiot and didnt realize you needed one. right now, i have outlet > fuse > transformer > rectifiers > caps > boards. i guess im supposed to have some sort of inductor or resistor? will this get rid of my hum?

i ask because when i was testing it outside of the case, i had no hum and it sounded fine. however, i was only using 2 caps, instead of the 8 im using in the power supply now. could it be a wiring issue?

any help anyone can offer would be greatly appreciated. Magura has helped me to better understand troubleshooting so i should be a bit better at following advice. i just want it to work so i can make the second mono block!

pictures are on my site if you want to see more of what's going on:

http://www.cowanrg.mesanetworks.net/projects/

 

[tobias_svensk]

Offtopic, just HAD to make some smileys about your 800VA trannie for your BoSoZ

 

[tobias_svensk]

Oh, and very nice projects there!

Like your volume pot.

800VA @ BoSoZ & 250VA @ Aleph 3

/Tobias

 

[cowanrg]

yeah, the 800va bosoz was a bit overkill, and the 250va aleph3 is slightly underkill, but both were easily available and the prices were right...

thanks for the comments. im learning a lot. its been a long road, but i at least feel comfortable building things now. the electronics are still a bit unknown.

anyone got ideas?

i downloaded duncan's PS designer, but i have NO idea how to use it. i dont know what the numbers mean. i can throw together what i have, and add resistors or whatever, but i dont know what the results tell me...

 

[jwb]

You have pictures of the interior but I can't tell from them how the power is connected. The star ground point should be on that huge copper bar which I assume is the ground, and the input connector should be grounded to that initially. You can try grounding the input to the chassis instead, sometimes it works out better that way.

Your assertion about that metal washer shielding the PCB is, as far as I can tell, unfounded. I think moving the transformer far, far away from the input amplifier is a good idea.

 

[Brian Donaldson]

Try moving your board a little further away from the toroid while it's on and see if it varies or goes away. I think a few traces on thew board may be acting as a partial wind on the transformer and inducing currents.
Also check to temp of the sinks with the case together. With the fins horizontal and encased, you may loose some of the majic smoke from the fets with it all closed up. But you may be ok as those sinks look like overkill for 2 fets each.

 

[cowanrg]

the metal washer isnt shielding the pcb... the pcb sits on a metal plate about 1/8th inch thick.

 

[cowanrg]

yeah, the sinks dont get all that hot, even when completely closed up. its bigger than it looks actually.

Try moving your board a little further away from the toroid while it's on and see if it varies or goes away. I think a few traces on thew board may be acting as a partial wind on the transformer and inducing currents.
Also check to temp of the sinks with the case together. With the fins horizontal and encased, you may loose some of the majic smoke from the fets with it all closed up. But you may be ok as those sinks look like overkill for 2 fets each.

 

[Brian Donaldson]

Is the plate magnetic? Aluminum won't corral the stray magnetic feilds. Also read about star ground. Does shorting the input stop the hum? if so, you may have a ground loop in the source chain.

 

[cowanrg]

hum, i didnt realize that. i thought metal stopped interference

i took some pics to give you all a better idea i hope.

so you dont think its the PS? its just bad wiring/grounding?

http://www.cowanrg.mesanetworks.net/projects/images/aleph3/DSCF0033.JPG

http://www.cowanrg.mesanetworks.net/projects/images/aleph3/DSCF0034.JPG

edit:

i ground the input by just connecting input to ground right?

 

[Brian Donaldson]

Very nice work by the way. Good star ground as far as I can see. The aluminum plate will stop RFI , but not maknetic flux. A faraday (sp) cage of magnetic metal would, but make sure it doesn't touch the bolt. That would make a shorted turn that would fry the primary over time. I think that the traces between the input pair and ccs are generating voltages from the magnetic flux. If any of the traces run perpendicular to the core, there will be a small voltage induced. and it doesn't take much to make a hum. Easy answer is to unbolt the board and move it about and see if the hum varies.


and yes, connect input to star ground or just short 2 terminals of the RCA

 

[cowanrg]

ok, sounds good. its a bit hard to move, but i could try. how far away SHOULD it need to be?

edit:

thanks for the comment btw. this is my first one thats really "together". its practice for the big ones that stare at me everytime i enter the shop.

