EL34 Rockster Help

[dorrisant]

Hey guys,
I just finished this new build... I'm having an issue and would love to get some input.
This is what I've got:
330-0-330 PT
GZ34 Rectifier
Hammond 158L Choke
Edcor 0720 OT
3-12ax7's - JJs
2-EL34s - ? (Radio Shack)
Also, I used two JJ can caps on this build... 500v rated. So the following values are slightly different:
C 17 & C 18 total 70 uf
C 16 = 50 uf
C 15 = 40 uf
C 13 & C 14 = 20 uf
R 26 = 16.4 k
R 24 & R 25 = 8.2 k

I went through the startup procedure with a 25 watt bulb... Two minutes after turning it on I switched it off standby. Bulb reacted just as expected. Same thing with a 60 watt bulb... No problems. Checked all voltages:
V1B = 461
V1A = 466
VB2 = 472
V3A = 468
V3B = 468
V4 Pin 3 & 4 = 475
V5 Pin 3 & 4 = 475
Cathode bias was set to 250 ohms... It is adjustable.
Put in all tubes and turned on power, still through light bulb and 16 ohm speakers connected for load. All filaments looked just fine. Left it this way for about five minutes, then switched it off standby. Bulb reacted normally again. Rechecked all voltages. Much lower this time around but looked like what I expected, so I unplugged the startup rig and went straight to the outlet. Turned the amp on for a couple of minutes, then switched it off standby with guitar plugged in and all controls turned all the way down.
I could hear a very slight hum from the speaker cab... so slight I had to put my ear inches from the speaker to actually verify it. I sat down, picked up the guitar, turned up the volume and a few seconds later saw a little smoke and noticed an ammonia smell. I quickly shut it off and looked at the power tubes almost simultaneously... Didn't see any red plating and the smoke was so faint I couldn't tell exactly where it was coming from. It was around the general area of the 50/50 can cap. I was getting tired so I unplugged it and went to bed.
This morning I did some checks... The choke seems to be fine as does the first filter. I still get 475 at the tube socket with the amp in standby. I'm pretty sure the smell was the 50/50 cap giving up the ghost. Also before my test, I had tested 3 EL34s that I had on hand. Two of them had very similar GM, one was very suspect... Somehow I got them mixed up and put the bad (?) one in with a good one. I'm sure this has never happened to anyone else. I don't know that the suspect tube is what might have killed the cap... but I guess it could have...
Does anyone care to take a look and make any suggestions. I tried to lay this out so it would be easy to trace. I can supply more pictures and info upon request.
Thanks for looking,
Tony
Ps: I used rj's kt66 rockster for the schematic.

[IMG768]http://i946.photobucket.com/albums/ad30 ... 61141_.jpg[/img]

[IMG768]http://i946.photobucket.com/albums/ad30 ... 61142_.jpg[/img]

[dorrisant]

The adjustable cathode bias...
I have pin 1 and pin 8 tied at both tubes. I have a jumper installed between both 8 pins. Two things tied to this jumper:
1- 100ohm 10watt cathode resistor in series with 300ohm 5watt pot tied to 1ohm 10watt bias point resistor, (gold anodized in image) then to chassis ground.
2 - 220uf 50 volt bypass cap tied to the same bias point resistor.
So the bypass cap bypasses both the static resistor and the pot for a range of about 100 to 400 ohms. I set it for 250ohms for startup.
I did kind of question my technique here when I read page 25 of the express build manual... Something about disconnecting pin 1 from pin 8 in order to get your bias... Does that apply here? This is cathode bias and the express is fixed... I only wish I would have taken a bias reading somewhere along the line... Duh! Lesson learned.
Anyway, just thought I should add that. It does deviate a bit from rj's schematic a little, but I wanted to be able to swap in some other power tubes and adjust the flavor.

Thanks,
Tony

[dartanion]

What are your operating voltages with tubes installed? What you have currently seems REALLY high for a Rocket, even with EL34s.

