Amiga 1200 composite video - Sony CXA1145 filter issues?

[nvana]

After recapping my valued Amiga 1200 I noticed the machine was working fine, except for the composite video output. It is completely dead it seems.

Why use composite while having RGB on the machine ? No reason for that, but being a purist to me the machine needs to work 100% not 99%

---

Test results:

On pin 20 (CV) the CXA1145 just gives 0.7 Volts and no signal at all on the oscilloscope.

On pins 16 -> delay line -> 18 I can see a luminance signal going out and in again.

On pins 15 -> BPF -> 17 I can see a very weak chrominance signal going out and being lost in the filter.

Measuring the Z221 (BPF) filter gives me almost 0 Ohm. Is that normal ? Could be a coil I'm measuring.

---

Long story short I ordered a new CXA1145 and soldered it in (which isn't too hard with normal soldering equipment).

Result: the same 0.7V on pin 20, all other luminance / chrominance signals look identical to me... Still no video out.

I de-soldered the 3 RGB caps (0.22uF) and measured them. Nothing strange. Soldered in 3 replacement 0.22uF caps. Same measurements on the in- and outputs of the caps and no comp video output of the CXA1145.

Tried removing the Z221 BandPassFilter and shorting the in and output of it (so no filtering - direct through), still no video out...

Only thing I could not compare to a well working CXA1145 where the voltages on pin 14 and 13 (Vref and Iref). Anybody knows correct voltages here ?

---

Somehow I have the feeling the CXA1145 does not recognize the horizontal or vertical sync and keeps the output off permanently. Pin 10 gets a composite sync signal, roughly 3.9V averaged. I can see the pulsed signal on the scope.

In my Amiga (rev 1d4) it has no capacitor from U23 going to the CXA1145. On the schematics it says 'C217 added Rev 2a'.

---

Do any of you guys here at the English Amiga Board have more experience at fixing the comp video out of an Amiga 1200 ?

[Keir]

Bypass the BPF and see if you get a video signal. If you do, BPF is the problem. However composite video signal will be extra crappy unless you replace the BPF (probably from a donor 1200 or 600 board).

[nvana]

No video after bypassing the Z221 BPF...

[Sandro]

if not worked after recapping surely was because the cap rated at 16v, 470mf, surely you reinserted it bad
such cap is located near the RF modulator

[nvana]

Comp video out was already broken before recapping.

[Zetr0]

@nvana

I suspect that you have a short that is syphoning off the input -

This could happen because of the electrolyte leak of the capacitors creating a damaged track or one that has managed to ground itself.

Have a look at the VIO circuit and start your inspection with a multimeter there. Should that it be a fault, you should be able to find it in continuity mode. I suspect that its possibly a damaged pad - a good look under some optics and a coffee is what I suggest =)

[bebek]

There is many things that can go wrong here. You can send it to me and you will have it repaired

[Havie]

Originally Posted by bebek

There is many things that can go wrong here. You can send it to me and you will have it repaired

BebEk just recaped my 1200 the other week and works great so +1 for repair.

[PARALAX]

I've exactly the same problem with my A1200, even before recapping. It drives me crazy as I already have replaced the CXA1145M, the delay line and the BPS without any change. When I connect Pin 15+17 from the CXA1145M I get a very crappy but coloured composite & HF signal but almost nothinh with Pin 16+18, see here:

[ Show youtube player ]

No trace cuts or short circuits have been found with help of PCB Explorer, everything seems to be okay but it isn't!

I already recapped a lot of boards which the same components and never had that issues, so it can't be the reason for sure. Is there any idea excpet permanentely connecting Pin 15+17 for such a shitty screen?

[nvana]

Yes, I have 2 A1200s which both have the same issue: everything works fine except for the composite output.

I replaced the CXA1145m, no go, I replace the bypass capacitors, no go. The delay line I cannot replace as the coils are stuck quite well in the PCB. Cutting them would be the only option I think and I didn't like that solution.

I went another way in one of my A1200's, I installed the CXA1645m into it. It doesn't need the stupid delay line at all. It needs a few modifications, but most of the signals are identical.

Voila a working composite PAL signal comes out of it now :-)

Many more people will have this composite video issue I presume, but who tests their composite outputs ? Probably not so many.

Wow the tip for Amiga PCB Explorer is an excellent one, man what time did I spend looking at the schematics and finding the right part and traces on the board !

