1959 Wurlitzer 2310S Restoration

I bought this in March 2012 as part of a load of Seeburgs I bought from a fellow in San Antonio. It looked pretty rough but I loved the styling so I bought it. Most people would consider it a parts machine at best but I like fixing up stuff that I call "a diamond in the rough". I often wonder about the history of things like this; the tales they could tell. About all I could really glean was from a barely legible sticker on the turn table backing panel that indicated "Sarbin Novelty & Cigarettes, Lafayette, LA".

The 2310S was built in 1959, played 100 selections and had stereo sound (the "S" designation in the model number). Other models in 1959 included the 2300 and 2304. The nice thing about the 2310 is that some of record magazine is still visible. For some reason, the juke manufacturers were starting to hide the mechanisms. Here is a picture prior to restoration:

Poor, poor Wurly! Who would put vine wallpaper on the sides? Maybe to hide a blemished original finish, but vines?

This picture shows the broken dome hinge panel. Not a good design and apparently many of this model suffered the same fate.

Dome Hinge Panel Repair:

I looked for a replacement on eBay for two years with no luck before deciding to try some form of repair. In the following picture, I removed the hinge panel and top to get a better idea what I was facing. I built a bracket out of 1/2" flat plate. I fastened this at the end bosses of the casting. I then attached the broken hinge bar to this, as well as the unbroken section.

And here's a picture of the repair with the hinge panel re-installed. The cracks will always be visible but the dome hinge is now solid. I'll probably fill the cracks with a silver JB type resin eventually, as this all has to come apart again down the road.

Now that the hinge bar has been successfully repaired, I need to put this project on hold while I take care of other projects. Up to now I have accumulated parts such as the speakers, new side glass, the "Wurlitzer" glass pane in the dome hinge panel, and chrome side pieces that are in better shape. The project once I get back into it will include:

Well it's July 2014 and time to get back to this beast. I started studying the manual and realized I had a problem...the manual has exploded parts diagrams (good), wiring diagrams (good), simplified schematics of the wiring diagrams (good), and lots of info on the multitude of adjustments (good). What it didn't have was anything resembling description of operation (bad!). I am going to have to figure out from scratch how everything works. If I can't understand how everything works then I'm going to have a hard time troubleshooting it and it's guaranteed some troubleshooting will be required. I can hardly wait to get to that stepper unit!

Here is a picture of the backside...not pretty! Closer examination and some more homework indicates I am missing the turn table motor and mechanism motor. I'm sure eBay will come through. With the heat of the Texas summer upon us, now is a good time to rebuild some of the components in the comfort of the air conditioned radio repair room.

Amp Rebuild/Restoration:

This is the "534" stereo amp, utilizing 7025, 12AU7, 12AX7 and 6973 tubes, the latter being a relatively uncommon audio output tube. The amp is in rough shape and of course, missing the tubes. The only tubes that gave me a concern were those 6973's, and some research indicates the 6CZ5 will sub. I happen to have 3 NOS in inventory, I just need one more. It's a lot of work but I decided to strip down the chassis and paint it and the transformers, followed by capacitor replacement. The following two pictures show the top and underside.

Closer examination indicated fine sand/silt deposits...not good! Unfortunately the amp and who knows what other parts of the juke spent some not-so-quality time partially submerged in water. After labeling all the leads, I removed the output transformers (OPT's) and did an ohm check on the windings. On one, there was good resistance on one half of the primary (around 300 ohms) but unusually high on the other half (over 4 k ohms). It was worse on the other OPT...one half of the primary looked good but the other half was open. I removed the bell cover of one to discover lots of silt deposits, as indicated in the next picture.

Both OPT's now need to be replaced. This could be a big problem finding something exact as there are many leads on the secondary for numerous speaker options. However, my speaker options are one...I plan on using the 12" speakers that belong to this juke (missing of course upon arrival but replacements bought on good ole eBay). That makes OPT replacement a lot easier. Reviewing the 6973 datasheets and amp design operating conditions, I determined I needed something with about 6,600 ohms impedance on the primary class AB1 with cathode bias) and at least 100 mA capacity. The Hammond 1620 appears to fit the bill although I will likely have to re-drill the mounting holes. And shell out some cash...they're gonna run about $90 each. Instead of finding 6973's, I decided to go with 6CZ5's.

