Microcontrolled concert light will not begin reset process.

Thread Starter

condenser

Joined Nov 13, 2017
16
I have what's commonly referred to as a moving light or intelligent light used for stage lighting. Under normal conditions the light would be controlled by the protocol DMX512 via a control unit. Before the fixture can be used it should automatically go though a reset process.
The reset process cycles the stepper motors until they arrive at a known position.
One fixture does not process the reset proceedure. If I manually turn the fixture head until it contacts the limit switches it will then finish resetting and work perfectly . The unit should perform that on its own, move to the limit switches.
I have swapped out mainboards from a known good unit. After swapping the mainboard the unit works correctly. That verifies the issue is on the original mainboard.

Video of the issue

Schematic of unit attached.

Steps I have taken.

Downloaded firmware from manufactures site and uploaded a fresh copy into the unit via a proprietary uploader.
Removed 93LC66 EEPROM and reinstalled files copied from a known good unit.
Kinda backwards but verified correct power supply voltages.
Searched for like issues in cyberworld.

Tools to aid in the repair of this unit on hand. Older analog scope, several multimeters

Not well versed on a repair of this type. How do I determine if this is a hardware issue or software issue? Should I consider a LOGIC analyzer similar to a SALEAE unit to possibly help? Several properly operating units to compare findings against.
I've time on my hands at this time to fool with this and learn. That's actually my main goal, to learn from this.
Thanks in advance for any and all help.
 

Attachments

Thread Starter

condenser

Joined Nov 13, 2017
16
does anyone have a guess where the startup point to make this thing reset properly is? It has two processors from what I've seen. At least two storage locations. A watchdog circuit which seems to work.
I have been able to reset the main processor by making the reset pin go low. It then starts to boot again. All the clocks appear to be working. Voltages at differing chips are there.

Is there someplace else on the site I should have this posted? MIcrocontrollers, programming, robotics?

Can I referance this post in one of those areas without getting my fingers wacked with a ruler ?
 

Alec_t

Joined Sep 17, 2013
14,280
If you have two nominally identical lights, have you tried measurements at identical points in both lights to see if/where there is a discrepancy?
How old are the lights? Capacitors are likely suspects for changing their properties with age.
 

MrChips

Joined Oct 2, 2009
30,720
This is a difficult one.
If you want to tackle this and there is nothing to lose, I would suggest replacing the main processor SAB 80C165.
 

Thread Starter

condenser

Joined Nov 13, 2017
16
I have actually considered replacing the main processor. I even considered swapping with a proper working unit. If I swap the processor with another, I could end up with 2 units with issues. That wouldn't be the first time ive been there. Sometimes I end up going backwards before making any progress.

As far as caps go, most of the caps are in the power supply. The supply tests as good. Not a switching supply. Big ole transformer. There are certainly some smaller SMD caps scattered about. Some of my guesses did center on a few of them. One of which is across the EEPROM. It checked just fine.

I did attempt some checks and comparison at differing areas with a scope. Made checks at main processor xtal pins 7/8 identical on two like units. Checked same on IC 201 AT89C2051 processor pins 4/5 same between two units. Signals at the EEPROMs look the same on both units. That's close to the limit of the scope as I know it. All voltages I've checked thus far have matched on both units.

I'm actually wondering about the function of IC305 as its on a line from the main processor marked boot. I am assuming the coding that starts the unit is in the EEPROM ??? I don't know that answer for sure.

These lights are very high end stuff most of the items used to build them are of high quality. Fiberglass boards and such. Greater than 15 years old though. This problem has been present with this light for many years. Sat in a dark corner because of it.

Something to note, the unit seems to operate the same with the EEPROM removed and or present I notice little difference in operation. Except that it shows a memory error when removed.

Does anyone think a logic analyzer would help? I've come across a few cheap ones that record to a PC.
 
Top