Speedball Upgrade Smoking!

[iskess]

After using (and loving) my Bottlehead Crack for a year with HD 650s, I decided on the Speedball upgrade.
The stock Crack would occasionally begin making “dying squirrel” noises which could easily be remedied by quickly turning it off and on again. I’ve always assumed this is just a tube irregularity but thought I’d mention it here in case it serves as a clue.

Anyhow, here’s the actual subject of this post:

After a successful “small circuit board” voltage check, I assembled and installed the “large circuit board” so as to complete the Speedball upgrade for my Bottlehead Crack. It passed both resistance checks, and then came time for voltage tests. In total I applied power to the amplifier three times in an attempt to conduct the voltage tests, and each time an alarming issue would arise. Provided below is a timeline of said issues:

1st time power applied (not caught on video):
Sparks and smoking between the “base” and “emitter” legs of the TIP-50 transistor in the Q2A socket after ~30 seconds. Assuming this was due to flux arcing from improper cleaning between the transistor legs, I wiped the area thoroughly with rubbing alcohol and decided to retry the test.

2nd time power applied (not caught on video):
Sparks and smoking between two of the legs (not sure which) of the TIP-50 transistor in the Q2B socket (NOT the Q2A socket) after ~4 minutes. Assuming this was also due to flux arcing from improper cleaning between the transistor legs, I wiped the area thoroughly with rubbing alcohol and decided to retry the test.

3rd time power applied (caught on video):
Smoke from an unknown location (appeared to be under the large circuit board) after ~15 seconds. Please take a moment to observe this on video: youtube.com/shorts/bDFA4OeiegE?feature=share

YouTube has decided to convert the video I posted into a “YouTube short”, which disallows slow-mo playback. This happens automatically to every video under 60 seconds posted in vertical video format, which I find very frustrating.
If you’d like to watch the video in slow motion, please just download it so you can scrub through it slowly with your computer’s built-in playback software.

If anybody needs photos, please don't hesitate to let me know.

If you’ve read this post up until now, I commend you. What an incredible community this is!
Thank you all so so much.

Jordan B.
Amateur Audiophile

 

[Paul Birkeland]

What kind of flux and solder are you using with your build? 

 

[iskess]

I used basic electronics solder for the Speedball. Not sure exactly what the percentages were but it's the same exact solder I used to assemble the stock Crack, and that's worked for over a year.

As for the flux, an image of the container has been attached to this post.

 

[Paul Birkeland]

Acid paste flux is for plumbing, not for electronics.

Your Crack needs to be run (without the wood base obviously) for a full cycle in the dishwasher with hot water to get that cleaned off.  Please do not use any extra flux in the future.

 

[iskess]

Thank you for bringing this to my attention; how remarkably embarrassing!

I only used the acid paste flux on the large board of the Speedball upgrade (not the small board or the stock crack), so do you reckon it'd be okay to thoroughly clean the entire board (and all associated contacts) with a toothbrush, hot water, soap, and isopropyl? Putting my Crack in the dishwasher would likely ruin its beautiful paint job.

 

[Paul Birkeland]

You need to run your amp in the dishwasher on a full cycle with hot water.  You can buy a 55 gallon drum of alcohol and drop the amp in there and the plumbing flux won't come off.

It is extremely likely that the flux has migrated off that PC board, so you really need to do the entire amp.

I have run into this several times before, and using the dishwasher is the only method I've found that's effective.

 

[iskess]

After two bouts of vigorous cleaning, (followed by an ample drying period) I prepared for another power-on test. Upon measuring the resistance between OA and OB, my multimeter displayed "OL". Super high resistance? Perfect.

Fortunately, I was thorough in setting up multiple camera before this plug-in test. Unfortunately, the Speedball's large board started smoking and sparking again. Please watch the following video:
youtu.be/1oW1AOgn8zQ

Any advice would be very much appreciated.

 

[Paul Birkeland]

If your vigorous cleaning didn't include putting your amp in the dishwasher and running it for a cycle, then I probably can't help you. That burned spot that you showed in your video is from the acid core flux you used.  You need to run your kit through a cycle on the dishwasher to get rid of that stuff.

 

[iskess]

Got it. As I'm currently living in a college dorm, I don't currently have access to a dishwasher. I suppose I'll try to make friends with a local homeowner, haha.

On another note, do you reckon that I've seriously damaged my amplifier (considering that one of the channels' LEDs went dim after the smoking & sparking)?

 

[Paul Birkeland]

Yes, you have seriously damaged your amplifier.  There's absolutely no point in trying to debug what you have until you've run the dishwasher cycle, as that acid core flux will migrate around and create new shorts if it's not all gone.  You may want to look around the common areas of your dorm to see if there's a kitchen with a dishwasher.

 

[iskess]

Okay, my amp has been dishwashed.
What should my next steps be? How should I go about troubleshooting to figure out what the short damaged? I'm really hoping its not the amp's transformer.

Also, as can be observed within the video from my November 20th post, there now seems to be a small scorch mark in the PCB between two of the TIP-50 transistor legs. Has this compromised the entire "Large Circuit Board"?

 

[Paul Birkeland]

Try turning the amp back on and measuring the DC voltages on all the PC board pads.

 

[iskess]

Okay, here's an update.

