Yea I thought easy things the front panel going dead. Checked all the supplies on the main board, all perfect, no change in voltages, no change in the very low noise on the rails. I monitored the voltage supplied to the front panel board as well, stays the exact same and no increase in noise. I do not have a connector pinout for that front panel board and that would help. *think* there are I2C like data lines that go up and interface with the main CPU somehow. But this connector has a lot of pins considering there is a controller chip on the front panel board. You would think the whole board could communicate its data on one I2C bus of 2 lines and gnd.
Tracing the lines on the board is no easy matter, they change sides a lot. They seem to terminate up below the CPU where there are a number of resistor/cap combos like pull up and filtering. So sure somethere there could go wrong, but none of that gets warmer. The only thing that gets warmer it the CPU and of course power supply stuff. On a really deep level std chips like logic arrays could get warmer and do weird things, I have never seen that, but, maybe.
I am awaiting freeze spray next week. By using spraying and dropping little tiny drops of freeze liquid I should be able to locate better the issue.
I did use a heat gun on that area. Like seriously. It caused it to work for 15 mins. Which was weird.
I can't easily probe the area I need to. So its hard to see exactly what is going on.
Thermal is almost always a cap power supply related or a solder joint. I have seen a few transistors get weird thermally. I just never remember seeing a chip get weird thermally like this.
I really need to see the datalines on a scope. I may work out a way to better look at this. I have never done this, but, I think I can remove the whole power supply and maybe feed the main board directly from 15V ? If I dont need that ribbon cable from the power supply to the main board, I think its ethernet, then I think I can get the power supply away from the main board and still have it work.
Its acting like the data-lines are somehow dropping in voltage getting sketchy and then gone completely. OR the CPU is less sensitive and cant see them well.
If I had a schematic, this would be a 10 mins fix. But Nooooooooooooo.......
And then this other board issue.. I have not even begun to think that thru yet. Only positive spikes away from zero. Random but low frequency.. However the spike seems infinatly short. Its a weird symptom. Its both channels evenly. The spikes occur on both ch1 and ch2 exactly the same in time. 500mv scale and below no issues and in calibration. There may also be a high noise floor on the 1V up scales.
Not sure what would affect only the 1V up scale AND only on the positive side AND both channels. That seems deep enough to be some freakyness in the ADC chipsets. Might be something analog maybe but hard to explain the zero width pulses. Those look like data errors or something.
Both problems are hard. For now I have the boards leaning towards scrap.
