6G3 clone issue

[Jeremy_FK]

Hi,

I've just built a 6G3 (bright channel only) in a 5E3 kit.
The amp is tremendous... big crunch sound with a LP or a Tele... I was really impressed. Early ZZ Top sounds... really.
But there is one issue... (of course !) : when the volume is beyond 4, if I start to increase the tone control above 10 (it goes to 12  ), the sound disappears, I hear a "poc" in the speaker, and a little hiss... moreover the B+ (390V) decreases about 8V and the bias (21mA) increases about 10mA 
If I lower the tone, everything becomes normal.

Do you know what is going on ? I've checked all that I can but without success...
If you need more voltage readings, please do not hesitate.

Thank you in advance !

Jérémy.

[sluckey]

Disconnect the NFB loop from the speaker jack. Does it get better?

[PRR]

> sound disappears, I hear a "poc" in the speaker, and a little hiss... etc

Supersonic oscillation. As Slucky says, try disconnecting NFB from speaker jack to driver. But that does not include the gain/tone control, so I'm dubious.

Picture of layout!! Big signals around power tube plates and OT can sneak-back to before gain/tone control and be re-re-re-amplified. Through-air sneakage is better for higher frequencies (how radio works). Fenders can amplify some above the audible band. When supersonic feedback makes the amp flip-out, yes the current can change (up or down), the flipped-out amp can't amplify audio right (or at all) but can amplify its own self-hiss to large level (superregenerative receiver). You have to see the circuit as many tiny antennas, some sending, some receiving. Sometimes moving a wire an inch makes enough difference to quell the super-squeal. (Like moving your wi-fi over 2 feet in/out of dead zone.) Sometimes a layout is just begging for trouble.

> bright channel only

Stock, there are two 220K resistors to mix the channels. Did you keep at least one of these? They also (with grid capacitance) tend to roll-off at the top of the audio band. Little things like this can turn sour if neglected.

[PRR]

On the other hand-- if it plays "early ZZ Top" you have no real complaint. Mark the 4-10 zone "Do Not Use" and ZZ the night away.

[Jeremy_FK]

thanks for your replies !

> Disconnect the NFB loop from the speaker jack. Does it get better?

I will try asap

> Picture of layout!!
here it is :
[source=http://img4.hostingpics.net/pics/770501P1020448.jpg]Source[/source]

The only difference with it : I've separated the V1 R/C cathode (now, each triode has its own R/C).

> Stock, there are two 220K resistors to mix the channels. Did you keep at least one of these? They also (with grid capacitance) tend to roll-off at the top of the audio band. Little things like this can turn sour if neglected.

Yes, I kept it.

> On the other hand-- if it plays "early ZZ Top" you have no real complaint. Mark the 4-10 zone "Do Not Use" and ZZ the night away.

 

[Jeremy_FK]

NFB loop disconnected : same issue with tone/vol pot 

[Jeremy_FK]

Here is the schematic :



I just put screen resistors (470 ohms, 5W) on both 6V6GT + bias pot (10K) with 22K resistor instead of 22K resistor on the schematic.

[tubeswell]

...: when the volume is beyond 4, if I start to increase the tone control above 10 (it goes to 12  ), the sound disappears, I hear a "poc" in the speaker, and a little hiss... moreover the B+ (390V) decreases about 8V and the bias (21mA) increases about 10mA 
If I lower the tone, everything becomes normal....

Use a V-meter and take some DC voltages on the Tone and Vol pot lugs and on the grid of V1b, when you have the tone pot at different (cut and dimed) settings, including the point at which the vol disappears.

[Jeremy_FK]

Hi tubeswell,

thanks. I will do this... and also double check all ground points.
Do you think it could be a faulty pot (tone or volume) ?

EDIT : I remember that even with amp volume on zero, the guitar is a little bit amplified... sounds strange to me !

Jérémy

[SILVERGUN]

- In the first pic, where does the circled green wire go to ground?

- In the second pic, how do you have those diodes wired?... direction of stripe and pin #s is good enough

[Jeremy_FK]

Hi Silvergun,

> - In the first pic, where does the circled green wire go to ground?

it goes on the ground bus that you can see on top of the board.

> - In the second pic, how do you have those diodes wired?... direction of stripe and pin #s is good enough

like this :

 

[SILVERGUN]

OK, all good
Good luck Jeremy...you've got great help here. Keep pushing and you'll get it

[Jeremy_FK]

OK, all good
Good luck Jeremy...you've got great help here. Keep pushing and you'll get it

thanks... I've just re-wired tone and volume pots... unfortunately, still the same issue 

[pompeiisneaks]

Unless I've missed something, I seem to see a missing anode resistor for the second channel bright, not sure if you're suing it, or not, but look here.

