Problem biasing for correct voltage

[dude]

I'm running 7591's in an EA-65, fixed bias and they are running way too hot.

I checked the neg voltage after the diode and it's -6.5 way too high, should be -22.

The ac voltage at V5, rectifier tube is 345vac, after the 390K I get 150vac after the diode I get -6.5vdc should be around -22vdc.

I lifted all resistors in bias and they are correct values.

The schematic calls for 390K should this be higher as the voltage after it is 150vac?

Or could the diode be bad? Edit: changed the diode, no difference, change the rectifier tube basically the same.

Edit: Some background, a guy got this amp at a yard sale, fired it up and played it till it blew, smoke and all. Then I get it to fix. A 1K 10 watt R off the choke was wiped out. burn up some wires that I fixed, all the power filter caps wer leaking glubs of goo, replace all filters, five. After that the voltages seem OK so I played it, sounded good but the 7591's were running hot, way too hot, so I checked the bias way off, here I am. The schematic calls for one 20uf and four 10uf's, I used a 32uf for the 20, and two 16uf for two 10s and two 22uf's for the other 10's. I can't see the differences would cause this problem. But for a freebe, I used used stuff I had around.

Maybe the 7591's are toast but they play well just that the bias is way off and their red plating. replaced everything in the bias circuit except the bias filters..?

 al

[LooseChange]

Lift the wire the feeds the 220k resistors and read the voltage on that wire. Did it go back to the correct bias voltage value?

[stratele52]

Your diode look good and your new caps value is OK. 

Change the 51k for a higher value resistor in the bias circuit.  This will help to have more bias negative voltage. Or you can also put a lower resitor in place of the 390 K, try a 220 K

You can also install an adjustable bias . Easy to do.

With a pot in serie with the 51 K. I'll use a 10 to 25 k pot in serie with  the 51 K.  Only the wiper of the pot must go to the two 220 K. You may have to try some resistor if bias is not enough negative for your tube.

Multi turn Cernet is the best and easy to use and cheap ( $2.50 )

[HotBluePlates]

Let's have a method to the madness.

First off, get the output tubes out of the amp, then measure the negative voltage at the output tube grid pins (pin 6 for a 7591).

Your resistors measure correct, and you replaced the diode. What about the cap across the diode? If that's leaking, it will drag the bias voltage down. Check it by replacing it, noting correct polarity.

Remeasure the voltage at the grid pins. If it is still way low, we'll move forward with other changes.

[LooseChange]

Psst HBP... I think the 47n's from the PI are leaky.

[dude]

What a dumbo I am, I'm almost afraid to say what I just did. I fried the grids pins 4 and 6 of one of the 7591s by putting in the rectifier socket by mistake. I can't believe I did something so stupid. A nice GE 7591 that I just tested on a tube tester and both had no leaks now the grids are leaking and the tube is shot. I have no other 7591 and they are expensive.

Well, isn't always when you're doing a friend a favor for nothing you screw-up  .

Keep with me, I'll just rewire the sockets for 6V6's I have a bunch of old stock USA, then go from there.

The pin-out is close for 6V6s just have to change the wire from pin 6 (G1) to pin 5 and the wire that was on pin 5 to pin 8 (K/G3) leaving pin 6 open. Sound correct?

What other changes do I have to make, grids to the bias off the 7591 were 120 ohm so that's Ok. I'll have to change the resistors i nthe bias, see schematic what do you recommend there? Schematic is posted above.

I'll also change the bias caps that was probably the cause of the high negative voltage. I guess 20uf at 50v is OK.

BTW, without the power tubes I still have -6.5 after the diode and of course nothing on the feed to the 220 or 120ohm since I blew up the 7591

stick with me, I 'll get this thing fixed, what a dumb A...

al



 

[Shrapnel]

What a dumbo I am, I'm almost afraid to say what I just did. I fried the grids pins 4 and 6 of one of the 7591s by putting in the rectifier socket by mistake. I can't believe I did something so stupid. A nice GE 7591 that I just tested on a tube tester and both had no leaks now the grids are leaking and the tube is shot. I have no other 7591 and they are expensive.

Well, isn't always when you're doing a friend a favor for nothing you screw-up  .

Keep with me, I'll just rewire the sockets for 6V6's I have a bunch of old stock USA, then go from there.

The pin-out is close for 6V6s just have to change the wire from pin 6 (G1) to pin 5 and the wire that was on pin 5 to pin 8 (K/G3) leaving pin 6 open. Sound correct?

