Tube supply question

Started by rankot, November 15, 2017, 02:48:06 AM

Previous topic - Next topic

Rob Strand

QuoteWires on the left are original ones - red is positive rail, black is negative rail and blue is ground. This is the star point of the amp. Wires on the right are mine (OK, I know I connected colours incorrectly, but they go to right places on my PCB), and they go to the connector on PCB. Yellow wire goes from this star point to red marked ground point on preamp PCB.

OK, it looks like you are mimicking what was there before.

What about how the preamp connected to the power amp? How is the grounding done?  Try to copy that as well.
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

antonis

To get a little further, I tend not to trust ground lifted voltage regulators..
(although it worked well on your prototype breadboard..)
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

rankot

OK, I will use LM317 to precisely set 6.3V :)
  • SUPPORTER
60 pedals and counting!

antonis

Only after verifing no grounding issues...  :icon_wink:
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

merlinb

Looks to me like hum picked up on the cable because the circuit is unshielded and your volume pot is waaay too high resistance. Use something in the range of 5k to 20k (and use a correspondingly larger output cap, like 10uF or something).

rankot

I'll try that too.

I have tried to disconnect heater from main amp's PS and to connect it to it's own PS. Now rail balance is almost perfect, but I still have the hum (and it changes with volume pot, with maximum on max, lowest hum at 10-12 o'clock).

I will try to connect to ground at different place and see what happens. Old preamp was connected at different point, so I will try to connect there and see if there is change. I will also try to use shielded wire for signal.

:icon_frown:
  • SUPPORTER
60 pedals and counting!

rankot

#26
I have also tried to remove IC1 and see what happens if I connect C6 to amp. It works of course, but there is the same hum. And I have also noticed that if I connect amp's input to input jack (thus bypassing preamp at all), there is almost the same hum, but not so strong (because it is not amplified by preamp). Less hum if I don't plug guitar in. So it must be picking up some noise from input jack to preamp input, I will try to use shielded cable here.

I tried to remove tube itself and disconnect heater PS. And the hum is not present when I connect premap out to amp in, but if I connect input jack to amp in, bypassing preamp completely, the same old hum returns.

:icon_mad:
  • SUPPORTER
60 pedals and counting!

Rob Strand

#27
QuoteI have tried to disconnect heater from main amp's PS and to connect it to it's own PS. Now rail balance is almost perfect, but I still have the hum (and it changes with volume pot, with maximum on max, lowest hum at 10-12 o'clock).
Situations like this there is usually more than one problem and they cancel out.  This is why I think you should focus on the hum when the pot is at zero.  This case depends on ground issues (and power routing) and not the circuit. It doesn't matter what you do to the circuit that one won't go away until it is fixed.  And if you don't fix it first you can arrive at false conclusions (due to complicated interactions.)

QuoteI tried to remove tube itself and disconnect heater PS. And the hum is not present when I connect premap out to amp in,
Can you run all wires for the heaters back to the +rail and the ground point at the caps? 

Quotebut if I connect input jack to amp in, bypassing preamp completely, the same old hum returns.
OK this is a very common issue.  I'm assuming the input sock connects to chassis ground?   What other ground connections are there to the chassis?  If you have more than one it is much more difficult to arrive at a solution.
[edit: what about the speaker output jacks?]

Another issue, not related to the grounding, I noticed your tubes a fed from the 50V rail.   You might need to add an RC filter to remove ripple from the 50V rail.   Maybe 1k and 100uF to start.






Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

rankot

Quote from: Rob Strand on November 17, 2017, 04:26:54 PM
Another issue, not related to the grounding, I noticed your tubes a fed from the 50V rail.   You might need to add an RC filter to remove ripple from the 50V rail.   Maybe 1k and 100uF to start.
Oh, I thought it is already filtered enough through main amps PS, so I will try this too. But first things first - to solve grounding problem. :(
  • SUPPORTER
60 pedals and counting!

rankot

I am trying to figure out differences between old SS preamp and my new hybrid, regarding connections.

There are two differences:

  • Original one's input jack is connected to chassis, while my new preamp used jack just hanging outside, because I am still experimenting with it.
  • Original is grounded through the point where input jacks connect to PCB, while my amp is grounded to the star point between main PS capacitors.

Original (not exactly original, because that's another construction by myself, only SS) preamp gets noisy (but very little) only when volume is at 70% or more.

So I will try to ground this new one exactly as the old one was grounded, and report what happens.
  • SUPPORTER
60 pedals and counting!

Phoenix

Quote from: rankot on November 18, 2017, 05:17:09 AM
I am trying to figure out differences between old SS preamp and my new hybrid, regarding connections.

There are two differences:

  • Original one's input jack is connected to chassis, while my new preamp used jack just hanging outside, because I am still experimenting with it.
  • Original is grounded through the point where input jacks connect to PCB, while my amp is grounded to the star point between main PS capacitors.
Original (not exactly original, because that's another construction by myself, only SS) preamp gets noisy (but very little) only when volume is at 70% or more.

So I will try to ground this new one exactly as the old one was grounded, and report what happens.
That could definitely be part of the poblem. Have a read through Merlin's article on grounding.

Rob Strand

Grounded input sockets can be a pain.

Maybe like this,



You put a break between the low level signals and high current signals.
This way you have two star points with a bridge between them.  One for quiet stuff and one for noisy.
The wire bridge must not carry significant current or ripple.  The bridge wire can be the chassis.
Generally you don't want power supplied across the bridge at all but in some cases you can have
the preamp power running across it.   (Also it's not good if the low signal side looses is 0V when if you disconnect the input socket from the chassis.  As a precaution, you can put a 10ohm resistor between the power supply 0V and the the preamp 0V.   The resistor doesn't do much in normal operation.)



Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

rankot

#32
I have tried two more things:

1. connect my preamp to ground at input jack only (grounding scheme exactly like the old SS preamp)
2. connect above + ground from triple power connector on my preamp's PCB, to star point on amp's PS, thus definitely creating a ground loop

Both behave completely the same. :(

I will try to change volume pot right now, as Merlin suggested.
  • SUPPORTER
60 pedals and counting!

rankot

Quote from: Rob Strand on November 17, 2017, 04:26:54 PM
Can you run all wires for the heaters back to the +rail and the ground point at the caps? 

It was connected that way for the first time I connected preamp, but the hum was the same as with later connection schemes. If I understood you well - ground for preamp connected directly to star point between filter caps on amp PS, and power wires for preamp connected directly to +/- rails at filter caps? But I later decided to separate heater supply, because it dropped rail voltage too much. However, that change didn't affect hum at all.
  • SUPPORTER
60 pedals and counting!

rankot

#34
There's a schematic in PDF: http://www.cad.rs/d/ECH83v1.pdf
and PNG:


I have changed output volume pot to A50k (didn't have smaller A type at the moment) and capacitor before it to 10uF and it helped to fix one thing - hum is not maximizing when volume pot is at 0, but gradually grows as I turn it to maximum. Maybe I could leave C18 to 100n, cause it forms HP filter with knee at 31Hz with 50k pot, but never mind, I can restore it later, just to get rid of this hum.

I have also added one 470 ohm resistor before filter cap C8, but it didn't seem to alter the hum a lot, then I removed ground path from star point to power connector (red gnd) on preamp and it decreased hum a lot, but not completely.

I will now try to add another 470 ohm resistor between caps C8 and C9, thus forming another ripple filter, and I will also try to completely isolate input jack and connect ground to star, to see what's happening in that case.

Lots of things to do :)
  • SUPPORTER
60 pedals and counting!

Rob Strand

#35
I have a feeling about this one.
Can I suggest a simple test. 

You have two wires, or a coax between the input of the power amp and the output of the preamp - yes?
Without having any preamp present, but with those two wires left in circuit,  short those two wires together.
Do you hear a hum?

If yes that means the power-amp input has the hum and you need to find a new ground point which has the least hum.

Look at this circuit,
http://sound.whsites.net/p3a-f1.gif

Often the best ground point is to connect near the ground side of C3.
[Edit:  Just to be clear, some circuits will have C3 and R4 reversed. In that case you want R4's connection to ground.]

Many power amplifiers will keep the wires for C3, C2, R2 on their own PCB track and the output from the preamp (ie. the power amp input) connects to there.   Some power amplifiers will even put a 10 ohm resistor from that point to the 0V in order to force good grounding and avoid ground loops.


Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

rankot

#36
Quote from: Rob Strand on November 18, 2017, 03:50:04 PM
I have a feeling about this one.
Can I suggest a simple test. 

You have two wires, or a coax between the input of the power amp and the output of the preamp - yes?
Without having any preamp present, but with those two wires left in circuit,  short those two wires together.
Do you hear a hum?
No. I've removed new preamp, and when I did what you suggested (short amp input and gnd), amp was absolutely silent. There was some hum when those wires (ground and amp input) were NOT shorted. I will remove old preamp completely and see what happens then.

I have also noticed that chassis is connected to circuit ground by input jacks, and there is no chassis-to-earth direct connection, see image below. So I will first add earth-chassis connection and replace all the jacks with plastic ones, as described in Merlin's script.



Original preamp is powered form + rail only, and it is connected to the rail at the very beginning of it. I will try to fix all the grounding, but if I don't succeed with eliminating hum with preamp, I will build separate PS for pre only, with additional transformer.
  • SUPPORTER
60 pedals and counting!

rankot

And one more question: if I add hum-loop block network, as proposed in Merlin's script, may I use 18 ohm 5W resistor instead of 10 ohm, because I have some of those in stock?
  • SUPPORTER
60 pedals and counting!

Rob Strand

#38
QuoteAnd one more question: if I add hum-loop block network, as proposed in Merlin's script, may I use 18 ohm 5W resistor instead of 10 ohm, because I have some of those in stock?
The resistor value isn't that important.

However,  that circuit doesn't apply directly to you case.  It is used when all the *sockets are isolated* from ground.  Then you bridge *circuit ground* and mains ground using that  network.  It prevents ground loops when you connect devices to the amplifier.   If you have non-isolated sockets that connect the ground to the chassis then that circuit won't work.

Look at figure 4 on this page, the circuit doesn't connect to the chassis and connector (in fact all connector) is isolated from the chassis:
http://sound.whsites.net/earthing.htm

You could change all your sockets to isolated ones, remove all circuit ground connections from the chassis, then add one of those networks Merlin proposed between the circuit ground and the mains ground.   I would say this configurations is *the* most successful configuration in removing all types of grounding issues.   In fact when I was young I experienced many issues like yours on systems with grounded sockets.   I changed to using the isolated sockets and those ground networks and never had any problems (that was for guitar and HIFI systems.)


Edit: I just realized you are in the process of changing to isolated sockets.  However, this is still a bit weird to me:

The thing I'm puzzled about is, you don't get any hum with the "shorted wire" experiment but you do get hum when you connect the power amp input to the preamp PCB and turn the output volume knob to 0.   What other connections are there to the PCB when you connect the preamp?  because one of those connections is the cause!

Something else I'm not 100% clear about:  For the input socket are you using separate input sockets hanging off your preamp  PCB?  Are they mounted to the chassis?  Or, are you wiring the preamp input to the existing sockets mounted to the chassis.


Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

thermionix

I haven't read through the thread, so forgive me if this is totally irrelavent, but did the solid state amp work well (hum free) before adding the tube preamp?  Those filter caps look original and decades old.