Ice Tube converter issue

[Mundoo]

Heya!

I sent an email to you using your contact-us form, and I was instructed to post it on the forums. So, here goes:

About a month ago I bought an Ice Tube clock kit as a birthday present for myself. I just got around to installing the VFD display and plugging it in to see if it worked. My soldering on the PCB looks sort of bad, due to my lack of having a proper fine-tipped soldering iron at the time. However, it did pass all of the voltage tests, so all the solder points should be working.

Two days ago I was able to solder the VFD to its respective riser-board, now that I've got my new iron with the fine tip (so much better, I wish I had one before >.<). I plugged the clock in, it beeped, but nothing appeared on the display. I knew the firmware was set to the lowest brightness setting on first-power-on's, so I turned the room lights off, plunging everything into darkness. The only light I saw was that of the screen on my laptop.

I figured something wasn't soldered properly, so I removed the VFD and checked the solder points along the connector. I reflowed them all with the new iron, so they look a lot better than they used to. Thinking this solved the problem, I reattached the VFD and plugged the clock's converter into the wall again.

Something happened here, and I'm not really sure what. When I plugged the converter into the wall (it was not plugged into the clock at the time, I unplugged the converter both from the wall and from the clock) there was a orange-ish spark in the wall outlet. That's not too uncommon around my house, I usually will see a small spark inside the outlet when plugging in my laptop charger. That spark is always blue though, not orange.

I shrugged off the spark (wasn't too big) thinking it was just a spark like what I see my laptop charger do sometimes. So I continued testing, plugged the converter into the clock. Problem: it didn't beep this time. I thought, maybe while I was reflowing the solder, I messed something up? Unplugged converter, flipped PCB over, checked solder... everything's fine.

Huh, weird. Got out my volt meter and checked the voltage on the converter: 0 volts. Gah, the converter died! Alright, no biggie, I've got another 9v converter I can use. I'll just try that one.

Still no dice. I verified the second converter does output the proper voltage, (it's labeled 10v @ 500mAh, however the voltmeter reads 9v).

I'm at a loss with what to do now. Could the converter have killed my PCB? I don't see any problems with it, nothing appears burned and nothing feels hot...

[adafruit_support_bill]

When I plugged the converter into the wall (it was not plugged into the clock at the time, I unplugged the converter both from the wall and from the clock) there was a orange-ish spark in the wall outlet.

Sounds like possibly a defective converter. They are rare, but out of the thousands we sell, we do see a bad one now and then. If you contact support@adafruit.com with a link to this thread we can get you a replacement.

In the scenario you describe, it is unlikely that the converter damaged the clock. It was not connected to the clock at the time of the spark & it now outputs zero volts.

I verified the second converter does output the proper voltage, (it's labeled 10v @ 500mAh, however the voltmeter reads 9v).

Are you sure it is a DC converter?
[edit - fixed quote]

[Mundoo]

I'm very sure it's a DC converter.

The label says:

Code: Select all

INPUT: AC 120V 60Hz 14W
OUTPUT: DC 10V 500mA

I've made sure the polarity is the same, the inner ring of the connector is positive while the outer is negative. The same as the original converter.

I'll try another one to see if it's any different. Perhaps the jack is a slightly different size and it's not making a connection inside.

Is it possible to re-order a new PCB and the parts? I found a link to a replacement PCB via Google but the page had 404'd :/


Edit again: Yep, it was the new converter not making a contact inside the jack. PCB still alive, woohoo! Still nothing on display though >.<

[adafruit_support_bill]

I'm very sure it's a DC converter.

OK then. Lets have a look at the board before we power it up again. Post photos of the front & back.

[Mundoo]

The PCB from left to right:






The back, left to right:






Closeup of VFD display header (female)



Closeup of CPU solder points:


You'll notice C1 is missing on the front, it's on the back because I made a booboo and couldn't quite figure out how to fix it, so I stuck it on the bottom.



I'll need to fix that so the thing fits in the case.


Closeup's of the VFD header (male) and the solder points for the display itself. These are a lot cleaner than the ones on the board, because of that new iron.





