Help with tube preamp build :)

[SantaClaw]

Hello, I've been assembling a dyi kit from Ebay https://www.ebay.com/itm/112685340207 Circut board is branded DIY More.

I thought I'd get a few upgrades in the thing, so I ordered Rubicon caps for it, and replaced the 35v 470µf and 16v 470µf caps with Rubicons of the same value.

I also bought upgraded tubes, as the tubes that came with the thing had corrosion on the pins, and I refuse to believe they where matched. Oh, and I changed the Blue LED's for Orange ones.

https://www.ebay.com/itm/112546651965



After initally trying to power it on using the Bench Powersupply I have, and only getting shitty sound from 1 channel, I noticed it was rated for 12v AC ... Doh.....

So I jumped back on ebay and got one of these: https://www.ebay.com/itm/152906128518

But...

I only have sound on 1 channel, I've tried swapping tubes, and inputs/outputs, one of the outputs is dead all 5 tubes work in both sockets.

I've measure the three points marked on the bottom side of the board:

14.4v DC at the 12.6v test point (16v 470µf cap)  and on the + and - 28v test points I am measuring 37.2v DC On the center caps, I was measuring up to 39v! On a 35v rated cap
The tubes have normal glow, and look to be working.



And yes, the filament on top is glowing, not just the LED on the underside.. Difficult to capture on photo, but the color seems to match nicely




So, any idea what is wrong?

[schmitt trigger]

The saving grace of a stereo amplifier with at least one working channel, is that you can proceed checking the DC bias points in the functional channel, write it down, and compare that value to the non-functional channel.

Yes, it is time consuming. But back in the day, I fixed several stereo amps that way.
Do you have a schematic? Most likely not, but the circuit is not that complex anyways.

[SantaClaw]

I guess that's my only option at this point.

[ebastler]

Agree; comparing the signals in the two stereo channels is probably the way to go. A couple of things you could check first; maybe you are lucky and one of these will provide a shortcut:

• Double and triple-check that you have populated all components correctly.
• Specifically: It's hard to see on the somewhat dark photo of the top side, but are R10, R11, R16, R21 the correct values?
• Check the stereo pot with an Ohm-meter.
• Check the tube sockets for continuity on all pins.

Good luck, hope you find the sore spot quickly!

[SantaClaw]

I am 100% confident that the resistor values are correct, I measured them before installing them to make sure they matched the silk screen, even changed out a few that had the wrong values... I'm thinking maybe the DC incident could have damaged some components though..

[Hextejas]

There are 2 points on the back that I would hit again with the solder. I had a board once that had a solder connection like the larger hole connection and Re soldering did the trick. Let me know how it sounds.

[floobydust]

This may be the schematic for your 6J1 tube pre-amp.

Check the +/-28VDC rails, sometimes these kits have fake parts like transistors with no markings, reversed pinouts, bad semiconductor die etc.

If you have good rails, measuring the tube voltages will show the trouble.

edit: is R21 100k ohms? It should match R16.

edit2: Fever/Aoshike 6J1 schematic has an error, resistors R16/R21 100k connect to audio ground, not -28VDC.
Also confusing is the 6J1 is a pentode but sch shows it as a triode. So the screen-grid is pin 6 (connected to plate pin 5 for triode-mode)

[SantaClaw]

It's a china tube chart tube, Russian made Tube 6J1P and it's a completely different tube to the 6J1. The replacement tube I am using is the 6J1P-EV <-- last two just means that it's got gold plated pins and glass dome from what I gather.

All the transistors where clearly marked on the components themselves.

There are 4 transistors, 2x are B647 AC 7H2 and 2x D667 AC 7H2

I will try to re-solder all the pins again.

[ebastler]

You are a bit too quick in answering "no, that can't be it" to all suggestions, for my taste. Have you actually re-checked R16 and R21 (and R10 and R11, for good measure)? That's what "double and triple checking" means -- we are all 100% confident that we have done it right first time, of course. 

Also, I am not sure I understand your comment regarding the schematic which floobydust found. It seems to be the right one; all components I spot-checked do match. Did your comment imply that you think  the schematic is not relevant?

[SantaClaw]

I am bringing out my magnifying thingy and my soldering station, will check everything suggested, ASAP.



