Hittin’ A Lick

I was down to the final issues on this SMC3 I’ve been working on for three months. One of them being a very common problem with the mechanism at rest, when activated, just goes the wrong direction. This one was doing that. Whenever the mechanism was at rest on the right side with the box turned off and then powered up the mech would scan left, reset, then back right to home. Selections were fine.However, if the mechanism was at rest on the left side when the box was turned off and then powered back on the mech would try to go left and hang up. I would have to turn the jukebox off, use a finger on the motor coupler and roll the mech to the right a few inches then power it back on. When I sell a jukebox to a customer it can’t come with funky directions in case this or that happens. At least not in this shop.

So, I thought I would order this new replacement PCC interface board that Victory Glass has recently came out with. I got it and all was good as far as the problem just iterated. Board looks good, is typical of today’s boards in that it is processor and software driven rather than TTL. All was right in the world until I turned the juke off while playing a record. Let me tell what is supposed to happen when that situation comes up. There is a small board in the CPA unit that will reject the record and then the mech goes merrily upon its way if the juke is powered on while in Play position. It does this because when the mech is powered back up with a record clamped it might be turning in the wrong direction for the record side in which case it will play backwards and never trip. I imagine it didn’t take too many operators hollering at Seeburg before they came out with the retrofit reject board.

I turned the juke off in Play position for some reason. When I turned it back on the mech tripped, I could hear it pop, but the motor did not spin up. It took several messages and a video to convince Victory what was happening. Turns out their reject board doesn’t work and they never thought to check for that particular operation. I was told he’d get right on it and find a solution. In the meantime another problem cropped up. The record position started coming up 3 spots off. If I played 160 the mech would pick up 163. If I put the Seeburg board in then the mech picked up correctly. I adjusted the detent point spacing and the new board worked fine for a day then went 2 spots off. Apparently the detent debounce circuitry needs work as well. Well, this was enough. No telling what other problems will crop up. I waited three weeks for a response hoping I would be sent a board that worked properly only to be told he was working on something else, no time for the interface board. This was time my customer had to wait as well as the box is sold. He has been patient as can be. Thank you Tracy. I’m sending the board back.

I take “hittin’ a lick” to mean I got something done well in a short time. The first “lick” was my original interface board. These interface boards can be a bear to troubleshoot in that I would have to drag the scope, meter, and schematic to juke-side and spend a lot of time signal tracing. There is a shorter way that I have touched on before. The time honored practice of “shotgunning” or just replacing chips willy-nilly. I did a little better than willy-nilly. Looking at the schematic for the PCC board I could see that the motor direction control was dependent upon the detent switch and of course the limit switches. These signals fed through U1,U4, U7, and then U3. A bunch of NAND and NOR logic gates. Basically all this is used for scan direction latch, scan control, and finally play control as to which direction the motor spins. I replaced these four chips as an educated guess and the board worked! No more off/on motor problem. This made me very happy. To say the least.

The second lick was the MCU3 I got off ebay for a song. I got lucky. No one else bid on it. The 3’s came out when Stern started making the SMC2’s and was used in all the SMC3’s. I manage to repair these MCU’s enough that this was a good chance to take. The case was clean, there was no battery acid damage which is practically shocking. I dis-assembled the three boards and went over them looking for problems. I tested the transistors, cleaned the legs on the processor. No real problems found.Clean boards. When I tried the computer the mech actually scanned and it reset to 00. Was I going to get this lucky????

No, I was not. I actually was able to make a few correct selections then the MCU went off the rails stopping at 100 and playing every record like a tormat that has been charged. Mick sad. Something was heating up apparently. I let a few hours go by and practically the same thing happened again. I have several of these MCU’s that I haven’t been able to repair but one thing I did do was check all their processors in one good MCU that I had so had a pile of good processors. Again, going through schematics I couldn’t find any particular circuit responsible for making a selection. It’s there, just couldn’t see it. So I did the next best thing and swapped the 40 pin processor out. And By Jimminy it fixed the MCU. Mick glad!

One more problem now that I got it running. The buttons and display were acting weird. Any buttons I pressed gave either a one or a five with only a few digits displaying. Luckily I have several working display boards which also house the selection buttons in the same pile of boards previously mentioned. I swapped in a good board and all is well. My $40 MCU3 is good to go. These are hard enough to find working. One can pay almost $400 for a Gen2 CPU and I have had to do that a few times. It feels sweet to add a working MCU to the stockpile. I hit a good lick!

