I am Biased out

[capmando]

Having some challenges with my Traynor YGL 3 mark 3.
 4 X 6ca7 80 watts.
 Started to blow fuses, sometimes. Pulled preamp tubes, Variac and light bulb limiter one preamp tube 12AX7 at a time, different tubes same socket,
 different tubes different sockets all 12Ax7 positions. slowly pops fuses like 10 seconds. Somethings not right.
In my frustration I suspected transformer, 304AC secondary on Transformer seems good.

I  also changed out 1N5062 Avalanche diodes for BYW95C fast recovery  diode 3 Amp 800V, (what i could find at the time locally and stronger).
As a good measure I disconnected death cap and switch. I accidentally ran the amp for 30 seconds without the ground buss to chassis wire, in place of Death cap ground circuit. Oh Crappers!

 Drained the caps directly (had 120V on them). checked the fuse OK. Huh?  installed missing ground buss wire to chassis. star washer all the grounds to chassis.

 Put 115.3 Vac Into amp from Variac getting  B+ 440dc+- slightly higher than spec 436Vdc Run switch on,   B+ 330dc in standby. Fuses ok no Tubes installed


 Multimeter Neg probe connected to chassis ground, Pos to 64uF Bias cap Neg lead.
. I get 0V! with Run switched in. Standby switch in I get -3.77
 Do I need power tubes installed to get -37 Vdc Bias on Schematic.?

My question is am I measuring this right?.

 I am hesitant to install and possibly damage power tubes?
Anyone have experience with Avalanche Diodes are they better
worse the same? than gen purpose. Did I silently reverse pop 2 of 4 diodes in a full wave Bridge? set up. I tested diodes and resistors with Multimeter and 64uf cap all seem to test Ok.
Any advice is greatly appreciated.


I hope everyone is safe and sound during this Challenging time
Wash your hands often, keep 6 feet away, Self isolate, support our local Medical staff.
Thanks 
 
 

[PRR]

> Do I need power tubes installed to get -37 Vdc Bias on Schematic.?

No.

And not bonding chassis to circuit only means they may find your body next to a hummy amplifier.

Someone with more experience on big Traynors may spot what is wrong.

[Latole]

 Try pull outputs tubes, one or few could be short. Plug amp and report.

[sluckey]

Pull the output tubes. Do you have -37V now?

If not, refer to attached pic...

[jjasilli]

Use a circuit breaker, at  to at least temporarily  to stop wasting fuses while testing.  Use a repair flowchart.


See www.geofex.com > Tube amp debug page > Pops fuses


Always suspect tubes first.

[capmando]

Yes the Amplifier will be humming the death March! Ha ha,  PRR

After replacing the 1n5062 Diodes, and the chassis ground wire, the 4 Power tubes were pulled. I will test them in place once I get the Bias circuit repaired.  Thanks for ideas Latole       

I don`t have -37dc Bias, unless I am measuring it  the wrong way.

 If I reference the chassis with my Multimeter Neg
 probe and probe the B- of the full wave bridge I get 0V, but i get
 B+440dc, on the other side of the rectifier. I buzzed out my Diodes
getting forward .44 and backward OL +-on each of them tested single.  in circuit connected together.
Could the 3 Amp, BYW95C diodes be not opening up at this low a current?
Could they test fine at low voltage and high voltage not working properly?
The .047uf checks ok too.
 According to my El cheapo Chinese wonder circuit board meter
I get: Bias Diode out of circuit readings of Uf=685mv, C=42Pf,Ir=42na.
 Not familiar enough with these type of readings
The Bias Cap: out of circuit 64uf 64V, shows 70.6uf Vloss=1.0% ESR 2.5 Ohms. 

Thanks for pointing out the test points, and my suspicions about the   -37dc  Slucky I will swap out some parts and see what i get!

Some good Ideas at the Pops Fuses Geo website page, Thanks Jjasilli
started to look at it.

[shooter]

According to my El cheapo Chinese wonder circuit board meter

I would not trust old dead tubes to that meter, and brand new tubes!!  get a meter that IS reliable, your amp depends on it

[Latole]

Use a circuit breaker, at  to at least temporarily  to stop wasting fuses while testing.  Use a repair flowchart.


See www.geofex.com > Tube amp debug page > Pops fuses


Always suspect tubes first.

Always suspect Power Tube first

Better than a breaker is a bulb lamp curent limiter ; fuse will never blow and you can work on amp.
Bulb lamp will glow "high" with short circuit. circuit Ok = glow low.

