Amiga 1200 CXA1145 and composite/rf output

[Solo761]

I had problems with getting my A1200 to show picture on my LCD TV via either RF out, composite or RGB to SCART. Then I tried it on CRT TV and picture was fine over RGB to SCART, RF out and composite were still faulty.

In the meantime I recapped it and replaced RF modulator (old one was kinda busted) but it didn't help in composite or RF modulator output (although RF out is no longer just snow, you can notice kickstart diskette in the background). I thought it could be bad caps that were causing this, but recapping (I used panasonic caps from rsdelivers, not "panasonic" from ebay ) changed nothing.

So now I'm wondering what exactly CXA1145 does? Does it handles all video outputs, RGB, compsite, RF out, or only composite and RF out? In other words, could faulty CXA1145 be doing this? This is how it looks like over composite, on both, LCD and CRT tv's. I don't have picture of RF-out, it looks less snowy, you can see purple kickstart color but it jumps up-down and left-right. I tried finetuning on both TV and RF modulator but it didn't help.

If it could be the culprit I saw them on ebay so I would replace it.

[roy bates]

the cxa1145m is a video encoder.it handles rgb and composite.
take a rgb lead and see if you have rgb through scart.which might work.

the rf modulator looks like its fed from the composite output on the cxa1145,hense the comment about rgb cales.


just look up the pinouts of the device youll see what it does from that.

http://www.google.co.uk/url?sa=t&rct...,d.d2k&cad=rja

[Solo761]

According to this datasheet and A1200 schematics from http://www.ianstedman.co.uk/Amiga/sc...chematics.html RGB goes in CXA1145 and goes out as composite and to RF modulator. RGB port lines goes directly to video DAC.

I replaced capacitors a month ago and it didn't help with this issue. I doubt resistors could create this type of fault, although I'll check them just to be safe. I guess chances are pretty big that CXA1145 is at fault here :/. Unfortunately I don't have an oscilloscope so I can't check it's logic, I guess it would be easier to just desolder it and solder new one.

[roy bates]

yep just desolder the old one and replace it,see how you get on.

[hooverphonique]

I have a used CXA out of a sat tv decoder that you can have for free, if you like.. I can't guarantee that it works, however..

[Solo761]

@roy bates

It seems that's the easiest way to do try it. And probably faster than testing via oscilloscope .

@hooverphonique

Thanks but it'll probably cheaper (postage wise) to order them from China . Plus, you just gave me an idea, I probably have some old tv stuff I can desolder it from. I wonder where my old analogue sat tv decoder went, maybe there's one in there

[roy bates]

http://www.ebay.co.uk/itm/CXA1145M-S...item2c664c2a78

i found these on ebay,the seller also sells them in packs of ten,ive had them of them before there new i think.

[Solo761]

Yup, I saw that one too, that's why I thought about replacing it. I also think I ordered some components from that seller before.

[hooverphonique]

that's up to you... postage for 50g from denmark to croatia is like 1.5€,100g 3€, which is so cheap, I won't bother to charge you :-)

[Stedy]

Hi,

Check the delay lines, Z221 and Z222 + support components, R223, C222, R224.
The image shows a video signal where the chrominance information is interfering with the luminance and should not be. The aforementioned components affect both the RF and composite outputs.

Try cleaning the PCB using IPA around the CXA1145, this may help.

Ian

[Solo761]

Originally Posted by hooverphonique

that's up to you... postage for 50g from denmark to croatia is like 1.5€,100g 3€, which is so cheap, I won't bother to charge you :-)

Thanks, postage wouldn't be a problem. It's just that if I'm going to replace it then it's better to put a brand new one inside . Thanks again.

Originally Posted by Stedy

Hi,

Check the delay lines, Z221 and Z222 + support components, R223, C222, R224.
The image shows a video signal where the chrominance information is interfering with the luminance and should not be. The aforementioned components affect both the RF and composite outputs.

Try cleaning the PCB using IPA around the CXA1145, this may help.

Ian

Thanks, I'll check them and try cleaning again. I recapped it a month ago and cleaned parts of MB where capacitors were.