~Phil

[pompeiisneaks]

I'm also seeing no cathode ground resistor/capacitor for V2?  Or did you change that from a 25uF to a .1? I'm just comparing to the schematic you showed, but you may have deviated on purpose.

~Phil

[Jeremy_FK]

Hi Phil,

> Unless I've missed something, I seem to see a missing anode resistor for the second channel bright, not sure if you're suing it, or not, but look here.

actually, there is no "second channel". There is only one channel, which is the bright one from 6G3.

> I'm also seeing no cathode ground resistor/capacitor for V2?  Or did you change that from a 25uF to a .1? I'm just comparing to the schematic you showed, but you may have deviated on purpose.

V1a and V1b are the 2 gain stages. At the beginning they were sharing cathode resistor and cap (1K / 25uF). It seems that it was a bad idea (even if the sound was terrific !) so I installed a resistor+cap (1.5K / 25uF) for each cathode of V1a and V1b.
V2 is the LTP PI as shown in the schematic.

Just another precision : I have no 1.5K grid resistors on 6V6GT (like the original)... maybe it could help.

Jérémy.

[pompeiisneaks]

Adding the resistors will help with the overall life of the tubes, but I don't think it should cause the problem you're seeing. 


edit sorry life of tubes not amp.

~Phil

[tubeswell]

Hi tubeswell,

thanks. I will do this... and also double check all ground points.
Do you think it could be a faulty pot (tone or volume) ?

EDIT : I remember that even with amp volume on zero, the guitar is a little bit amplified... sounds strange to me !

Jérémy

I was just wondering whether you have DC voltage appearing on the grid of V1b with the tone pot at a certain setting. If you did, it could indicate miswiring or some other problem, but it would give a clue. Take the VDC readings from grid of V1b to ground (with test clips), and turn the vol pot as you go

[Jeremy_FK]

Hi tubeswell,

thanks. I will do this... and also double check all ground points.
Do you think it could be a faulty pot (tone or volume) ?

EDIT : I remember that even with amp volume on zero, the guitar is a little bit amplified... sounds strange to me !

Jérémy

I was just wondering whether you have DC voltage appearing on the grid of V1b with the tone pot at a certain setting. If you did, it could indicate miswiring or some other problem, but it would give a clue. Take the VDC readings from grid of V1b to ground (with test clips), and turn the vol pot as you go

tubeswell,

I did it... and that's pretty interesting !
When I take Vdc reading from V1b grid to ground (with test clips because it is too noisy), there is a big hum with an periodic oscillation. The V-meter indicates 0Vdc whatever tone and volume settings (from zero to dimed). And moreover, the issue with tone pot at 12 does not appear (well, in fact I don't hear it because of the big hum...)

Jérémy

[Jeremy_FK]

I got news :)

I was wrong... when the "thump" occurs, the sound does not disappear... it is completely reduced with nasty overdrive

Then I tried some modifications :
I removed the 220K between tone stack and the grid of V1b : the thump in the speaker is quieter but it occurs as soon as the volume pot is >2 (1 = zéro on this amp) and the tone is > 10.

I tried to put a 15K between tone stack and the grid of V1b : when the volume is low (1-2) and the tone high (>10) the thump is quite strong. When the volume is high (>4), and the tone high (>10), the thump is really weak. However in any case, as soon as you get the thump, the volume is reduced with nasty overdrive...

Finally, if the volume is full up (on 12), the thump appears with tone on 3...

A friend of mine told me that the grid of V1b takes to much gain which may cause a sort of blocking distortion... so we tried to fix it ... but no success... the thump was there... and the amp was way too loud (impossible to put volume above 2).
So I re-wired as it was... I will make a pause because I have a a show in 3 hours now... I have to rest a while ;)

See you tomorrow !

hope this helps 

Jérémy

[tubeswell]

I think I see what the problem might be.


I only see one Rk||Ck for the V1 tube. If you are trying to use a shared Rk||Ck for cascaded V1a and V1b stages, you will likely have oscillation problems. The solution is a separate Rk||Ck for each stage

[kagliostro]

Post #4

Seems that V1a and V1b has separated cathode R/C

Franco

[tubeswell]

Post #4

Seems that V1a and V1b has separated cathode R/C

Franco

Yes the schematic shows seperate Rk||Ck for V1a and V1b, but I can't see it in the picture

[Jeremy_FK]

Post #4

Seems that V1a and V1b has separated cathode R/C

Franco

Yes the schematic shows seperate Rk||Ck for V1a and V1b, but I can't see it in the picture

Yes I'm sorry, I took the picture as soon as I finished the amp, but meanwhile I separate Rk//Ck for V1a and V1b. I also removed Ck for V1b for testing purposes.