Silk Screen


R10 - 2.193 K-ohm



R16 - 101.8 K-ohm



R21 - 101 K-ohm



R11 2.2 K-ohm




Continuity on all the tube socket pins.

[iMo]

Looking at pictures will not help you much in electronics, even with tube technology.
You have to measure all components and voltages/currents to move forward.
Switch the power source off, discharge all the caps and double check (by a measurement) all resistor's values are ok (write down their values).
Also check the 1uf coupling caps (if you do not have a Cmeter handy do check at least their resistance is >10M).

When done, switch the power on, wait 30secs, and do measure all the DC levels (no input signal) at all possible places.
Report the DC values (write them down into the above schematics picture).

[SantaClaw]

You are a bit too quick in answering "no, that can't be it" to all suggestions, for my taste. Have you actually re-checked R16 and R21 (and R10 and R11, for good measure)? That's what "double and triple checking" means -- we are all 100% confident that we have done it right first time, of course. 

Also, I am not sure I understand your comment regarding the schematic which floobydust found. It seems to be the right one; all components I spot-checked do match. Did your comment imply that you think  the schematic is not relevant?

I don't mean to be arrogant, I am a dumb, it's not something I can control myself...

I have re-checked the suggested resistors, and I will start checking the rest now.

 I'm not exactly experienced by any means, and now that I look at the schematic, and disregard the tube name, it does look to be this circuit indeed.

[ebastler]

Thanks for the evidence, SantaClaw -- that should stand up in court! 

Not much luck with the shortcuts, it seems. Hence, the somewhat tedious comparison of bias voltages and signal amplitudes between the two branches of the stereo circuits will probably be required... I'm sure you will find the culprit!

[SantaClaw]

Component values with caps drained, in circuit:

D1 - 0.566v
D2 - 0.563v
D3 - 0.560v
D4 - 0.561v
D5 - 0.557v

R1 - 2.2 ohm
R2 - 4.58 kohm
R3 - .996 kohm
R4 - 47.1 ohm
R5 - 47 ohm
R6 - 4.67 kohm
R7 - .997 kohm
R8 - 46.8 ohm
R9 - 46.7 ohm
R10 - 2.193 kohm
R11 - 2.200 kohm
R12 - 4.69 kohm
R13 - 0.460 Mohm
R14 - 199 ohm
R15 - 4.61 kohm
R16 - 101.2 kohm
R17 - 4.60 kohm
R18 - 0.470 mohm
R19 -199.2 ohm
R20 - 4.67 kohm
R21 -  100.1 kohm
C9 - 1.491 µf
C10 - 1.612 µf
C11 - 1.516 µf
C12 - 1.595 µf

I have re-soldered just about everything.

I will re-connect power and measure the voltages

+- 28v = -37.4v and +37.4v

12.6v - 14.44v
Input jack - 14.77v ac
C1 - 20.52v
C2 - 38.8v
C3 - 38.6v
C4 - 37.21v
C5 - 18.46v
C6 - 38.8v
C7 - 38.8v
C8 - 37.8v
C9 - started at 200mv and is dropping towards 0
C10 -  20.2v
C11 - 36v
C12 - 19.7v
C13 - 14.33v
C20 - 19.7v

I am seeing a problem at C9, no voltage to speak of...

[iMo]

Also double check the Volume potentiometers resistances (total - between audio-in and gnd, and - wiper at low, mid and high positions against the gnd). Mechanical parts (potentiometers, switches, sockets, contacts, relays) - are always the biggest sources of troubles in electronics.
Therefore check the contacts inside the ceramic sockets are clean and have good contacts with the tube's golden legs.

[SantaClaw]

I think there may be a bad solder connection on C9 as the traces are on top of the board, and the solder hasn't filled the pad properly



Nope. That did't do it

[iMo]

What are the voltages at R14/R19 and R13/R18?
I think the 36V at C11 is wrong.. There is no way to get 36V at the grid when the anode is 20V.
The grid voltages have to be around 0V DC, and the cathode a few volts plus.
That creates "a negative voltage at grid" against the cathode - that sets the operational point of the tube.

[SantaClaw]

What are the voltages at R14/R19 and R13/R18?
I think the 36V at C11 is wrong..

c11 is rated at 50v

R13 - 37.2v
R14 - 37.7v
R18 - 37.4v
R19 - 37.4v

Grounded at the +28v cap negative. vs the positive over 70v

not really sure what I can use as ground, there is both ac and dc going around...

