Fixing an Amiga M1438S that only showed red
Last week I picked up an Amiga M1438S CRT monitor for just £20 from Wampa Cave near Norwich. Bargain! These Amiga-badged monitors have a proper Amiga video connector rather than VGA, making them a perfect match for my Amiga 1200. They sell for a lot more on eBay (if working).
The catch was they told me it had a fault. It only showed a red image rather than full colour. I hoped it would be an easy fix so I took a punt on it and spent a few days working on it.
The M1438S is a rebadged Microvitec, from the same family as the Acorn AKF50. I couldn’t find a service manual anywhere. The M1438 page on CRT Database has the specs, the owner’s manual (which includes the connector pinout) and instructions for opening the case, but no schematic. Everything below was done without one.
When opening these up you need a star-shaped screwdriver bit and a paperclip to push down on some clips.
Ruling out the obvious
I started by checking the Amiga and the cable. The Amiga wasn’t the problem, and the cable passed a continuity check on every pin. I also cleaned the contacts on the neck board connector.
The neck board has six trimmer pots: a gain and a B/L (black level) for each colour. I tried adjusting those next, but whatever I did, only red and a faint blue ever showed.
Next I tested every electrolytic capacitor on the neck board with my new FNIRSI LCR-P1, and reflowed every joint, including the tube socket. All the capacitors were healthy. One 30-year-old capacitor from the board even measured better than a new generic one from my parts box.
The capacitors weren’t the problem. The green and blue guns still weren’t firing, so I started testing every component on the neck board with my multimeter to find out why.
Testing the neck board
Luckily, the parts on the board are numbered by channel, and each colour has an identical layout.
| Channel | Part numbers | Status |
|---|---|---|
| Red | 80x (TR801, R807, VR802…) | Working |
| Green | 82x (TR821, VR822…) | Dead |
| Blue | 84x (TR841…) | Weak |
Since the red channel worked, I could compare each green and blue part with its red twin. Any part that didn’t match would be the fault.
The big 4K7 5W power resistors all read about 4.7k. The transistors all tested fine, either out of circuit on the LCR-P1 or in circuit in diode mode. The diodes matched too.
One resistor in the green channel, R829, looked burnt. However, it measured the same as its red twin, so it wasn’t the cause.
The cause
In circuit, the readings on the smaller resistors were unreliable. They drifted, climbed past 140k, then showed OL. So I lifted one leg on a small blue resistor in each channel (R811, R831 and R851) and measured them out of circuit.
- Red, R811: 155.5k
- Green, R831: open
- Blue, R851: open
Both the green and blue resistors had failed open. They are 150k resistors (brown, green, black, orange, brown). They were clearly faulty and worth changing.
I replaced all three with new 1W 150k metal film resistors, including red, so the channels match. Note that this part of the circuit runs at 150-200V, so I didn’t want to use a standard 1/4W part.
I powered it up and green was back. Phew! I had photographed the pots before adjusting them, so I set them all back to their original positions.
I’m very happy I fixed it. The LCR-P1 tester came in very handy for testing the caps, and I made good use of my desolder station again, which was a good purchase.