ice cube clock stops working when menu button pressed

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

Re: ice cube clock stops working when menu button pressed

Post by jarchie »

crispytwo wrote:...my display is dark.
The default display brightness is fairly dim, and in a bright room, the display can appear as if it is inactive. I recommend reducing the amount of ambient light when testing the tube. I do trust your observations, but I have also witnessed intelligent people mistake a working clock for one with no display in a well-lit work environment. Just something to be aware of.
crispytwo wrote:I also attempted the long jumper wire in case that would help -- It didn't in my new circumstance. :-( -- However, I did see " .. b. b. " faintly. I've remove the wire and spent time devising the above.
I have not personally tried this solution, but remember another forum user trying it successfully about a month ago. If I can find the time this weekend, I will do this myself, and post the results here.
crispytwo wrote:I went through the standard trouble shooting and came to this measurement:
Low Boost Voltage.
With no tube or VFD chip installed, you should measure between 40 and 70v on the striped end of D3. With a VFD chip installed, you should measure approximately 14-16v.
When I tested this and I'm getting 12.6v at D3 -- which seems potentially too low.
I think 12v is fairly typical, despite the troubleshooting information. I can say with certainty that 12v is enough to drive the display.
crispytwo wrote:@PhilD13
In reference to 'Russell 27' method, it seemed to me that he is testing the voltage drop across the resistor R3 to ground.
So, my in-circuit voltages (with the tube connected):
  • connection at Q3-R3 is 1.97V
  • connection at R3-tube is 1.19V
His values were 4.26V and 2.96V respectively (although his are measured at 70 brightness -- which I expect that mine is the default 30)
This looks like the problem. The Q3-R3 connection is where the MAX6921 VFD driver chip draws power, and the MAX6921 needs at least 3.0v for reliable operation. There's a good bit of manufacturing variation among IV-18 VFD tubes, but I would put the absolute minimum filament voltage in the 1.8v to 2.5v range.
crispytwo wrote:Is there anything I could test to show that the VFD chip is damaged?
Or test that the voltage coming into or out of it is incorrect?
Is it possible that the tube failed?
For the moment, there still seems to be a problem with Q3. I would recommend taking Adafruit up on their offer of a free ZVP2110A, and then test that with your solder-free test setup. That would also rule out the possibility of Q3 heat damage from all the board rework.

I know you've already tested your repaired board traces for continuity, but perhaps you could check them for resistance also. A dodgy connection or damaged trace with high resistance could explain the unusually low voltages you see coming from Q3.

I would only look for a different problem once you know Q3 is working properly, and right now, it doesn't seem to be.
crispytwo wrote:I'm wondering if I should start a new thread since it is no longer the original problem?
Since it's the same clock with the problem, I think it's appropriate to continue posting to this thread.
crispytwo wrote:Now the jumper has no positive effect. This leads me to believe that I screwed up something else.
When you jumper Q3 now, what is the voltage at the jumper? The R3-tube connection?
crispytwo wrote:It is good to know that the transistor voltage is only the filament.
If you look at the schematic (below) or board design (posted earlier), you'll see that output of Q3 powers both the tube and VFD driver chip:
Q3 powers both the tube and VFD driver chip
Q3 powers both the tube and VFD driver chip
Q3.jpg (79.31 KiB) Viewed 1005 times

User avatar
crispytwo
 
Posts: 16
Joined: Sat Jan 24, 2015 4:40 am

Re: ice cube clock stops working when menu button pressed

Post by crispytwo »

@jarchie

Thanks for all your detailed advice!

I will focus on the Q3 and iron out the details you've outlined later tonight.
I will try to jumper the circuit with the Q3 installed - I did not think of that!
I have a spare zvp2110a and a spare zvp3306 so I don't have to wait to get an extra. ;-)

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

Re: ice cube clock stops working when menu button pressed

Post by jarchie »

jarchie wrote:
crispytwo wrote:I also attempted the long jumper wire in case that would help -- It didn't in my new circumstance. :-( -- However, I did see " .. b. b. " faintly. I've remove the wire and spent time devising the above.
I have not personally tried this solution, but remember another forum user trying it successfully about a month ago. If I can find the time this weekend, I will do this myself, and post the results here.
I removed Q3, and soldered the long jumper wire as described before. The clock worked fine and kept time during sleep. So I can confirm that this should work in practice.

The clock I tested did have a few mods; none should make any difference here. But in the interest of full disclosure, the mods were as follows: (1) The microcontroller is an ATmega328p running my xmas firmware. (2) An Advanced Photonix Inc PDV-P8001 photoresistor and 5.6k are installed to support the autodimmer mod. (3) Q3 and R3 were replaced by a ZVP2110A and 11 ohm resistor, respectively.
long-jumper.jpg
long-jumper.jpg (26.38 KiB) Viewed 999 times
clock-display.jpg
clock-display.jpg (43.61 KiB) Viewed 999 times

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

Re: ice cube clock stops working when menu button pressed

Post by russell 27 »

