Author Topic: The “VOLKSAMP”  (Read 11881 times)

Offline ghostinthemachine

Re: The “VOLKSAMP”
« Reply #45 on: June 03, 2009, 09:10:23 AM »
Thanks Ludo... I have a simple motto "Anything man-made can be fixed when it breaks!" (well, most of the time it is true  ;D)

Rudi, I came across the recommendation about the transistors but I would like to leave the amp as it is because it works simply beautiful as it is.

Thank you Hi-Ph... I guessed it was something like that. Like the "AC/DC" switch on 70's Marantz power amps....


Conquer yourself rather than the world - Rene Descartes

#77mfgco - www.77mfg.co.za

Offline ludo

Re: The “VOLKSAMP”
« Reply #46 on: June 03, 2009, 10:51:46 AM »
Check the schematic. There's high and low frequency filtering on both inputs. (Much) more filtering on the Normal input.

If it 's going vrot, put it on the buffet ! - Fats

Offline ghostinthemachine

Re: The “VOLKSAMP”
« Reply #47 on: June 22, 2009, 09:24:24 AM »
So the transformer is finally in the post on its way to me. The guy also sent me the dead 3020B's PCB, power switch and bottom cover plate.

I think when I am going to do the re-cap thing I might want to consider doing it on the "new" donor board, if its in good shape, and simply swop boards in the amp. 

Since my bottom cover plate has holes drilled through it I now can replace it with an "intact" cover.

I have a suspicion that the power switch in my 3020A is not original and thats the reason why I requested the power switch from the donor unit as well.
Conquer yourself rather than the world - Rene Descartes

#77mfgco - www.77mfg.co.za

Offline ghostinthemachine

Re: The “VOLKSAMP”
« Reply #48 on: June 28, 2009, 04:21:56 PM »
I got the "rare" part and installed it on Friday. Wow!! Its like the 3020 is now turbo-charged from to sound that I used to get with the under-rated transformer.

Its pre-amp is very, very quiet and I am now appreciating the quality of the 3020A's sound. Even the phono-stage is quite remarkable!

It makes me happy when a project turns out like this. Finding an original part (transformer) is great!

Okay, the PCB is a bit buggered because of the part that the previous "owner" sliced off. Fortunately no "absolutely vital" parts where lost.

The 3020B PCB that I got with the transformer from PE is too much corroded to be put to use in my 3020. I might rework it and use it as a headphone amp or such. A kinda skunk-joppie....
Conquer yourself rather than the world - Rene Descartes

#77mfgco - www.77mfg.co.za

Offline ghostinthemachine

Re: The “VOLKSAMP”
« Reply #49 on: July 01, 2009, 11:06:59 AM »
TIME OUT!

I am now totally confused by something a bit odd.

This morning, fiddling with the "scrap" 3020B pcb I have here I noticed that I missed something on my 3020A's board.
I had noticed that the previous owner's DIY efforts did not only remove the fuse holder section but also severed a supply line to a small network connected to the output of the pre-amp section. This supply line feeds a very odd network that in turn feed the FET's [Q509 & Q510] (I think they act as a buffer stage) on the pre-amp output.

When I restored the connection the amp behaved slightly different but not in a good way. When powering on there is now a violent "zonk-buzzzz-hummmzzzz" that stays a while until it fades away.

The oddity is that the amp functioned just fine without this connection.

See attachment. Can anybody explain what this circuit does? Ludo? Ampdog?
Conquer yourself rather than the world - Rene Descartes

#77mfgco - www.77mfg.co.za

Offline ludo

Re: The “VOLKSAMP”
« Reply #50 on: July 01, 2009, 03:43:03 PM »
Dew, that's a 3020i or 3225PE PCB you have there. The jfet is used for muting the output, the original 3020 just had a mechanical switch there. Get the manual from the page I linked to in post #20 of this thread.

If it 's going vrot, put it on the buffet ! - Fats

Offline ghostinthemachine

Re: The “VOLKSAMP”
« Reply #51 on: July 03, 2009, 10:10:42 AM »
Dew, that's a 3020i or 3225PE PCB you have there. The jfet is used for muting the output, the original 3020 just had a mechanical switch there. Get the manual from the page I linked to in post #20 of this thread.



No, no... my version "A" features this circuit and the version "B" board I have in my possession. I got that diagram from the version "B" schematic. Both identical on the two boards.

Conquer yourself rather than the world - Rene Descartes

#77mfgco - www.77mfg.co.za

Offline ludo

Re: The “VOLKSAMP”
« Reply #52 on: July 03, 2009, 08:23:26 PM »
HA! So I'm learning one thing every day. I've also found the more relevant schematics now. :-[

It's not so simple to explain what's happening here. But given a bit of time I'll compose a little diddy that will hopefully make it clear to you. The jfets mute the pre-out signal when you switch the unit on, for about 5 sec. Also when you switch off.