Timeline of events:
- Power applied to Bottlehead Crack.
- All LEDs on the small circuit board turn on.
- The pair of LEDs nearest the OA terminal turn on. Near-simultenously, the pair of LEDs nearest the OB terminal turn on, then flicker as a thin whisp of smoke comes from the underside of the large circuit board. (It would appear that this smoke originated from the pins of the TIP-50 nearest the OB terminal.) As this happens, the pair of LEDs nearest the OA terminal AND all LEDs on the small circuit board dim substantially. However, the pair of LEDs nearest the OB terminal seem unaffected following their intitial flicker.

I only noticed the thin whisp of smoke and associated dimmings after reviewing the voltage test footage (I'd set a camera up ahead of time), and consequently did not cut power to the board immediately after the smoke occured (keep in mind that smoke was only produced for a few seconds). Instead, I went ahead and obtained voltage readings for the large circuit board's main terminals. See below.

Voltage readings:
OA: 63.5V
OB: 33V
G: 0.08V
B+: 67V

On another note, bear in mind that the whisp of smoke appears to have originated from the same TIP-50 solder joints between which the large circuit board's PCB had sustained a visible scorch mark during previous testing (this can be observed at the end of the video from my Nov 20th post). I assume that this isn't the byproduct of residual plumbing flux, as — in addition to dishwashing the entire crack — I also dishwashed the large circuit board itself around 5 times. Perhaps the scorch mark is enabling some sort of conductivity, or perhaps the issue lies elsewhere? I have no idea.

Thank you again for your continued assistance and support.
Jordan Bouret

 

[Paul Birkeland]

I would throw away the Speedball and start over.  The dishwashing doesn't always work, and if you can't get that stuff off, then you have to replace everything it touched.

 

[iskess]

Okay, update time.
I removed the large circuit board and replaced it with the two 3kΩ resistors which were originally installed. I didn’t feel the need to remove the small circuit board, as it was never exposed to the plumbing flux.

Moments later, I made a horrifying discovery. the long black wire connecting 12L to the headphone jack WAS NOT SOLDERED at 12L. This had gone unnoticed for an entire year when I was using the standalone Crack, as I suppose the loose attachment was still supporting continuity. During the Speedball upgrade, I must’ve knocked it loose such that — as I’ve just noticed — there is no continuity between terminals 12 and (for example) 3. This is quite possibly what's been causing the sparking issue this entire time (whereas previously that plumbing solder had been the primary suspect). Needless to say, I soldered it. Also, the 270KΩ resistor (supposed to be attached to terminals 12 and 13) didn't look soldered to 12L either. Needless to say, I soldered that too.

After this, I conducted resistance measurements as specified by the standalone Crack manual:

1 - Open Line
2 - Open Line
3 - 0.09Ω
4 - Open Line
5 - Open Line
6 - 2.45kΩ
7 - Open Line
8 - 1.1Ω
9 - Open Line
10 - 2.45kΩ
12 - 1.0Ω
13 - Open Line
14 - 1.0Ω
20 - 0.09Ω
22 - 1.6Ω
B3 - Open Line (1.3 megaΩ)
B6 - Open Line (1.2 megaΩ)
Center pins of RCA jacks - ~98kΩ

Obviously, the majority of the measurements above are not as the standalone Crack manual says they should be. Do note that at this point the small Speedball circuit board is still installed. Is that what’s throwing off the resistance measurements, or are the funky measurements indicative of component damage? If so, I really really hope it’s not my transformer.

Thank you so much for your help,
Jordan Bouret

 

[Paul Birkeland]

I don't know how a defective transformer would cause problems with the resistance readings.  I haven't run into that before...

If terminal 3 is 0 ohms and you have 3K resistors between 3 and 7 as well as 3 and 9, then you should see about 3K at 7 and 9.  Not seeing that there would have me asking to see some fresh build photos.

 

[iskess]

Alright, after testing I discovered that the 3kΩ resistors between 3 --> 7 and 3 --> 9 weren't supporting continuity, as — in the process of initially removing them for installation of the Speedball — I'd practically severed their leads. Needless to say, I've now replaced both 3kΩ resistors altogether. After doing so, I went about taking another round of resistance checks, and this time discovered only a few erroneous readings. See below:

----------------------------------------
1 - 13.14MΩ
2 - fluctuating
3 - 0Ω
4 - fluctuating
5 - 13.17MΩ
6 - 2.49kΩ <--- outright bad
7 - 2.96kΩ
8 - 0Ω
9 - 2.95kΩ
10 - 2.48kΩ<--- outright bad

12 - 0Ω
13 - fluctuating
14 - 0Ω

20 - 0Ω
22 - 5-9MΩ <--- outright bad

B3 - 2.96kΩ
B6 - 2.95kΩ

Center pins of RCA jacks - both within acceptable range (red and black were 100kΩ & 95kΩ respectively)

Ground tab of RCA jacks - 0.9Ω <--- seems a little high but still acceptable
----------------------------------------

(While considering the readings above do consider that the small speedball circuit board is still installed. Not sure if that changes anything but thought I'd mention it.)

I'm happy to be making progress. Any advice on what might be causing the erroneous measurements at terminals 6, 10, and 22?
Thank you so much for your help,
Jordan Bouret

 

[Paul Birkeland]

Do you have a 1/4" to 1/8" adapter plugged into your amp by chance?

 

[iskess]

Oh my word, I do. Has that affected measurements? I hadn't been aware it could.

 

[Paul Birkeland]

It will make terminals 6 and 10 measure 2.49K rather than 0.