Falcon Csync video fault.

[exxos]

I had a email a few days ago from Steve asking about a capacitor bulging problem after re-capping his Falcon motherboard.. But this actually gets more interesting later...

This particular capacitor is in the video circuit, or more explicitly, it is the coupling capacitor for the composite sync signal.

1.jpg

My first simple test to check the voltage across the capacitor to see what was there...

2.jpg

The particulate meter does register HF pretty well, though for a quick test, we can see the voltage is backwards across the capacitor and we are measuring negative 3.89V!


Next up I swapped the capacitor around and retested..

3.jpg

So now we have a proper polarity and the voltage is now 3.49V.


I then went on check the voltages on the scope...

cap1.png

Above we see a maximum voltage of 4.44V and a minimum of 2.08V.

cap2.png

On the other side of the class to we see 0.64V max and (negative) 1.48V!

So between the two negative voltages, we actually have a 3.56V difference, which basically matches the meter readings of about 3.49V.

So in both tests, the voltage across the capacitor is reversed!

I went back in the ST designs and looked at the STE schematic....

ste.jpg

Where the same capacitor also seems in backwards! I even checked the MSTE and that is also the same circuit.

I went further back to the original STFM design...

stfm.jpg

Where the composite video output as the capacitor in the opposite direction of STE/Falcon based machines! So in actual fact, the STFM capacitor seems to be the correct direction.

As to why capacitor ended up being backwards I really don't know. Though for some reason Atari decided to swap this capacitor around, either on purpose, or by some huge screw up and carried on that mistake to or from other machines.

The interesting thing is that I had a Falcon in for repair a few months ago which I mentioned in this thread...

https://www.exxosforum.co.uk/forum/viewt ... 70&p=11759

Where it was exhibiting some video sync fault, which I ultimately gave up trying to diagnose.. At the time, my test Falcon was working fine but recently started exhibiting the exact same problem as in that thread. At the time I could not find any real difference between my working Falcon and the Falcon I had in for repair. But basically I was comparing 2 faulty machines. Only at the time, my own Falcon was actually working! So this was a huge spanner in the works of trying to diagnose this problem as I did not realise my own machine is actually faulty as well! I actually thought it was something to do with my monitor going faulty at one point, but I never had the sync issues with my STFM's or STE's in fact. So it actually remained one of those mysteries to look at another time.

Though since swapping around the capacitor in my test Falcon, my monitor has not had any sync faults in the 2 or so hours it has been powered up for . Previously it would lose sync every 30 seconds or thereabouts easily.

So I am actually thinking now is going to be a common fault on Falcons and other machines that use the composite signal (or sync). I have seen it happen on two machines myself. Though Steve reported this as a issue that the capacitor was bulging after just three days use. Steve did not mention any video issues, though clearly there is a issue that is capacitor was self-destructing after just a few days. and 3V or higher backwards at a high frequency would certainly not do the cap any good!

Oddly the original capacitor (wangs caps etc ) seems to be mostly happy in working like this, though of course my test Falcon has recently started developing issues which seem to be cured by turning the capacitor around.

So I thought I would document this as I assume this will be a problem for the people in the future. Of course thanks to Steve for bringing this to my attention!

[dhedberg]

Ouch, ouch, ouch! This is an awesome find! Seems completely insane that a mistake like this has gone undetected since the STE though. Have to have a look at it myself...

[Steve]

I'm really grateful that you spent the time to investigate this for me, it had been bugging me for a while. I knew something was up, when every time I replaced the cap it would bulge - I ended up putting the Wang back in but still it bugged me so much, I'm so glad we have an answer.

At the moment I route pin 9 on the MC1377 chip in the video circuit to my composite out plug (As I manually removed my RF modulator and fitted a composite plug instead) but someone on AF recommends that it is a bad solution and instead I should route my composite from pin 12 of the Falcon monitor port.

Chris do you have any thoughts on that? About where I'm routing my composite signal from.

Cheers

[exxos]

Chris do you have any thoughts on that? About where I'm routing my composite signal from.

Why are you not using the proper CSYNC output ? That is driven via that capacitor and its where I drive from.

I did route pin 9 via a 1K and then via that cap to drive it that way, but didn't see any point as there is a driver transistor there, so using CSYNC is the proper and correct method.

[Steve]

Chris do you have any thoughts on that? About where I'm routing my composite signal from.

Why are you not using the proper CSYNC output ? That is driven via that capacitor and its where I drive from.

I did route pin 9 via a 1K and then via that cap to drive it that way, but didn't see any point as there is a driver transistor there, so using CSYNC is the proper and correct method.

Because originally I followed a guide on removing the modulator and replacing it with a composite output plug connected via that chip. I didn't know this was incorrect. So you're saying if I route my composite output plug via a patch cable underneath the motherboard to pin 12 of the falcon monitor output - this is the best solution?

[exxos]

Because originally I followed a guide on removing the modulator and replacing it with a composite output plug connected via that chip. I didn't know this was incorrect. So you're saying if I route my composite output plug via a patch cable underneath the motherboard to pin 12 of the falcon monitor output - this is the best solution?

Would think so yes. I mean pin 9 is only buffered by Q3 , which also drives CSYNC so full composite signals should actually be on CSYNC. If you look at my scope images in the top post, its clear there is a lot more than just syncs in there, so think its full video..

Though why are you using composite and not RGB anyway ?

[Steve]

I do use VGA mostly, I only plug in composite when I want to watch certain Falcon demos that are composite only.

[exxos]

I do use VGA mostly, I only plug in composite when I want to watch certain Falcon demos that are composite only.

Didn't know such demos existed which only worked on composite.. :shrug:

[Steve]

Well maybe I mean RGB. But I don't have an RGB cable as I always use VGA, a lot of demos don't support VGA.

[dhedberg]

Well maybe I mean RGB. But I don't have an RGB cable as I always use VGA, a lot of demos don't support VGA.

If your VGA monitor can handle 15kHz, the 3in1 monitor adapter in exxos store would provide what you need.