Meantime if you're in a hurry, a couple of possibilities: First replace all the broken tracks/wires as per the schematic, to undo the previous mods. Check if switch 1c does in fact still open, not close, when you press the power switch "ON". I think
it's located behind the power switch but part of the same assembly. Two wires go to it from main PCB?

Replace caps C531,C532, C533. I think one of those might be flaky if it's making noises. Also check diodes D505 & D506 with a multimeter for shorts/open circuits. While you're there check the resistor values, you never know.

I think the jfets are still fine.

It should work then.

I'll post a more proper explanation of what happens there a bit later. Quite a story to write down, but you might find it amusing. ;)

If it 's going vrot, put it on the buffet ! - Fats

Offline ludo

Re: The “VOLKSAMP”
« Reply #53 on: July 04, 2009, 02:32:12 AM »
Dew

I had to choose between writing this and drilling PCBs in the garage. Probably 1deg C in there at the moment, so here you go. I hope my explanation doesn't make any of the real experts' hair stand on end. It will not pass any exam, but the objective is just to explain it to you. :)

The muting must prevent nasty turn-on and turn-off thumps from the speakers. So it must mute immediately at turn-on, wait a bit for the pre-amp regulated psu and the pre-amp to settle, and then un-mute. Then at turn-off it must immediately mute again. The poweramp itself is by some magic of clever design not given to thumpin'.

So from the beginning.

When the gate of a Jfet is at the same potential as the source, the channel (Drain-Source) is conductive.

For an Njfet, when the voltage on the gate is lower (more negative) than the voltage on the Source, the channel becomes non-conductive, like with valves. (For Pjfets, when the channel is at higher voltage, more positive than the Source, the channel becomes non-conductive. There are no valves like this though.)

So in this setup we have J111 Njfets. Their Source pin voltages are at Gnd (at least at DC), because of C519/C520 on the preamp outs. The voltage on the pre-amp output line doesn't stay at ground, but will go up and down between something like +1V and -1V in use.

Now if we take the gate voltage on the Njfet and make sure it will always be at the same voltage as the Source pin, the jfet channel will be a practical short-circuit and signal will pass through it. It conducts equally well in both directions then, much like a wire. But if we drag the gate voltage down way low negative, no signal will pass through the channel, as the channel develops a huge resistance then.

So we push the gate voltage up by taking the +29V supply, dividing it by 2 with R547/R548, and applying it to the gates via a diode and 10Meg resistors. The gate of the jfet forms a diode onto the channel. It is a very tiny diode and cannot sustain much current when we drive it more positive than the source pin, but the 10Meg resistor limits the available current to something like a microamp or two, which is safe. The voltage drop across the gate-channel diode is about 0.5V then, as indicated on the schematic.

If we keep pushing a microamp in there under all signal conditions, the D-S channel stays "open" and conducts. We have about 14.5 Volts with which to push, and signal swings will not be that large on the pre-amp output, so the channel always conducts enthusiastically. The voltage on the gate stays at about 0.5V (diode forward drop) above the pre-amp output line. The microamp or so of direct current that gets added to the signal line is not audible. (Let's just leave it at that for now.)

If we replace the positive driving voltage to the gate with a negative voltage, the jfet channel shuts off. No signal gets through then. The negative voltage is generated here by a voltage doubler fed from a transformer secondary. The D505/D506 and C532/C533 combination is the voltage multiplier that gives us -62V from the transformer secondary. The electronic equivalent of climbing on top of your own shoulders works very well, also when done upside down like here... ;) A multiplier like this makes voltage higher/lower quite well, but usually limits available current a bit. You can say it has a high-ish output impedance.

In the circuit, the -62V gets blocked by a cap (C531). So it just sits there, unless switch 1c is closed. When the switch is closed it sees (via the 390 Ohm resistor) the +14V or so we have at the junction of R547/R548, and the current that flows swamps the current from the 29V side coming in via the 150k resistor (R548). So we get -55V there at the anode of D507. (Both sides that are driving that point have output impedance, but the -62V wins, you might say.) The "+20.8V" indicated there on the schematic is a mistake. There will never be +20.8V there. ???

When this point is at low voltage, nothing is keeping the jfet gate voltage equal or above the signal voltage (on the source pin) on the pre-amp output line any more. So the jfet shuts off and the signal is muted. More on this below.

Some things to note:
1) The -62V points on the schem will reach -62V in practice. So C531 will see 62V plus 14V.5 across its terminals most of the time. Replace with a 100V part?

2) The mains switches on the schem are open, with voltages nevertheless indicated on the schem. They should have been drawn closed.

3) Sw 1c is OPEN when the amp is on, not closed, as on the schematic.

4) C533 should also be uprated, as it has 62V across it. Rather go for 100V there too?

I must also admit that I have never seen a failure in this section in any of the 3020s I've fixed. Under-rated caps or not. Can I be that lucky or am I just wrong as usual?