[iMo]

Ok, I was wrong, the R14/R19 is connected to -28V, not gnd.

We need to measure voltages at tube's pins, volume at minimum (wiper at gnd), against -28V (GND for this measurement). I refer the schematics above.

[SantaClaw]

Ok, I was wrong, the R14/R19 is connected to -28V, not gnd.

We need to measure voltages at tube's pins, volume at minimum (wiper at gnd), against -28V (GND for this measurement).

Tubes in or out ?
See measurements below

[iMo]

DC Voltages at the black points against black GND at each tube (all pins 1,2,3,4,5,6,7)..

[BrianHG]

Terrible soldering job on one of the RCA connectors signal pin, not to mention it looks like you used the wrong soldering temperature throughout, or no flux whatsoever.

[iMo]

Tubes inserted.

Could you also check if the R16/R21 is connected to -28V (our Black GND) or the gnd (the gnd triangle) - see the schematics.
My guess is those resistors must be connected to the gnd triangle. Otherwise input/output grounds will have 28V difference.. Maybe wrong schematics.

Is the Tube 1 pin 1 -53.8V (cannot be)?

[SantaClaw]

(nearest R1)
 Tube 1
1 - -37.6v
2 - 35.7v
3 - 7.08v
4 - 14.03v
5 - 19.3v
6 - 19.2v
7 - -35.7v

Tube 2
1 - -37v
2 - -36v
3 - 0.007mv
4 - 7.08v
5 - 20v
6 - 20v
7  - -35.8v

[SantaClaw]

Terrible soldering job on one of the RCA connectors signal pin, not to mention it looks like you used the wrong soldering temperature throughout, or no flux whatsoever.

Thanx, I did use 1.2% flux core lead solder, didn't use any extra flux though prolly the wrong temp I usually keep my iron at 380c. The top of the rca inputs all look like this, and I added a little more solder to 1 pin earlier.

This is my third kit build. cut me some slack..

[iMo]

(nearest R1)
 Tube 1
1 - -37.6v
2 - 35.7v
3 - 7.08v
4 - 14.03v
5 - 19.3v
6 - 19.2v
7 - -35.7v

Tube 2
1 - -37v
2 - -36v
3 - 0.007mv
4 - 7.08v
5 - 20v
6 - 20v
7  - -35.8v

Are the above voltages with Tubes inserted and against the Black GND (bold gnd in my picture)??
Cannot be..

[SantaClaw]

Could you also check if the R16/R21 is connected to -28V (our Black GND) or the gnd (the gnd triangle) - see the schematics.

I get 37.3v between r21 and the -28v test point

Cut power and I get 20v between the two, gradually reducing.

"Is the Tube 1 pin 1 -53.8V (cannot be)?"

No I posted the correct values, measured between r21 and the various pins on tube 1 and r16 on tube 2

[iMo]

Then the schematics above is not the valid one.
We cannot get 37V between R21/R16 (cold side) and -28V because the schematics says the cold side of R21/R16 is connected to -28V.

[iMo]

Could you also check if the R16/R21 is connected to -28V (our Black GND) or the gnd (the gnd triangle) - see the schematics.
My guess is those resistors must be connected to the gnd triangle. Otherwise input/output grounds will have 28V difference.. Maybe wrong schematics.

OK you have verified my doubt
The schematics is fortunately wrong..

[SantaClaw]

Could you also check if the R16/R21 is connected to -28V (our Black GND) or the gnd (the gnd triangle) - see the schematics.
My guess is those resistors must be connected to the gnd triangle. Otherwise input/output grounds will have 28V difference.. Maybe wrong schematics.

OK you have verified my doubt
The schematics is fortunately wrong..

I wouldn't know.. lol

Well, I am off to bed, It's late and I didn't get much sleep last night, thank you for your help so far.  As I mentioned earlier, there is a difference between the 6j1 and 6j1p tubes, one is standard, the other is Chinese\Russian spec, figures if the schematics would be different too.

[iMo]

Ok, btw, so far - when the above measurements are ok then the pin2 of Tube 1 is somehow shorted to positive voltage rail (+28V), or floating.
The pin2 of Tube 1 has to be at -36V against signal ground (triangle gnd in the schematics).

