MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

TR-606 Restoration - No Pattern Write Mode?
MUFF WIGGLER Forum Index -> Music Tech DIY  
Author TR-606 Restoration - No Pattern Write Mode?
Sunshine Jones
Hi,

I've got ahold of a TR-606 and I've opened it up and given it a good and thoughtful cleaning and inspected it.

Initial behavior:
Sounds function
Battery wires have been disconnected (patterns are silly)
Patten Play Mode / Track Play Mode appear to be functional
No response from Pattern Write Mode

First step:
Inspected the board for corrosion, broken jumpers, bad wires, cut traces.
Everything looks pretty nice actually.

Second Step:
Replaced the main CPU with a pulled one (from a working 606 which was upgraded with the QS mod and passed down to me.)

Results: No change.

Third Step:
Replaced the RAM IC's with replacement IC's (same source as the CPU, working perfectly when pulled and in great condition.)

Results: No change.

Fourth Step:
Read up about other people's troubles with the 606 and double checked the device for cold solder joints, continuity between all points on the Mode potentiometer, all jumper wires, the DIN Sync Jack etc. Everything appears to check out.

Fifth Step:
Took notes of all the IC's (logic, etc) and ordered replacements

At this point, since it doesn't appear to be the CPU or the RAM, I'm guessing it's going to be the TC4011 or other such logic, but I'm not exactly sure why this might prevent the entire Pattern Write Mode from working or coming on line when selected.

So I worry I'm off on the wrong path...

Has anyone encountered a complete omission of write mode on a TR-606 repair or restore?
What was your solution?

I'm open to any ideas to follow up on.

Thanks!
guest
both pattern and track write do not work? how does it respond when you put it into write mode? does it act like play mode? or do the buttons respond as if it were write mode, but nothing gets written? or is it just completely unresponsive?
Sunshine Jones
guest wrote:
both pattern and track write do not work? how does it respond when you put it into write mode? does it act like play mode? or do the buttons respond as if it were write mode, but nothing gets written? or is it just completely unresponsive?


The device acts as if the mode hasn't been selected.
So play and write modes are identical.

More symptoms:
Can not tap the sounds
Selecting different instruments does not change displayed Led's
Led presentation of entered steps does not display (led's are functioning)
Scale mode does not change
Led's do change in track play, and patterns do switch

Seems like the write mode is just not there at all...

Next thing to try is to replace the IC's - Get a new NAND gate and Hex Trigger IC in there and see if that doesn't help, but I think the thing I am suspecting that I don't know is why wouldn't the mode at least be accessible?
It's as if the mode is just gone.

Again, I have tested the pins of the selector potentiometer (SRM 1034) for continuity, and I get a beep at the next (and then the next) pin for the whole thing.

Crazy-pants!
Sunshine Jones
Further down the line:

I had a 4011, 4013, and 4001 on hand, so I removed them from the pcb and socketed them, and replaced them with fresh, new IC's.
Still no pattern write mode. No change in behavior.

I'm going to proceed and replace all the IC's (I got my eye on that AN6912) and we'll see... but I can't really figure out why any of this would prevent an entire mode from being entered into. This is the neural pathway I need to create for myself here.

I'm open to following any leads, or if anyone else has encountered this, or something like it, I'd love to hear your story.
guest
so, its possible the switch or diodes or leads from the swtiches are busted, and its just not getting into write mode. if it was a memory issue, it would still switch modes. a few things to check real quick:

are the powersupply rails present on all ics? the memory chips are fed from the batteries, and if a diode is gone in the powersupply, that wont switch over from the wallwart supply.

is anything else wonky in pattern or track play mode? do those work 100% as should?

do you have an oscilloscope? the next thing i would check is if the switches are getting scanned and read correctly. but, this is hard to do without a an oscilloscope as the lines are pulled high for a very short period of time.
Sunshine Jones
guest wrote:
so, its possible the switch or diodes or leads from the swtiches are busted, and its just not getting into write mode. if it was a memory issue, it would still switch modes. a few things to check real quick:

are the powersupply rails present on all ics? the memory chips are fed from the batteries, and if a diode is gone in the powersupply, that wont switch over from the wallwart supply.

is anything else wonky in pattern or track play mode? do those work 100% as should?

do you have an oscilloscope? the next thing i would check is if the switches are getting scanned and read correctly. but, this is hard to do without a an oscilloscope as the lines are pulled high for a very short period of time.


Amazing. Thank you.
Will investigate all of this carefully and reply with results.
Sunshine Jones
Battery Report
I connected the battery wires to the bottom case, and installed fresh batteries.
The behavior is the same. No change.

In terms of the other modes, I find it challenging to tell with the strange scaled, short length parts that the bateryless and unclearable state leaves this in, but it certainly appears that the Track play functions (moves through a series of patterns) and Track write appears to function (allows me to change those patterns) and pattern play appears to be just fine (select bank A or B, pattern 1-16 and it will play what's there.)