[dorrisant]

Dartanion,
I wrote them down but can't seem to find the piece of paper. Should have written them in my composition book.
Mark Fowler has a similar configuration here:
https://tubeamparchive.com/viewtopic.ph ... tart=0http
I even asked him about the "higher than rocket" voltages. No offense to Mark... I just figured I wired something wrong. New cap is on the way, but I'm kinda gun-shy about putting a new cap in there yet.
Maybe I could put some diodes on the CT to drop the B+ down before I try it again?
What would you suggest for the B+?

Tony

[ampgeek]

Tony,
No need to drop the B+ before moving on to see if the amp functions.

The voltages that you have posted are without tubes. Right? They will drop down once you get the tubes in there.

I would focus on what burned for the time being. You mentioned smoke in proximity to the JJ cap can. Did a close inspection in that area indicate that anything was awry?

You also mentioned that you had tested the EL34s prior to installation into this amp. Can you re-test the same way to sort them out again?

Any reason why you wouldn't do that, install them and fire it up keeping a close eye on the suspect area along with the normal things (e.g., redplating, overheating PT, etc..etc..)? Probably ought to start over again with light bulb limiter as well just to be darned sure.

I would keep my meter across the 1ohm bias measuring resistor(s) at all times until you are confident that the power tubes are not biased too hot.

Sometimes you just have to run until an all out componant failure occurs to know exactly where the problem is (and...do your best to make sure that it isn't an expensive one!). Otherwise, you could end up hunting and pecking with guesses that, in the end, are more frustrating/expensive than replacing a cheap, burned up resistor or similar.

Good luck,
Dave O.

[dorrisant]

Dave O,
Yes, the voltages I listed are with the tubes out... I remeasured the voltages with the tubes in, but I can't find them. They were lower, as they should be, nothing to be alarmed about though. As a matter of fact, I used the original PT and rectifier tube that powered the pair of EL34s that were in it for years. I didn't see any reason why it wouldn't work.
Yes I did retest the tubes this morning... I have my own tester, so its too easy to eliminate any question. That's how I noticed that I had mistakenly put in a weak tube. They are all sorted out now. I just should have known better than to try to finish this so late at night. Should have went to bed and tried it with some fresh brain cells in the morning.
I usually have a cheaper meter just for monitoring the bias while I take the other measurements with my go to meter. Didn't have it... Comedy of errors!
I'm pretty sure at least of the cap cans is toast, maybe both. I will pull them loose and check. I don't see anything suspect at all. The cap doesn't even look that bad. Maybe I didn't let it cook long enough to leave visable signs, but I sure smelled that familiar smell of a cooked cap... Maybe rare instead of well done?
I will post what I find tomorrow. Thanks for the responses so far.

Tony

[ampgeek]

LOL! Rare vs well done. Love it!!

I am at a bit of a loss as to what could cause a (new?) cap can(s) to fail that wouldn't have shown up during the light bulb test.

I have used many (50'ish) JJ cap cans over the last few years and never had a single problem. Or..any other large filter cap for that matter. Of course, that probably doesn't represent a hill of beans statistically.

Hoping that somebody with more experience will weigh in and share their thoughts as I have always wondered about large cap failure modes.

May I ask where/when you bought it?

Cheers,
Dave O.

[dorrisant]

I got them from tubedepot.com... but like I said, I'll probably find something that I did wrong. Though, I don't recall getting through the light bulb procedure and having a failure like that. I will check these caps tomorrow and order new ones as needed. Probably rewire the whole area just to be sure. May find something then. I will continue to post what I find.

Tony

[Cliff Schecht]

Not trying to be rude here but there are a few things that you have done lead-dress wise that may be at fault. I've never used the JJ can caps but have read only good things about them. Furthermore I've used many NOS and used old can caps and have yet to have one go bad on me. Granted these get tested and reformed before I shove them in an amp (at least nowadays ) but you're using brand new stuff that shouldn't be giving you problems.

I'm looking at the way you wired up the power section with the double-twisted wires. This seems like a bad idea because the smaller gauge wires insulation is probably only rated for 300V. If that's the case then any nicks in the insulation could be causing something to arc. Maybe this is far fetched, but it's something that comes to mind..

Also running the transformer wires through the chassis holes with no grommets is a no-no. I used to do this too but now I'm weary about doing so after it eventually wore the insulation down on multiple transformers and a choke (luckily nothing of value). If you move the amp around those wires rub the chassis, especially in your case with the choke where the wires barely clear the hole.. You might need to drill the holes a bit bigger and/or put some grommets in there.