>>> Plz doublecheck the voltage at the pin2 Tube1 tomorrow..

PS: otherwise it seems the both tubes are set to Ia=3.5mA (0.7V/200ohm), with Va=20.5V (37V-3.5mA*4k7) without any signal. That is measured against virtual gnd (the middle of +/- 37V).

Measured against cathodes the Va=58V.

Thus your problem could be a) crappy contacts in the audio connectors, or b) issues in the volume potentiometer (ie broken wiper).

Also make the fix in your schematics above, the R21/R16 resistor's cold side is connected to the gnd, not to the -28V rail. Therefore both audio-in and audio-out grounds are at the same potential as expected.

[SantaClaw]

One side of R16 (closest to pin 1) vs pin 2 = -35.9v

The other side  on R 16 (Closest to TR1)  = -36.7v

The audio in - output jacks have common ground

How do I Check the pot? It's an 8 pin pot with an on/off switch that cuts power to the amp.

Ohms, across what pins and see if the value changes  (I am guessing 50 ohms at one end?)

Hm.. The two leftmost pins vs the ground pane on the audio jacks, show 48 K-ohms

The two center pins differ, the bottom one shows 48 K-ohms (bottom of the board with the audio jacks pointing up) The top one starts out at 46 M-ohm and goes to open.

[ebastler]

Hm.. The two leftmost pins vs the ground pane on the audio jacks, show 48 K-ohms

The two center pins differ, the bottom one shows 48 K-ohms (bottom of the board with the audio jacks pointing up) The top one starts out at 46 M-ohm and goes to open.

Aah -- that might be an issue right there. Aren't the two inputs supposed to see 50 kOhm to GND (via the potentiometer), and the two outputs 100 kOhm to GND (via R16 or R21, respectively)?

[iMo]

Also double check the Volume potentiometers resistances (total - between audio-in and gnd, and - wiper at low, mid and high positions against the gnd)...

Since the last late night debugging exercise we call GND this triangle (see below). The schematics has an error in it (see above). And yes the input shall show 50k against GND, wiper 0-50k against GND, and the output resistance 100k against the GND.

[SantaClaw]

The output/input jack grounds share the ground with the 50k pot from what I can see ?

[iMo]

Yes.

[SantaClaw]

Well, that's the ground I was using when measuring the pot.

I guess I need a new pot ?

[iMo]

When you set the potmeter in the middle and you touch the "wiper" pin (or the input side of the coupling 1M5 capacitor) on the pcb you have to hear a noise/hum in the speaker. When you touch the audio input connector (center pin) and you hear nothing there is no contact in the pot.

What about Tube1 pin2 ??

[schmitt trigger]

Which channel is the one that is operating OK? Tube 1 or Tube 2?

Because it is very obvious the DC voltages are VERY different.

[SantaClaw]

Pulling tube 1 kills the audio slowly fades it away, like my headphone amp. (closest to R1)
Pulling tube 2 kills the audio  instantly, with an extremely faint sound on the non working channel

[SantaClaw]

What about Tube1 pin2 ??

One side of R16 (closest to pin 1) vs pin 2 = -35.9v

The other side  on R 16 (Closest to TR1) vs pin 2  = -36.7v

[floobydust]

All your voltages look reasonable to me. They do not perfectly match but 6J1 really would like >120V but here it runs at about 1/2 that, so mismatches are exaggerated.

Voltage measurements here are confusing because the AC voltage multiplier used generates +28V, -28V but the tube is referenced to 0V as the audio ground with R13/R18.
In a more expensive power supply, you would have 0V and +56V rail, instead of everything shifted by 28V.


When you pull a tube, you should lose audio immediately in that channel, while the other channel fades away as its heater cools off. This is because both heaters are connected in series so any one tube removed stops heater current flow in the other. This is also why the voltages (to GND) on heater pins 3,4 are different in your measurements.
It's easier to swap tube 1 and 2 and see if the problem follows a tube.

I suspect the potentiometer right now, it should not go open circuit.
You can short the two channels inputs together (=mono) at the pot wiper (middle pin) or at R12/R17 so see if signal cannot make it into one tube's grid.

