USC2 - narrowing list of suspects for selection problems

[danno]

Been working on our USC2 (Firestar) for a while now, and am currently diagnosing some selection-related issues.

I started with the "scratch" tests involving the J3104 and J3101 connectors on the DCC, and all of the required

"key #s" played successfully (with an add'l selection on a few of the tests).  I believe this is a good sign that

the DRD-1 is OK, and possibly the DTP-1 (not as sure about that assumption).

Then, I stepped through "Phono Procedure Chart #3A from Ron Rich's trouble-shooting guide, and well....I think it

shows how the jukebox seems to have a mind of its own sometimes.  The test results are posted below.

When I first started chasing this problem down last year, there were some seemingly-random problems like I'm

having now, but there was one static problem in that I could never get the 4 key to register.  I went through

Ron's service notes at the time for the DES, and sure enough, there were a few bent/broken "fingers" on the switch

for number 4.  To remedy that, I swapped with fingers from one of the "reset" buttons (per Ron's guide).  I also

cleaned the PCB boards with Lava soap, applied Luberex to pads, etc.  Afterward, the issue with the 4 key was

fixed, but there continued to be some flakiness with random selection errors.

A question I have:  Is there an easy way (using multimeter) to test the contacts at the DES to try and further

narrow the problem down?  It seems possible that a bad component (or several bad components) on the "buffer

board" (311134, I believe) could also be involved, and I saw another post in this forum where a bad transistor

was blamed in that case, but I think ruling out the DES first (if there is a solid way to do that) would be a logical first step.

I appreciate any feedback, especially if you've had luck chasing one of these problems down before.

Selection testing results:

Selection entered          |   What actually happened / played

--------------------------------------------------------------------

100                              |  1 - 0 - <Re-set and Re-select> light lit up

100                              |  1 - 0 - <Re-set and Re-select> light lit up

100                              |  100 played OK

110                              |  1 - 1 - <Re-set and Re-select> light lit up

110                              |  110 played OK

122                              |  122 played OK

133                              |  1 - 3 - <Re-set and Re-select> light lit up

133                              |  1 - <Re-set and Re-select> light lit up

133                              |  133 played OK

144                              |  144 played OK

155                              |  153 played instead

155                              |  155 played OK

166                              |  161 played instead

166                              |  116 played instead

166                              |  111 played instead

166                              |  111 played instead

177                              |  111 played instead

177                              |  111 played instead

178                              |  118 played instead

278                              |  218 played instead

203                              |  203 played OK

236                              |  231 played instead

[jukenorman]

Hi Danno,
I don't work on these and I don't have Ron's guides so can't help much at all, sorry. All I can suggest is to look through the posts on the forum to see if you can find something similar. There is quite a lot of history here, especially for Seeburg models.

Good luck, Norman.

[danno]

Thanks at least for the reply, Norman. Even if an answer can't be offered, it is at least assuring that others are still watching this forum!
And, believe me...I've been through many of the Seeburg threads in this forum already, and have a few of them bookmarked.

Right now, I have the DES (keypad assembly) removed from the jukebox and am going to dis-assemble again and take a closer look. I'm wondering
if the timing of the main trigger switch might be off. The DTP-1 ("Black Box") apparently may still be faulty, but the "Re-set and Re-select"
lights happening seem, to me, to lean more toward the DES.

Anyway, the learning adventure continues! I'll post back here with any interesting updates as I figure things out.

[e093116]

Hi,

If you've cleaned the DES and the contact fingers are in good condition and are all at the same "height"/level with each other, that portion is probably ok...don't overthink it.

When I've had the reset light come on randomly, I've been able to trace that back to poor connection between the connectors and circuit boards on the back of the DES.  I think the guide talks about adjusting the contacts in the connector to provide a bit more contact pressure, and also make sure the connectors are seated fully and squarely.  

Also, I think there are diodes or some similar component on one of the boards on the back of the DES (I can't remember but the boards are pretty simple), if those are "bad" you can also get the behavior you've observed.  I think there are some notes on this in Ron's guide too.

You're probably closer than you think to getting the problem sorted, keep after it and good luck.

Bill

[tonyj]

For kicks I wrote a little Arduino program to hook up to a DES and will display what key is pressed based on the status of the 4 data lines A through D.  If you're interested, I can send you the code and some wiring notes on how it connects to the Arduino.  The Uno model is what I did this one on, and you'll need an edge connector to make up a harness to your DES.  I ended up having a 22 position card edge connector with .156" spacing that I used.