Now you might ask why on Gods Green Earth is diode D507 in the circuit? It may be true that as long as the gate voltage on an njfet is not more than a diode drop positive than the channel, that gate doesn't draw current. But in order to get the voltage on the jfet gate negative from it's positive state, one has to push at least some electrons (current) in there, if only to charge the gate capacitance negative. So if the diode blocks the electrons that have to flow to the gate, why should it be there? The circuit will work just as well without it surely? Not so. There is a limit on how negative you can make the gate voltage on a jfet. For J111 the gate must not be more than 35V below the source pin, or the thing breaks.

So the idea is when you switch the amp off, and switch 1c closes, the -55V pulse that gets applied to the anode side of D507 only discharges the jfet enough (via the 10Meg resistor) to kick the muting in. The muting only has to stay active until the amplifier psu has drained in order to prevent turn-off thumps. The current required to pull the gate to a low voltage flows via the diode's internal capacitance. You can imagine it as a little cap sitting in parallel to the diode. As the reverse voltage across the diode is increased to -55V, the cap effectively gets smaller. (All diodes behave like varactors/varicaps to some extent. The symbol for a varicap has the diode in series with a cap, but that's just to confuse us a bit. :))

This is probably the reason there is a voltage multiplier in there at all. To get the voltage spike low enough (to -55V) to be able to pull the gate voltage of the jfet low, via the reversed diode's parasitic capacitance, when you switch the amp off and switch 1c closes. Also geting the gate voltage low enough quickly at turn-on, requires a lower voltage than is already available from the amplifiers power supplies.

Make sense?

« Last Edit: July 04, 2009, 02:37:02 AM by ludo »
If it 's going vrot, put it on the buffet ! - Fats

Offline ghostinthemachine

Re: The “VOLKSAMP”
« Reply #54 on: July 04, 2009, 10:15:04 AM »
Ludo!! Uhmm wow... thank you for the excellent explanation of the muting circuit. I will look into getting at least the caps replaced. To my ears it does not seem to work as it makes more noise when the circuit is employed.

I did some Googling regarding the "tur-on" noises of the 3020 and many guys reply that the amp makes a "welcome noise" (hum/buzz) when turned on.

In my honest opinion it works just fine when the circuit is not employed. The turn-on thud is not that harsh.

You drilling PCBs? I am working on something else... see attachment... (looks familiar??)
Conquer yourself rather than the world - Rene Descartes

#77mfgco - www.77mfg.co.za

Offline ludo

Re: The “VOLKSAMP”
« Reply #55 on: July 04, 2009, 09:59:43 PM »
If the caps around the multiplier have gone bad there might be enough voltage ripple to turn the muting "less on" sporadically. Which is why I suspect the caps are vrot. With 1V ripple on the drive to the gate you could get almost something on the output of the preamp, because there's current going through there. Good old hums and buzzes.

Jeez. I like it when these old things still work, but shouldn't you rather be cloning a Krell? ;D

If it 's going vrot, put it on the buffet ! - Fats

Offline Hi-Phibian

  • Commercial Member
  • Trade Count: (+48)
  • *****
  • Posts: 12,632
  • Total likes: 279
  • I really prefer email, see my banner for address..
    • Croak Audio Exploration
Re: The “VOLKSAMP”
« Reply #56 on: July 05, 2009, 10:32:12 AM »

Jeez. I like it when these old things still work, but shouldn't you rather be cloning a Krell? ;D



Would that sound better?
Proprietor of Croak Audio Exploration.
Fair, not crazy, cash paid for turntables and tonearms from Rega, Linn and Thorens.          http://www.croak.co.za

Offline ghostinthemachine

Re: The “VOLKSAMP”
« Reply #57 on: July 05, 2009, 12:35:02 PM »
Jeez. I like it when these old things still work, but shouldn't you rather be cloning a Krell? ;D

Cloning? Who said anything about cloning...? I am merely attempting to "fix" my 3020's mangled PCB...  :)  ;)

Krell? To hell with Krell...  8)  ;)
Conquer yourself rather than the world - Rene Descartes

#77mfgco - www.77mfg.co.za

Offline templetune

Re: The “VOLKSAMP”
« Reply #58 on: May 11, 2011, 07:13:53 AM »
I experienced repeated failure of the muting circuit - the output stayed muted because the voltages were way out of spec - after doing a complete electrolytic capacitor replacement of my 3020A. The cause turned out to be dust that had built up around switch SW-1c. It seems a little unlikely, but presumably the dust caused some leakage across the switch contacts, enough to throw the circuit out.

Offline Uriah

Re: The “VOLKSAMP”
« Reply #59 on: May 11, 2011, 08:45:49 AM »
Hi guys,
I have a nad 3020B amp and it also has the funny thump / buzz about 5secs after swithing on,i have owned this amp from new and i think it has always had this noise at least i am sure it has had it for the last ten years as i use this amp almost every day
regards
Trev