Plexi 50 troubles

[uki]

Back to this amp.

After little use of this amp, maybe played it 4-6 hours tops total since it was done, something went wrong and amp is like failing, sound volume goes down and back up...
Then I decided to look at the power tubes while in operation, V4 is like flashing or sparking inside... 

Probably related to the screen voltage 460 volts, that I couldn't find a way to reduce. So since the OT have UL taps and it have reduced voltage, it could be a solution for the problem. Haven't read the UL taps voltages yet.

Is it a good idea, does the UL taps need a resistor ?

Related topic:
https://el34world.com/Forum/index.php?topic=23039.msg246988#msg246988

[shooter]

wait for a consensus, but I don't believe UL taps are meant to tame grid voltage.  The primary reason I've found, they keep the tube in a happier place, more linear? in relation to the signal, not DC idle characteristics.

you should be able to add a tap for screens, but you might have to re-balance the downstream ones
 

[PRR]

460V on EL34 screens in UL is not your problem. (When the tube conducts max, G2 dips to 300V.)

Bad tube. Bad joint. Failing cap or resistor.

[uki]

460V on EL34 screens in UL is not your problem. (When the tube conducts max, G2 dips to 300V.)


Bad tube. Bad joint. Failing cap or resistor.

My concern is the data sheet says 425 max, Its an EH tube.

Sorry sir I think I wasn't very precise there, the screens are in the B+ bus. I thought moving it to the UL could be a solution.
What caps or resistors should I look for? Its the Hoffman Plexi 50.

[jjasilli]

UL taps need no caps or resistors.  It's a direct connection UL tap > screen grid. 

[shooter]

My concern is the data sheet says 425 max

the UL tap is essentially another "choke" in-line with G2, chokes (inductors) are used to "stabilize" current, not really used for voltage dropping.  you can see this by measuring volts DC on either side of your PS choke, compare it to the voltage drop across a PS resistor.

If you really want to drop the volts, as a test, disconnect the choke in the PS and replace with a 5k -10k, 5-10W R.  NOTE: everything to the right will also drop and may need to be tweaked back up.

I agree with PRR, the G2 volts is probably  not your dead tubes issue

[uki]

My concern is the data sheet says 425 max

the UL tap is essentially another "choke" in-line with G2, chokes (inductors) are used to "stabilize" current, not really used for voltage dropping.  you can see this by measuring volts DC on either side of your PS choke, compare it to the voltage drop across a PS resistor.

If you really want to drop the volts, as a test, disconnect the choke in the PS and replace with a 5k -10k, 5-10W R.  NOTE: everything to the right will also drop and may need to be tweaked back up.

I agree with PRR, the G2 volts is probably  not your dead tubes issue

I see, if the voltage isn't the issue then:

1st thought, maybe I just got a bad new tube, the other PA tube looks fine. Swap the tubes for checking good or bad idea?
Took the amp out of the cab and examined the components for visual damage, all looks as before.
I did measured all the resistors and all have the expected reading.
Can't get a reading on the caps, is it because they are connected?
Read all voltages?

[shooter]

Can't get a reading on the caps,

with amp off for long time, put a gator clip from ground to + side of cap, IT WILL SPARK, SOOO be quick, be confident 
now, while shorted, gator clip your meter, OHMS, black - ground, red - cap +
Now, disconnect the shorting gator, your meter should run up pretty fast to at least >100k ohms, hopefully infinity - open.

if so;
 

Read all voltages

preferably with tubes in, just grab BIAS, plate b+, and the last tap.  compare to old readings then decide what's next

[uki]

with amp off for long time, put a gator clip from ground to + side of cap, IT WILL SPARK, SOOO be quick, be confident 
now, while shorted, gator clip your meter, OHMS, black - ground, red - cap +
Now, disconnect the shorting gator, your meter should run up pretty fast to at least >100k ohms, hopefully infinity - open.

if so;
 

Read all voltages

preferably with tubes in, just grab BIAS, plate b+, and the last tap.  compare to old readings then decide what's next

The filter caps already had zero charge, zero voltage.

The caps across the circuit read open, except one in the preamp connected to the mid pot , the electrolitic caps does read some resistance, i did also read in farads for the filter caps, the caps still connected, 3 cap cans:

cap 1  -  A 157.9  both positive pins connected
cap 2  -  B 126.2 ???
cap 2  -  C 55.5
cap 3  -  D 60.7
cap 3  -  E 55.9

Haven't read the voltages yet.