I first decided to rebuild the circuit board. There were about 23 wires connecting to the board. I labeled them and disconnected them.

The board was then removed from its' mounting frame.

A visual inspection indicated some of the capacitors had already been replaced with Sprague "Orange Drops" so I decided to keep those. There is one potentiometer on the board and it was guaranteed to be full of silt. I removed it, cleaned the element, confirmed it was in good condition and then installed a new housing. I found a charred section of board between a B+ trace and pin 3 (cathode) of V2, a 7205. It seemed like an odd place to arc but regardless, conduction had taken place and had to be repaired. I removed the socket, dremeled out the burnt phenolic material, placed a piece of masking tape on the component side, and filled in the void with JB Weld. Then a new pin hole was drilled and the socket re-installed. The following pictures show the repair.

This board uses two "couplates". These were forerunners to the integrated circuit and contained an assortment of resistors and capacitors. They're also hard to confirm they're good, as individual components cannot be tested, at least with this type of couplate. My experience with Predictas and their couplates suggested they be replaced. I used 1" x 2" perf board, mounted the appropriate components with leads corresponding to the original pin-out and re-installed. The following is a picture of the couplate replacement process.

All resistors were checked and those that were 20% plus out of tolerance were replaced. The board was cleaned, solder joints reflowed and the board was ready for installation. I had problems reflowing some of the tube socket connections so I replaced 4 of the sockets (ceramic).

With the circuit board out, it was time to strip down the remainder of the chassis in preparation for sandblasting and painting. It's a lot of work but the chassis will look much better when done.

Here is a picture of the finished product with the Hammond output transformers mounted on the gold hammertone painted chassis, new power cord, and labels designed and installed.

The output transformers were wired in ultralinear mode. Without getting into a lot of theory, ultralinear transformers provide pentode screen grid voltage via taps on the primary side. This changes the plate voltage/current relationship, making them more like triodes. And distortion is reduced. The original circuit used a separate screen grid B+ supply via 47 ohm resistors. I disconnected this supply from the output tubes and connected the screen taps via 270 ohm resistors. I couldn't find any definitive values for these resistors but decided I needed them to protect screen grid meltdown when the amp was running hard (i.e. plate voltage lower than screen voltage). I simply used what I used in a homemade 6L6 amp. I also went with 330 ohm cathode resistors.

Finally it was time to power everything up. I initially used 25W 8 ohm resistors as speaker loads as it was quick and easy. I just wanted to pass the smoke test at this point, and it did! I quickly checked for output tube redplating. No redplating but it looked like one pair (B channel) might be under load as I was getting purple glow on the inside of the glass envelope. I took a few B+ readings and noted a healthy 400 vdc or so but it kept slowly declining which I thought was odd. At this point I decided to do a full test hook-up (speakers, audio generator) and disconnected everything. I inadvertently touched one of the electrolytic cans on the power supply and it was HOT!!! This was a real mystery because these were not connected to anything. It took a while to find the problem but I knew voltage was somehow being applied to one or both of the terminals. To aid in electrolytic can replacement I designed something I call a "terminex". It is simply a small rectangular piece of copper PCB etched so that there are two small copper squares separated by a thin line, maybe 1/16". A hole is drilled in one square to mount and solder old wiring and the new capacitor lead. The solid square is soldered to the old capacitor terminal, thus providing an extension. One particular terminex had a thin line between the two copper sections...sure enough, a thin copper line. It was enough to apply B+ voltage to the old capacitor and like many, it was leaky and started heating up big time. Just another lesson for why I never mess with old capacitors, I replace them.

Here is a picture of the test set-up. I used two 15" 8 ohm speakers, an EICO 377 audio signal generator and appropriate jumpers installed in the power supply socket to the stepper box. I powered up and shortly afterward I got a nasty howl/oscillation from one of the speakers. I experienced the exact same problem with my home brew 6L6 amp and knew what it was...a phasing problem at the output transformer. This circuit uses feedback and the phasing was wrong, creating positive feedback. I had to reverse the output transformer speaker leads and that solved the problem.