[Amiga1992]

nvana, perhaps you want to post a guide on how to use that other video decoder?
Would it be possible to get SVideo out of it?

Originally Posted by nvana

Wow the tip for Amiga PCB Explorer is an excellent one, man what time did I spend looking at the schematics and finding the right part and traces on the board !

Holy shit, who made this? Amazing work! Hoping they add more PCB models like A600 and A500!

[mpattonm]

Originally Posted by nvana

Yes, I have 2 A1200s which both have the same issue: everything works fine except for the composite output.

I recently got my very first A1200 board, bought on Amibay from respectable seller. The board was basically virgin, with the exception of 3 broken track from electric screwdriver mayhem. However, these were only CPU to edge connector lines. Well I fixed them anyway.
First think I noticed that I got very distorted, rolling flashing picture via composite-out. RGB seems to be working fine tho.
I fully recapped the board hoping that it would help, very clean job, no go. Still no decent picture on composite out.
I checked all the pasives components around CX1145, they seem to be fine. I am out of ideas, just not ready to give up yet. Did anybody actually find the root cause?

[mpattonm]

An observation: if I short out pins 16 and 18 with jumper wire, rolling/flashing distortion is gone, but a picture is still blurry and colors washed out. All the RGB component colors are there in correct ratio, just not vivid.
Any clue anyone?

[sean_sk]

Sorry for resurrecting an old thread, but have the same issue with composite out on my rev 2B motherboard. In fact the video posted by PARALAX is very similar to my issue. It appears to be a common issue. Again, capacitors have been replaced and other surrounding passive component "seem" to be fine. On inspection cannot see anything untowards with any of the traces. RGB video out is also fine. Any ideas? Thanks in advance.

[nvana]

I solved this problem in my Amiga 1200 with the composite video encoder out years ago by using a Sony CX1645 (newer type of encoder). This chip doesnt need the inductors to function correctly.

If I remember correctly it needed a tweak with 2 or 3 external resistors which needed to be soldered close to the CX1645 mostly on the bottom side of the A1200 mainboard.

If you need a photo of how I soldered the CX1645 onto the mainboard including the resistor mod please let me know. I should have the photo somewhere and can look it up for you.

All mods were done in a removable way, so a next owner can undo it. As so the A1200 keeps its value.

[demolition]

This kit also uses the CXA1645 to fix any faulty video circuit on A600s and A1200s:
http://www.amibay.com/showthread.php...for-A600-A1200

[nvana]

12 Euro seems like a reasonable price tbtorro set for this fix. It took me more time for researching on how to assemble this fix in a reversible way. In 2016 this fix wasnt available yet.

[e5frog]

... it's not THAT old, this thread. ;-)
Have flicker sometimes in blue and red graphics (on an A600), delay line is suspect.
I have recapped it with the expensive kind of hybrid caps, RGB works fine. RF and composite has this problem.
So I'm planning to do this:
https://www.ikod.se/cxa2075m/
Replacing it with a $2 chip and a 2.61kOhm resistor - CXA2075M:
http://www.demand.nu/retrotech/cxa2075m.html

Hope that solves the problems.

[sean_sk]

I'm hoping to find someone that can shed some light on a small issue:

The band pass filters on my A1200 rev. 2B motherboard failed which necessitated the need for me to perform this modification found here:

https://www.ikod.se/cxa2075m/

I followed the guide extremely closely and double, triple and quadruple checked my work. By and large, the mod works and I get a picture. Unfortunately the downside is that the CXA2075 chip gets ridiculously hot to the touch while others are reporting that theirs only gets mildly warm. I measured the temperature with a heat camera which reports a temp of about 95 - 98 degrees Celcius. I'm concerned that the chip will get damaged through extended use at that temperature. I've got a few of these chips and have tried 3 with the same results.

Has anyone tried this mod, particularly on a rev. 2B motherboard and had similar results or may be able to shed some light on why it might be getting so hot or if there is something about the rev. 2B motherboard as opposed to a 1D.x that might be causing this issue?

[Bruce Abbott]

Originally Posted by sean_sk

The band pass filters on my A1200 rev. 2B motherboard failed

Did you remove them?

the CXA2075 chip gets ridiculously hot to the touch while others are reporting that theirs only gets mildly warm.

This suggests something is drawing too much current from it. Can you measure the voltages on each pin and tell us what they are?