Ice Tube Clock Help

For RTC breakouts, etc., use the Other Products from Adafruit forum

Moderators: adafruit_support_bill, adafruit

Please be positive and constructive with your questions and comments.
Chris04920
 
Posts: 4
Joined: Sat Mar 15, 2014 1:39 pm

Ice Tube Clock Help

Post by Chris04920 »

Hello,

I have recently built the kit, everything went together easy, and the clock works, well for about a few seconds. When I plug it in, the display comes on, the time flashing, but then after a few seconds, the digits start turning off one by one starting from the left, until there are no more left on, and they never come back on. If I unplug it I see the whole display light showing all the digits for a split second. Also I have to wait for a little while before I can plug it back in and all the digits are working again.

I have checked all the voltages specified here (http://forums.adafruit.com/viewtopic.php?f=41&t=27032) and they all check out. I have looked around the forums for anything similar happening but can't seem to find anything.

Any help or insight would be great!
Attachments
IMG_20140315_124658.jpg
IMG_20140315_124658.jpg (190.77 KiB) Viewed 699 times
IMG_20140315_124635.jpg
IMG_20140315_124635.jpg (398.21 KiB) Viewed 699 times
IMG_20140315_124615.jpg
IMG_20140315_124615.jpg (265.42 KiB) Viewed 699 times

Chris04920
 
Posts: 4
Joined: Sat Mar 15, 2014 1:39 pm

Re: Ice Tube Clock Help

Post by Chris04920 »

I found this topic that appears to be similar, something to do with Q3. Can anyone confirm this?

http://www.adafruit.com/forums/viewtopi ... 41&t=50437

User avatar
jarchie
 
Posts: 615
Joined: Sun Jun 24, 2012 2:16 pm

Re: Ice Tube Clock Help

Post by jarchie »

That seems to be exactly what's going on. I call it the flaky segment problem. There is a more detailed description of the problem in this thread. The symptoms are different, but the problem is the same. Here's another thread describing your symptoms exactly, but it's quite a bit longer so you'll have to read through more to get the essential information.

User avatar
russell 27
 
Posts: 242
Joined: Thu Sep 12, 2013 3:59 pm

Re: Ice Tube Clock Help

Post by russell 27 »

Is your MOSFET installed correctly. It looks inverted from your image. Here is an image of the part, source connection is toward outside of board. You can jumper outer leads together to test and see if display works properly with jumper.
MOSFET JUMPER.png
MOSFET JUMPER.png (6.16 KiB) Viewed 687 times

Chris04920
 
Posts: 4
Joined: Sat Mar 15, 2014 1:39 pm

Re: Ice Tube Clock Help

Post by Chris04920 »

I tried jumping the outer leads, the display stays on when I do this. But it is installed properly, which leads me to think that Q3 is in fact damaged.

User avatar
russell 27
 
Posts: 242
Joined: Thu Sep 12, 2013 3:59 pm

Re: Ice Tube Clock Help

Post by russell 27 »

The rounded side should be towards the switch, unless you have a different MOSFET.

Chris04920
 
Posts: 4
Joined: Sat Mar 15, 2014 1:39 pm

Re: Ice Tube Clock Help

Post by Chris04920 »

I installed it as per the instructions which stated that the curve should line up with the silk screen, which I did. Also the reference picture shows the rounded side facing away from the switch.
http://learn.adafruit.com/assets/4639

User avatar
russell 27
 
Posts: 242
Joined: Thu Sep 12, 2013 3:59 pm

Re: Ice Tube Clock Help

Post by russell 27 »

All I can say is, take a look at the data sheet, that's where this image came from. Source connects to +5 rail. Been there.

User avatar
russell 27
 
Posts: 242
Joined: Thu Sep 12, 2013 3:59 pm

Re: Ice Tube Clock Help

Post by russell 27 »

There's a body diode inside there, when connected in reverse, diode will make a connection. But I think there is enough voltage drop through the diode that it doesn't allow filament enough current, hence the odd characteristics of your display. If you have tried battery backup and your battery is discharged, it's a sure sign, I wouldn't until you double check.

User avatar
jarchie
 
Posts: 615
Joined: Sun Jun 24, 2012 2:16 pm

Re: Ice Tube Clock Help

Post by jarchie »

Russell 27 wrote:All I can say is, take a look at the data sheet, that's where this image came from. Source connects to +5 rail.
I agree, but also think that you may have the orientation backward. The +5 rail is on the bottom--not near the edge of the board. So I think that Q3 should be mounted according to the Adafruit instructions in the same orientation as printed on the silkscreen. By my eye, Chris has Q3 installed correctly.