I took a few voltage measurements. The 6CZ5 cathodes were running about 25 volts and no voltage drop across the screen grid resistors, i.e. no current. I calculated a plate current of nearly 75 mA...too high for this tube, it needed to be around 45. I added 68 ohm resistors in series with the existing 330 ohm resistors and got a cathode voltage of

Power Supply Rebuild/Restoration:

The power supply (PS) for the 534 amp is supposed to use a 5U4 full wave rectifier for B+ and a selenium rectifier for dc working voltage (various coils). The amp that came with the juke had neither. Great...I have some oddball amp that doesn't belong to the juke. Some quick research though indicated it was likely the 538 amp used in the 2400 stereo version, utilizing diodes for all rectification, plus a voltage doubler circuit for B+. Like the PS that is supposed to be there, two transformers are used: one for B+ and a portion of the amp filament supply and the other transformer had a 40 vac CT winding, two 26 vac windings and the rest of the filament supply winding. Wurlitzer wired these two filament windings in series to provide the desired 6.3 vac but why? Condition-wise, it was rougher than the amp as the following two pictures show. And since it looked rougher than the amp, I was even more concerned the transformers were bad and if so, they wouldn't be as easy to replace.

Upon removal of the transformers, I did an ohm check and everything appeared to be OK! Whew! To be on the safe side, I applied 120 vac to the primaries and then did voltage checks on the secondary windings. Things continued to look OK. I left them powered up for about 15 minutes and felt them to see if they were getting warm...no temperature increase! That was a good sign but the true test will be when they're under load.

Here is a picture of the unit before and after sandblasting, painting, and re-assembly. I still need to install a few diodes and procure some missing fuse holders.

Coin Registration Unit Rebuild/Restoration:

This was Wurlitzer's fancy name for the credit unit. It has two credit modes: 5/10/25 and 10/25/50 cents, determined by a slide switch on the back. The number of credits can be set for dimes and quarters in the 5/10/25 cent mode and quarters and half dollars in the 10/25/50 cent mode. Nickels (and dimes in the 10/25/50 cent mode) produce only one credit. There was also a "dual pricing unit" but this juke uses the conventional credit unit.

Here is a picture of the credit unit. It seems the components are getting uglier and uglier.

This one had even more silt in it than previous components and I decided at this point to name the juke "Katrina". I compared it with the credit unit from my other (complete) 2310S to discover parts had also been cannibalized. I quickly decided it was eBay time again and one was listed at $50 BIN plus shipping. I slept on it and then thought maybe I could rebuild this one. I first checked the condition of the two relays (pulse relay, timing relay #1) and determined they were OK but needed serious cleaning up. The missing parts included two "coin coils", the bracket they were mounted on, some springs, and the cancel solenoid. Looking at the good unit, the coin coils looked like simple electromagnets and relatively easy to fabricate. I ohmed the two coils and got .3 ohms and 4.4 ohms. Something wasn't right here and I figured the .3 ohm coil was bad as it did show signs of heating. The wire looked like #30 gauge so that is what I used. When I was done winding, I had about 6.6 ohms each. Using Ohm's Law and a supply voltage of 24 vac, I came up with a current of 4 amps; way too high for #30 AWG. However, these coils are powered only for a fraction of a second as a coin paddle completes the circuit when a coin is inserted. And the manual did mention something about using a slo-blo fuse and that the fuse would blow if excess current occurred for more than 3 seconds. A stuck coin paddle and the wrong fuse and poof...these coils will burn out! I was going to put in some dropping resistors but I will wait and see how they perform. Here are two pictures of the mechanical portion. The first one is with the missing coin coils, the second one shows the new coin coils and new springs.

Next up was the missing cancel solenoid. I was going to fab one from scratch but remembered the Seeburg Selectomatic mechanism used a solenoid called the trip solenoid. Could it possibly work? I had a spare mech so I pulled the coil out. The plunger was about the same length but slightly larger in diameter. The Seeburg uses 25 vac to energize this coil and the Wurlitzer uses 24 vac to energize the cancel coil so that was a good match. I drilled and tapped some new holes in the mounting brackets, found a suitable link pin in my supply of hardware, found a spring to retract the plunger and voila...the Seeburg coil works! Here are some pictures: the first is a side-by-side with the complete unit showing what was missing and the second shows the installed Seeburg coil.

The inside of the credit unit looked very cruddy so I completely removed all remaining components, sandblasted the inside and bottom, painted with silver Krylon and re-assembled everything. The two plugs were repaired and reconditioned and the unit is ready to go. Here are pictures of the painted housing/bracket and everything re-assembled.