What other changes do I have to make, grids to the bias off the 7591 were 120 ohm so that's Ok. I'll have to change the resistors in the bias, see schematic what do you recommend there? Schematic is posted above.

I'll also change the bias caps that was probably the cause of the high negative voltage. I guess 20uf at 50v is OK.

BTW, without the power tubes I still have -6.5 after the diode and of course nothing on the feed to the 220 or 120ohm since I blew up the 7591

stick with me, I 'll get this thing fixed, what a dumb A...

al

The 7591 is kind of a cross between a 6L6 and EL84. It has near the power handling of the 6L6, and the sensitivity of the EL84, and is considered a pentode, not a beam-tetrode. This makes the 6L6 closer than a 6V6 to the specs of the 7591(A). Using the prices found at Tubes and More (AES): Yeah, the new production 7591s (that act like the old productions, and don't need changes to the circuit to use) are more expensive than the 6L6, 6V6, EL84, and EL34... BUT they are cheaper than the 7868, 6550, and KT88.

AFAICT, the only new production choices are EH and JJ. IF you are truly thinking of converting, the 6L6 has usually been the prime choice in the past.

[dude]

Thanks for the info, I'll remember that if  I ever convert but maybe the PT can't handle 6L6s.

Anyway, I just finished rewiring the sockets and I'm working on the bias. The amp does have 220ohms r off the power tubes but they were very high at 260 each, I replaced them wit ha matched pair at 219 ohms. Put in grid resistors too. also put a 1 ohm 1% r off the cathodes to ground for biasing. I checked out a Fender deluxe and wired the sockets the same, I'm sure I got it right.

The only thing is the PT doesn't have a low voltage bias tap so I figure I'll use the 390K like in the schematic.

Can I wire it like the layout attached using one 25/50 cap with the 390K (instead of the 470ohm 1W) to the rectifiers tubes ac voltage?
This circuit does not have tremolo so can run from the diode to bottom leg of the 10K-L bias pot and wire everything else like the layout?

So maybe I'm back to where I was before just that I have 6V6 now. Any changes to PI because of the 6V6's? Remember I'm using two 6Eu7s as in EA-65 schematic above.

al    

[sluckey]

Can I wire it like the layout attached using one 25/50 cap with the 390K (instead of the 470ohm 1W) to the rectifiers tubes ac voltage?

390K will likely be too big and cause the bias voltage to be too low. I'd start with a 100k first. Adjust the size up/down to give the bias voltage range you want.

[Shrapnel]

Thanks for the info, I'll remember that if  I ever convert but maybe the PT can't handle 6L6s.

The 6L6 is about 100mA higher in heater current, each... might not be much of a problem, unless they got super cheap with the PT iron and demanded every ounce it had.

The only thing is the PT doesn't have a low voltage bias tap so I figure I'll use the 390K like in the schematic.

Can I wire it like the layout attached using one 25/50 cap with the 390K (instead of the 470ohm 1W) to the rectifiers tubes ac voltage?
This circuit does not have tremolo so can run from the diode to bottom leg of the 10K-L bias pot and wire everything else like the layout?

So maybe I'm back to where I was before just that I have 6V6 now. Any changes to PI because of the 6V6's? Remember I'm using two 6Eu7s as in EA-65 schematic above.

Those 6EU7s are practically a hi-fi version, with a different pinout, of the trusty 12AX7. Because of that, those tubes can be expensive too.

The Epi EA-65 "rivoli" PI should Be Okay as-is,  What you got is one gain stage followed by a split-load and then each half of that split-load output amplified with another gain stage.

Note that the 7591's expected about 425VDC while the 6V6's expected 415VDC... probably close enough not to worry about, even though it pushes a little harder than Leo pushed the 6V6's in that AB763.

[dude]

Thanks all,

I just recheck my wiring again on the 6V6 conversion and check those 6EU7's pin outs too. Thanks Shrapnel for your advice.

I look at the amp's wiring and schematic for it (EA-65 above) and it's not wired like the schematic. Someone has made incorrect changes, no wonder it has strange voltage off the bias and no wonder it blew up in the first place. It's a true PTP rats nest but pretty much a simple amp. The voltages are way high as the choke is not even wired correctly.

I'll rewire the power supply correctly and see what happens, work the rail and bias voltage then put the tubes in and go from there.

I learned one thing today, no matter how simple a mistake can be, anyone can do it. Still can't believe I put a power tube in the rectifier socket. anyone else do that and turn the amp on...?