[ebastler]

Well, that's the ground I was using when measuring the pot.
I guess I need a new pot ?

Looks like it, and it is not an implausible fault. In my experience, user errors and (electro-) mechanical faults are the most likely categories when a project does not work correctly. The incidence of user errors may be related to me , but mechanical failures being more likely than electronics issues can be generalized, I think.

Before you go hunting for a pot with the proper characteristics and pinout, you could try to solder two fixed resistors between the open wiper pin and the signal and GDN pins of the pot. (Use similar values to what the other half of the pot gives you when set to a decent amplitude.) If that makes the bad channel work, you have confirmed that the pot is at fault.

[iMo]

As discussed earlier, there are 2 possible sources of trouble:
a) crappy input audio connector (short or open) or crappy soldering
b) volume potentiometer - open or short.

I suggested a simple test - touching (and thus injecting a noise) the pcb tracks along the signal path (around input connectors and the volume pot) and listening the speakers will show immediately where the issue is located. It needs to be done, however.

PS: "touching" means you take a screwdriver (with plastic handle) or a test lead and you touches with its tip the pcb tracks or the component's leads.
Your amplifier is ON (tubes are of course in the sockets, and heaters on). Therefore be careful.
Decades back I did it with my finger 
The screwdriver (or a test lead) works as your signal generator

Again:
Set the volume potentiometer into the middle position and
Left:
1. touch the central pin of the input audio connector - what do you hear?
2. touch the central pin of the input audio connector (at the pcb side) - what do you hear?
3. touch the left (input) side of the 1.5uF input capacitor (at the pcb side) - what do you hear?
Right:
1. touch the central pin of the input audio connector - what do you hear?
2. touch the central pin of the input audio connector (at the pcb side) - what do you hear?
3. touch the left (input) side of the 1.5uF input capacitor (at the pcb side) - what do you hear?

Look into the schematics and follow the signal path. You have to hear a noise or 50/60Hz hum (or a nearby radio transmitter) in the respective speaker (left/right) when you touches the signal path. When you stop hear your signal (or the signal is weaker) along the signal path the circuitry close to that point are either short or open (it depends on the situation). Always start at the end of the signal path, proceed from output to input of your amplifier. Easy.
Some multimeters have got a signal output (ie 1kHz square wave) - you may use that as well. With 1kHz square wave you may even test radios (its harmonics reach megahertz).

[SantaClaw]

I kinda lost hope when I de-soldered the pot, and tore out all but 1 solderjoints

Oh... well... I've got new pots on order, and I guess some "I'm an idiot wires" can fix just about anything... 

[SantaClaw]

Ok, I found a bad cap too... One of the yellow ones is open... I quadruple checked...

1500uF 50v, I guess I have to look for an axial film capacitor ?


Any cue where I can find a suitable cap ?

[SantaClaw]

hrmf... I re-ordered the kit

hehe... this time I'll try to do a better soldering job, since i saw voltages over 39v on thew 35v caps, I ordered some rubycon yxa 50v 470uf caps... though i suspect I might have gotten duds ?

470uf +-20% I still expected to see them around 470 uf, but all seem to be around 404-406 uf

They won't work for this ?

[SantaClaw]

I put together the second kit I received, used 50v caps, works a treat, wonderful clean sound Awesome...

[Zero999]

Good, I'm glad you got it working, just one question:

Hello, I've been assembling a dyi kit from Ebay https://www.ebay.com/itm/112685340207 Circut board is branded DIY More.

I thought I'd get a few upgrades in the thing, so I ordered Rubicon caps for it, and replaced the 35v 470µf and 16v 470µf caps with Rubicons of the same value.

I also bought upgraded tubes, as the tubes that came with the thing had corrosion on the pins, and I refuse to believe they where matched.

Why would they need to be matched? I doubt matching is critical here as they're not used as a differential pair, but separate amplifiers: one for each channel.

[SantaClaw]

Good, I'm glad you got it working, just one question:

Hello, I've been assembling a dyi kit from Ebay https://www.ebay.com/itm/112685340207 Circut board is branded DIY More.

I thought I'd get a few upgrades in the thing, so I ordered Rubicon caps for it, and replaced the 35v 470µf and 16v 470µf caps with Rubicons of the same value.