Mar 28, 2023 20:01:40 GMT danno said:

A question I have:  Is there an easy way (using multimeter) to test the contacts at the DES to try and further

narrow the problem down?  It seems possible that a bad component (or several bad components) on the "buffer

board" (311134, I believe) could also be involved, and I saw another post in this forum where a bad transistor

was blamed in that case, but I think ruling out the DES first (if there is a solid way to do that) would be a logical first step.

[danno]

tonyj - Whoa...that sounds awesome!  Do you happen to know if Arduino programs can easily be ported to work with Raspberry Pis

(since I have a Pi 3 model B available)?  I'd be interested in pursuing that, if so.

e093116 - Thanks for your input.  I checked the pins on the Molex ("eggshell") connectors early on, and they all seemed to

be positioned properly.  I believe the 'daughter' PCB board on the back of the DES is for Album vs. Single selections (the label

on mine is worn off pretty badly, but from what remains that's what it seems to involve).  Ron said that the switches being

'in between' positions could cause issues, but mine seem solidly to the Single selection position, and visual inspection of the

board & components "looks OK".  And, thanks for the encouraging words and the caution about 'not overthinking things'...it's

like you know me well  

As a quick update:  I took this opportunity to disassemble everything with the DES:  I took each keypad button out, soaked them

in Dawn dish detergent and used a Q-tip with 91% rubbing alcohol to clean the interior of the buttons.  That didn't totally get

rid of the nicotine and fading, but I was feeling the light coming through from the fluorescent bulb wasn't illuminating the keys

as much as it should, and I think that will help. 

When inspecting the DES components, I was encouraged that no new 'fingers' appeared to be bent / broken.  Before disassembly,

I checked the timing of the keys firing the main switch, and the first alignment notch was sometimes not visible yet for some of

the keys.  After I reassembled everything, I adjusted the 2 'alignment screws' (it was a bit of a finicky process) so that all the

keys now seem to activate the switch with the first notch showing completely, but before the second notch comes out.  I think

I missed that the first time I dis-assembled the DES, and it seems reasonable to hypothesize that caused some of the selection

problems due to the fingers not making sufficient contact, leading to Re-set errors or wrong quadrabit translations.

I got everything back together in the jukebox, but right now we are having thunderstorms here in western NY, so I am going to

wait before plugging it in and testing things out.  I'll post back with results (and a few photos) once I get the chance.

Mar 31, 2023 15:57:10 GMT tonyj said:

For kicks I wrote a little Arduino program to hook up to a DES and will display what key is pressed based on the status of the 4 data lines A through D.  If you're interested, I can send you the code and some wiring notes on how it connects to the Arduino.  The Uno model is what I did this one on, and you'll need an edge connector to make up a harness to your DES.  I ended up having a 22 position card edge connector with .156" spacing that I used.

Mar 28, 2023 20:01:40 GMT danno said:

A question I have:  Is there an easy way (using multimeter) to test the contacts at the DES to try and further

narrow the problem down?  It seems possible that a bad component (or several bad components) on the "buffer

board" (311134, I believe) could also be involved, and I saw another post in this forum where a bad transistor

was blamed in that case, but I think ruling out the DES first (if there is a solid way to do that) would be a logical first step.

[danno]

The "TL;DR" version:
Well, there is both good news and bad news. And, they're essentially the same thing....the jukebox is (once again) apparently working as expected, but
this has happened before with our USC2, so I am reluctant to declare "Mission Accomplished."

The longer version:
After reassembling the DES and waiting for all thunderstorms to clear out, I tried selecting some random entries off of the same list (above). But,
only 2 of them played as expected. The '<Re-set and Re-select>' light lit up a time or two, and most of the selections just resulted in the mech scanning
twice and stopping.

So, I unplugged the jukebox and the DCC, then went through (again) and checked the split contacts on all of the relevant connectors (DES, black/grey boxes,
buffer board in DCC). I made sure my paper clip could fit easily under each pin, and reseated each connector (again). Firing up the jukebox again, things
seemed to go from bad to worse. Now, none of the selections played, and each time the mech scanned twice and stopped.

