by markyk » 13 May 2012 09:46
6 weeks after my last post, I finally got my SL-1800 working (without wobble) - hopefully no problems will develop. Schrei and I have had an arduous journey and I post this in the hope that it can help other people as they search this forum.
So, in my last post, I had replaced the 2 transistors and some low-ohm resistors. I then replaced half of the electrolytic capacitors (those that were available at the local electronics store). The wobble actually got worse. I had placed an order for capacitors so in the meantime, I decided to put back the old 470uF capacitor to see what the difference was.
Well, from that point on, the platter either didn't spin or would spin at high speed no matter what the speed selector switch was set at. The likely culprit when this happens is one of the IC's (1800 only has one). In these models (those with the AN630U chip), and it looks like other Technics models as well, whether or not power is sent to the motor is controlled by the chip - so I guess the moral of the story is to never turn on power without the platter attached - the motor could spin regardless of what the speed selector is set to (I didn't do this - just mentioning it as a caution).
In any case, I figured something happened to the chip but I hoped that some other component was bad and replaced the electrolytic and polyester capacitors (except 2 polyesters, 20 in all), the resistors (17) and the 3 OA90 diodes. At that point, there was only the chip, the 2 polyester capacitors and the 2 power diodes (hard to get) left of the original components.
The platter still wouldn't spin (I didn't think it would) so I began the quest for the AN630U chip. I couldn't find it new but found a parted-out PCB from a similar model. That did the trick. It works - without the wobble.
Some issues I came across:
- It's not easy finding an exact replacement for the 470uF. The one I first used was almost certain +- 20%, not the -10 +50% specified. A direct replacement was available on back order with a long lead time and would cost $20. I ordered 680uF and 1000uF Nichicons, but they were both radial, not axial. Luckily, in the 16/17/1800 (non-MK2) models, there's quite a bit of space near the capacitor and the radial cap can be laid down flat. There's also a large contact area under the PCB so I drilled holes to place the leads. I ended up using the 680uF.
- Did the chip actually burn out? Out of the PCB, I used the diode mode on the multimeter and placed the + lead on pin 16 (ground) and the - lead on all the other pins. The voltage readings were pretty much the same for both the old and replacement chips. I would have expected to see some differences if the old chip was damaged. I didn't feel comfortable trying to do tests with the turntable on - as I said I think the platter needs to be on while the power is on - making it tough to test. I'm not throwing away the old chip.
- If the chip was bad, what made it go bad? As I said, it went bad after putting back the old 470uF capacitor. My best guess is that the power wires started to fray and may have caused a problem. Shortly afterwards, one of the wires did come loose from the PCB. On the 1800, the 6 wires to the transformer and strobe are soldered in. In order to work on the PCB, you either have to do it while connected to the upper chassis or desolder/solder the 6 wires. I put quick-connect terminals on these wires to make it easy to work on the circuit board. If anybody thinks somehow the old 470uF cap couldn't have caused chip damage, I'd like to hear about it.
- I did have a short and I lost 5 fuses as a result. One of the capacitor leads wasn't soldered. It wasn't easily visible with the naked eye - I finally saw it with a magnifying glass. The short occurred after the platter stopped turning so that wasn't the cause of the chip damage.
- At the time, I couldn't find a supplier for the OA90 so I used NTE110A. I've seen people advising against this if possible. The NTE110A's seem to be working fine for now. On the multimeter diode test, there is a slight difference between the OA90 and NTE110A. These seem to be pretty durable so they probably shouldn't be changed unless required. I was just hoping that some other component than the chip was bad.
- Resistors are also pretty durable but as Schrei mentioned, they can go out of spec over time. In my case, I did find 3 resistors out of spec and another 3 that were close to being out of spec. Resistors are cheap and to properly test them means you need to remove at least one lead from the PCB so why not replace them all? (Easy to say for the 16/17/1800 models, I notice some other Technics models have a lot more resistors.)
- None of the polyester capacitors I tested had a capacitance reading out of spec. They could still be a problem (ESR). But again, the ones I chose to replace them because they were cheap. I didn't have exact replacements for the last 2 so I decided not to change them.
