Plexi 50 (done)

[uki]

sometimes it is better to skip a cup of coffee and go with new, signal path caps are one of those coffee sacrifice places.

I got some 22n new caps in hand just in case ! 

Need to replace volume pots, they are like granny of a granny  Did took it apart and cleaned it inside, was dirty man!!! It is working better now but feel like the volume goes to max before 10...

The amp sounds really good although it goes on overdrive really soon at volume 2 or so, knob at like 8-9 o clock.   

Thanks @Tubeswell for the file good reading !!

I did the readings and the results are:

C11 cap have 24.2n
 dc mV is jumping between .01 and 11.x every few seconds, 10-20 secs interval or so.  The negative symbol keep showing up and disappearing while reading.   

C12 cap have 24.4n
 dc mV is also jumping just like C11

How much the dc voltage on a cap have to be to consider it a leak?

Thanks !

[sluckey]

How much the dc voltage on a cap have to be to consider it a leak?

I would not worry about it unless it is more than 500mV

[uki]

How much the dc voltage on a cap have to be to consider it a leak?

I would not worry about it unless it is more than 500mV

Thanks Sluckey, so those caps are good then! Where else to look at ?

[tubeswell]

I did the readings and the results are:

C11 cap ... dc mV is jumping between .01 and 11.x every few seconds, 10-20 secs interval or so.  The negative symbol keep showing up and disappearing while reading.   

C12 cap ... dc mV is also jumping just like C11

How much the dc voltage on a cap have to be to consider it a leak?

That will be random tiny amounts of signal nose


What about C8 and C9? Did you measure those for leakage?

[uki]

What about C8 and C9? Did you measure those for leakage?

If I did this right, lift cap side that is connected to the pins 2 and 7 yes?
Got those readings and looks much better than the other two, it stays at .001 - .002  on both caps with the meter at 200mV.
The cap values read good as well !
Hmm where to look at next? 

[shooter]

where to look at next? 

Is it a chicken or egg?, is your distortion at 2;
overdriven signal or
bad (wrong part, etc.)
a scope comes in handy here.  short of that, put a 12AU7 in V1.

what are the 2 gain knobs on when vol @2 overdrives?

[uki]

where to look at next? 

Is it a chicken or egg?, is your distortion at 2;
overdriven signal or
bad (wrong part, etc.)
a scope comes in handy here.  short of that, put a 12AU7 in V1.

what are the 2 gain knobs on when vol @2 overdrives?

Amp is going on overdrive with volume at 2-3, I think those pots for volume are part of the problem, they are super old, and faulty, I got the impression the volume get to max before turning the pots to max. I did tried the bias adjust to fix the distortion that is when the red plate happened, I tried a lower gain tube(dunno which model it is, no label) but I only got one and its already in the other amp, not much difference thou.

Look the pictures of the knobs:

Is this amp supposed to overdriven like this ?

[shooter]

sorry, I added a mv on mine, got confused 

scratchy pot can also be dc through you caps so if you change a pot check the voltage where the couplers connect to the top of the pots

[tubeswell]

What about C8 and C9? Did you measure those for leakage?

If I did this right, lift cap side that is connected to the pins 2 and 7 yes?

No, for those LTP grid caps, the grids at at a DC voltage (because of the current through the tail resistor).

You need to test the other end of those caps.


But before you do, What are the idle voltages on the LTP plates?

[shooter]

What are the idle voltages on the LTP plates?

got 'em in reply 47, I'm not a PI guy, but they seem a smig low 

[uki]

What about C8 and C9? Did you measure those for leakage?

If I did this right, lift cap side that is connected to the pins 2 and 7 yes?

No, for those LTP grid caps, the grids at at a DC voltage (because of the current through the tail resistor).

You need to test the other end of those caps.

But before you do, What are the idle voltages on the LTP plates?

I did the plate readings and today it is a bit higher than the other day

P1  217 (214)
P6  206 (203)

Is this the expected voltages? Shooter says it is a bit low.