User avatar
russell 27
 
Posts: 242
Joined: Thu Sep 12, 2013 3:59 pm

Re: Ice Tube Clock Help

Post by russell 27 »

Well, if that's the case, then I'm wrong. It appears the +5 trace does come from the Atmega and that would make your mount correct. My apologies. I looked at an earlier post, and it appeared to have the same mounting as yours and after discussion it was flipped, go figure. Recently I've been using a different circuit configuration, Guess I need to stop helping. When I built the kit, I did put the MOSFET in as explained, and it was wrong, it happens. At least you know the problem.
Last edited by russell 27 on Sat Mar 15, 2014 3:47 pm, edited 2 times in total.

User avatar
jarchie
 
Posts: 615
Joined: Sun Jun 24, 2012 2:16 pm

Re: Ice Tube Clock Help

Post by jarchie »

Russell 27 wrote:Well, if that's the case, then I'm wrong.
No worries there... But since I'm wrong more than my fair share of the time, I'll look it up to be sure...

...okay, the attached image shows the upper pin of Q3 going to R3 and the VFD driver chip, so the +5 rail is on the bottom pin (pin furthest from the edge).
Chris04920 wrote:I tried jumping the outer leads, the display stays on when I do this. But it is installed properly, which leads me to think that Q3 is in fact damaged.
I agree with this assessment. Now you must decide if you would like a free replacement from Adafruit (and I assume they will offer eventually). Or if you'd like to upgrade Q3 to a ZVP2110A for the reasons I mention in this post. Or a fair number of people replace Q3 with a PN2907A&resistor as Russell suggested. Both the ZVP2110A and PN2907A/resistor solutions are good, in my opinion.
Attachments
Adafruit board showing orientation of Q3
Adafruit board showing orientation of Q3
Q3.png (42.4 KiB) Viewed 644 times

User avatar
adafruit_support_bill
 
Posts: 88087
Joined: Sat Feb 07, 2009 10:11 am

Re: Ice Tube Clock Help

Post by adafruit_support_bill »

Thanks jarchie & russel27 for the diagnosis.

Chris - You can contact [email protected] with a link to this thread for a replacement Q3.

User avatar
jarchie
 
Posts: 615
Joined: Sun Jun 24, 2012 2:16 pm

Re: Ice Tube Clock Help

Post by jarchie »

Russell 27 wrote:Recently I've been using a different circuit configuration, Guess I need to stop helping.
If that's true, then I need to stop helping also. If you search the forums, I think you'll find several instances where I have been very, very wrong... But if you help enough people, you're going to be wrong from time-to-time. It goes with the territory.
Russell 27 wrote:I looked at an earlier post, and it appeared to have the same mounting as yours and after discussion it was flipped, go figure.
I may be able to explain this. When the flaky segment problem first started cropping up, many users reported that flipping the direction of Q3 fixed the problem. So Adafruit support and other users (including myself) started telling people to flip Q3 to solve the flaky segment problem, but that turned out to be bad advice.

Flipping Q3 allowed conduction through the body diode, and that did fix the display problem... just like jumping the outer pins of Q3. A couple months after the flipping-Q3-solution became accepted, Phil (forum user PhilD13) had the brilliant insight to ask people if sleep worked after flipping Q3. Sleep never worked with Q3 backward, of course, because conduction through the body diode did not allow the clock to turn off the VFD filament and driver chip during sleep. The battery simply could not provide enough current and would be completely drained within minutes.

So this is a case where even Adafruit support gave bad advice, and they're the experts! If you're beating yourself up over one mistake... STOP!

User avatar
russell 27
 
Posts: 242
Joined: Thu Sep 12, 2013 3:59 pm

Re: Ice Tube Clock Help

Post by russell 27 »

The first post I ever made was about an inverted MOSFET, right after I built the kit last September '13. I noted my discharged battery because of this situation and posted result of this as well. Phil D built his kit about the same time and had already left a post about MOSFET mounting techniques. Much of the original information of these posts has been intermixed into many posts and has gotten a bit stretched unfortunately. That's part of the reason when I looked at the photo I mistakenly thought the P channel was inverted, my over thinking, not reading the other post carefully; and too many BUD lights, not yesterday in life. This situation was rectified in the kit directions, but ironically this part has become the source of even more odd display behavior. Every time I think I know something about electricity, or how that works, something else comes along and WHAAAAAAT... Any part made of CMOS construction is static sensitive, seems like almost all modern parts. I've pulled some real blunders and never destroyed a microcontoller or MOSFET that I know of, from ESD, and I'm not fanatical about it. I would say that P channel MOSFET's are harder to turn on than N channel, especially in this package size, and circuit voltage. If the tube was switched by an N channel instead, I don't think you would see this problem. I have a P channel J FET that I should probably try sometime. While I understand the circuit, knowledge of the program is needed to fully understand the circuit. So it's very irritating to me, that I only understand parts of the program.

Locked
Please be positive and constructive with your questions and comments.

Return to “Clock Kits (discontinued)”