Next, I checked (again) the RCA cable and connections for the Tormat (through the jumper connection up to the DCC), and reseated those. Then, I made a few
test selections, and the jukebox now proceeded to play selections that I had entered previously (when the mech scanned but didn't play). Once those finished,
I decided to switch on the "Selection Pass" service switch, start the mech (which scanned twice and stopped), then switched off Selection Pass. When I
entered a new selection, the jukebox first played a few more selections that had previously been selected (but not played yet), and then successfully played
the record I had picked (YAY!). Then, I entered 5 more selections (one-at-a-time), and they all played as expected - no Re-set/Re-select lights, and selection
numbers covering all the different numbers.

So, I decided to unplug the jukebox and quietly walk away to give us both a break for now. These quirky, randoms selection issues have plagued the jukebox
since day #1, so I think there is still an underlying root cause (maybe more than one?) to be discovered and fixed. Problem is, the target seems to keep
moving, so it makes things hard to diagnose. I'm starting to lean (slightly) toward a flaky black box (DTP1), but not confident in that suspicion. I've
seen a few black boxes available at eBay (ePay, as Ron disparagingly called it), but Ron's guide mentions something about the factory having to do something
to it (calibrate?), and said swapping with a different box "may or may not work".

If anyone has any ideas for next steps, I'm open to suggestions. My plan is to run through the entire list of test selections I listed at the start of the
post in a day or two, just to see what happens.

BTW - the DES keys are definitely cleaner now, but I don't think the cleaning helped allow any more light to pass through them.

[jukenorman]

Hi Danno,
Reading through your last post, especially that the faults appear to come and go, I wonder whether you might be having issues with poor (or less than ideal) grounding?

Norman.

[tonyj]

Apr 1, 2023 21:00:29 GMT danno said:

tonyj - Whoa...that sounds awesome!  Do you happen to know if Arduino programs can easily be ported to work with Raspberry Pis

(since I have a Pi 3 model B available)?  I'd be interested in pursuing that, if so.

Mar 31, 2023 15:57:10 GMT tonyj said:

For kicks I wrote a little Arduino program to hook up to a DES and will display what key is pressed based on the status of the 4 data lines A through D.  If you're interested, I can send you the code and some wiring notes on how it connects to the Arduino.  The Uno model is what I did this one on, and you'll need an edge connector to make up a harness to your DES.  I ended up having a 22 position card edge connector with .156" spacing that I used.

I'm not sure of porting the program over to the Pi, it might be just as easy to write a small program in Python as well.  Basically what the program in the Arduino did was monitor the data lines from the DES as well as the "trigger" switch on the side.  When the trigger input was activated, those 4 inputs were read as a 4 bit value and the corresponding value was used to determine which number (or reset) key was pressed and displayed on the serial terminal.

[danno]

Thanks, jukenorman and tonyj, for the continued attempts to help!

Norman - any ideas or likely candidates to check with regard to grounding?  I've definitely checked the components with which I've

been working (grounding wire on the DES, etc.), and I remember Ron's guide mentioning some places (on the mech, I think?) which

had 'poor screw heads' that sometimes led to grounding problems. 

I've re-read through selected parts of the service manual and Ron's guide, and since the write-in process and detent seem to be

working (though sometimes in a delayed fashion), I'm wondering if the issue might be bad / insufficient voltage with the"write source",

as described by Ron in 'Phono Procedure Chart #4' (section #1).  He does say in the last sentence that if the DRD (gray box) is defective,

the defect can be intermittent, "making it very difficult to find."

Does that seem reasonable?  Right now, we have thunderstorms again in the area ( ), so I may not get to testing the voltages

for a day or two.  I don't want to get too far ahead of myself, but if the indication is that the DRD is bad, I don't know what I'd

be able to do.  I could get Tony Miller's book, but I don't feel very confident in my ability to handle component-level repairs even

armed with that info.  I didn't see "DRD" or "Gray Box" mentioned in the '"PARTS" (Used, new) /"BOOKS"/ "REPAIRS"--Where to find !'

FAQ forum topic. 

I'll post back once I get a chance to test again.  Thanks for hanging in there with me.