[tubeswell]

plate readings and today ... the other day

P1  217 (214)
P6  206 (203)

That amount of difference isn't unusual for a LTP, which suggests the grid voltages are probably okay.


What type of tube are you using for the LTP? a 12AX7? or a 12AT7?


And how big is the bias resistor? 470R? 820R? 1k? 1k2?

[uki]

What type of tube are you using for the LTP? a 12AX7? or a 12AT7?

And how big is the bias resistor? 470R? 820R? 1k? 1k2?

12ax7 JJ in the LTP

and the bias resistor is 15k I think, the amp is the Hoffman Plexi 50.

The readings for those caps are:
C8 22.1nF   (200range)mV 0.0 - 0.1
C9 96.9nF   (200range)mV 0.0 - 0.1

What next ?!

Thanks !

[tubeswell]

I'm talking about the LTP bias resistor.

[uki]

I'm talking about the LTP bias resistor.

I've found this!

...You have a largish tail resistance (39k) because KF wanted to balance the plate resistors. And the HT voltage for the LTP stage is only 270V. Both factors reduce available plate-to-cathode voltage and bias voltage. Compare that to the LTP for (say) a Fender Brown Deluxe HT = 325, Tail = 6k8, much cleaner

Same situation yes? Just like in my amp!

Oh , the one connected to the cathode? 470 Ohms 1watt, and the tail resistor is 10k.

I did found out that R31 and R32 are a bit off of each other. R31 read 1215 ohms and R32 read 1025, gonna replace R31 with a closer one to the other. Could this be part of the red plate problem when changing bias?

[tubeswell]

470R is on the hot side for a 12AX7 LTP. You want more like 820R to 1k. If the LTP bias is too hot, if can muck up the LTP duty cycle, resulting in red-plating on one side of the output stage under large signal conditions

[uki]

470R is on the hot side for a 12AX7 LTP. You want more like 820R to 1k. If the LTP bias is too hot, if can muck up the LTP duty cycle, resulting in red-plating on one side of the output stage under large signal conditions

The amp still going on OD at low volume, although it sounds really loud!! The tone is nice and the amp is very responsive, but I would like some more clean tone at low volume, and get that nice OD when volume is about 7-8.

The PA tubes are swapped to see if the difference in mV get fixed, it still about the same. Does that means the tubes are slightly unbalanced?
I did replace the LTP bias resistor with a 997Ohms, I did notice a voltage change across several points in the amp after replacing the LTP bias resistor, and  I got readings for everything I could think of:

B+ nodes
A  461
B 460
C 326
D 279
E 262

V1 P1 150
     P2    0
     P3    1
     P6 201
     P7    0
     P8    1

V2 P1 161
     P2    0
     P3    1
     P6 279
     P7 161
     P8 162

V3 P1 245
     P2   17
     P3   28
     P6 242
     P7   18
     P8   28

V4 P1 across 1ohm resistor 17.5 mV
     P3 458
     P4 456
     P5 -43
     P6 -43

V5 P1 across 1ohm resistor 20.4 mV
     P3 458
     P4 456
     P5 -43
     P6 -43

Thanks !!

[shooter]

what happens when you 0 one pot, does the other have more range, try that with both

tack a 1Meg R to one pot, outer pins, leave the other pot 0 what happens

anyone you know, work with have a scope
troubleshooting the signal path without knowing if it's a gain thing or a circuit flaw is a big crap-shoot that usually leads to self installed problems

[tubeswell]

V4 P1 across 1ohm resistor 17.5 mV
     P3 458
     P4 456
     P5 -43
     P6 -43

V5 P1 across 1ohm resistor 20.4 mV
     P3 458
     P4 456
     P5 -43
     P6 -43

If you have 2 well-calibrated V-meters for measuring the tube current simultaneously, you'd soon be able to see whether the difference in tube current between V4 and V5 is due to slight fluctuations in the supply at different measurement times, or because the output tubes aren't absolutely evenly matched. (I suspect the latter, but having said that, they're not that badly mis-matched.) 


