200 Watt All Tube Bass Amp

[SupLar]

Hello everybody. I am brand new to the forum. Besides my introduction post, this is my first post.

 Back in 2009 I built a 200 watt all tube bass amp for a friend.  (Links to schematics are below).  It was an ambitious project, with lots of challenges, but it turned out great and the owner has been constantly gigging with it for the past five years (he loves it) and it has worked flawlessly and faithfully... until now.

 The other weekend the amp went dead.  The owner shut it down as quickly as possible... but I have a bad feeling.  At least one 6550 power tube red plated. I can tell by scorch marks inside the head enclosure.

 I disconnected the PSU from the circuit board to isolate the PT, rectifiers, and bias circuit.  The HT secondaries read 420 VAC each leg (120 VAC line voltage).  However, I'm only getting around 380 VDC at B+ and I should be getting approx 520 VDC.  The diodes check out OK - I even replaced them just to be sure.  The tubes all seem to be OK. (I tested them in other amps I have.)    I'm perplexed already.

 Any thoughts?  I could use suggestions on good testing procedures for both the OT and PT to determine if they are damaged.  This is my biggest worry.

 I didn't design the amp (my engineering and design skills are still developing) but you may recognize that the amp is based on a Sunn quad 6550 run UL from the schematics.

 Any help or suggestions would be greatly appreciated. I'm anxious to get the amp back on-line since the owner gigs with it almost every weekend and he is missing it already.

 Thanks in advance! schematics are here:

Pre Amp:
http://docdroid.net/gx61

Power Amp:
http://docdroid.net/gx64

PSU:
http://docdroid.net/gx66

[sluckey]

I disconnected the PSU from the circuit board to isolate the PT, rectifiers, and bias circuit.  The HT secondaries read 420 VAC each leg (120 VAC line voltage).  However, I'm only getting around 380 VDC at B+ and I should be getting approx 520 VDC.

I'm thinking you don't have any filter caps connected to the B+ line for this test? You need at least one cap connected to see a good dc reading. Without a cap to charge to the peak of the rectified AC pulses your meter will lie to you.

[SupLar]

Hi Slucky,

Thank you for your response.  That makes sense.  I will give that a try and report back.

[SupLar]

Oops... sorry... I misspelled sluckey!

[PRR]

The PT seems alive. The OT won't cause a red-plate condition.

_I_ would be looking for NEGative bias at 6550 grids before dropping bottles back. 6550 at super-Dyna conditions without G1 bias will bust a gut.

[SupLar]

Thanks for your post!  The gut on at least one 6550 definitely got busted LOL.  Your assessment makes perfect sense. Other than the bias circuit itself, what other components, if they fail, can cause negative bias voltage to change?  Could it be as simple as one 30 cent PSU diode wearing out? Thanks again!

[HotBluePlates]

Other than the bias circuit itself, what other components, if they fail, can cause negative bias voltage to change?  Could it be as simple as one 30 cent PSU diode wearing out? Thanks again!

You might test the 6550's and determine whether the remaining 3 are working fine Try testing the redplating tube and see if it acts up by itself in other devices. Measure pin 5 of the output tubes (without the 6550's in the amp) and see if the same negative voltage is present at each socket (and the value you expect). Check the redplating tube for simple catastrophic failure like a pin 2 to pin 3 short (you can do that with a meter and the tube out of the amp). Inspect the sockets for stray bits of wire, solder blobs, carbon tracks from arcing, excessive dust, etc.

Did the fuse pop? What else happened when the amp went silent?

If only 1 tube appeared to have an issue, you may be able to rule out elements which are common to all output tubes (like the overall shared bias supply). That might lead you to look at the grid resistors or the coupling caps feeding the output tubes (because those are unique to each tube/each pair). If you measure correct negative voltage at every socket (which seems to verify bias supply, coupling caps and grid resistors), then maybe look to see if the sockets themselves are holding the tubes tightly. If yes, then it would seem time to pop tubes in the amp and give it a listen to see what you note about its operation.