Sorry if the quality isn't too great, I couldn't find my good camera so I had to use the one on my video camera :/

[adafruit_support_bill]

There are a few suspicious looking joints there:

A lot of the joints on the headers for the tube-board look look solder starved. That could explain the dark tube.
There are also a lot of long leads that could cause a short. These should be clipped close to the board.
The joints for the capacitor on the back look 'cold'. These (and any other joints that are not smooth and shiny) should be re-touched with a clean, hot iron.

The ideal joint is shown in the diagram below:

[Mundoo]

Hmm, after reflowing some points (and moving that capacitor I had mounted on the bottom of the board to the proper place on top), the display still won't light up.

I understand it's very dim when the clock first starts and you're supposed to set the brightness manually... but it's not dim, it's just... not even on. Turned the room lights off and I didn't see anything.

But I did notice something, when holding the PCB to unplug the power from the jack, the PCB felt quite hot. Around the LM7805C chip. Could it be the fuse heating up?

The new converter I'm using is as close as I can get to 9 volts @ 1000mA, it's 12v at 1250mA. The 9v converters I have laying around are AC or don't have enough power to operate the clock.

I tried cutting the connector off the 12v and putting it on the 10v, but it wouldn't beep when I did that (verified polarity too), so, dunno.

Pictures:




It might be blurry, but this was the best I could get of the C1 capacitor I moved.


This area of the PCB gets very hot after being plugged in. Could it be because the converter is sending it too much current?


Reflowed these points, they look somewhat like the diagram you linked.


However, the display still remains inactive. Not even dim, it's like it's not plugged in or anything. Perhaps the fuse is shutting off before I'm able to see anything on the display?

Edit: meant to say mA instead of Ah.

[adafruit_support_bill]

Those solder joints on the header connections are looking much better.
A 12v supply is going to make the 7805 work a bit harder and get warmer than usual. Warm is OK, but it should not be painfully hot. Is the fuse getting warm too?

I tried cutting the connector off the 12v and putting it on the 10v, but it wouldn't beep when I did that

Does it beep with the 12v supply? Do the other voltages check out OK?

[Mundoo]

Yes, it beeps with the 12v supply.

I just powered it on again to find the source of the heat. The fuse gets quite hot, so I guess something isn't working properly.



While doing this voltage test (picture is from the directions) I get a reading of 12.3 volts.



This test, I get 6.02.



I got a reading of 20.8, but I have the VFD chip installed, so I assume that is why the voltage is lower.

[adafruit_support_bill]

This test, I get 6.02

That is very suspicious. The output of 7805's usually measure within a few percentage points of 5v. Before jumping to conclusions though: Does you meter have fresh batteries in it? Can you check it against a known source? (e.g. a 9v battery)

[Mundoo]

The low-battery indicator has been on for ages, but it's never prevented it from reading voltages properly.

I don't have a 9v battery to test, I'd have to take it out of the volt meter. But then it wouldn't work without it. So I tested it with a AA battery.

It read 1.73 volts. It reads the same as the volt meters in my old high school's "principles of technology" class (Really fun class, I wish I could have taken it a second time.) when I tested the batteries we were using for the fan-powered car project.

[adafruit_support_bill]

It read 1.73 volts.

Typically those put out a bit less than 1.6v open circuit when fresh. I suspect that your meter is reading high. Best to heed those "low battery" indicators.

[Mundoo]

Something's still wrong though. That fuse shouldn't be getting hot and shutting everything off after a few seconds. I've checked the solder points, none of them are touching. The speaker beeps, so the CPU is working, and the VFD controller and its voltage-booster are working as far as I can tell. Just, the VFD never lights up and the fuse shuts everything off >.<

Between each voltage test I had to unplug the power and wait a few seconds before starting the next test, otherwise I'd get 0 volts :/

[adafruit_support_bill]

Something's still wrong though.

I don't doubt that. But we need to make sure our test instruments are accurate in order to get to the root of it.

[Mundoo]

I'll look around and see if I can find another 9v battery.

Yep, found one that's pretty full. The battery indicator is gone now, anyway.

Retested that same battery, got 1.34 volts. Huh, never knew the battery level would affect voltage readings so much o.o I could have used it to make my wind turbine project's voltage output seem higher than it really was (We were supposed to drain the batteries before charging them with the turbines... I didn't drain them hehe)

Alright, I'll do another voltage test. Gimme a minute :3