If you're concerned about losing bias voltage, make sure that:
1) The bias supply circuit is functioning properly by checking (and re-flowing any bad) solder connections:
2) The bias supply filter caps are functioning properly (which you would usually notice sound-wise if they aren't);
3) The output tube grid pin clamps are sufficiently tensioned (you can re-tension them with a small jeweller's flat-head screwdriver - with the amp off and unplugged, and the caps discharged)


Always check you have the requisite negative DC voltage on the output tube grid pins before plugging output tubes in.

[uki]

what happens when you 0 one pot, does the other have more range, try that with both

tack a 1Meg R to one pot, outer pins, leave the other pot 0 what happens

anyone you know, work with have a scope
troubleshooting the signal path without knowing if it's a gain thing or a circuit flaw is a big crap-shoot that usually leads to self installed problems

Well I don't know anyone that have a scope.

I did the 1M resistor in the first volume pot but it didn't change the behavior, those pots are very old and I will replace then when I get some new ones.

If you have 2 well-calibrated V-meters for measuring the tube current simultaneously, you'd soon be able to see whether the difference in tube current between V4 and V5 is due to slight fluctuations in the supply at different measurement times, or because the output tubes aren't absolutely evenly matched. (I suspect the latter, but having said that, they're not that badly mis-matched.) 

If you're concerned about losing bias voltage, make sure that:
1) The bias supply circuit is functioning properly by checking (and re-flowing any bad) solder connections:
2) The bias supply filter caps are functioning properly (which you would usually notice sound-wise if they aren't);
3) The output tube grid pin clamps are sufficiently tensioned (you can re-tension them with a small jeweller's flat-head screwdriver - with the amp off and unplugged, and the caps discharged)

Always check you have the requisite negative DC voltage on the output tube grid pins before plugging output tubes in.

Very good advices that I'm taking notes and will keep in mind. Thank you for that!

Only one meter atm that can measure mV, this is what I got: https://media.dooca.com.br/raisa.com.br/produto/zoom/multimetro-10a-linha-viva-md-1500-icel-14946192255304.jpg

Would be a good thing to add a bias balance pot to the amp, besides the bias adjust one? That way the current could be adjusted even yes?

This may be very important, in the EL34EH data sheet it says: no more than 425v in the screen, but there is 456v, could this be the cause of the excessive distortion?

Check out those pots 

[tubeswell]

Screens should be kept around 420V on EL34s IMHO, especially modern production ones. You should use decent Rg2 (1k to 2k2, rated at 5-7W) on the screen pins. The 'design problem' with EL34s is the signal grid is wound close to the cathode, in order to achieve maximum gain. But this makes the signal grid prone to collecting emissive coating material from the cathode over time, which then means the grid starts charging up, and losing its bias voltage. If the grid leak resistance are too high, this can further contribute to loss of bias, leading to thermal runaway and red-plating.


Also there are opinions out there that modern production EL34s should be nominally de-rated to 20W (instead of the advertised 25W) due to reliability problems.
(see the 'Lord Valve' recommendations at the bottom of this page: http://www.duncanamps.com/technical/lvbias.html


This means setting them up to idle at 30mA max if you're running the plates at 460. This often mean modding the bias supply circuit to get the grid voltage more negative to the point where it will reliably stay put under normal operation.


Bias balance shouldn't be necessary if you have the tubes reasonably closely matched.


Without getting a scope onto the signal path, its a guess as to where 'distortion' is originating from.

[uki]

I got two new CTS pots for the volumes and 2.2k 5W resistors for the PA screens.

It did take the amp to a new level  , now does have some clean sound, and it is much more pleasant to hear, but it is still breaking up somewhat too earlier, about volume 4-5.

