On 8/9/2017 8:37 PM, Its Me wrote:
On Wednesday, August 9, 2017 at 4:46:30 PM UTC-4, Mr. Luddite wrote:
On 8/9/2017 4:32 PM, Its Me wrote:
On Wednesday, August 9, 2017 at 3:59:53 PM UTC-4, Mr. Luddite wrote:
On 8/9/2017 2:15 PM, wrote:
On Wed, 9 Aug 2017 13:50:30 -0400, "Mr. Luddite"
wrote:
I think the problem is on the PC board in the control unit. It's
supposed to turn on, light up blue and display all the membrane switch
areas and labels. It doesn't. Has 12 volts but doesn't turn on,
That is the problem with most new things. Everything is on a single
board that may only cost a few bucks to produce in some 3d world
country but sells for hundreds of dollars. They seldom/never provide
any documentation about what is on that board. With a little knowledge
of what usually goes wrong, you might get lucky and find a bad part
you can replace on the board but they are bad about using parts with
no helpful markings on them.
I would start with the most rudimentary thing that is not working and
trace back to the 12v ... like those back lights. It may just be a
diode or regulator on the front end. (the actual controls may be 5v)
Are there any burned traces? Parts that look like they got hot?
I usually try to draw out what I see going along the traces and hope
it is not a multilayered board.
I got into this trying to fix gate controller boards. Fixed a few,
tossed a bunch. The chip that seemed to be the failing part on most
was not documented and the manufacturer was no help.
So far I have only seen one side of the control panel board. I plan to
remove it completely and inspect it. There are a bunch of tiny diodes
on it that are about the only things I can check (forward and reverse
bias) if I scrape away some of the conformal coating but nothing else is
testable that I can see so far. I think the problem *is* on the PC
board but a replacement is not available.
Once you get to where you can see both sides of the PC board, follow the power where it enters the board and look for a fuse onboard. It may not even look like a fuse. There are auto-resetting and one-shot current limit devices. If there is one it should be very close to the wire (or header) where the power attaches. Hopefully the board isn't multi-layer.
Thanks. I'll look. This panel is supplied with battery voltage in
three different places. All three supplies are good, at least on the
terminal strip before it connects to the board. Don't know what happens
after that and without removing the PC board, I can't see how it
connects. Hopefully I won't spill a handful of membrane switch parts on
the floor. I hate working on "other people's stuff".
After thinking about it a little more, here are some other thoughts:
- Many times the positive input to a circuit board in this kind of application will have a power diode in series. It's to protect the board from being hooked up backwards to battery power. Alert, bad ascii art!
+V -||- board
- The backlighting you described may not be turned on/off by the board's controller, unless one of the membrane switches is an on/off and is supposed to light the board up and give you control. A membrane switch doesn't toggle on/off on its own, and can't carry much current. It would have to activate something that likely turns on a relay that then supplies power to the lighting and controller. Or the controller chip is always powered on. Maybe a bad relay?
- The controller chip will likely not run off 12 volts. There should be a regulator somewhere that takes it down to probably 5 volts. It could be bad. Look for a three legged device, probably on a heatsink, with some capacitors around it.
- Hmmm... remove all power from the board, and then re-apply after 20-30 seconds. Maybe a re-boot is needed?
- Hopefully, the membrane switches are a "sheet" switch, so there's nothing to spill. You can hope. 
In any case this will be a tough nut to crack, especially without a schematic and parts list, unless you can find an obvious smoking gun. Good luck!
Good thoughts and ideas. I had thought about the possibility of a
protective diode and it's one of the reasons I'd like to take the board
completely out and check them. They are small diodes though, but it may
be all it needs.
There *is* a main power on membrane button. Obviously right now it does
nothing (doesn't turn the panel on) which made me suspicious of a diode
as you suggested since it appears 12 volts is available to the PC board.
I've tried the "alt-control-delete" thing :-) I actually disconnected
the positive leads off off both house batteries just in case the panel
power was not removed when the main battery switch is turned off. In
fact, it's one of the suggested fixes to this problem. Unfortunately,
it didn't reset or fix anything.
The owner is away right now, so I won't get to try anything for a couple
of days.
Thanks for the suggestions.
---
This email has been checked for viruses by AVG.
http://www.avg.com