Last thing is the amount of parallel wire runs you have in there is...excessive. Just around the preamp you have all of the wires running long and bunched together. I try to avoid this, plates with wide voltage swings can couple into the cathodes and screw with your biasing (unwanted negative feedback). The length of the wires that go from the can caps to the plate resistors are also really long. This can make the decoupling ineffective and the stages will interact in bad ways.

I hope you don't take this offensively, I'm not trying to insult your build. I figured you wanted some constructive criticism and this is what I saw.

[Ian444]

dorrisant, can you check the can caps to be sure they are faulty? At this stage, there seems to be no real evidence. How is the 220uF 50V cathode cap, is it OK? And the 300 ohm 5W pot in the cathode circuit, is it still OK? I would be inclined to check and inspect components as far as possible, then power it up again with good tubes in it. 70uF straight up for a GZ34 is pushing it. I have read its good practice to take the B+ feed from pin 8 of the rectifier. Over and out for now

[M Fowler]

Do the bleeder resistors get really hot and start to burn up? I would suspect that before bad can caps.

Mark

[dorrisant]

Cliff,
No rudeness taken... I'm glad you took a look and have all the reason to look forward to your comments. Yeah, it may deflate the ego a bit, but its worth it to get this thing going.
As far as the wire, it is 600v rated. It is small wire so I was doubling for current capacity only. It was just what I had on hand and I'm still waiting for some spools of some good stuff on the way. This amp was really a challenge from a couple friends of mine. "What could you do with this piece of $@%#?..." I was trying to get it done with what I had on hand and spend the least possible.
When I pulled the caps out, I couldn't see any signs of nicks, cuts or arcing. Since it is halfway disconnected and I agree that it could be better, I'm going to rewire what you have mentioned. Watch for a future post with updated pics. Also, I have grommets around here somewhere... once I get my new bench setup, I will find them and install.

Ian444,
I checked the cap cans, the bypass cap and readjusted the cathode pot. Everything looks fine... I have an ESR meter and they looked good on that as well. I'm going to recheck all the components as I rewire, so maybe I will find something then.

Mark,
Nothing was getting hot at all, except for the tubes... But just normal heat. I could have sworn that I blew the 50/50 cap but I really can't find anything wrong with either of them.
Could it have just been a bad tube? Idk... Oh well, I will just rewire some and check as I go and it probably should work out.

Thanks again guys for all your time and consideration.
Tony

[M Fowler]

Does the amp work? If your able to take voltage readings I would assume your able to plug a guitar in and play as well?

Put a set of power tubes in and watch carefully for problems.

Mark

[soma_hero]

There could be several issues here. The bleeder resistors could be getting warm and burning a little. I don't see any reason the can cap should have gone up. If you were able to measure voltages then the cap was stable right? And I have used JJ's in almost every amp build and never had a problem.

You might have had a little flux or dirt on a solder joint that burned up too. Or all the current going through the heater windings could have heated up some insulation....

I'd fire it up and watch closely.

[dorrisant]

Ok...
I put the caps back in and rewired the whole power supply, installed the grommets and checked everything. I turned the bias all the way down. Went through the whole light bulb startup with 20, 60 and 75watt bulbs. Everything was just fine... pulled the bulb out of the circuit and turned it on. Left it on, in standby for about 15 minutes, wrote down voltages. No problems. Took it out of standby and watched everything like a hawk. Then I saw it. Smoke ot of the cathode bypass cap. It was blowing a small jet of smoke right over top of the cap can I had been questioning. I shut it down, drained the B+ and pulled it out. Just enough smoke to show the problem. The voltage rating must not be high enough at 50v I guess.
Nothing ever got too hot and the amp was working... I could hear it amplifying the signal when I was probing for voltage at the input stage... I know this may sound stupid to some, but the sound from probing with my DMM sounded right. So I know it is working, just need to up the rating on a new cap there. I have a 350v cap, but it is too large physically. Could clip-lead it in so I could see if that voltage was too high. I will hit it again tomorrow.
Thanks,
Tony