Open question to others with more experience on 6550 amps: the data sheet usually says "50kΩ grid resistance" for a fixed-bias 6550. Any real-world issues with using 100kΩ to feed bias to a pair (instead of the more typical ~33kΩ)?

[sluckey]

the data sheet usually says "50kΩ grid resistance" for a fixed-bias 6550. Any real-world issues with using 100kΩ to feed bias to a pair (instead of the more typical ~33kΩ)?

My only hands on experience with 6550 amps has been four old Sunns. They all used 100K grid resistors. I've looked at most all the old Sunn 6550 schematics and they all use 100K grid resistors, even the big 4 X 6550 amps.

I don't recall seeing any 6550 amp schematics that used a 33K grid resistor. I didn't know that was a typical value?

[SupLar]

I performed the sequence of checks Hot Blue Plates suggested (thank you so much). No shorts in the 6550 tubes and normal voltage readings everywhere I checked inside the amp. I popped the tubes back in and fired her up - and the amp is working. The only components that are different from before are the new set of diodes I installed as I mentioned earlier.
I'm relieved but still a bit mystified.
Are the 6550s a little wiggly in their sockets? Yeah, they kinda are. I will install retainer clips to hold them in better.  Excessive dust?  Not really.  No visible signs of arcing (I have pins 3,4,5 shrink tubed at each 6550) or scorched components (except the nasty scorch mark inside the head cabinet near one 6550). 
The tubes themselves are showing their age and I'm going to replace them. Maybe new filter caps, and phase inverter coupling caps, wouldn't be a bad idea either?

[sluckey]

The schematics provided indicated 100K resistors are being used in R22 and R23.  I saw a 33k resistor at R19, please confirm amp contains 33K resistors at R22 and R23.

why?

[6G6]

Glad it seeems to have worked out.
I'd retension the sockets, just because.

[SupLar]

R22 and R23 are both 100K, not 33K.  R18 and R19 are both 47K and not 27K and 33K respectively as shown on the schematic. 

[sluckey]

Questions were raised by posters in this thread regarding 33K and 100K resistors in the power tube  portion of the amplifier.  By confirming the values of the resistors,  we know amp tinkering has not taken place, at least not in these locations.  Just eliminating variables in troubleshooting.

That was a side bar generic question not related directly to this thread.

[Ritchie200]

What op tranny are you using on this?  Note sure what your friend is using between bass and amp but UL designs are not happy being pushed with distorted high gain inputs.  You end up with a TV flyback type condition with some very big voltage spikes - typically 3X B+.  You mentioned evidence of an arc near a tube but no tracking at the socket?  It had to have come from somewhere?  How do you know a red plate condition occurred?   Usually they will arc at an output tube socket between pins 2 and 3.  It can happen above or below the socket, outside or INSIDE the tube base.  Eventually the output tranny breaks down and you have arcing in the windings.  In the Major we call it the Smoke on the Water effect.  It's like falling down a hill...  Dime it, play through it hard and see what it does.  I don't think the problem is fixed.

Jim

[SupLar]

Hi Jim,
I apologize for my late response. I am just seeing your post now. The output Transformer is a Heyboer, 200 watt RMS, primary impedance is 2K with 40% UL screen taps, secondary impedance 8R, 4R, 2R, frequency response is 20Hz to 20kH. It is as big as an infant's head.  (The entire amp weighs 75 lbs +)
This bass player might occasionally use a distorted high gain pedal for selected sounds during their song set. Not 100% sure but I will ask him to clarify/confirm this.  (I don't think he dimes the amp - he uses the line-out at every gig).
I haven't heard of the possibility of HV spikes with UL amps that you bring up, but I don't have any past experience with UL platforms prior to building this amp. Thank you for bringing this to my attention.
The reason I suspect a red plate condition is that the rear panel (the rear panel of the head cabinet attached behind the power tubes) is melted (the head cabinet is clear plexiglass) - but only melted directly behind one tube (I can post a photo for clarification). I am going by circumstantial evidence, since I didn't see the occurrence, and I could be wrong about the red plating. Are you saying the OT could be damaged now or it eventually will arc between the windings if the problem isn't fixed?  A damaged OT would be very bad.
I have pins 3, 4, and 5 shrink tubed at the pins (but not pins 2 and 7) on all 6550s. 
The 6550s don't have tube retainers (not yet anyway) and it is possible that one tube was slightly popped up enough (exposing the pins above the base) for the arcing that you are have described (can this happen with all tubes snug in their sockets?).
You might be right.  The problem might not be fixed. This is staring to keep me up at night.  I can play it hard and wide open for awhile to see what happens...
Thanks!!

