TR 606 going insane

[magnus]

Hi, hope there are some x0xperts out there, sorry about the long post...
Today I took possession of a rather neglected 606 and found that it worked almost perfectly at first (cymbal sometimes seemed to file itself, and all of the patterns needed wiping), but soon went haywire.

I understand that memory can be a problem with the 606, but all of the problems seem to be related to the sequencer on this one. The first big problem i noticed was that in write mode, the keyboard acted as eight measures repeated once rather than 16 measures with no repeats. So, i for example I place a kick on measure one, the LEDs for 1 and 9 both light up, and the pattern played corresponds (most of the time).

Then write mode went completely nuts. It started playing unwanted sounds that I could only remove sometimes, with the LEDs lighting up at strange intervals. Finally write mode stopped playing the notes that I entered completely, marching only to the beat of it's own drummer, as it were.

Through all of this, pattern play mode has been largely unaffected besides the eight measures repeated thing, although it does occasionally misfire too.

Any ideas? I really surprised me that it deteriorated so quickly (within an hour) As I say, I don't think memory is a problem - in fact, I tried to reset the 606 by removing the batteries, and found that it retained its memory for far longer than it should. Also, I had a quick look inside - it was a bit grubby, and one of the screws by the battery was corroded by battery acid, but I couldn't see any damage to the board.

Does anyone know I'm on the right track thinking this is a sequencer problem, and if so, which IC i need to replace? I noticed that DINsync.com sell a RAM IC. Would this end my woes?
Cheers
Magnus

[nathanscribe]

I seem to remember mine doing something similar a while back. From what I recall, I cleaned the switchboard and it was fine. Mine also had a rough old individual-outs mod, which I removed.

For what it's worth, I never use batteries in my 606. I just write down any programs I want to use again, and re-enter them when necessary. Is it playing up when you run off batteries, or are you using an adapter? If you take out the batteries, then use only the adapter, does it still malfunction?

I wouldn't run out and buy any expensive old ICs just yet. It's probably something simple.

[magnus]

Thanks! It does seem to be a lot better without batteries -how bizarre. It's still refusing to let me write a full 16 measure bar though - just 8 measures repeated. Could it be turned on to a particular function that I'm not aware of?

[justinvm]

In write mode, press function and 16 simultaneously. that will extend your pattern to 16 measures.

[magnus]

thanks, but it's already set for a 16 measure sequence - it's just that whatever i write in the first 8 measures lights up and plays in the second 8 too, and vice versa.
Also, despite removal of the batteries, the sequencer has gone mad again. now all the sequences are full with every sound, and can't be cleared, so any more tips would be much appreciated

[Solderman]

Sounds to me like you have corrosion on the pins of either the CPU or memory. Battery acid leakage, even if it does not settle on circuit traces or components, can leave a corrosive atmosphere inside the case if left for very long periods unchecked.
I've never tried to do anything but replace components with corroded contacts, so maybe a much more experienced technical person can suggest what to do.

[justinvm]

Solderman is right. There is a much higher probability that there is a problem with circuit contacts/traces than there is faulty ics. My suggestion would be to test all the traces between cpu and memory and surrounding logic chips. Check that they are conducting properly. Check all solder joints.

If there is a faulty trace between the cpu and ram then the cpu could be writing to incorrect memory locations. Thus doubling up on your sequences. Check the traces to the ram first.

[magnus]

Thanks a lot. I see ICs 7 and 8 are involved in memory. They are also the closest ICs to the leak. Are these the ones I need to test around? I'm no expert in electronics - how would I test the current running through a trace rather than the pd between two solder points?
cheers

[justinvm]

Well if there is battery acid leakage on the pcb then I would strongly recommend cleaning the pcb as your first priority. You want to neutralize the acid to prevent further damage.

Testing traces is straight forward enough. You'll need a multimeter to do this. Make sure the 606 is powered off, disconnect power adaptor (this is only so your readings are accurate). Set your multimeter to measure resistance. Identifiy a trace you want to test. Find 2 solder points that the trace is connecting. Place the multimeter terminals on each of the solder points. If the meter measures less than say 100 Ohms across the trace (It should say 0, but traces can provide some resistance) then the trace is working properly. If you're getting infinity ohms or a number higher than 100ohms then the trace is broken/damaged and is not conducting. The easiest way to repair a trace is to take a wire and solder it onto the 2 points it's supposed to be connecting.

Checking solder joints themselves is tricky because it's hard to see if the solder is actually connecting the pin to the circuit board. Sometimes you can see that a joint is loose/broken. Other times the pin can appear to be soldered looking at the exterior of the join, but the interior of the solder is making no connection to the pin at all.
You can do similar tests as the trace test above. But this time put one terminal of the meter on the leg/pin of the ic and the other on some point along the trace that the leg/pin is connecting to. If it is not conducting resolder the joint, check the trace.

I'm pretty certain that the problem is PCB related and not IC. So just hang in there. Test everything. Take breaks every fifteen minutes. You will find the problem(s) eventually. I found 20 broken traces in my tr-909 before I got it working again.

[magnus]

Okay, thanks so much - I was beginning to despair. I reckon I've got a long and arduous task ahead of me, but at least I know what I have to do. Things haven't started too well - after cleaning the residue from the switchboard PCB, the 606 is refusing to run sequences. (I suspect the run/start switch has got water in it; is there a way to test this switch? is it push to break?)
Anyway, I'll just need to slug on I guess. Thanks again

[justinvm]

Why would the switch have water in it? Use water sparingly on pcbs as it promotes rust and can lead to short circuits. Always apply with a cloth and let it dry before you turn anything on (a few hours at least). I use compressed air to dry things quicker where appropriate.

You can test the switch by making a short circuit across the pins that it would be acting upon. Check the schematics for the switch. locate the appropriate traces. Simulate the short circuit that the switch creates using a wire between 2 solder points on the switch.

Theres probably another broken traces in there now, hence patterns not running.