JTM45 - Broken power tubes?

[jwarlock]

Hello guys, my name is José and recently I was modifying/rewiring my old JTM45 build in order to take out some annoying hum noises that (I think) were product of a faulty build when I was a teenager. The next 3 paragraphs serve as an introduction/context, for the juicy stuff jump to the voltage charts at the end.

Years and years have passed, not playing a single note on my amp, mainly because it’s too loud for my room/house, but last week I felt the need to finish that amp.

I'm attaching the layout that I’ve used (I’ve made the amp around 2010, so this is a newer layout with very minor component modifications), it has a ½ power switch on the power tubes that I found very interesting. All measurements have been made with this switch on the full power position.

When I was first trying to measure voltages, I’ve made a terrible mistake, connecting all tubes and turning on the power and stand by switch without plugging the speakersl, then I’ve seen some weird lights and heard some noises on the power tubes, I turned off the amp immediately.

I’ve checked 7 times all the connections, looking for possible short circuits, and everything seems to be ok. The following voltages are DC unless otherwise noted. All control pots are turned to min, and bias pot is turned to the most negative voltage (-61V) at the end of the 56k next to that pot.

After plugging the KT66 tubes, when measuring the bias it reads 5/6 mv, and it does not change if I turn the bias pot.

I suspect that the power tubes are broken. If you guys need me to take more measurements I’ll take them, I’m even willing to take measurements of the whole board if you think it’s necessary!.

Any help will be most welcome and appreciated, thanks in advance guys!.

[pdf64]

You may have wiring errors around the rectifier / standby switch / reservoir can cap.
Good photos of the chassis and on those areas would be helpful.

[jwarlock]

Thanks for the reply!

Here are some photos, sorry for the mess, originally all cables were soldered under the board, but in this rewiring (after checking several times) I just thought that it would be better soldering on top of the board, it just helps to visually check everything. Once I have everything running, I’ll make things pretty :)

2 clarifications. As you can see, I have a 16+16 cap, one end is going to the 8.2k (as per layout), and the other is replacing the axial capacitor that would have been installed on the board connecting to pin 6 of V2. Another thing, my indicator is a 220v one since I couldn’t find a 6.3v for the filaments, and is connected to the ON switch.

[bmccowan]

You are right that things are way off! I suggest treating this like a new build and going through a systematic test procedure like the one suggested here on Doug's site. Do you have a variac or a lightbulb limiter. If not, I suggest making up the latter so that your tests do not fry something valuable. I also suggest rebuilding it to known successful layouts, as what I see in the pics is a bit confusing.

[pdf64]

700VAC at the rectifier anodes indicates that the HT winding’s CT may not be connected to 0V chassis common.
Can you investigate how voltage is getting to the power valve anodes whilst in standby? And how’s it getting to the anodes but not the screen grids?

[Latole]

All solder look very bad.
If I was you, I'll practice solder on olds pieces of scrap.

[jwarlock]

I'll definitelly try to improve my soldering skills!, I know it looks very ugly :(.

I've made some further measurements, I don't know if this is what you asked me pdf64. The first thing that I did was check again if I've made some mistake in wiring the std by. Everything is fine there. Then I've removed every single tube, and did some measurements without load.

• Measured the voltages on the rectifier tube socket, I think it looks fine (rectifierSocketVoltages.png).
• After that, I've plugged the GZ34 and measured again, remember that no other tube are plugged. A voltage chart available online (I'm attaching that too) says that I should be getting 400 VDC on pin 2-8, I've got 462 VDC. And 320 VAC on pin 4-6, I've got 464 VAC. Maybe my readings are fine if we account that I have no other tubes connected, I don't really know.
• When reading the voltages coming to the power tube sockets (kt66 unplugged) I have around 70 VDC more than the reference of 400 on pin 3, pin 4-6 are very strange since I've got nothing.

[2deaf]

Taken at face value, the original voltages posted are not within the realm of possibility.  You state that the standby is "ON" which must mean closed because there are voltages on down the line.  I would prefer that you state whether the standby is open or closed as opposed to on or off so that I don't have to play detective.

Starting with 316Vdc then going through a choke and two dropping resistors to end up with 335Vdc isn't going to happen.  Measuring 316Vdc at the rectifier with the standby open and then measuring 335Vdc down the line with the standby closed could happen.