Chasing My Tail

This could also be titled ” How I Learn”. Been working on a couple of Digital Control Centers for a fellow. This consists of a visual inspection, component testing, then replacing electrolytic capacitors on the power supply board and testing the SCS write-in component. These were duly done and the DCC’s taken to the garage to test on the STD4. I quickly noticed the scan add coil was staying down. This is a good way to cook a coil. I started subbing boards from the DCC just taken out of the juke. The DCC that’s been in there several weeks had been working just fine. I started with the obvious one-the smallest board is the scan control. No joy. Tried another gray box, no joy. It was very hot even out of the sun. I figured I must have a wire grounded somewhere and called it a day.

That evening I started looking through the manuals for a schematic for the service switch and lines to and from it, the scan coil, and the gray box. I ended up getting the information needed a couple hours later from the service manual for the USC1 Bandshell. These 70’s jukeboxes are so similar that this was directly related to the STD4 made six years later. Armed with this and a few notes I hit it again this afternoon after lunch.

This is when I started jumping down rabbit holes apparently to see where they went. This is a simple circuit. The gray box sends a pulse to the scan control board. A thyristor on there sends a scan start signal or ground really, to the Scan Add coil. It is temporary so that the coil pulls in and then releases. 28 volts AC is constantly supplied to both the add and subtract coils. It goes to the service switch and this is where I got lost. I would make a selection, watch the coil pull in and stay in and then start trying various things. I could pull P3102 off and the coil released. I thought this meant that something was staying grounded in the service switch assembly.

This lead me to start ohmming out wires and testing for shorts to ground. No joy. So, I started pulling plugs off the control center to see if any had an effect on the energized coil. None did. But I did notice quite a few wires on the various colored plugs that seemed to be almost pulled out from the molex pins. These plugs have been pulled off and re-plugged hundreds of times testing all the control centers I’ve done over the years. I never noticed how some wires were slowly being pulled out of the molex pins until I really looked closely today. My first thought was to pull the harness and work on it inside where it’s nice and cool. I had to give up that idea pretty quickly as the harness for this box is one piece. It would be so much hassle and time spent to get it all out much less re-install it.

I managed to get the two DCC cables loose from the top of the box and over the side where I could work on them. I got a TV tray for a bench topped with plywood for protection, my old roller mechanics seat, gathered my soldering iron and tools. This would be an excellent time to try the new crimpers and get used to them. For years I crimped molex pins with a pair of needle nosed pliers and then added a drop of solder for insurance. I started doing them like this because back in the day a crimper for the molex pins was almost a $150!! And in the early days I really had to watch my expenses. I got good at crimping with pliers. And you know how habits are-ya just keep doin’ what ya do. I’ve done thousands of molex pins and connectors like this. It is a pain in the butt though with an occasional pinched finger. One day I found out crimpers were available for much less now and got a pair.

I would pull a pin, cut it off, strip about 3/16 insulation, roll it, put a molex pin in the crimper, crimp it, admire the crimp, and then add a drop of solder for insurance. Those wires will not pull loose by gosh. I do like the way the new crimper works. And the crimps are so professional looking. I repaired about 10 wires hoping that one of these would magically fix the problem I was experiencing. However, no joy.

Time to go back to square one. As I said this is a simple circuit. The thyristor turns on and off supplying a temporary ground. I tried a different scan control board and it worked! I was so astonished I put the old one back in, watched the coil stay energized, swapped the new one in and watched it NOT stay energized. The thyristor on my DCC scan control board had gone bad. And the first DCC under test had a bad one as well. I didn’t realize it and chased my tail for quite awhile. No wonder I never found any wiring problems in the service switches or elsewhere. I connected my meter to pin 6 of the scan control board, saw the 28vac, pushed 111 for a selection, watched the voltage dip and then go back to to 28vac. My Fluke is just quick enough to be able to see this dip as the thyristor does its temporary grounding.

I gathered every scan control board within a hundred feet and started testing them. Came up with four bad ones. I’ll fix these in an evening session on the bench and retry them at a later date. So, I do know the really simple circuit inside and out now. Even figured out the service switch ties in because of the Motor Start wire. It is its own circuit. And this is how I learn, spending a lot of time familiarizing myself with a circuit, doing more work than necessary, and then finally fixing it. I’ll have to relate the time I tripped a mechanism about 300 times while figuring out what I was seeing.

RIP Stevie Ray

Oct 3, 1954-Aug. 27, 1990