It's just Pattern Write that's missing...

I'll move on and look at the diodes, and try my best to follow the leads out from the switch and repost back more.
guest
well, if you dont have an oscilloscope, you can shut the machine off, and use your multimeter to make sure the switch is making good contact through all of its positions. can you select the various drum sounds individually? does that rotaray switch work? in particular, can you select the bass drum?

you can also go through with your multimeter set to diode check mode, and check all the diodes. do them in both directions.
Sunshine Jones
Diode Test Results

I tested all of the diodes and got some surprising results.

Most of the diodes are IS2473 (cute, tiny little things) and they measure either 0.59v - OL or the other way around OL - 0.59v which seems normal to me.

D43 is a different type of diode and it reads 0.357v - OL and that seemed normal to me too.

Here are the exceptions:

Position Positive/Negative Negative/Positive
D3 0.53v 1.43v
D6 1.69v 0.59v
D8 1.7v 0.59v
D9 1.8v 0.59v
D10 1.8v 0.59v
D11 1.8v 0.59v
D36 0.59v 1.59v
D39 1.35v OL
D41 1.13v 0.57v

Is this possibly 9 bad diodes? Dang.

I don't currently have any IS2473 Diodes, it seems I can still order them, but I do have a decent supply of the usual 1N4148 I could try swapping (1S2473 is compatible with the ECG519 diode and I found out the ECG519 is compatible with the 1N914 diode, which is compatible with the 1N4148 diode so unless I was attempting to preserve the original parts for aesthetics then I would order the correct, exact parts. However if what I have on hand will work then what the heck?)
Sunshine Jones
Well it's fixed! I can hardly believe it, but it actually works!!!

Here's what did the trick:



After that diode test I started wondering if I couldn't just start swapping the weird ones with 1N4148's.

It's been said that 1S2473 is compatible with the ECG519 diode and the ECG519 is compatible with the 1N914 diode, which is compatible with the 1N4148 diode...

To avoid the heartache of either madly ordering more components in the middle of the night and waiting a couple weeks for them to arrive, OR further wear and tear to discover that no, the 1N4148 is not an acceptable replacement for the 1S2473, I moved on to the next test.

The Mode Selector switch only had 4 positions.
Many of the pins go to little islands.
Of the pins which travel out into traces to be followed only the last pin was strange...

The pin beeped for continuity at point 2, but it did not at point three.
So I added a little jumper between points 2 and 3 and plugged it in and like an exhilarating force of nature the TR-606 has Pattern Write Mode again!!!

Last Step functions
Pattern Clear functions
Scale Selection functions
Quarter note display functions
And most importantly the instrument select and step input functions.

I'm so grateful to you for your help. Thank you.

Now to order a set of switches. I only have 2 spare here, and it seems like I'm going to need 10 of them. May as well order a set of 21 and replace them all.

This is just great!!!!
guest
im glad to hear you found it. lifted/cracked traces are pretty common on those old boards. my 303 had a similar issue.
synchromesh
guest wrote:
im glad to hear you found it. lifted/cracked traces are pretty common on those old boards. my 303 had a similar issue.


I think this is what's wrong with my 606; it mostly works, but if you flex the case a little bit then it goes berserk. eek!

I figure the PCB is a bit "rotten" so I've been waiting for Paul Barker to decide to finish his RE-606 PCB(s), but then I'll have to transplant each part to the new PCB. I can't decide whether this would be more or less tediously unpleasant than trying to track down the cracked traces in the original PCB, particularly when you have to keep bending the board to get it to fail.

Can I ask you guest what your advice would be? How did you isolate the location of the problem(s) with your 303 PCB? Cheers!
meatcliff
Happy that Sunshine Jones got it running smoothly - bad RAM, switches, and traces are definitely most common for functionality issues with the 303 & 606 from the dozen or so I've worked on.

Synchromesh - Which part of your 606 goes berserk?

With harline fractures, my advice is to isolate the issue to the area that's misbehaving and go through it trace by trace. One thing that helps a lot is either loading the schematic into graphics software and highlighting traces and checking them joint to joint, or doing the same with a print out. On the computer it's easier to color code the power and different signal paths.

Working this way will also help you isolate if the issue is on a specific component instead of the traces. Something may look fine, but be damaged internally. The catch 22 with swapping everything to a repro board is that you could end up with "restored" 606 with the same issues.
Sunshine Jones
synchromesh wrote:
Can I ask you guest what your advice would be? How did you isolate the location of the problem(s) with your 303 PCB?


Yeah, visual inspection of this one didn't reveal any obvious trace issues, nor visually observable problems. To be fair, it was seriously toast and headed for the bin when I intervened and began to look into restoring it.