[jukenorman]

Hi Danno,
My main reason for thinking about poor grounding was your mention of the tormat related connectors which is so typical of poor grounding. Also the fact that you have said that the problem comes and goes would appear to rule out insufficient voltage which you might not expect to be intermittent. I don't have Ron's guide though. As far as checking goes, there is an issue using digital meters on continuity because of the miniscule current applied. Ron always advocated an analogue meter but my one is so big and unwieldly that I rarely use it. I use a homebrewed battery and lamp for running these types of checks so you could think about using something like that?

Norman.

[e093116]

If you think you have moved past DES issues, particularly with no recurrence of the re-select/re-set light when making a normal selection, clean/adjust the detent switch.  Some of the behavior you described (missing selections on the first pass, but playing them on pass #2 or "randomly" later) is symptomatic of a flaky detent switch.

Also clean the RCA jacks (female side) and reform the wings on the RCA plugs (male side) for a tight fit/good contact if you haven't already.

I've had good luck with "LA finest degreaser"...yellow stuff in bottles I think in 20 oz or 32 oz sizes... from the dollar store for getting rid of nicotine residue without scrubbing.  I generally mix it 60% water/40% degreaser.  If you use it straight it's very powerful and will remove ink stamps and the like.

[danno]

Thunderstorms had cleared out overnight, so I spent a few hours testing the jukebox this morning.  Here's an update:

First, here are a few photos to show the adjustment I was describing with the DES main trigger switch.  The photo in Ron's guide

doesn't really show this:

- photos.app.goo.gl/fFt1NwGWFH6xVgjLA

- photos.app.goo.gl/ZUZL7GHnLpiqHhQB8

I do think this helped, but still have the re-set/re-select light at times...more on this below.

jukenorman - Regarding your comment about insufficient voltage and intermittent problems:  Ron's guide described a possibility

with a 'defective' DRD could cause intermittent (hard-to-detect) problems with the write-out processing, so I went ahead and went

through the steps.  I used a good quality analog voltmeter (GB Instruments), and checked the 'write source' test point with J3101

unplugged, and it was good (around +120 to 125 volts).  I plugged J3101 back in (from DRD) and checked again - still OK, and

voltage recovered the same when turning power off/on. Then,I did the next step of making selection #111 several times very quickly,

and the voltage recovered quickly after the 3rd digit (which seems to indicate DRD is OK).  However, the mech never actually played

#111 - it just kept making passes (I didn't count, but am guessing 2 per selection of #111) and stopped.

So, I thought I'd try making selections again (see initial post - per Ron's guide).  For the first 3 (100, 111, 122), the mech made 2 passes

and stopped.  Then, entering 133, the Re-set/Re-Select light came on (3rd digit on first try, 2nd digit on second try).  So, since it seemed

to make a difference before, I set the "Selection Pass" switch on, activated the Start Mech switch, waited for the 2 passes to finish, then

turned the "Selection Pass" back off.

After doing that, I re-entered 133 and it was accepted successfully (no Re-set/Re-Select error), and the jukebox proceeded to play

that record, along with the 3 I previously entered.  Then, I entered 5 more random selections (150, 109, 264, 173, 128), all of which

played perfectly...no Re-set/Re-select errors, either.

So, I'm still not really sure what to conclude.  It does seem that once the jukebox 'warms up', that the errors seem to subside.

e093116 - I think your suggestion (checking detent switch) seems like a good one.  I haven't sought that one out yet, so will have

to consult the service manual.  I'll plan to follow up on that soon.

I did visually check the RCA jacks on both cables (DCC to jumper box, Tormat to jumper box) - I had cleaned them before, and the ground

wire soldering looks OK.  I also looked at the jumper box connections on the back side - those seem OK, too.

[danno]

One other thing I meant to include regarding the DES:  Looking at the traces on the main PCB board,

it seems that the gold flashing is fairly worn at spots. 

photos.app.goo.gl/ohXWKDGzeMSBVNob6

I have no idea if this is normal and expected with age, or (if it isn't normal) if it could be why the

"Re-set and Re-select" error intermittently occurs.  I followed Ron's guide for cleaning the PCB

strictly, and only did the pads (didn't touch other parts of the board / traces).  Is it anything to

worry about?

[jukenorman]

That PCB board looks pretty good to me. As has already been pointed out by e093116, your recent problems have been with the read-out side and these types of issues would be common to all tormat Seeburgs - detent switch, tormat itself, RCA connectors and the read-out "frog" and Ron has put up a lot of info on checking/cleaning etc these items.

Norman.