 

[highbias]

How did you configure the AC ground? Did you use a CL-60 as the original did? Or just tie AC ground to the star point? The $2.00 CL-60 in series with the AC ground can help alot when there is a ground loop issue which this could easily be. Does the amp hum with the inputs just shorted? Also, I agree that the star center point should be on the copper strip which is power supply ground this might be another problem to rectify. I agree with Brians suggestion below. Either move the board farther away or place a piece of ferrous steel sheet metal between the board and tranny and see if the hum is reduced at all. I usually place the board at one ond of a case with the filter caps located in- between the tranny and the board. Never have had a magnetically induced hum problem building them this way....

Mark

 

[Brian Donaldson]

As much as the wires allow. Doubling the distance from the tranformer should give you a 4 fold reduction in induced hum (I think) (carefull, I'm now pulling stuff out of my ***. don't get any on you). I thought it wouldn't be hard to unbolt and lift up a inch or 2 to see if that's the culprit.

 

[cowanrg]

How did you configure the AC ground? Did you use a CL-60 as the original did? Or just tie AC ground to the star point? The CL-60's can help alot when there is a ground loop issue. Does the amp hum with the inputs shorted?

Mark

let me try to short the amp first. is CL-60 a wire type? its shielded right? none of the cable is shielded. i could do that though. i will try shorting first though.

As much as the wires allow. Doubling the distance from the tranformer should give you a 4 fold reduction in induced hum (I think) (carefull, I'm now pulling stuff out of my ***. don't get any on you). I thought it wouldn't be hard to unbolt and lift up a inch or 2 to see if that's the culprit.

ok, sounds good. i just have all my wires JUST long enough, so i dunno if i can get 2 inches out of it, but from what it sounds like, moving it at all could possibly tell me if thats what it needs or not.

so, should i assume i dont need a power supply filter, or is it just wiring then???

 

[jwb]

I see from your photos that your star ground and the ground bar of your capacitors are in different places. I would move the star ground to the screw in the middle of the copper bar.

It doesn't seem like much, but once I had a terrible hum just from having the star ground 1/2" of thick PCB trace away from the filter caps. Moving the star ground as close as possible to the filter caps made a huge difference.

 

[cowanrg]

ok, i can try that. phew, a lot of things to know/try!

 

[Brian Donaldson]

The original doesn't use a psu filter, but it can quiet noise floor. Try the shorted input first, then move around the board. just a little shift should tell you if this is the right path.

 

[highbias]

The CL-60 is a thremisistor which normally is at 10 ohms resistance. If you look at the original A-3 schematic Nelson put one in series with the AC ground path. If current flows through it the resistance decreases and provides an effective ground path, normally it looks like its out of the circuit because of its cold resistance value of 10 ohms. I put them in all the amps I build. They look like a disc capacitor but are black. Many use them in series with the primary of a toroid power tranny to limit inrush current. You might even consider doing that while you're at it. You canuse the same CL-60 in series with the AC hot lead.

Here is a link to the DigiKey page......
http://dkc3.digikey.com/PDF/T042/0935.pdf
Its part number KC006L and they're located in the lower right hand corner of that page.

Mark

 

[cowanrg]

tried the shorted input, no dice. its still got same hum. so it must be the amp itself. i will have to completely rewire to do any moving of anything, so ill devote this to tomorrow night.

i think i can move the transformer back towards the back of the amp, and the board closer to the front. that could separate them at least a couple of inches more.

about this shielding stuff... where can i get it? if i could just encase the transformer or something, that would be easiest.

of course, if its a ground thing as jwb suggested, i cant do that i guess.