[Solo761]

Today I finally checked these components and they seem fine, no breaks in lines also. Only thing that I've noticed is that R223 doesn't have resistance. It should be 1.2k or 910 ohm, but I got only 0.7 ohm (definitely not 0.7k)...

On a sidenote, I found out that my current PC monitor supports 15 kHz horizontal frequency . It's Dell UltraSharp U2311H LCD monitor, after a bit googling I found that this series seems to support 15 kHz. Although colors were off. But it may be cable's fault. I'll have to check the pinout. Here's how it looks, kickstart and workbench.

[TomCrazy]

A basic rule of thumb for successful electronic troubleshooting is to always start off by checking the supply voltage(s) at a certain suspected active device.
Start off by checking the 5V supply pins for the CXA1145.
If everything seems ok, desolder and check the ceramic capacitors that are AC coupling the RGB signals into the video encoder chip.

Like Stedy mentioned, it's also a good idea to have a look at the delay line and the band pass filter, although I think it will be pretty difficult to come to any conclusion about these components without having access to an oscilloscope.

According to my experience, it's much more common that ceramic capacitors fail rather than ICs. Chips can of course also fail, but if I were in your situation and wasn't able to check the signals with a scope, I would then definitely start off by replacing the three AC coupling input caps at the video encoder, C211, C212, C213. If it didn't solve the problem, I would then probably replace the CXA1145, and lastly the DL and BPF since these two components can be rather difficult to get your hands on as spare parts.

Edit:
Just saw the updated pictures you posted. So I take it that you're feeding RGB into the monitor here?

[Solo761]

Yes, simple homemade DB23 to DB15 adapter. It's connected like this schematic.

[TomCrazy]

The VSYNC (pin 12 Amiga side) is going into pin 15 on the VGA connector in the diagram, although it should go to pin 14. But I suppose you hooked up the wires correctly when making the cable or you would be having sync problems.
Except for that, I see no problem with the diagram you posted.

Another thing you could do btw is to check the DC resistance of some of the ceramic caps in the video section while the power is turned off. A short between the two terminals seem to be a quite common failure mode for ceramic capacitors. You can even check the DC resistance of capacitors that are acting as AC coupling caps while they are still in circuit, as long as there is no other circuit element hooked up in parallel with these capacitors. C211, C212, and C213 are examples of capacitors that can be quite easily checked without having to desolder them from the board.
With decoupling caps, however, it's a completely different story... Although I don't really believe that a decouling cap is what's causing this problem for you.

[Solo761]

Yup, i drew it wrong, cable is connected correctly. Pin numbers above the diagram are correct. Why this monitor displays color funny is something else. It's mostly peculiar that there are rather new monitors that support 15 kHz. I bought it cca 2 years ago and it wasn't something special. My reason for buying it was that it had IPS panel.

I checked them yesterday, they weren't shorted, I don't remember exactly, but they had at least 20+k resistance. Only thing that was kinda off was that R223 resistor which had 0 resistance.

[Stedy]

@Solo761

Can you remove R223 from the PCB safely and then measure it?

If the bandpass filter, Z221 has shorted, it would null R223. Between pins 3 & 4 of the device, should be a few kilo ohms, and between pins 1 & 5 should be 600-700 ohms in circuit.

Also check C221, it should be high resistance to DC.

RGB output works as this uses the TTL HSYNC/VSYNC signal from Alice and bypasses the CXA1145.

Ian

[Solo761]

Well, I never desoldered SMD components, but it doesn't look too hard. But I think I'll order some new ones first. If I'm going to desolder it I might as well solder new one back. It should be an 910 Ohm one (code on it is 911, which should be 910 ohm) And while I'm doing it I might as well change these ceramic capacitors.

Yup, I saw that in A1200 schematic on your site. My main suspect for this issue is CXA after all, one of electrolytic capacitors that's near it leaked. Not much, there was nothing like corroded lines, just that solder around that cap lost it's shine, plus it had that characteristic stink when I desoldered it.