[uki]

preferably with tubes in, just grab BIAS, plate b+, and the last tap.  compare to old readings then decide what's next

V4
p3 464v
p4 459v
p5 -44v
p6 -44v
cathode resistor  20.8 mA

V5
p1 464v
p2 460v
p3 -43
p4 -43
cathode resistor 18.3 mA

It is about the same as last readings

[sluckey]

Brrrr. That's biased very cold.

[uki]

Brrrr. That's biased very cold.

I did try while ago to bias it hotter but its when one of the PA tubes start to red plate, and come to this state. Its biased at 20mA as PRR suggested.

What I got is a bad tube yes? Is the circuit ok ?

[shooter]

What I got is a bad tube yes? Is the circuit ok ?

without changing bias for now;
does it play without a tube arcing?
if you can play for an hour, bump up the bias a tad, if a tube gets ugly swap it, does the problem "follow" the tube?

[uki]

What I got is a bad tube yes? Is the circuit ok ?

without changing bias for now;
does it play without a tube arcing?
if you can play for an hour, bump up the bias a tad, if a tube gets ugly swap it, does the problem "follow" the tube?

The tube that red plated is in V3 position now, it was in V4 position before, when it did red plate.

Yesterday the amp did play while it doing the arcing as you said, but the volume sound goes down and up.
About 40-50% less volume when it drop.

Just about 10 mins ago: ...
I'm puzzled now, I did as you said, "play for an hour" , about that, the problem didn't happened   

Man this thing plays loud rock'n'roll sound like !!
By the way the OT gets somewhat warm, it is possible to keep touching it. Temp is about 40C or over.

So bad solder joint?

[shooter]

Man this thing plays loud rock'n'roll sound like !!

that's what we LIKE 

did you verify that the 1.5k grid R's were ok?
If so, you're probably ok
I'd have a spare tube or 2 just in case it is a bad one

a quick check if it's on the bench, pull the 2 PA tubes, unplug speaker, ohm the outside 2 primary OT wires to the middle wire, they should be reasonably close, wiggle the wires while the meters hooked up just to "see" if something is loose.
I like to put one probe in the socket where the tube goes, that way you catch all the connections, then clip the other probe at the choke that feeds the CT OT

[uki]

did you verify that the 1.5k grid R's were ok?
If so, you're probably ok
I'd have a spare tube or 2 just in case it is a bad one

a quick check if it's on the bench, pull the 2 PA tubes, unplug speaker, ohm the outside 2 primary OT wires to the middle wire, they should be reasonably close, wiggle the wires while the meters hooked up just to "see" if something is loose.
I like to put one probe in the socket where the tube goes, that way you catch all the connections, then clip the other probe at the choke that feeds the CT OT

Did play another 5-10 mins all good.

The resistors read good, they are 2.2k 5W, you mean screen resistors yes?

Been thinking about a couple spare tubes... that money thou is for the KT88 amp.

The amp is in the bench still, did the readings , no PA tubes, no speaker.
Choke read 106.8 ohms

OT does read 101.6 ohms(V4) in one side and 131.9 ohms (V5)in the other.   
Reading across the choke, 207 ohms in one side and 237 ohms in the other side
30ohms off, that may be the cause of so much distortion? Also the cause of red plate?

... In a perfect OT each half of the primary winding would be exactly the same. But because of the way OTs are usually wound, one side will have a little more resistance than the other side. 150Ω and 172Ω are close enough, ...

I did use a cheap meter at that time.

[DummyLoad]

if you have an arcing tube, throw it away before you damage the OT. The resistance readings are not exactly what you'd want, but i've measured OT primaries that worked fine with a similar difference. That said, I have also had a couple of amps with similar differences of the secondary DCR incite red plating of the tube connected to the lower R side of the winding with matched tubes. if the red plating follows the tube, no matter the socket, then the tube may be gassy (bad) or grossly miss-matched.


--pete

[uki]

if you have an arcing tube, throw it away before you damage the OT. The resistance readings are not exactly what you'd want, but i've measured OT primaries that worked fine with a similar difference. That said, I have also had a couple of amps with similar differences of the secondary DCR incite red plating of the tube connected to the lower R side of the winding with matched tubes. if the red plating follows the tube, no matter the socket, then the tube may be gassy (bad) or grossly miss-matched.