Google amp bulb lamp limiterÙ:

 

[Pietro]

"Multimeter Neg probe connected to chassis ground, Pos to 64uF Bias cap Neg lead.
. I get 0V! with Run switched in. Standby switch in I get -3.77
 Do I need power tubes installed to get -37 Vdc Bias on Schematic.?"


Try the positive lead to the positive pole on the cap.
keep one hand in the pocket mate !

[Latole]

Pull the output tubes. Do you have -37V now?

If not, refer to attached pic...

Sluckey you show a GL3 / 3A not GL-3 Mark III

GL-3 Mk III

[Latole]

1-  Read AC volts here ; arrow IMO you may have 90 to 125 AC
     Probe each side of the secondary.

    No need standby to play, no need any tube in. Keep power tubes out until amp is fixed

2- Read negative DC bias caps side ; + probe to ground / chassis, negative probe to filter cap, diode side.
     You wrote read 0 volt probe wrong side ! Probe at right side must be same if 0 ac volts at number 1

Report 1 and 2

[sluckey]

Sluckey you show a GL3 / 3A not GL-3 Mark III

I show the schematic he posted. What was I supposed to do? He's in good hands now. You fix.

[Latole]

Sluckey you show a GL3 / 3A not GL-3 Mark III

I show the schematic he posted. What was I supposed to do? He's in good hands now. You fix.

I'm sorry, I didn't mean to blame you and it's a bit my fault, I hadn't seen his link to the schematic and which you used.

[capmando]

I am Biased Out:    Part 2

After receiving and installing new .047uf cap, 68uf Cap, 33K Resistor,and 1n5062 Diode. Most of the Neg Bias Circuit. I am now getting -34Vdc No Tubes installed,  -34Vdc Pin 1 on all 6Ca7 sockets. Fuses are holding. Finally Yeaah!  Thanks Everyone

I suspect it was the .047uf cap but I changed everything to be on the safe side. I got a intermittent reading on cap so, into the bad parts bin.

Another Challenge:
I also changed out the Neon indicator Light voltage drop resistor, it had burnt area hiding the color values. And Neon was out.
I read the schematic as 100K, or maybe its 100H   R51 ?, I am pretty sure this is an incorrect Value. Kinda Blury on schematic.

  I changed it and set up with current limiter light bulb.
Dialed in my Variac at 114.9Vac, and used my Greenlee DM 860A, (My Fluke series 73 is broken), I get 193 Vdc (127Vac)  probed at the Neon connector referenced to  chassis ground,  I shutdown drained caps,  then unplugged the Neon. Its rated at Max 125V!

I was going to try 100 Ohms, is there a way to check this?
Thanks again for everyone help!

[tubeswell]

Capacitor-coupled bias supply needs a good coupling cap.


Increasing the capacitance of that cap will enable the bias circuit to yield a higher negative voltage

[2deaf]

I also changed out the Neon indicator Light voltage drop resistor, it had burnt area hiding the color values. And Neon was out.
I read the schematic as 100K, or maybe its 100H   R51 ?, I am pretty sure this is an incorrect Value. Kinda Blury on schematic.

Several other models call for 220K.

[sluckey]

DO NOT REPLACE WITH A 100Ω RESISTOR !!!

R51 is 100K and that is correct. A 100Ω resistor will destroy the neon. I'd rather see a 220K or even 470K for R51.

[2deaf]

Pete may have put the neon lamp across the secondary so that the same lamp would work with domestic as well as "English" models.  A bunch of his amps had a neon lamp with a built-in resistor across the primary.  The 125V neon lamps available today also have a built-in resistor, so you could put your 125V neon lamp directly across your primary instead of fooling with secondary resistor values.

[capmando]

Thought I would try SLucky suggestion of replacing 100k resistor to something higher. After replacing the 100K R51 Resistor for the Neon Panel light with a 330K 1/2 watt (That`s what I had) my Panel Lamp AC came down to 107 great!, But still had 175DC on it measured Multimeter Neg to the chassis ground.

Without the Neon complaining or being brighter, it went out. I think the old one was rated for 220/125 for export. Had a 2nd one handy no worries, 125V stamped on it like the one i am replacing.  So I took a thought from 2 Deaf, Installed direct to 120AC coming into Amp. Installed it, seems to be happy now. Great thanks Everyone.