Selector Switch (Keyboard Unit) Restoration:

This particular model has 10 letter buttons and 10 number buttons, a reset button and a "Select" indicator. Compared to the Seeburgs and Rockolas I have worked on, this appears to be the most complicated of them all: two banks of sliding switches, a latch solenoid with a linkage system, and about 5 leaf switch units.

Before starting work on this unit I anticipated trouble: I could see tons of debris on the underside of the unit. I suspected this was the doing of some rodent and that there would be significant damage from mouse pee corrosion and chewed wiring. Well that was not the case. The debris appears to be an indication of the flooding "high water mark". The debris floated on the flood waters and settled or got trapped under the keyboard. All in all, the keyboard unit so far is in the best shape...no damage, no pillaged parts. The first picture shows the "high water mark" debris and the second is a picture of the selector switch assembly after removal.

There wasn't much to do with the unit. The major work was removal of the switch board assemblies and cleaning of all the sliders.

Cabinet Rebuild:

I decided to fabricate a new cabinet based on the water damage. The plywood sides were delaminating, especially on the inside. Many joints were loose (originally screwed and glued). The base plywood was warped. And of course, the outer surfaces of the sides were covered with that sticky vinyl vine material. Restoring the faux finish after vinyl removal was likely futile and re-veneering something this large is always a tricky affair. So, the entire cabinet was to be disassembled and brand new finished oak plywood used for the sides, stained and finished appropriately. About the only parts that were re-usable were the speaker board and the rounded corner solid wood pieces. Here are pictures showing the cabinet strip-down:

A description of the following pictures: first, one corner of the front lower panel incurred wood rot on the bottom. The piece looked too complicated to fabricate from scratch so I repaired it with epoxy. The second picture is the underside of the new base. The third picture is the topside of the new base. The fourth picture shows the inner surface of the new sides with all necessary holes pre-drilled. The fifth picture is the outer surface of the new sides, with holes drilled and stain applied. The sixth picture is a temporary assembly to ensure the major pieces fit.

At this point, lacquer has been applied to the sides and the sides permanently attached to the base. The next challenge was the aluminum trim along the bottom edge. What I had was badly damaged and unrepairable. I couldn't find this piece as a reproduction part and I didn't want a banged and scratched piece from another jukebox (even if I could find one) so it was time to make one from scratch. I was familiar with flexible aluminum trim used in the fabrication of my mini trailer so I bought a piece from an rv teardrop trailer parts supplier. With a little cutting and trimming, the piece worked quite nicely as shown in the next two pictures. There is also three aluminum panels along the bottom of the front and sides. These panels had an embossed pattern (small squares) and was likely anodized aluminum. While the front panel may have been salvageable, the two smaller side pieces were in rough shape and I didn't think I could ever get them looking perfect again. It was again time to fabricate something from scratch. A quick search on eBay yielded a supplier of embossed aluminum sheet paneling. I couldn't find a small square pattern but I did find a small diamond pattern that I thought would look good. The piece I bought will provide enough aluminum to do two machines. After some careful measuring, cutting and bending, I had new panels. The new panels are shown in the following two pictures.

For the cabinet interior, I took a piece of the original delaminated plywood to Lowe's and had them do a color match. I don't know the official color Wurlitzer used but a lot of jukes seemed to use a similar color for the interior. The following picture shows the interior painted.

The cabinet is nearly ready to move in the house for installation of the mechanism but there were three things I wanted to do first: install the curved aluminum trim along the bottom edge of the sides, install the straight aluminum divider trim along the sides (between the sides and the curved grille panel), and install the tweeters in the speaker board.

Unfortunately the curved aluminum trim pieces have the usual assortment of dings and scratches, as do the divider trim pieces. My first thought was to get rid of the blemishes and then sand and polish. The aluminum would eventually tarnish/oxidize though. I thought about anodizing or chrome plating but Houston was the nearest city that could do this kind of work and let's face it, plating is just flat out expensive. I did some research and found a brush plating system specifically designed for aluminum! I ordered it and began the process. Unfortunately, the aluminum plating exercise was a dismal failure and I ended up filing, sanding and polishing the aluminum trim.