--pete

Is there any way to balance the OT ?  Bias balance does help with it ?

Amp is on, did play a bit, no bad behavior so far after about 40 minutes.

Got a clip for you guys !

[shooter]

nice sound!

I been thinking that for a small imbalance you might??? be able to use a 10ohm under 1 tube, and a 1 ohm under the other kinda thing, but I'm not good enough to actually know before hand 

I'd take DL's suggestion, new tube, but I would label and keep the suspect for experimenting only.

[uki]

I'd take DL's suggestion, new tube, but I would label and keep the suspect for experimenting only.

I did let the amp on yesterday for about 6 hours played about 3, no problems happened...

nice sound!

Thanks !! 

I been thinking that for a small imbalance you might??? be able to use a 10ohm under 1 tube, and a 1 ohm under the other kinda thing,

Lets say that is possible, how to do that, resistor in series with OT tap to tube pin ? or resistor to ground?

 

[shooter]

Lets say that is possible

paraphrased; "ALL things are possible, but not ALL things are beneficial" 

I have done it to self biased PP, never fixed bias, that said, IF I was to try it would be with the 1 ohm R's used to monitor current, and I probably would chicken out anything > 20 ohms.

the "better" way would be bias the "hot" tube to your happy place and play

[uki]

I want to add a balance bias pot to the amp for better control over the PA tubes, would this work ?

[sluckey]

no

[uki]

no

Is it totally wrong or is it missing something?

[sluckey]

Totally wrong. It will do nothing. But you have fixed bias amps. Try it and see for yourself. Meanwhile, think about it and I bet you will figure out why. 

[uki]

Totally wrong. It will do nothing.

Argh       

But you have fixed bias amps. Try it and see for yourself. Meanwhile, think about it and I bet you will figure out why. 

There is no ground, that is one thing. I don't fully understand it.

Bias adjust +  bias balance is the goal, I it found at Robinette page.
It is going to be trick to adapt it.

How about the UL taps ?

[sluckey]

Your circuit simply changes the resistance value of the two 220K grid resistors. For example, with a 100K pot you could adjust the value of one resistor to be 320K and the other would be 220K. You could even make those two resistors have exactly the same value (approx. 270K each). But you cannot change the bias voltage that appears on the grid of the output tubes because there is no current flow through your pot or those 220Ks and with no current flow you do not drop/change any voltage. Bias remains the same.

Rob's circuit works but requires that special pot with a tap. The simplest solution I've ever seen is to simply have two separate Fender style bias adjust pots. For example see page 2...

     http://sluckeyamps.com/sunn/sunn_sceptre_1971.pdf

[uki]

Your circuit simply changes the resistance value of the two 220K grid resistors. For example, with a 100K pot you could adjust the value of one resistor to be 320K and the other would be 220K. You could even make those two resistors have exactly the same value (approx. 270K each). But you cannot change the bias voltage that appears on the grid of the output tubes because there is no current flow through your pot or those 220Ks and with no current flow you do not drop/change any voltage. Bias remains the same.

I see, it is just another resistor in combination with the 220k ones.

Rob's circuit works but requires that special pot with a tap. The simplest solution I've ever seen is to simply have two separate Fender style bias adjust pots. For example see page 2...

     http://sluckeyamps.com/sunn/sunn_sceptre_1971.pdf

Ooops that was the wrong picture.

How about this one ?

[uki]

Got schematic for the above.

Have to figure out how to implement in the circuit, not much space there, maybe a small board just bellow the two 220k resistors.

Would this work ?

[st]

Imho, you're overthinking/overdoing the balanced bias thing. A slight mismatch, caused by either tubes or output transformer won't (necessarily) hurt your tone. Bias the hotter tube correctly, as someone already said, and leave the colder one, well uh, cold. Better still, with your transformer, to use a slightly mismatched pair: colder tube on the hotter side and vice versa. This way you don't have to get finnicky with the bias balance trimpot nor pay for platinum premium tube matching, ha!

[uki]

Imho, you're overthinking/overdoing the balanced bias thing. A slight mismatch, caused by either tubes or output transformer won't (necessarily) hurt your tone. Bias the hotter tube correctly, as someone already said, and leave the colder one, well uh, cold. Better still, with your transformer, to use a slightly mismatched pair: colder tube on the hotter side and vice versa. This way you don't have to get finnicky with the bias balance trimpot nor pay for platinum premium tube matching, ha!