 So 80 Watts of Light bulb limiter, Variac at 115V Reading 441DC at Rectifier Good. -34DC at Bias Trim Pot, Trim pot wiper has 2/3 more movement to go. The Schematic asks to measure for -37DC,  Also on the schematic:  Adjust Bias supply to 8 volts DC across R57 No signal applied.

Is this too cold a bias setting Tubeswell and that's why you suggested changing capacitor to a higher value.
 

 So without power tubes installed no bias measurement. Power tube sockets have 441DC and 6.2 AC on heaters looks Good.   I am willing to Change Bias circuit Capacitor, and reading up on how to measure power tube bias. I will buy Dougs Bias Measurement Kit! Whats a good value for the Capacitor to change to?

Yeah!  I kept going. So I measure Preamp Sockets voltages on V1 to V6.

 V1       Socket  1,6 to Lug 9:  420DC , 420 DC
 V2        Socket 1, 6 to lug 9:  417, 417
 V3        1 and 6 to 9,  Lug 1:  347 Lug 6: 427
 V4                                  1:  414,  6: 419
 V5  Pins 7,8,9 connected together 433DC
 V6                                  1: 430  6: 432

Seems Good so in go the preamp tubes and power up with Limiter and Variac at 115 AC.
orange yellow glow, no red plating, that I can see, Light bulbs 40 Watt X 2 = 80Watts, glowing a little.

 R39 15K 10 watt heats up, burn your finger to 140deg F feeds V5 6BQ5, as well as R58 5K 5 Watt Temp 125 Deg F, feeds other tubes. All other 12AX7 slightly warm, 6BQ5 is heating up I can touch it Max 1 Second on tube near plates. I can see a nice orange glow through center of tube socket into bottom of tube. Normal with this tube? Is tube Toast?, something else
Any thoughts is greatly appreciated.

[mresistor]

I don't understand your voltage on tube test or terminology. Tube sockets have pins. Once you have determined that the filamemts are fine,  on pins 4, 5 & 9,  then the pins we would be interested in would be 1, 2, 3 and 6, 7, 8 for 12AX series tubes. Plate, grid and cathode.


Like this   V1   pin 1 ----
                      pin 2 ---
                      pin 3----   and so one 


Those plate voltages all look too high. What is lug 9? V1 and v2 plates should be +150v according to that schematic.  Pin 9 is the middle of the AC filaments, I would use chassis ground for the meter.

Test the pins with the black lead on the meter grounded to chassis and use the + red lead to probe the pin.


If your limiter is slightly glowing  it indicates that there are no shorts..   and you can take the amp off the limiter..  when the 12AX tubes plate voltages are confirmed to be correct and not high like you reported.


V5 is the Reverb Driver and the temps you indicate are not alarming to me.   It is one of the hardest working tubes in the amp.   How are you measuring temperature?


The important concern here is that you need -37 volts bias to the grids of the 6CA7/EL34 power tubes..  do not install them untill you have it..   
Why are we running at 115VAC?  Is that what comes out of your wall? 

You can take Tubeswells advice to increase the bias cap value, say to 75uf 100V IF we're talking about C52. What say you Tubeswell?   - you need a good "range of adjustment" for power tubes..  have you tried adjusting the 22K bias pot?  If so what was the range of voltage? 


there are ways to increase bias voltage (make it more negative)   but first try adjusting it?

[mresistor]

I would adjust the bias adjustment pot (R77 is it?)  to near maximum negative voltage.  When you get to the point of adjusting the bias then you can adjust it back to where it needs to be.


Did you know you can also install 1 ohm resistors on the cathodes of the power tubes to ground on this amp.   Then you can measure total tube current and it is shown on the meter as milivolts and directly
correlates to the milliamps of tube current.   Then you don't need a bias checker, although if you plan to bias alot or amps the checkers come in handy.

[tubeswell]

For capacitor-coupled bias supply, its the (first) coupling cap from the HT winding that needs increasing* (in order to increase the voltage). See Steve's schematic in post #3 above.

* or adding another cap in parallel, as in Steve's schematic

Or look at the bottom of Merlin Blencowe's page here: http://www.valvewizard.co.uk/bias.html. Or just hack Merlin's suggested bias supply circuit and increase 'C1' if needed.

The size of this coupling cap determines the amount of charge that is present in the following (bias) circuit. Bigger cap = more charge being passed on, so you get more negative voltage out of the other end of the bias supply circuit.


As for the type of cap, you could use a bipolar e-cap, but simplest to stick with film caps for reliability. (And if the exiting coupling cap is 0.047uF, there's plenty of scope to increase this using film caps).