I'll get back with the results.

al 

[Danskman]

"Still can't believe I put a power tube in the rectifier socket. anyone else do that and turn the amp on...?"
It happens to me one time, w/o further consequences (GZ34 vs 6V6, JJs) 

[Shrapnel]

Thanks all,

I just recheck my wiring again on the 6V6 conversion and check those 6EU7's pin outs too. Thanks Shrapnel for your advice.

I look at the amp's wiring and schematic for it (EA-65 above) and it's not wired like the schematic. Someone has made incorrect changes, no wonder it has strange voltage off the bias and no wonder it blew up in the first place. It's a true PTP rats nest but pretty much a simple amp. The voltages are way high as the choke is not even wired correctly.

I'll rewire the power supply correctly and see what happens, work the rail and bias voltage then put the tubes in and go from there.

I learned one thing today, no matter how simple a mistake can be, anyone can do it. Still can't believe I put a power tube in the rectifier socket. anyone else do that and turn the amp on...?

I'll get back with the results.

al 

Dude,

 You are quite welcome on the advice. As you discovered many wiring errors already, I can't blame you for wanting to check the wiring on those tubes too... I'd want to double-check myself.

As far as making the mistake that caused a tube to die? Well, I haven't done that yet,  but I have made stupid errors (where I should have known better, and did know better) with other things and damaged them. <sigh>      

Now I have to laugh though... I missed you posting the EA-65 schematic in your first post, and here I posted it again. At a glance it looks to be exactly the same schematic.

[stratele52]

Your diode look good and your new caps value is OK. 

Change the 51k for a higher value resistor in the bias circuit.  This will help to have more bias negative voltage. Or you can also put a lower resitor in place of the 390 K, try a 220 K

You can also install an adjustable bias . Easy to do.

With a pot in serie with the 51 K. I'll use a 10 to 25 k pot in serie with  the 51 K.  Only the wiper of the pot must go to the two 220 K. You may have to try some resistor if bias is not enough negative for your tube.

Multi turn Cernet is the best and easy to use and cheap ( $2.50 )

Dude you forget to read my post ?

[dude]

I've been working all day, got to have a day job the amps are for "fun"

Well, things have come a long way. I did wire the 6V6s correctly and the amps power supply was wire correctly accordingly to the schematic EA-65 above. What threw me off was one: the rats nest, not use to that, two: a lot of the unused tube pins are used for stand-offs points. You have to follow the schematic closely when dealing with true PTP but a good leaning tool.

I got it running with the 6V6's. I had a problem with the ac bias supply off the GZ34 before the diode, the schematic call for a 390Kohm and with that I couldn't get the voltage negative enough. Sluckey said to use a 100K but I thought that would made it worst, I should have listened to him right away as as soon as I put the 100K, bingo I had -35 volts. Thank you Sluckey, with a five level what was I thinking...?

I'm have two problems at this point before I add a few caps to those cathode resistors for some spice.

one:  Some 6V6's I can't get the correct mV for biasing some tubes, the lowest I can get is 28 mV a little too hot (between a 1 ohm r on the cathodes}. I used Sluckey's recommended 100K off the acv to diode and a 10K L pod with a 15K on the leg to ground. Just like the Deluxe Reverb AB763 schematic above, except they had a 10K, didn't have one handy. How can I get more range using the 10K L pod and the 100K r? Change the 15K to what?

two: the EA-65 has some kind of photo cell LDR off the OT and volume Pot, see schematic, one leg broke and won't stay soldered to the photo cell (amp cuts out when the leg comes off) so I cut wire leading to it (green see pic)which came from the volume pod and that cap from the LDR to cut that cell from the circuit now where do I attach that sweeper off the volume (green wire) and what about the ceramic cap?  See EA-65 schematic above look for the LDR.

I'll have to deal with adding a grounded plug  but the switch is a three way and in standby it supplies power to the screen grids o nthe power tubes, what's up with that?

Here's some pics of the wire cut and ceramic cap cut, cutting these changed the volume control and amp is sounding weak.  

EDIT  I just figured out one answer on the volume and the cap, just attached the greed wire to the end of that ceramic cap..?

[DummyLoad]

is that electrical tape holding the two ceramic power resistors strapped in series together? 

 

--DL

[dude]

is that electrical tape holding the two ceramic power resistors strapped in series together?  

 

--DL

Yeah, that'll hold them together, I ran out of solder. At least I twisted them together. It's called a mechanical connection

OK, I used a little solder, just a little, how did you like the nice hole I cut for the cap can?  