The DC voltage at the rectifier doesn't change when the load changes, which isn't going to happen.  It will if the standby is open, however.

350Vac on the transformer and only 316Vdc at the rectifier isn't going to happen.  Unless the standby is open and there is no smoothing capacitor and your meter interprets it to be 316Vdc.

The voltage drops across the OT primary with tubes installed isn't going to happen with a 500mA fuse in the line. 

What meter do you have that can measure over 1000Vac?   

[pdf64]

To avoid confusion, just eliminate the standby switch, eg link its terminals together or connect the wires together at one of the terminals, or whatever.

[DougGuy]

I have a Metroamps kit JMP 50w model 1986 bass head that after a couple of years in constant use, started grenading one of the two power tubes and it turned out to be a cold solder joint on one of the speaker jacks!  After sucking solder off and reflowing all of the joints on the speaker jacks, the problem is fixed and never happened again.  To look at the workmanship in this amp is to appreciate excellence, and to look at the solder joints on the speaker jacks they looked perfect.

[jwarlock]

Taken at face value, the original voltages posted are not within the realm of possibility.  You state that the standby is "ON" which must mean closed because there are voltages on down the line.  I would prefer that you state whether the standby is open or closed as opposed to on or off so that I don't have to play detective.

Starting with 316Vdc then going through a choke and two dropping resistors to end up with 335Vdc isn't going to happen.  Measuring 316Vdc at the rectifier with the standby open and then measuring 335Vdc down the line with the standby closed could happen.

The DC voltage at the rectifier doesn't change when the load changes, which isn't going to happen.  It will if the standby is open, however.

350Vac on the transformer and only 316Vdc at the rectifier isn't going to happen.  Unless the standby is open and there is no smoothing capacitor and your meter interprets it to be 316Vdc.

The voltage drops across the OT primary with tubes installed isn't going to happen with a 500mA fuse in the line. 

What meter do you have that can measure over 1000Vac?

What I mean by standby switch "on", is that it's on the closed position, sorry for the ambiguity.

[2deaf]

The only AC voltages that should be on your charts are pins 5 and 7 of the GZ34 and the heaters of the other tubes.  Pins 2 and 8 of the GZ34 and all of the non-heater pins of the other tubes should only be DC on your charts.

Keep the standby switch closed for all measurements.

[2deaf]

... pin 4-6 are very strange since I've got nothing.

Is that turquoise wire connecting the two top lugs of the 1/2 power switch connected to the 8.2K dropping resistor (with or without the 1K resistor)?

If you switch to half power, is there voltage on pins 4/6?

[jwarlock]

... pin 4-6 are very strange since I've got nothing.

Is that turquoise wire connecting the two top lugs of the 1/2 power switch connected to the 8.2K dropping resistor (with or without the 1K resistor)?

If you switch to half power, is there voltage on pins 4/6?

Thank you for pointing that out, I dont know how I missed that in all the checks that I've done, thanks!

After soldering that 1k resistor I've taken those measurements again. GZ34 is the only tube connected, standby switch is closed, "half power" switch is on the full power position, and the amp is without load.

Edit.
I tried to measure everything with all tubes plugged. Out of nothing I decided to test the cable that I was using to connect the speakers (an old Fender, about the age of my amp, 10 years), it was broken... So the times that I've taken readings with power tubes on, were without load.
Put another cable on, and when I power on the amp (with standby switch closed) to take a reading of the power tube voltages, I hear a "hum" (not too loud, but not low either), and about 5 seconds later the 500mA fuse blows.

[2deaf]

Put another cable on, and when I power on the amp (with standby switch closed) to take a reading of the power tube voltages, I hear a "hum" (not too loud, but not low either), and about 5 seconds later the 500mA fuse blows.

The prime suspect would be a KT66 or two because it blew the fuse with them in but it didn't blow the fuse without them.  A KT66 can draw a lot of current if the bias isn't sufficient or if the KT66 itself is shorted to some degree.     

[jwarlock]

I think i'm going to buy a paired match of TAD KT66, I can get them from tubesAndMore for just under 65 bucks. From what I've read they're maybe better than the JJ that I have.
For a last check, I've measured the resistance of the OT primary and it reads 170 ohms, I don't think it's broken. What else could I do to make sure that I dont need to buy another "unusual" component aside from the power tubes, and by unusual I mean Tubes, transformers and choke.

Thanks for all the help that you've provided me guys.