I stopped commenting here because no matter what I said it was argued, words have to be perfect. Yes the transistor does disconnect the tube filament and maxim chip when power is off. This is so battery does not power these during power outage, microcontroller only to keep time. In your case I was referring to the tube, filament Vs anode boost. I feel at least 4 volts should be on the transistor drain for for decent circuit operation. When measuring boost output, what you have is close to what low brightness boost output would be. When you adjust brightness up, boost voltage goes up, max may be in low to mid 20 volts. It's difficult to test anode tube pin voltage, it's multiplexed. The transistor is nothing more than a switch, connecting vcc to maxim chip and tube filament. Make sure you have good vcc trace connected to both of these points, either thru jumper or transistor. You can always fix the traces with small wire and light solder. This transistor does nothing but turn on and off a part of Vcc rail. But that switched part of rail does a multitude of things. Use your meter to very carefully check things over. I doubt your maxim chip is at fault, or tube. In short, if jumper originally worked, voltage was good at both these points, hopefully this long version is of help. The long jumper wire is a slick remedy, it effectively eliminates the transistor, while isolating the battery. If one of your traces is broken, may not work any better.

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

Re: ice cube clock stops working when menu button pressed

Post by russell 27 »

By the way, don't be upset about breaking a trace. I've opened new devices to find a broken trace right from the factory, fixed with a small jumper.

User avatar
crispytwo
 
Posts: 16
Joined: Sat Jan 24, 2015 4:40 am

Re: ice cube clock stops working when menu button pressed

Post by crispytwo »

@jarchie
Thanks for validating what I did with the long jumper wire. That is really appreciated.

@Russell27
Thanks for the description of the long jumper wire. A board with that as a design would be simpler and cheaper. That's cool.

So, I figured, I should return to the long jumper wire solution since that should be giving best connection for the circuit at the point of Q3-R3, and attempt to see what can be done about the clock display.

The voltages for R3 for my in-circuit voltages (with the tube connected, long jumper wire, no Q3):
  • connection at Q3-R3 is 4.9V
  • connection at R3-tube is 3.5V
The tube is displaying '.. .d d.'

One interesting tidbit with this current setup: After about 45 seconds, the display of the double d-s starts to slowly fade and at about 60 seconds the clock beeps like it just lost power, and the display goes dark. The voltage drop at R3-tube remains stable (3.5V).

So, the puzzle:
if Q3 is eliminated and the R3 has good voltage levels, what would cause this failure?

User avatar
phild13
 
Posts: 247
Joined: Mon Sep 10, 2012 1:05 pm

Re: ice cube clock stops working when menu button pressed

Post by phild13 »

Does part of the processor get hot after start up? I had one I messed up and it acted in a similar way.

User avatar
crispytwo
 
Posts: 16
Joined: Sat Jan 24, 2015 4:40 am

Re: ice cube clock stops working when menu button pressed

Post by crispytwo »

@PhilD13
I just retested to double check, and no, it is cool to the touch. The start beep occurs normally, and the display starts just as it first beeps.

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

Re: ice cube clock stops working when menu button pressed

Post by jarchie »

It looks like you fixed the Q3 issue, so I think you are correct to look for another problem now.

With the Adafruit firmware, a bad oscillator sometimes causes display problems when the clock is in the time display mode. Are you able to toggle through the menus with the menu button?

Also, try touching the fuse when the clock restarts. Is it warm?

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

Re: ice cube clock stops working when menu button pressed

Post by russell 27 »

Could you post photo of bottom of the board.

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

Re: ice cube clock stops working when menu button pressed

Post by russell 27 »

One other thing, do you have the correct diode and polarity used with the inductor?

User avatar
crispytwo
 
Posts: 16
Joined: Sat Jan 24, 2015 4:40 am

Re: ice cube clock stops working when menu button pressed

Post by crispytwo »

@jarchie
The menus seem to work.
from startup to 3rd press, it seems to 'dim/brighten' the display
from startup to 4th press, I can select the loudness of the buzzer.
the display does not change at all -- it remains in the '.. .d d.' state
the fuse seems to be warm after operating for a minute or so. Surprisingly, it 'operated' for a couple of minutes without resetting during this test.

@Russell27
long jumpered bottom of board image:
bottom board with long jumper
bottom board with long jumper
long-jumpered.jpg (722.7 KiB) Viewed 932 times
the diode follows the schematic -- polarity bar (cathode) is furthest from the inductor. The anode is attached to the transistor.

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

Re: ice cube clock stops working when menu button pressed

Post by jarchie »

crispytwo wrote:the display does not change at all -- it remains in the '.. .d d.' state
Since toggling the menus did not change the display, it sounds like the crystal oscillator is just fine.
crispytwo wrote:the fuse seems to be warm after operating for a minute or so. Surprisingly, it 'operated' for a couple of minutes without resetting during this test.
This seems like a huge clue to me. Something is pulling more current than it should and that's tripping the fuse. I've seen burnt rosin create a high impedance short, and given all the board rework, that's one possibility.

Burnt rosin is an easy thing to rule out. Remove the battery and scrub the board thoroughly with alcohol and a toothbrush. After all visible traces of rosin are gone, wait for the board to dry completely before trying to power it again.

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

Re: ice cube clock stops working when menu button pressed

Post by russell 27 »

If you haven't squared things away yet. Try reflowing solder on 32k crystal and it's capacitors. If you need to clean that small area of flux, a Q-tip with some acetone works great.

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

Return to “Clock Kits (discontinued)”