Tell me what resistor to put on the bias pod, that's hang'n there too, remember this is a freebe, that's what you get. And no grounded plug, don't touch that mike stand...

Stratele52 I did read your post, idea was good but the values changed as I blew up one of the 7591s and converted to 6V6s, thanks

al

[Shrapnel]

two: the EA-65 has some kind of photo cell LDR off the OT and volume Pot, see schematic, one leg broke and won't stay soldered to the photo cell (amp cuts out when the leg comes off) so I cut wire leading to it (green see pic)which came from the volume pod and that cap from the LDR to cut that cell from the circuit now where do I attach that sweeper off the volume (green wire) and what about the ceramic cap?  See EA-65 schematic above look for the LDR.

EDIT  I just figured out one answer on the volume and the cap, just attached the greed wire to the end of that ceramic cap..?

That Photocell is part of a feedback circuit that I'm not 100% familiar with. About all I can tell is is the feedback off the output runs through this, odd to me, cap-diode & resistor circuit before going to the LDR. (Perhaps PRR or someone else can enlighten me on it.) The LDR then controls the varistor part of itself to adjust the resistance to ground, I presume attenuating the signal.

I was going to ask about the yellow and green wires, but it appears that the yellow wire leads off to the tone and volume controls (with their associated parts) and then returns via the green wire in the same grommet. Connecting the green wire to that ceramic cap with the broken lead will restore your volume. Personally, I probably would have replaced that .005 µF ceramic cap and called it good, but if you don't mind no feedback circuit, your fix should work fine.

[sluckey]

one:  Some 6V6's I can't get the correct mV for biasing some tubes, the lowest I can get is 28 mV a little too hot (between a 1 ohm r on the cathodes}. I used Sluckey's recommended 100K off the acv to diode and a 10K L pod with a 15K on the leg to ground. Just like the Deluxe Reverb AB763 schematic above, except they had a 10K, didn't have one handy. How can I get more range using the 10K L pod and the 100K r? Change the 15K to what?

Read my first post again. Decrease the 100K to decrease the 28mv reading across your 1 ohm cathode resistor.

I'll have to deal with adding a grounded plug  but the switch is a three way and in standby it supplies power to the screen grids o nthe power tubes, what's up with that?

No it doesn't. Look at the schematic. It puts a short across the drive lines to the 6V6 control grid. This kills the signal going into the 6V6s which effectively puts the amp into standby mode WITHOUT actually turning off any power supply voltages. Sears did this on their twin twelve amp too.

[HotBluePlates]

That Photocell is part of a feedback circuit that I'm not 100% familiar with. About all I can tell is is the feedback off the output runs through this, odd to me, cap-diode & resistor circuit before going to the LDR. (Perhaps PRR or someone else can enlighten me on it.) The LDR then controls the varistor part of itself to adjust the resistance to ground, I presume attenuating the signal.

There was another thread on this amp a short while back. The LDR feedback circuit is a compressor that limits the output of the power section. The output is sampled and converted to a d.c. voltage to adjust the gain of a stage earlier in the amp.

The downside of the plan is the level is set by an internal trimpot, so you have to open up the amp to adjust it. It really is supposed to be "set-n-forget."

[dude]

 

I'll have to deal with adding a grounded plug  but the switch is a three way and in standby it supplies power to the screen grids o nthe power tubes, what's up with that?

No it doesn't. Look at the schematic. It puts a short across the drive lines to the 6V6 control grid. This kills the signal going into the 6V6s which effectively puts the amp into standby mode WITHOUT actually turning off any power supply voltages. Sears did this on their twin twelve amp too.[/quote]



Yes, I was having a hard time figuring that out, I've never come across an amp wired this way any clues on how to add a three prong power chord. Do I just run the ground like on the schematic and fuze the hot side eliminating the switch part that has the .022 cap?

Also the voltages are too high for the 6V6s, B+ 472, Plates 466, screens 461. Could I just use a 5Y3 or would a zener off the HV center tap be better? Or use the Zener before the choke? Would you use a bigger screen grid value then 1K5 1/2 watt.

I'm appreciate your help on the bias set up, I do remember you saying to work that 100K, I didn't have a 1 watt or better anywhere near it so I wanted to work the r to ground of the pod, I tried a 20K there and it worked OK. 

Thanks again all,

al

[sluckey]

any clues on how to add a three prong power chord. Do I just run the ground like on the schematic and fuze the hot side eliminating the switch part that has the .022 cap?