The following picture represents work done over about a year and shows some of the aluminum trim pieces installed, the grille straightened and painted with chrome paint. I attempted to plate the "W" but it was another dismal failure so I painted it. Maybe some day I will find one in great shape. The back panel was stripped and painted anodized blue using Duplicolor's system. It turned out pretty good. I took the side spears, diffuser panel and selector panel to Speed & Sport Chrome in Houston. Pricey but they did an excellent job. In the picture, the left side spear and light diffuser panel are permanently installed. The selector panel needs to be removed so that the push button assembly can be installed. Before that, I need to install the glass diffuser panel and the fluorescent light.

There was some work to do before re-installing the mechanism in the cabinet, as follows: 1. fabricate and install speaker wiring, 2. fabricate and install junction box and power supply hold-down clamps, 3. Sandblast, paint and install the coin box and credit unit box, 4. install coin mechanism, 5. Clean and paint mechanism support brackets. Here is a picture of the interior of the cabinet with key components installed:

As mentioned above, the light diffuser panel needs to be installed before the keyboard unit can be installed. A while back, I ordered the glass light panel. It arrived broken. I contacted the vendor and the sent another...it too arrived broken. Getting tired of this, I made a two piece panel from the broken panels. My glass cutting skills suck but I did manage to make it work. The glass panel is held in place with a vent screen hold down bar and by the fluorescent light socket brackets. These were cleaned and painted. I used rubber tape on the underside and top of the glass to keep it from moving around or vibrating.

Next up was the keyboard fluorescent lighting. The ballast looked ugly but after temporarily hooking it up to a bulb and installing a new starter, it actually was good!

I cleaned and painted the ballast and installed new wiring to/from the ballast.

In this picture, the keyboard unit has been installed in the keyboard panel, along with new instruction windows and reconditioned "Select" indicator and "Reset" buttons. The plastic panels on either side of the record basket have been repainted and new vinyl pattern labels applied on the top. With everything lit up, it looks nice!

"Mechanism" Rebuild/Restoration:

This towering hulk of motors, gears, levers, cables, switches, cams and other stuff I call the "mechanism". It is actually a collection of sub components: the electrical selector unit at the bottom, the pin bank unit, the crank arms that are used to detect and cancel a selection pin, a control or timing cam with it's attendant link/pivot arms, a drive motor and gear, arms to raise and lower records into the play position and at the top, the record carousel. Most if not all of this unit had also been submerged in flood waters so it all had to come apart, for a couple of reasons: first, to repair/restore all the components and second, by taking it apart I would hopefully learn how the thing operated. I find it much easier to troubleshoot when I know how something is supposed to work. The first order of business was to build a cart to place the entire unit on. Nothing fancy, just some scrap lumber laying around the shop and a set of small caster wheels from my local Lowe's. The cart also allowed me to drag the filthy thing into my radio repair room in the house where I could work on it in air conditioned comfort.

The first order of business was to drop/remove the electric selector unit at the bottom. It hangs from three supports: two legs and a bracket at the back. Big letters on the bracket indicated "do not remove lower bolts, remove upper bolts to remove unit". For some reason this didn't register and I removed the lower bolts. I suppose I messed with some intricate adjustment that will bite me in the backside later...time will tell. Here are a few pictures of the electrical selector unit in varying stages of dis-assembly; before, pin bank removed, letter/number coils, coils removed.Note the large amount of clay/silt; many letter coils were almost submerged in it. No surprise, being the lowest point and a perfect location for silt to deposit.

Restoration begins. I decided the base needed to be stripped down and sandblasted which required the removal of all the letter coils and the number coils and "quadrant" they were mounted on. The first picture below shows half of the coils removed, the wires and coils cleaned and the plungers loosened to ensure they moved freely. I first checked resistance and they all came in around 6 ohms...good news! Just to make sure, I applied about 24 vdc from my bench power supply to ensure the coils operated as they should. The second picture shows the sandblasted base; much nicer!

With everything cleaned, it is time to re-assemble the unit, as shown in the following pictures:

The last component of the electric selector unit is the small "selection motor" that rotates the big disk that houses the selection lever rocker arms. The motor case is riveted so it was likely never intended to be serviced. Since it had been underwater, it needed to get opened up. Sure enough, that nice fine silty clay had worked its way inside. Following cleaning and lubing, the motor was reassembled. The following three pictures show the motor rebuild.