The resistor didn't drop much the voltage for the screen of the PA tubes, from 456 it went down just about 2-3 volts , there is a little improvement of the mV in the tube that was higher now is a bit lower, dropped about 0.05mV.

Going higher with the screen resistors would do the trick ?

[shooter]

I'm not a player sooo....
have you played through a Marshall plexi style amp, it seems to me the players I know like it for it's "crunch"
it's "good" distortion, I don't recall anyone wanting clean 

I know the default for mine was 2 dark channel 4-6 clean side, BUT they each had "their own" sound, bad, imho.  what made the amp is the mix.  Not saying to quit tweaking, just maybe know what the amp was designed for, otherwise you might wind up tweaking it into a fender 

[uki]

I'm not a player sooo....
have you played through a Marshall plexi style amp, it seems to me the players I know like it for it's "crunch"
it's "good" distortion, I don't recall anyone wanting clean 

I know the default for mine was 2 dark channel 4-6 clean side, BUT they each had "their own" sound, bad, imho.  what made the amp is the mix.  Not saying to quit tweaking, just maybe know what the amp was designed for, otherwise you might wind up tweaking it into a fender 

On a Plexi nope, but I did play on a Marshall full tube once long time ago, it was a JCM900 it had a clean tone, I didn't fiddle with it enough thou to find out what it could do. I do play on a hybrid model quite often with tube preamp(dunno the model) but it isn't the same thing as a full tube amp...
Man the amp sounds much better now, I do like the crunch !! At 50% volume it rocks !!! And I do like the volume mix thing!

The thing is: in the data sheet for EH EL34 it says 425volts max for the screen, the screen of the tube is at the moment at 454, it will stress out the tube yes, how to drop the screen voltage down to the specification levels?

[tubeswell]

Its probably not going to be easy to drop screen voltage without installing a complicated regulated screen voltage supply.


It would be easier to instead increase the -ve voltage range on the bias supply circuit, in order to ensure you can get the cathode current down to about 20mA. This may mean having to going up 'one notch' swapping out the bias range resistor. Being able to bias the output tubes on the cold side like this will also help increase the headroom all round, even if it means the bias ends up deep towards Class B.

[shooter]

it will stress out the tube yes,

yes, but you're within 10%, and it rocks, and it's loud, and it makes you smile when playing, budget for a new set next year 

[uki]

It would be easier to instead increase the -ve voltage range on the bias supply circuit, in order to ensure you can get the cathode current down to about 20mA. This may mean having to going up 'one notch' swapping out the bias range resistor. Being able to bias the output tubes on the cold side like this will also help increase the headroom all round, even if it means the bias ends up deep towards Class B.

The cathode current is already at 20mA on V5 and V4 at 17mA.
Leave as is now ?

Well the overall sound is very good, no noise or hum, I did play on it with a different guitar that have a more clean tone and I was able to get some clarity on the notes, also plugged in a 4ohms speaker cab(2x12 150w each) but the tone is somewhat dry and metallic, on the 8ohms cab(1x12 80w) it does sound much better, more round and warm. In the EL34 data sheet says plate resistance 15k ohms so I guess a 16 ohms cab would have a better range for the PA tubes.

I still need to finish the cabinet, only the skeleton is done...

Thank you !

[shooter]

plate resistance 15k

my sheet show Ra-a @450vdc ~7k

the tone is somewhat dry and metallic

2X12 will make her happy, but you need to roll a bunch of speakers to find the sweet spot, maybe check out the cabs n speakers section of the forum for ideas
I run a modern eminence re-issue with an '80s creamback, AND it sounds best with the cream-back on the "right" looking at it 
I get sweat-clean from the left and the right takes care of the rest

[uki]

Those PA tubes are from Electro-Harmoniks, here is their data sheet.

[shooter]

I believe plate resistance is NOT the same as plate load.  Need checking but plate resistance stays basically the same while plate load is dynamic

[uki]

Oh the load resistance in the data sheet says 'nominal'.

my sheet show Ra-a @450vdc ~7k

You might be right about that, then if that is the case then putting the 8ohms cab in the 16ohms set in the amp will do the trick? Or is it a bad idea?