[capmando]

I feel a bit more at ease with taking the power off of the light bulb Limiter. Thanks Mresistor
I realize my preamp tube readings seem a little unclear and incomplete.

At 119VAC, on Run off standby B+ is 450Vdc. Direct to line. All readings DC, except noted
I used a Greenlee DM860A Fairly new a year old, shows AC and DC at the same time, and used the thermocouple supplied to make Temp.readings.  Chassis ground to Neg Multimeter probe, Pos Probe to each tube socket Pin. Hand in pocket HIP. I was using Fast Blow fuses in case there was a problem
 to save transformer it held at 1 amp, can this affect my readings?

V1: 12AX7  Pin (1) 158  (2) 0   (3) 1.23  (4+5) 3.13AC  (6) 154  (7) 0  (8) 1.19  (9) 3.19AC

V2: 12AX7  Pin (1) 158  (2) 0   (3) 1.23  (4+5)  3.23AC  (6) 156  (7) 0  (8) 1.13  (9) 3.17AC

V3: 12AX7  Pin (1) 262    (2) 40.24AC to 45.14AC and -1.15 to 3.29+- goes up and down slowly.

V3:12AX7  Cont: Pin (3) 17DC to 28.7DC and 25Vac. (4+5) 3.18AC  (6) 187DC to 222DC, 64AC to 80AC +-
V3: Cont: Pin (7) -.277DC to +.140DC+-, 1.13AC to 1.36AC+-  (8)  1.57DC+-, .179AC+- (9) 3.26AC

V4:12AX7 Pin (1) 107  (2) 0  (3) 1.75  (4+5) 3.15AC  (6) 165  (7) 0  (8)1.71  (9) 3.23AC

V5: 6BQ5  Pin (1)NC  (2) 0  (3) 3.83  (4) 3.16AC  (5)3.22AC  (6) NC  (7+8+9) 121

V6: 12AX7  Pin (1) 217  (2) 38.9  (3+8) 53.9  (4+5) 3.19AC  (6) 230  (7) 38  (8)    (9) 3.29AC

I pushed my Bias Pot to the Max I got -39VDC, No Power Tubes

Temp of tubes was: V1: 118F,  V2: 119F, V3: 106F  V4: 115F  V5:162F  V6:126F

Temp of big ceramic Dropping Resistors:
 15K Ohm 10Watt, 170 Deg. F wired to 6BQ5,
 5K Ohm 5Watt, 160F. wired to other preamp tubes.

 I am surprised the resistors get so hot can they take the heat?, and is the heat an indicator of a bad tube or something else?, Also could I put a larger wattage if it fits?, I tested 6BQ5 on my Eico 635 Tube tester (don`t laugh), it was in the Green zone at 70% no shorts.

Tubeswell    I will look at changing the the Bridge Film Capacitor,  I started looking at Merlins
Bias Article you cited. Any Guesses on a value?
 Do you feel biasing the tubes hotter
will make the sound give more breakup earlier, more warmth more cream?

MResistor    I actually was thinking of installing 1 ohm 1% resistors I had ordered some already. I may try the long way just to know and feel how bias is working. 
 My understanding is getting a little clearer with each post  thanks again Everyone!
 

[tubeswell]

you could double the bias supply coupling cap (i.e. solder tack another one of the same value in parallel) and see where you land. the difference in voltage vs capacitance of this cap is not a linear relationship.


Re the '15kOhm 10W resistor', are you sure its 15,000 Ohm?, and what is the voltage across it?

[capmando]

I added another .047uf in parallel to the 1st .047uf cap, the Bias voltage range of trim pot increased to the 30 to 60 volt DC range, from the 30 volt to 40 volt DC range. I adjusted the bias trim pot a few times and measured.  I hope my numbers and calculations are in the ballpark. These are my 1st experiences biasing power tubes.

 All Measurements done on warmed up Amp ref to chassis 120.7 AC steady, 2 pairs of 6CA7

 Output Transformer: CT 445 DC , drifted up to 446.90 then down again.   
 Prim 52.18 Ohms (Plate Green wire) and 47.77 Ohms (Plate Red wire) Measured CT to Primary caps drained.

Ground to Pin 3: 445.83 DC   green Wire .  Ground to Pin 3: 445.03 DC red Wire . Also measured Plate to CT:  -.606 Green, -.623 Red Wire. I get 12.65ma green, 13.04 Red,   for each pair of tubes
-45.07 Neg Bias V. Kinda cool. I mean Bias haha.