The best direction I received here (anywhere) was to be patient, and go over the traces and points until faults are found.

The tricky part for you and your machine is that it doesn't do it unless you stress the board. Please don't demonstrate that too much.
Have you followed the schematic and the traces with a multimeter for continuity on the bottom of the pcb?

That's how I found this problem and was able to jump it no sweat.

I'm absolutely going to undertake the RE-606 (and the RE-808) when they're ready, but there's something about restoring and repairing the original that has a lot of value for me.

Worth a cursory once over for sure.
Sunshine Jones
meatcliff wrote:
The catch 22 with swapping everything to a repro board is that you could end up with "restored" 606 with the same issues.


I agree. Personally, I would hunt down the original components and create a brilliant, fresh, new RE-606 rather than remove the components from the TR-606 PCB. They DO come out nicely with speed, patience, and focus. I was actually really surprised at how nicely they came out, but a brand new edition with fresh electrolytics, and ll made by my own hand would be super thrilling...
Sunshine Jones
One thing I can't seem to determine:

I have no screws for this unit. It was (as I said) headed for the dumpster when I intervened and paid the guy off to let me at least try to save it.

The service manual says 3 x 10mm, 3 x 18mm B1 Fe Cr Binding Self Tapping screws...

For the life of me, I can not determine the B1 screw format. What's the diameter? the threading? Seems to be gone from the world...
guest
Sunshine Jones wrote:
For the life of me, I can not determine the B1 screw format. What's the diameter? the threading? Seems to be gone from the world...


B1 refers to the point (its a flat point - ie no point). all you need is a "self tapping screw for plastic" with pan head. they look like this:



your local hardware store might have them. they have a really course thread, no point, and a pan head. an english #4 screw should be close enough to work as well.
guest
synchromesh wrote:
Can I ask you guest what your advice would be? How did you isolate the location of the problem(s) with your 303 PCB? Cheers!


id echo meatcliff here. i had the same issue, where it only failed on bending, or sitting at a certain angle. basically, the output would cut out, so i narrowed it down to the VCA/headphone section. i then visually inspected, but didnt see anything as it was a hairline fracture in the ground plane. so i went through with a multimeter while flexing the board.

one other thing you can try, i just reflowing all of the solder joints in the affected area, maybe applying a bit of flux. sometimes its a pad that has lifted, but you cant see that because the break is right where the solder ends, so there is a visual line there already.

is it just board flexure that causes it to fail, or do certain pots/switches/jacks cause it? there are multiple boards inside, so you can flex them individually if you open it up. you can also press locally on sections of the pcb and see if that makes a difference. the stuff that you touch (pots/switches/jacks) is where the most stress is, so around these areas is the highest probability of failure.
synchromesh
meatcliff wrote:
Happy that Sunshine Jones got it running smoothly - bad RAM, switches, and traces are definitely most common for functionality issues with the 303 & 606 from the dozen or so I've worked on.


Yes, sorry Sunshine Jones, it was rude of me to ignore the OP... oops Congrats on tracking down the problem, I admire your perseverance. It's great that you've saved this 606 from oblivion! screaming goo yo w00t

And thanks to meatcliff, Sunshine Jones and guest for your advice and suggestions.

meatcliff wrote:
Synchromesh - Which part of your 606 goes berserk?


I have recorded some video to demonstrate the issues (at 4:42 I press RUN/STOP FWIW). I don't think the issue is with the rotary switches particularly. Once I stop the unit playing I still get some triggered notes out of it, so perhaps the problem is on the switch board.



The switches, pots and LEDs were replaced after I bought the unit; as I recall everything was more or less OK, but pretty soon it was acting up pretty severely. I have had a look at the PCBs but I clearly didn't approach it in a sufficiently diligent and systematic manner.

I do have a nice copy of the service notes so I could copy the schematic & PCB layouts and start working my way through.

If anyone can sit through my video and offer any more comments/advice, that would be most appreciated. SlayerBadger!
guest
there used to be replacement switch pcbs you could buy, but i dont think they are available anymore. that would be a pretty quick way of testing it out.

id reccomend opening the 606 up, and spreading the pcbs out a bit. then set it running in pattern play mode, and apply pressure to the various boards individually. maybe you can narrow it down to a single board that way. you can also try pulling on the cables between boards.

its looking like something on the switchboard.
synchromesh
guest wrote:
id reccomend opening the 606 up, and spreading the pcbs out a bit. then set it running in pattern play mode, and apply pressure to the various boards individually. maybe you can narrow it down to a single board that way. you can also try pulling on the cables between boards.


Thanks guest, I'll try that first.
meatcliff
Weird behavior... I'd check very carefully at the ribbon connectors and wires going between the switch board and the main board for frayed strands that might be shorting.
MUFF WIGGLER Forum Index -> Music Tech DIY  
Page 1 of 1
Powered by phpBB © phpBB Group