I also bought upgraded tubes, as the tubes that came with the thing had corrosion on the pins, and I refuse to believe they where matched.

Why would they need to be matched? I doubt matching is critical here as they're not used as a differential pair, but separate amplifiers: one for each channel.

I guess you'd get different output level on each channel if the tubes where way off, but I absolutely have no clue. I am an amateur with little understanding on how electronics works, except the basics.

[Zero999]

Good, I'm glad you got it working, just one question:

Hello, I've been assembling a dyi kit from Ebay https://www.ebay.com/itm/112685340207 Circut board is branded DIY More.

I thought I'd get a few upgrades in the thing, so I ordered Rubicon caps for it, and replaced the 35v 470µf and 16v 470µf caps with Rubicons of the same value.

I also bought upgraded tubes, as the tubes that came with the thing had corrosion on the pins, and I refuse to believe they where matched.

Why would they need to be matched? I doubt matching is critical here as they're not used as a differential pair, but separate amplifiers: one for each channel.

I guess you'd get different output level on each channel if the tubes where way off, but I absolutely have no clue. I am an amateur with little understanding on how electronics works, except the basics.

The circuit should be designed so slight differences in parts, shouldn't affect the gain too much. I think the mismatch would have to be really bad, for you to notice: the difference in sensitivity between each ear, will probably exceed that of the amplifiers and if you've got this connected to a power amp, driving speakers, then it'll be even less noticeable.

Another thing you could consider is adding a balance control, to adjust for any perceived mismatch between channels.

[SantaClaw]

Good, I'm glad you got it working, just one question:

Hello, I've been assembling a dyi kit from Ebay https://www.ebay.com/itm/112685340207 Circut board is branded DIY More.

I thought I'd get a few upgrades in the thing, so I ordered Rubicon caps for it, and replaced the 35v 470µf and 16v 470µf caps with Rubicons of the same value.

I also bought upgraded tubes, as the tubes that came with the thing had corrosion on the pins, and I refuse to believe they where matched.

Why would they need to be matched? I doubt matching is critical here as they're not used as a differential pair, but separate amplifiers: one for each channel.

I guess you'd get different output level on each channel if the tubes where way off, but I absolutely have no clue. I am an amateur with little understanding on how electronics works, except the basics.

The circuit should be designed so slight differences in parts, shouldn't affect the gain too much. I think the mismatch would have to be really bad, for you to notice: the difference in sensitivity between each ear, will probably exceed that of the amplifiers and if you've got this connected to a power amp, driving speakers, then it'll be even less noticeable.

Another thing you could consider is adding a balance control, to adjust for any perceived mismatch between channels.

You should have seen the tubes... hehe, bent pins is one thing, corrosion is another, but the internals of the tubes looks bent out of shape on two of the 4 I now own.  Balance control is absolutely an option.

[vk6zgo]

Everything about this circuit gives me a headache!

(1)When tube preamps were the default, heater voltages were normally ac --- it would seem that the rage for DC heaters appearing in "modern" tube circuitry is because all the tubes available are either factory seconds, or "pulls" removed from equipment due to high heater/ cathode leakage.

(2)I fail to see why they don't return the "-28v" supply to signal common/ground, & simply "float" the junction of C4 & C8, so they have a +56v supply w.r.t ground.
Or, use another stage of voltage multiplication on the positive side.

By the way, there is no voltage regulation on the supply, hence 36 v instead of the rated 28v.

(3) What are the LEDs supposed to do?

[SantaClaw]

Everything about this circuit gives me a headache!

(1)When tube preamps were the default, heater voltages were normally ac --- it would seem that the rage for DC heaters appearing in "modern" tube circuitry is because all the tubes available are either factory seconds, or "pulls" removed from equipment due to high heater/ cathode leakage.

(2)I fail to see why they don't return the "-28v" supply to signal common/ground, & simply "float" the junction of C4 & C8, so they have a +56v supply w.r.t ground.
Or, use another stage of voltage multiplication on the positive side.

By the way, there is no voltage regulation on the supply, hence 36 v instead of the rated 28v.

(3) What are the LEDs supposed to do?

I think something might have blown in the first board when I accidentally connected 12v DC to it.

3 - They are there as an "on" light, they shine through the tube sockets, originally they where blue, but I've used amber/orange ones.