Just connect the black (hot) wire to the fuse. Connect the white (neutral) wire to the switch just like the existing power cord. Put a ring lug on the green wire and connect the lug securely to chassis using one of the power transformer nuts. Don't mess with the switch wiring. I'd leave the cap on the switch, but it won't hurt to clip it out completely.

Also the voltages are too high for the 6V6s, B+ 472, Plates 466, screens 461. Could I just use a 5Y3 or would a zener off the HV center tap be better? Or use the Zener before the choke? Would you use a bigger screen grid value then 1K5 1/2 watt.

Putting 6V6s in that amp is a mistake IMO. Those 7591s were much beefier than a 6V6. Rather than use 6V6s and kludging the power supply to drop the B+, I'd use 6L6s and diddle the bias accordingly. Or spend $20 for a 7591.

[dude]

any clues on how to add a three prong power chord. Do I just run the ground like on the schematic and fuze the hot side eliminating the switch part that has the .022 cap?

Just connect the black (hot) wire to the fuse. Connect the white (neutral) wire to the switch just like the existing power cord. Put a ring lug on the green wire and connect the lug securely to chassis using one of the power transformer nuts. Don't mess with the switch wiring. I'd leave the cap on the switch, but it won't hurt to clip it out completely.

Also the voltages are too high for the 6V6s, B+ 472, Plates 466, screens 461. Could I just use a 5Y3 or would a zener off the HV center tap be better? Or use the Zener before the choke? Would you use a bigger screen grid value then 1K5 1/2 watt.

Putting 6V6s in that amp is a mistake IMO. Those 7591s were much beefier than a 6V6. Rather than use 6V6s and kludging the power supply to drop the B+, I'd use 6L6s and diddle the bias accordingly. Or spend $20 for a 7591.

Yeah, makes sense, thanks again for all your help, sure wish I hadn't fried that 7591 but teaches me to be more careful, it could have been a nasty shock mistake. Makes me think...?

Anyway, what would your opinion be if I were to go with EL-34's, would the voltage be better suited for that tube? I believe the pin-out is basically the same as for 6L6s except grounding pin 1 to 8?

Thanks all again.

al

[Shrapnel]

Yeah, makes sense, thanks again for all your help, sure wish I hadn't fried that 7591 but teaches me to be more careful, it could have been a nasty shock mistake. Makes me think...?

Anyway, what would your opinion be if I were to go with EL-34's, would the voltage be better suited for that tube? I believe the pin-out is basically the same as for 6L6s except grounding pin 1 to 8?

Thanks all again.

al

You don't have to do anything to the sockets for the 6L6 since you already wired for 6V6. Just adjust bias appropriately. You can wire it for EL34s if you wish now, it just means taking pin 1 and tying it to ground, as you stated and bias appropriately. But, the 6L6 is the better swap for different tube type. The 6L6 has always been the go-to tube when the 7591s were really scarce.

The EL34 will definitely demand more heater current: 1.4A more... You can try, but you better keep an eye on how hot the power transformer gets. You may need another filament transformer just to handle the extra current if you go that route.

EL34: 1.5A each - 3.0A for two = + 1.4A
 6L6: 0.9A each - 1.8A for two = + 0.2A
7591: 0.8A each - 1.6A for two = + 0.0A

[sluckey]

I wouldn't mess with EL34s because of the higher filament current requirements. If this were my amp, and I didn't have any 6L6s, I'd just buy a pair of 7591 for $34 and be done with it.

[dude]

I put a pair of 6L6 RCA black plates in, biased to 35 mV the voltages on the tube plates dropped to 437v the amp sounds punchy and tight, very clean with almost all headroom on full volume.

Sounds much better with the 6L6s, but I noticed there are no screen grid resistors in this amp, how about two 3 watt 470 ohms?

Also no cathode caps on the 6Eu7s, first stage has just a 2.2Kto ground and second stage cathode doesn't go straight to ground so not sure where to use a cathode cap there. Any suggestions to get a little less headroom on the cap placement?

Thank you again for the advice sluckey, I'm really liking this amp and hate to give it back to the owner, he paid $10 for it at a yard sale.

sincerely,

al

[sluckey]

I noticed there are no screen grid resistors in this amp, how about two 3 watt 470 ohms?

good idea

first stage has just a 2.2Kto ground and second stage cathode doesn't go straight to ground so not sure where to use a cathode cap there.

Put a cap on the first stage if you want to, but leave the second stage alone. It's the PI and if you put a bypass cap on it's cathode, you'll kill one of the drive signals to the PA.

[dude]

Thanks, I promise no more questions on this amp.

al