I have done those calculations while ago for the OT, I think this will help now to get the best out of the amp.

32 * 32 = 1024    1024 * 4Ω = 4096Ω

25 * 25 = 625     625  * 4Ω = 2500Ω

19 * 19 = 361     361  * 4Ω = 1444Ω



32 * 32 = 1024    1024 * 8Ω = 8192Ω

25 * 25 = 625     625  * 8Ω = 5000Ω

19 * 19 = 361     361  * 8Ω = 2888Ω



32 * 32 = 1024    1024 * 16Ω = 16384Ω

25 * 25 = 625     625  * 16Ω = 10000Ω

19 * 19 = 361     361  * 16Ω = 5776Ω

[uki]

Slowly turn bias from -53V toward -30V. Watch the milliVolts on the 1 Ohm resistors. For now, stop at 20mV (=20mA), which is on the cold side, but *safe*, and should play OK with slight hoarseness in note decays.

35mA may be a good final goal, but I would play it a long long time at lesser current to shake all the bugs out. Those vintage parts may resent being re-awaken after a decade of sleep.

Talking about shaking the bugs out, weird behavior happening with the volume controls, the sound doesn't come up right the way as it is turned up, the sound is coming up when the volumes are at about 3, and suddenly it just appear like there was a fast spinning of the pot, to the actual amount set with the pot. What to do ?

Thanks !!

[shooter]

What to do ?

both ch or just 1

sketchy input jack, sloooow plate voltage, bad cable, solder connection...

[uki]

What to do ?

both ch or just 1

sketchy input jack, sloooow plate voltage, bad cable, solder connection...

Both are doing same thing. Could this behavior being caused by the screen resistor that was swapped previous to it?
Would it cause slow plate voltage, in which tubes(if that is the case)?

I did find the problem, the new pots are defective, the old vintage pots are back in place and both volumes are working as expected !! 

Now wait for the cabinet !!

Thank you  !

[shooter]

slow plate voltage

put your meter on the "D" node before you power up, jack in some canned music into the input jack before power up, set volumes ~~ 4 before power up.
power up, watching the meter, does it "react" with the sound, like, start at 200vdc the just to 300vdc when the music kicks in?  (bs numbers for example only  )

the schematic shows 1k screen R, what did you go to?

[uki]

slow plate voltage

put your meter on the "D" node before you power up, jack in some canned music into the input jack before power up, set volumes ~~ 4 before power up.
power up, watching the meter, does it "react" with the sound, like, start at 200vdc the just to 300vdc when the music kicks in?  (bs numbers for example only  )

the schematic shows 1k screen R, what did you go to?

2.2k 5w  I did the trick to fix the excessive distortion, it still does distort at high volume 
Thanks to Tubeswell, his idea !!

Guess what?!

I did find the problem, the new pots are defective, the old vintage pots are back in place and both volumes are working as expected !! 

Now wait for the cabinet !!

[uki]

Almost done now with the cab, only missing the back panel !

[uki]

Done !!

Thanks to all of you that came to my aid !!

[sluckey]

Looks good! Well done!   

[Big_Mike]

Great work!  I see you have one input instead of the standard 4 inputs on a Plexi, did you configure it so both channels are jumped?

[uki]

Great work!  I see you have one input instead of the standard 4 inputs on a Plexi, did you configure it so both channels are jumped?

Thanks !!

Here is how it is in the circuit, I did it just like in the Hoffman 6V6 version, there is one volume for each side of V1, so it is possible to mix the bright with the dark channels. My amp is the Plexi 50 with EL34s.

https://el34world.com/Hoffman/files/Hoffman_Plexi_6V6-V2.pdf
https://el34world.com/Hoffman/files/Hoffman_Plexi50.pdf