[SupLar]

Here is an update on the 6550 quad UL bass amp.  I installed these parts earlier in the week:

Replaced the power tube screen resistors to resistors rated for 5 watts (instead of 2 watts).
Replaced the phase inverter coupling capacitors.
Added tube retainers for the power tubes to keep them snug in their sockets.
Replaced all of the diodes in the power supply (1000V, 3A)
Replace capacitors in the bias supply circuit and the diode (1000V, 1A)

I have a bass playing friend who lives nearby and I had him take the amp home to play it for a few days.  I told him to play it hard. He said it worked great with no issues.

Today I installed a new matched quad set of Tung-Sol 6550s.  As soon as I turned the amp on the new tubes red plated immediately. I quickly powered down the amp when I realized what was happening.  It caught me off guard!  I went upstairs and made a cup of coffee while the amp cooled down. Upon my return I removed the power tubes and checked the bias voltage. It was right on (-50V to -65V depending on bias pot). B+ voltages, plate, screen, and bias mA all read as expected. Nothing stood out as an oddball reading. Hmmm.

I put the old power tubes back in and fired the amp back up (slowly this time using a variac).  No issues. I put the new tubes back in and fired her up again using the variac. No issues.  The dreaded word: intermittent.

Thoughts?

One clue: I'm 99.9% sure all the tubes (not just one) were red plating - maybe 2 of the 4 (on the same side) more than the others.  Maybe one tube more than the others.  Like I said, it happened so fast and I shut the amp down right away. I think this is a good clue...

Should I replace the tube sockets and double check all my wiring, and solder joints, going to the power tubes? Seems like a good idea since I can't think of anything else. (After Jim's post earlier I am a little worried about the OT.) 

Any ideas or suggestions would be appreciated.

Thanks so much! 
Larry

[PRR]

Before you rip-out, get a BRIGHT light and LOOK.

Yogi says you can learn a lot by looking.

Electronics joints should be Inspectable. Either the solder clearly WETS both the lug and the wire, or you put a mark on it and send it back to the solderer to re-do RIGHT.

If you think it was all-four tubes, it may not be the sockets, but the Bias Supply.

[Ritchie200]

THe Heyboer has been used in quite a few Marshall Major repairs.  This amp beats them up pretty good and I have not heard a bad word yet.  I would look elsewhere.

The docdroid site just sits and grinds when I try to look at your schematic.  What size are your screen resistors?  I'm surprised the 2 watt lasted - or did they?! 

Jim

[SupLar]

Screen resistors are 47R but I have 50R (5 watt) installed for both screen and plate resistors. It certainly makes sense that the bias circuit could be the culprit.  I've replaced both of the 16uf caps (and diode) already. 

Here is the PSU and PA schematic and layout:

[SupLar]

The amp work is done.  I finished up tonight with the installation of the new power tube sockets. I also rebuilt the power tube circuit board.  This board board keeps plate, grid, screen resistors orderly and neat, and makes taking voltage readings really easy.  I've either replaced, or rebuilt, almost every component that is related to the bias circuit. The amp is working perfectly (at least in my shop) and all readings are normal. It thunders.  The new Tung Sol tubes seem like they're well made. I'm going to deliver the amp back to the owner and hope for the best. Thank you everyone who contributed to this post!  I really, really appreciate it.

[Ritchie200]

That is great news!  Did you happen to check the parts you replaced to see if they had failed?  It would be nice to put your finger on the offender!  Also, amp porn is always appreciated around here and this was certainly an ambitious project.  Well done!

Jim