Broken clock circuit on Sycon thin film controller.

[macona]

So I am getting ready to hook up my thin film controller to my e-beam coating system so I hooked it up to make sure it still worked since I had not turned it on in years. Push the button. Pop sizzle. Uh-oh. Ended up being a tag tant cap on the main board. Easy fix, replaced it.

Or so I thought. The display on the crt is scrambled. Figured it was the crt board but after connecting a couple external displays as well as I had another logic board and verified my display is good. But I can see activity on the screen and make out some text. Unfortunately the other logic board is a newer Rev and does not have the extra options I need for the sequencer.

I put the scope on the composite output and it looked normal. Until I measured the period of the signal and found it’s refreshing at around 65hz instead of 60 which I verified with the other board.

The board runs off of an Intel 80188-12 and found the clock pins and looked at that signal and it is a total mess. Around 24Mhz but so much jitter. The cpu halves the frequency so I think it’s actually being overclocked a bit which is messing with the refresh rate. I noticed a small trim cap next to a crystal and played with that and it did effect the display and clock signal.

I managed to figure out the clock circuit a bit. It uses a 88.3354Mhz crystal to create a clock and then 74ac04 to square it up a bit and then a series of 74ac109s to act as clock dividers. The problem is the 88.3354Mhz signal is a mess. I check the signal at the quartz crystal and there appears to be nothing there. Heck, I pulled the thing out and it still boots so it seems to be getting that same crappy ~90mhz signal from somewhere, maybe the tank circuit for the oscillator. I compared it to the other board and the other one has a jittery signal on one crystal pin and a nice clean sine on the other. It has a circuit that is very similar.

Checking around that section it has a 8v reg which seems to be good and clean. That leaves the active components, the two MPS3563 transistor may be the problem? I need to find a cross and try replacing those. Could it be the crystal? This is not something I have messed around with before.

[srb1954]

It could be that the crystal oscillator is not oscillating at the correct frequency. If is supposed to be oscillating at 88MHz then it is probably a 3rd overtone, or even a 5th overtone type.

Normally, oscillators designed for overtone operation have a separate LC circuit attached to the oscillator to suppress the gain at the fundamental crystal frequency and enhance the gain at the overtone frequency. If the crystal activity is low the oscillator may be oscillating at the resonant frequency of that LC circuit or the oscillator may randomly switch between fundamental frequency operation and one of several overtone modes giving the appearance of an unstable frequency.

Please provide a sketch of the complete oscillator circuit so we can comment further. A scope trace of the oscillator output would also help but be aware that the scope probe load impedance may considerably affect the signal levels and wave shape unless you use a low-capacitance FET probe.

[macona]

Welp, it was a bad crystal. Swapped crystals between the two boards and old one works and new one don’t. Thinking of trying one of those Adafruit Si5351A Clock Generator Breakout boards and injecting the clock signal to repair the other one. The can be OTP to boot to a default freq.