Setting 2:
Then I Biased to schematic specs to see 8 volt drop across  470R R51  (460.6). I needed to set Bias voltage trim pot to -35.56 VDC.  CT to pin 3 -2.91, CT to pin 3 -2.87.  Also cathode voltage 434.42 DC Pin 3 to Pin 8 .  So I get 55.76ma and 60.07ma. per pair of 4 tubes  Kinda High, No signs of red plating. More hum coming from tranny. Played Guitar slight more fuzzy sounding.

last Goldilocks Bias Setting 3:
-39.34 Bias DC, Plates to ground 441.35, 441.47.   CT to Pin 3 -1.88 and -1.86.   Cathode Pin 3 to Pin 8:  442.51 and 442.38 DC.    Gives me 36ma, and 38ma  15.9 Watts and 17.2 Watts per pair of tubes.  Played guitar sounded more defined a little more chimey.
Also measured 15K  10 watt resistor was getting 120VDC across it.
I noticed a drift in my measurements as i made them +-  slowly going up then down approx 1.5 volts is this normal?.
The amp seems to be working, had it on for an  over hour at a time at idle. No redplating no poped fuses. What do you guys think time to break a bottle over the bow?.
Thanks for any advice.

[tubeswell]

An hour’s worth of stability is good enough

[capmando]

Thanks Tubeswell for the reply,
 the 170 Deg hot,  15K R59  10 watt resistors feeding the el84 reverb driver and the 160F hot, 5K R58 5 watt feeding all the preamp tubes both resistors are hot. Does this seem typical? if they are then the Traynor Beast lives!. Woo hoo very happy.

[tubeswell]

Thanks Tubeswell for the reply,
 the 170 Deg hot,  15K R59  10 watt resistors feeding the el84 reverb driver and the 160F hot, 5K R58 5 watt feeding all the preamp tubes both resistors are hot. Does this seem typical? if they are then the Traynor Beast lives!. Woo hoo very happy.

According to the schematic you loaded at the start of the thread:

• R59 is seeing 202V. (202x202)/15000 = 2.7W. 10W rating is plenty.
• R58 is seeing 114V. (114x114)/5000 = 2.6W. 5W rating should be fine here. If its a wire wound resistor, it should last and last.

In both cases, if they're mounted in 'free space' where it can radiate heat without cooking other components, even better.

[capmando]

Thanks again Tubeswell, all this is putting my mind at ease
You inspired me to take measurements and use your math.

 R59 measured at 212.10 VDC across 15K 10 watt resistor I get 2.99 Watts.
 R58 I measure 112.81 across 5K  5 watt resistor I get 2.52 Watts

My AC going into Amp is 121.2 tonight.  R58 seems a little low compared to original printed schematic voltages at 115AC that you pointed out.

R58 also measured before and after at 446.46 in, 334.17 out = 112.29   fed from 40uf Cap
If I tap V1 12AX7 it`s microphonic right now. Don`t know if this means anything.

[tubeswell]

The slight difference in mains VAC you're seeing isn't anything to get too concerned about. The heater (or rectifier filament) voltages are about the only really critical thing, and they will be fine as long as they're within +/-10% of the tubes' 6.3(or 5)VAC design target when the tubes are all plugged in.


The PT's High Tension winding can be easily 20 or 30VAC different* to what the 'stock' schematic voltages indicate, and the amp can still work fine, as long as the filter caps (and coupling caps) are rated in excess of the full unloaded B+ voltage (when there are no tubes plugged in), and component power rating is adequate etc. You manipulate the tubes' running conditions by way of biasing. Tube amp voltages are all kinda ballpark stuff. Exactitude is not as critical as the principles about biasing and load lines etc (Feel free to muck around with gain structure, load lines, biasing, Ck||Rk frequency shelving, inter-stage attenuation etc to your heart's content. 's where all the tone-fun happens)

[capmando]

Thanks again Tubeswell, for your view. I feel a little more again at ease.
Heaters are at 3.2, 3.18, supply Caps rated at 500 new F+T.

I guess the before transistors, with tubes it is more forgiving. I am starting to get a feel
about Biasing and i feel good about the work I`ve done on the power supply. I`ve  learned a lot
rebuilding a good part of this amp. I have a much better understanding of tube amps in general than a year ago.  Everyone at the Hoffman Amp Forum been super helpful. Always more to learn.
I will look into cascading one channel into the other with a switch, for a start. I feel the need to contribute to the site.  Thanks everyone