That is a good idea !!

Played yesterday with pedals and one or maybe both PA tubes are like flashing inside, and there a behavior like something is falling, the volume goes down and up ... a day before did play with no pedals, just the guitar in the amp, for about 3 hours no problem, the amp stayed on for about 6 hours under observation and the problem didn't happen.

I'm not worried about the tone it does sound good, althought break up too soon for my style of playing,
but the amp isn't working properly.  There is definitely, something wrong with the amp.

When biasing the hot tube hotter, it does red plate.

I think balancing looks to be one thing that will help to stabilize the amp?  I did stop writing and decided to look at the amp again...

10 mins later:
I just read again the OT primary, there is something wrong with it, at first one side wasn't giving any readings,   

then I shake the wires , fiddled, wiggled ... got reading

now it read

131 ohms in one side(V4) and 126 ohms in the other side(V5),  loose wire inside the OT?   bad joint ?

Then after re-soldering  the OT primaries

112 ohms in one side(V4) and 127 ohms in the other side(V5)

Keep changing?

 

[st]

There is definitely, something wrong with the amp.

I think balancing looks to be one thing that will help to stabilize the amp? 

I seriously doubt bias balance has anything to do with your problem.
Bad connections could be the source of the problem. Especially check anything in the bias supply up until the grids if the power tubes.
I also wouldn't discard some kind of oscillation as a possibility.

[shooter]

+1

Then after re-soldering  the OT primaries

the arc'd tube should be replaced now that your wiring is fixed, then just run it and watch.

[uki]

+1

Then after re-soldering  the OT primaries

the arc'd tube should be replaced now that your wiring is fixed, then just run it and watch.

First thing in the morning, did read again the OT

124.8 ohms V3     126.4 ohms   v5       why does it keep changing ?

Re-soldering all the bias joints as well.

The feedback loop resistor read 4.7k while connected to the OT , disconnected read 47k, is it expected?

One of the LTPI resistor is one 1k, but 470 ohms in the schematic, replace it ?

Going to get new tubes today(unmatched pair), And the parts for the bias balance, why? Cos the matched tubes are too expensive.

Is that bias balance circuit correct?

[sluckey]

> 124.8 ohms V3     126.4 ohms   v5       why does it keep changing ?
Not much of a change. Could just be the way the probes contact the leads. Make your probes needle sharp. Now always try to pierce the solder joint  and you will get more consistant readings.

> The feedback loop resistor read 4.7k while connected to the OT , disconnected read 47k, is it expected?
yes

> One of the LTPI resistor is one 1k, but 470 ohms in the schematic, replace it ?
yes. Why was it ever 1K???

[Tony Bones]

Copper has a pretty large temperature coefficient. 130 ohms while hot and 125 when cold is completely reasonable. It would only require about 20F increase.

[uki]

> 124.8 ohms V3     126.4 ohms   v5       why does it keep changing ?
Not much of a change. Could just be the way the probes contact the leads. Make your probes needle sharp. Now always try to pierce the solder joint  and you will get more consistant readings.

Today it read V3 103.3 ,  V4 126.4

V3 tap (blue tap of the OT) have a initial peak reading of about 130ish then rapidly drop to 100ish.

> One of the LTPI resistor is one 1k, but 470 ohms in the schematic, replace it ?
yes. Why was it ever 1K???

I dunno !! 
Why Fender uses in that part of the circuit 6.8k and Marshall 10k ?


Got two new Mullard tubes(unmatched), gonna do the bias/balance circuit .

Is the bias/balance schematic I draw correct? Remove the 47k (R35) resistor from original bias circuit?

[sluckey]

Is the bias/balance schematic I draw correct? Remove the 47k (R35) resistor from original bias circuit?

Sometimes you just have to build it and see if it does what you want.

[uki]

Is the bias/balance schematic I draw correct? Remove the 47k (R35) resistor from original bias circuit?

Sometimes you just have to build it and see if it does what you want.

The arrangement for the bias on the following schematic looks about the same yes? A bias adjustment pot and bias balance pot!

