Memotech MTX - PAL Video Board Fault

System Description : MTX500 Serial No.(26195), 4000-04 computer board with 32k DRAM chips.

Problem Description : Corrupted display output, Modulator and Composite Video

Resolution Summary :   Replaced failed C51

Ismo from Finland had an MTX500 with a Finnish language ROM that had unusable display outputs through both the RF modulator and the composite video output. I got him to do a few obvious checks for such things as bad ground wire connection or loose video wiring, we weren't able to find anything obviously wrong but it appeared to be a faulty video board that was the problem. Timo (also from Finland) was able to prove that there was a video board fault by swapping parts with his own machine, but was not able to do any further fault finding, so Ismo asked me to have a look at the video board for him. (To save money, only the video board was sent from Finland to the UK)

The first problem reported was that the screen "rolled" when using the composite video output, but the RF output was OK. This is illustrated in the small video that Ismo sent me. 

Subsequently, Ismo found that the modulator output also became very poor after the machine had been on for some time. This is typical of the behavior of many RF modulators of this era though, they degrade with age and very often drift off station as the system warms up so I didn't think that this symptom was related to the composite video problem.

My initial thought when Ismo mentioned a rolling picture was that there was a problem with the video sync signal, this is generated by the VDP and, as described in my MTX Video technical note, is fed to the video board through J13 Pin 5. This is the "Y" signal and comprises the luminance and composite sync signals. Given that the same symptoms were present with the video board plugged into another MTX, then the VDP itself was not the issue and as the sync signal is also used in the modulator circuit, then it was also working correctly in at least part of the video board. 

When I tried the Video board in one of my MTXs, I didn't see the picture rolling, but there were obvious problems with the composite video output. The colour shown here isn't due to the quality of the photo - the blue was much darker than it should have been and there is a purple band at the left hand side of the screen - something was definitely amiss!
The schematic of the PAL Video board is included in the MTX Operators Manual - the best copy being in the "Phoenix" manual available on the Manuals page on the site.
There is a link on the video board to select either colour or monochrome output. A simplified schematic of the video board is shown here, it can be seen that setting the link position to "b" bypasses all but a couple of components, taking the mono signal straight off the VDP and just amplifying it with a single 2N3906 PNP transistor.
The link is soldered but it was a quick job to re-pin it in the "b" position to see whether the fault was common to both the colour and mono positions. Not a great photo and the ghostly outline is my shadow :-) You may not be able to tell from the image, but there was definitely a paler band at the left side and some bleeding from the text, i.e., I was able to conclude that the fault was also present with the monochrome output.
As shown in the simplified schematic diagram above, the fault was going to be with R62 (390R), R64 (10K), C51 (10uF 16v electrolytic) or Q6 (2N3906). The resistors were unlikely to be the problem but in normal circumstances, I would have put a meter on them just in case. However, the video board is pretty congested and I could not easily identify the resistor locations, so just assumed that they would be OK. I didn't have a very good selection of transistors (in fact, there were only 3 in my multitude of component boxes!) and I didn't have a 2N3906 although I did have a BC557. A bit of "Googling" revealed that it was a  PNP transistor with similar specifications to a 2N3906, but the terminal positions are reversed. As I decided that I really should have had some spares for transistors on the video board, I placed an order for some, but for testing, I was going to temporarily install the BC557 "backwards" if the transistor was looking like the culprit. Similarly, although they were not going to be needed for this repair, I also took the opportunity to order up some CD4013B and CD4016B ICs that I also didn't have in my spares boxes in case they were needed for future video board repairs. C51 is an electrolytic capacitor that was now at least 35 years old, as electrolytic capacitors degrade with age, there was a fair chance that it was faulty and I did have a few capacitors of that type, albeit it that they were slightly taller than the original. I replaced C51 with a new one, leaving slightly longer leads so that I could bend it over to ensure that the video board would lie flat when installed.
C51 is located to the right of the vari-cap, just above the transistor. You can see how I angled the replacement in a similar way to the larger capacitors that Memotech treated the same way. The original C51 was much smaller and originally mounted vertically.
With C51 replaced, the video board output is now working perfectly - this photo doesn't do it justice, so . . . . .
A somewhat self congratulatory screen shot that shows the output from the fixed video board.
Although I didn't see the "rolling" screen reported by Ismo, I am pretty confident that the video board problem has been fixed so will be arranging to send the board back to him. Obviously this repair won't have any impact on his modulator issue, but I expect that he won't be using the RF output now anyway.