Low voltage oscillation ? what freq. ? Remove the input wire , do you still get it ?
OS
Still there with no input wire. It's only there when I attach my signal generator ground but the protection circuit it tripping as soon as I touch an input.
Attachments
Last edited:
Just run a normal music signal to it , pause the music- what do you see ??
what freq. is it ?? 25ms ... looks like 50/60hz - ground loop/hum ??
OS
Last edited:
It's around 200kHz. It's not there with music but as soon as I connect music from my PC I'm getting terrible 60Hz. I've definitely got some grounding issues.
Is everything plugged into the same circuit? I have two separate circuits in my workshop. I have to be careful that one piece of equipment isn't plugged into one circuit and another plugged into the other. The grounds and neutrals are all common but the hot can come opposite sides of the panel.
Last edited:
200K is too low for typical oscillation. OPS / IPS oscillations are local FB loops
oscillations are Mhz with modern devices.
The phase margin at 200K is massive and that freq. is in the passband of
the amp regardless ??
I suppose it's time to do some longer term listening to my Kyptons !
OS
The PC is on a separate circuit from the amp. I'm on three phase there but that shouldn't make any difference for 120VAC. I'll try a different grounding scheme in the morning and see if that helps.
A nice photo of your present grounding scheme would be good.
Of the supersym , badger , and this generation amp .... the slew
was the easiest to get REAL quiet (perfect grounding).
I had this problem , too. The 200k might be the PC's smps !!
I now have a radio shack loop isolator and the problem is gone.
PC is now on a different house circuit. The better long term solution
would be a optical Toslink "pocket DAC" or .... we can just design one
into a nice pre-amp project !
I have the Curl JC2 line stage ported into BJT right now 😀.
Edit - no isolator = PC/amp on the same power strip.
OS
There wasn't any sign of oscillation or hum before I assembled the chassis. I've changed the bias transistors to MJE340s which seemed to have made bias settings rock stable at every temperature I ran it through. I melted a couple .1uF film caps (cosmetic) while changing connector pins on the input board for a different mount so I changed them to ceramic X7R. Speaker returns, output board grounds and G2 grounds are all connected to a star ground. The star ground connects to earth ground through a monolithic bridge paralleled with a 10R resistor and a .1uF cap. I think AndrewT would approve of my hook up other than the crappy signal wire. It seems to me like the hook up is where the problem is originating though.
It was a strange wave form. Did you notice the step in the positive swing? It was even more pronounced at times.
It'll be hard to see anything in a picture. The power wires and speaker returns are all twisted together and put into wire loom and it's all in the bottom of the amp. I'll snap a picture when I get back at it.
For input wire, would balanced microphone wire work? I'm striking out trying to find any decent shielded RCA.
I'm in for a DAC/Preamp project. Remote control from an Android smart phone would be a welcome addition.
Do you think we could be creating a ground loop through the protection board? The 12V regulator is chassis grounded which would destroy the lifted ground on the input. Maybe this is the cause of my false triggering on my thermal sensing too?
That's a possibility.
I have my 7812/05 on sil-pads with shoulder washer.
my thermal uses the control board ground , the rail sense is an isolated
ground "sense".
OS
That makes sense now that I think about it. When the control board is on the bench I have rubber feet under it so wires don't pull it off the bench so it wouldn't be grounded through my bench.
Everything is on a common ground on the control board. I think I'll split the audio sense ground off in the next rev of the board. The Slewmaster is the only amp with a lifted ground input that I've connected it to.
Everything is on a common ground on the control board. I think I'll split the audio sense ground off in the next rev of the board. The Slewmaster is the only amp with a lifted ground input that I've connected it to.
My control board ground sense references the main cap bank's 0V terminal.
I thought the only purpose for the "rails" was as a reference for DC offset ?
OS
The rails are just for the DC sense. The circuit also uses ground to control the TIP120 for the speaker relays. It also sends a ground signal to the Arduino to tell it there's no DC present. I think I'd need to add optos to the base of one TIP120 and the Arduino to separate this ground from the control ground.
We're you having issues using the ground points provided on the boards, as originally designed?
I had no issues at all until I put it into the chassis. I remembered you looking at my grounded regulators.
Last edited:
There are two things I see different than mine.
My speaker output and speaker return are a separate wire run with the
speaker ground back to the main chassis star.
Your amp output seems to go in with the rail feeds ?
My MJE 340 goes right to the extrusion itself , away from the outputs.
I just want to measure the extrusion C , not an individual device. At 55C ,
that whole extrusion is going to be baking 😀 .
On a major level ... you just have 1 PS ? Mine is dual bridge , with
4 separate rails. I don't know if that is reason for my "110%" , but
it was the most logical way for me to proceed. Parasound (john Curl)
HCAxxxx amps do this exact thing. Super quiet , as well.
Most of your general layout is similar to mine.
OS
My speaker output and speaker return are a separate wire run with the
speaker ground back to the main chassis star.
Your amp output seems to go in with the rail feeds ?
My MJE 340 goes right to the extrusion itself , away from the outputs.
I just want to measure the extrusion C , not an individual device. At 55C ,
that whole extrusion is going to be baking 😀 .
On a major level ... you just have 1 PS ? Mine is dual bridge , with
4 separate rails. I don't know if that is reason for my "110%" , but
it was the most logical way for me to proceed. Parasound (john Curl)
HCAxxxx amps do this exact thing. Super quiet , as well.
Most of your general layout is similar to mine.
OS