[uki]

Your circuit simply changes the resistance value of the two 220K grid resistors. For example, with a 100K pot you could adjust the value of one resistor to be 320K and the other would be 220K. You could even make those two resistors have exactly the same value (approx. 270K each). But you cannot change the bias voltage that appears on the grid of the output tubes because there is no current flow through your pot or those 220Ks and with no current flow you do not drop/change any voltage. Bias remains the same.

Based on the circuit above got another idea, check it out!

Does it work ?     

[sluckey]

It will only take a few minutes to temporarily wire that mod to the board. You don't need any tubes plugged in to test it. No animals will be harmed during this test. Just do it and tell us what you think.

[uki]

It will only take a few minutes to temporarily wire that mod to the board. You don't need any tubes plugged in to test it. No animals will be harmed during this test. Just do it and tell us what you think.

What about the trimpot and resistors value? I got 20k trimpot, what value for the resistors?

[sluckey]

I don't know. Just experiment.

I don't like a bias adjust and bias balance system because the two pots are interactive. Tweak one and you'll need to re-tweak the other. May have to do this several times.  I've already shown you the simplest circuit to do this. And a bonus is the pots are not interactive. Set one to the bias point you want. Then set the other to the bias point you want. Done.

[uki]

I don't know. Just experiment.

I don't like a bias adjust and bias balance system because the two pots are interactive. Tweak one and you'll need to re-tweak the other. May have to do this several times.  I've already shown you the simplest circuit to do this. And a bonus is the pots are not interactive. Set one to the bias point you want. Then set the other to the bias point you want. Done.

Oh I see !  I thought it would be easier to fit in the board, but as I found out it is harder to tweak ! Now the way it is in the Sunn Spectre is much easier to tweak, but probably harder to fit in the board. Gonna study that !

It does work    !!   But the voltage increased a lot, to less than -10v   

I did put a jumper in the 47k resistor connected to the 50k trimpot, no changes hopping the voltage to go down... 

How to get the voltage down back to the range for bias ?

[sluckey]

Try 100Ks for those two resistors.

[uki]

Try 100Ks for those two resistors.

100k resistors did drop some voltage but just about -10v.

Then with 330k resistors, the bias adjust trimpot does -35 to -21
and the balance pot -36 to -34.


With 430k resistors bias voltage got close to -40 to -24, with the balance pot in one end, with the balance in the middle, bias voltage is just bellow -40, about -39ish.

Balance pot gives about 2 volts range,

Looks like still need about -10 volts to get in the desired range

How's that sounds like so far?

[sluckey]

Looks like you're heading down a deep rabbit hole. Don't talk to the smoking caterpillar.   

You're way too high with the resistor values. Go back to 100K and play with R36 value. Decrease value to increase voltage.

But before you do anything else, put the balance pot at center position. Set the bias voltage to some convenient level on pin 5 of one of the tubes. Not looking for any specific voltage. Now connect one meter probe to pin 5 of one tube and connect the other meter probe to pin 5 of the other tube. Ideally you want zero volts on the meter with the balance pot set to center. Now rotate the pot to one end and record that voltage. Then rotate the balance pot to the opposite end and record the voltage. One end should be a positive voltage and the other end should be a negative voltage. The wider this range of voltage, the more likely you will be able to balance tubes. But if you only swing from +1v to -1v, you will only be able to balance tubes that are already very closely matched.

[uki]

play with R36 value. Decrease value to increase voltage.

I got to buy some resistors, none to replace R36(220k 3W) at the moment with 3W or with smaller values like 200k 180k or so, I got tiny ones with 1/4W and 1/8W, maybe some with 1/2W gotta dig the stash.

Looks like you're heading down a deep rabbit hole. Don't talk to the smoking caterpillar.   

You're way too high with the resistor values.  Go back to 100K...

430k in the last test
What happen to the circuit with those high value resistors ?
 

[sluckey]

I think your balance voltage swing will just get smaller and smaller, making the balance pot less and less useful.

Do the test I suggested. What have you?

[uki]

I think your balance voltage swing will just get smaller and smaller, making the balance pot less and less useful.

Do the test I suggested. What have you?

I notice decreasing the balance resistors also decrease the bias range.

I'm still digging resistors...

In the Williamson, balance pot is 10k and bias 25k.   20k and 50k in place atm.
I thought the pot balance could be smaller as well, to increase negative voltage?