I played the amp during 4 hours yesterday for a gig... the sound was absolutely wonderful (even if the speaker is not broken in) and as I left the tone on 8, I had no issue.

But this bug teases me... 

[pompeiisneaks]

come to think of it, since you've removed the normal channel, you may want or even need a 220k resistor to ground off where the old channel comes in.  That channel had a slight path to ground in that side with the 220k to the volume pot, so it may have created a bit of a way for the signal to not come in too hot from the bright channel.  Just a wild guess.

~Phil

[Jeremy_FK]

come to think of it, since you've removed the normal channel, you may want or even need a 220k resistor to ground off where the old channel comes in.  That channel had a slight path to ground in that side with the 220k to the volume pot, so it may have created a bit of a way for the signal to not come in too hot from the bright channel.  Just a wild guess.

~Phil

That was a good idea Phil... unfortunately, it doesn't work :(
I tried with a 1M resistor from the V1b grid to the ground (knowing that I have a 15k resistor from tone stack to grid, not 220K).

Jérémy

[Jeremy_FK]

Any other idea ?

This thing drives me nuts ! 

[tubeswell]

Hi tubeswell,

thanks. I will do this... and also double check all ground points.
Do you think it could be a faulty pot (tone or volume) ?

EDIT : I remember that even with amp volume on zero, the guitar is a little bit amplified... sounds strange to me !

Jérémy

I was just wondering whether you have DC voltage appearing on the grid of V1b with the tone pot at a certain setting. If you did, it could indicate miswiring or some other problem, but it would give a clue. Take the VDC readings from grid of V1b to ground (with test clips), and turn the vol pot as you go

tubeswell,

I did it... and that's pretty interesting !
When I take Vdc reading from V1b grid to ground (with test clips because it is too noisy), there is a big hum with an periodic oscillation. The V-meter indicates 0Vdc whatever tone and volume settings (from zero to dimed). And moreover, the issue with tone pot at 12 does not appear (well, in fact I don't hear it because of the big hum...)

Jérémy

What size is the pre-amp filter cap? 10uF? Try 22uF or 47uF

[Jeremy_FK]

Hi tubeswell,

yes it is 10uF. I will try a larger one. Thanks.

Jérémy.

[Jeremy_FK]

Hi there !

I tried a larger pre-amp filter cap... nope... same issue.

Anyway ! I measured each node on the amp... I realized that something was wrong in the PI.
Every Vdc in the whole amp decrease when the thump occurs... except :
- cathodes of the PI (20Vdc => 24Vdc)
- grids of the PI (12Vdc => 15Vdc & 17Vdc => 19Vdc)
- at the center of the LTP (where the two 1M, the 6800 and the 820 ohms resistors meet) : 18V => 21V

So I tried to understand the origin of all this... and I realized that when the thump occurs, AC voltage appears from the OT, before the 56k NFB resistor (OVac => 128Vac). After the 56K, AC voltage appears but just 1Vac.

Does all this mean something for you ?

(I did not measure Vdc and Vac when I remove NFB loop but without NFB the thump occurs)

thanks

Jérémy

[Sonny ReVerb]

Sounds like it could be "blocking distortion"? (I see you did mention that in an earlier post)

http://www.aikenamps.com/index.php/what-is-blocking-distortion

I notice the 6G3 doesn't have any grid stoppers on the power tubes. Might try some 1.5k grid stoppers to see if it helps? Aiken has several suggestions. Re-check all your component values versus the schematic, of course.

OR

Found this on another forum (you did see increased voltage in the PI as mentioned here):

That LTP can put out way more signal swing than required to overdrive the power tubes.
 When overdriven, grid rectification acts to increase the effective bias voltage, which if taken to extreme can put the tube into cutoff, hence the signal seems to disappear.

 If you've a dummy load etc, you could test this theory on the bench by monitoring the bias supply voltage at 'full crank'.
 The bias voltage should be measured at the node of the 220k grid leaks and trem intensity wiper.
 You may find that the bias voltage increases significantly, eg 25%.
 The effect may be exacerbated at high settings of the trem intensity...
 
 The LTP output signal swing is best reduced by lowering the value of the power tube grid leaks, eg from 220k to 100k or less.

[Jeremy_FK]

Well, I finally found the solution ! 

I replaced the V1b grid resistor from 220K to 1M 470K : the thump totally disappear whatever the volume and the tone !

That's because I removed the normal channel... and I think the V1b grid needs to "see" a higher resistance.

Well... I'm relieved to see that everything is OK :)

Now I will tune some cathode resistors and/or caps ... and keeps the best for me.

Thank very much for your help !

Jérémy