Smoke while testing Crack OTL

[AtHeartEngineer]

So, to start, I'm pretty sure the transformer is dead, but I don't know what pins to test for continuity.

The Story: My buddy bought a Crack OTL a few weeks ago, and the last few days we have been spending an hour or two putting it together. He doesn't have a whole lot of soldering experience, and I have a lot of experience with digital circuits. AKA: I am not an analog guy, and while I LOVE AUDIO, I don't understand a lot of the analog world...give me an OPAMP, I am good to go...but I know those are frowned upon here (for good reason  ).

While we were building it, I was checking all of the solder joints and everything was going well. We finished, did all the resistance checks and everything checked out. Plugged it in, heard a hum, the large vacuum tube glowed but the small one didn't. I figured this was ok, a little hum wasn't alarming at the time. About 30 seconds in we test pin 1s voltage...its way off from what it should be, and I start seeing smoke. We immediately unplugged it and started looking at everything carefully.

It turns out, one of the diodes was put in backwards (shit...) and the transformer was...well I could probably cook an egg.

We de-soldered and resoldered that diode and started meticulously checking everything else. I think everything is good now, but we still aren't getting the correct readings, so I am pretty sure I killed the transformer. How should I check this? Where do I order a new one? Could it be something else?

I would rewire the transformer, but that is neither time friendly, nor fun.

Thank you guys for the help,

Tyler.

[kgoss]

Did you test the backwards diode before installing it correctly?  The transformer is pretty rugged and the smoke was most likely from an overheating resistor.  Don't worry, the guys at Bottlehead will help you troubleshoot the problem.
The first thing to do would be a full resistance check and report any terminals that are off. You don't want to apply power until the resistance test is passed.

[AtHeartEngineer]

The resistance check was passed both before and after the little incident.

We did test the diode and I am pretty sure it is good, continuity one way but not the other.

I normally would say its probably not the transformer, but it was a lot more than warm to the touch and I think that is where the smoke started coming from. It smoked for about 500 milliseconds before I unplugged it...thankfully I learned at a young age to have that reflex.

Thanks!!

[Paul Joppa]

If you didn't blow the fuse, you probably didn't overheat the transformer - it has a large safety margin. To check, the primary DC resistance is ~13 ohms, HV secondaries in parallel ~38 ohms, filament power ~0.1 ohms. Check the resistance of the large resistors, 270 and 3000 ohms, those are the likeliest ones. Also check the orientation of all electrolytic caps - I know you've done this more than once by now; do it again (and inspect for bulges).

I'll let the Crack experts take it from there.

[Paul Birkeland]

You could also post what your new voltages actually are at this point. 

Measuring the resistances of the power transformer windings is a good idea, but be sure to touch your meter probes together first to get a baseline for what your meter will actually read as 0 Ohms.  For example, if you get 0.2 Ohms with your probes touching, then measure the 6.3V winding, expect to see 0.2-0.3 Ohms. 

If you want to test the AC voltages of the power transformer, unplug both tubes, then remove either the black or the red wire that goes to the diode bridge of UF4007's from the power transformer.

Now, turn the amp on and you can measure the AC voltages according to the sticker on the transformer.  (Black probe on the "0" terminal for each AC pair)

The diode that was put in backwards should be replaced.  A 1N4007, 1N4006, 1N4005, or 1N4004 would do in a pinch.

I believe would be the first person to destroy one of our power transformer from a backwards diode, but certainly not the first person to install a diode backwards.

-PB

[muidem]

Hey guys!

I'm the buddy who actually owns the kit. Last night we got some work done and redid several of the components hoping the problem would clear up. Unfortunately, we're still seeing weird voltage numbers. First, I'll go over what we did last night and then the voltage numbers. Just in case someone is wondering I'm using a Fluke 113 True RMS Multimeter.

First we pulled all of the electrolytic caps and the four diodes off. We tested all of the caps and they were within range. We replaced all of the diodes with 3x 1N4007 and 1x 1N4004 (ran out of the 4007s) that we had handy. Then we soldered the caps and diodes back into place.

Here are the measurements from the caps:

Cap# | Rated | Actual
Cap1    220 uf   206 uf
Cap2    220 uf   205 uf
Cap3    220 uf   204 uf
Cap4    100 uf   92.6 uf
Cap5    100 uf   91.8 uf

Next, we pulled off the 270k 1W metal film resistor and tested it -- it came back good also. The 270 5W wirewound and the 22.1k metal film resistors were tested in place and came back good. I was unable to test the 3K wirewound resistors as my meter doesn't go that high but they seem to be intact upon visual inspection (no bulging, discoloration, scorch marks etc.).

We put everything back on the board (diodes in the correct orientation this time) and ran through the resistance check which we passed. Then we decided to go for the voltage check again. The hum is still gone, which is good and no excessive heat from the transformer, which is also good. The tubes don't light up very much at all (is that normal?). I can see the elements glowing in the larger tube but I barely see _any_ glow from the smaller tube.

Here are the voltage measurements. They correspond to the order of the first ten items in the voltage check section of the manual:

Point | Expected | Actual
1.        75-90v       25-28v
2.        170v          145v
3.        0v              0v
4.        170v          145v
5.        75-90v       30v
6.        0v              .9v
7.        100v          90v   
8.        0v              0v
9.        100v          91v
10.      0v              .9v


At this point we decided to stop testing any further because the voltages were clearly not where they were supposed to be and we became worried that we could damage some of the components by leaving it on too long. The .9v that should be 0v was particularly worrisome to us.

I'll try to get some pictures up soon in case that will help. Thanks for the help so far guys!

Edit: Here's an album of what it looks like -> http://imgur.com/a/WkEXN/layout/horizontal#0

[Paul Birkeland]

This is very odd.  Can you retest the terminal 1-5 voltages with just a 12AU7 in the amp?

With both tubes in the amp, can you measure the power supply voltages?  Of interest would be 21, 13, and 15.

The 0.9V on T9 can be ignored.

I'll have a better grasp on recommendations with that information.

-PB

[muidem]

Can you retest the terminal 1-5 voltages with just a 12AU7 in the amp?

1. 30V (LED at A8 goes out when I touch my probe to T1)
2. Starts at ~185V and rises to 220V
3. 0V
4. Starts at ~180V and rises to 220V
5. 45V (LED at A3 goes out when I touch my probe to T5)


The exact voltages of 2 and 4 were somewhat difficult to determine because they climb (that seems wrong). I measured three times and unplugged and let it sit for 10 minutes between each attempt. The "starting" voltages for T2 and T4 are the averages of those measurements.

With both tubes in the amp, can you measure the power supply voltages?  Of interest would be 21, 13, and 15.

21. 215V
13. 169V
15. 191.9V

[Paul Birkeland]

Hmm, this is interesting.  Can you be absolutely sure that your 9 pin tube says 12AU7 (or something equivalent on it, like 5814) on it?

What do you get for voltages at A2 and A7? 

We should get all this trimmed up before worry about running the 6080 in the amp.

-PB

[muidem]

Can you be absolutely sure that your 9 pin tube says 12AU7 (or something equivalent on it, like 5814) on it?

The smaller tube has 12AU7A printed on it near the top.

What do you get for voltages at A2 and A7? 

These were taken only the smaller tube (12AU7A) in the amp.

A2. 0v
A7. 0v

[Paul Birkeland]

1. 30V (LED at A8 goes out when I touch my probe to T1)
5. 45V (LED at A3 goes out when I touch my probe to T5)

This generally indicates a loose connection in the amplifier.  Can you poke around with a wooden object to see where you need to poke to make the lights go out?

[muidem]

This generally indicates a loose connection in the amplifier.  Can you poke around with a wooden object to see where you need to poke to make the lights go out?

I poked around points 1 and 5 with a wooden dowel and was unable to get the LEDs to turn off.

I think I have isolated the problem to my multimeter. To test, I probed point 1 and 5 with my meter (chassis as a ground). They both read 30V and one of the LEDs will go out as I said above. However, when I probe the same points with my meter off the LEDs stay on. I think my meter is doing something strange or smarter than I am. For reference, I have the meter linked below:

http://www.amazon.com/Fluke-True-RMS-Multimeter-Backlight-Alkaline/dp/B004Q05ODI/ref=sr_1_1?ie=UTF8&qid=1414858694&sr=8-1&keywords=FLUKE+113+true+rms+multimeter

[Paul Birkeland]

The Fluke 113 is not the problem.

The solder joint on the center post of the 9 pin socket looked like it could've been a little flaky. 

I would also poke around at more than just T1/T5. 

-PB

[muidem]

I just had a buddy bring over his multimeter and we ran through all the resistance and voltage checks again. We nailed the resistance checks and all of the voltages were within 5V of what they should be. We decided to do the ring test and it rose to 9 and then dropped back to zero as predicted.

I went ahead and plugged it in to a source and turned it on. It sounds great!

I do have two small issues that I'm curious about though.

1. It's very light on the bottom end, no real bass whatsoever. I think this might be due to my headphones. I currently have the MTH50s which have an impedance of only 38 ohms. I'm planning on picking up some Sennheiser HD600 or 650s unless you guys have a better recommendation?

2. There's a slight whine/buzz that I can hear through the headphones when nothing is playing. It gets louder as I turn up the volume. Are there any quick things I can do to try to clear that up?

[Paul Birkeland]

1. It's very light on the bottom end, no real bass whatsoever. I think this might be due to my headphones. I currently have the MTH50s which have an impedance of only 38 ohms. I'm planning on picking up some Sennheiser HD600 or 650s unless you guys have a better recommendation?

Yes, 38 Ohm headphones will be very rolled off on the bottom.

2. There's a slight whine/buzz that I can hear through the headphones when nothing is playing. It gets louder as I turn up the volume. Are there any quick things I can do to try to clear that up?

I'd wait till you have 100+ Ohm headphones in your hands, then listen to a track, pause, and listen for background noise. 

The amp will have some rush/hum with the volume knob all the way up into some headphones, which is well above any useful listening level.

-PB