Seeburg's HF100G was the 5th in the series of 100 Select-O-Matic jukeboxes beginning with the 1949 M100A. The HF100G is the first high fidelity jukebox made by Seeburg. More information available here.

Jukebox shortly after arrival Jukebox now

The pictures show the jukebox shortly after I acquired it and again at the time of writing this. I had the mechanism cover, but it was shipped separately and I had not yet installed it.

My wife and I decided to get a jukebox after seeing a working Seeburg M100C in an antique store (way over priced) and another still in service at an ice cream shop in Port Townsend, WA. At first we were thinking of getting an M100C, but after much deliberation about style we decided on the HF100G. In January 2009 we bought our HF100G in unrestored condition.

The story of this jukebox is that it had quit working many years ago and the owner never got around to having it fixed. Because it spent years in someone’s home instead of in a bar, it was cosmetically in excellent condition for its age.

Before working on a machine this complicated, I knew I would need to get the service manual. The guy who sold me the jukebox was kind enough to send me a manual for it. While waiting impatiently for it to arrive I looked for some part of the jukebox I could work on without the manual.

The Slug Rejector

Since I didn’t have the manual, I started by looking at the coin mechanism (commonly called the slug rejecter). I tried putting a quarter in the jukebox and it jammed immediately. I fished it out and tried a dime – same result. Any coin put in got stuck immediately and severely enough that the “coin return” button wouldn’t even eject it. The slug rejecter was filthy and filled with dust, cob webs and rust. I decided to remove it and take the whole thing apart and cleaned all the parts individually. After I put it back together, coins were still getting stuck in various places. I ended up having to work on the path of each type of coin one by one. Turns out some parts of the slug rejecter are made of cheap pot metal that bent slightly over the years due to the strong springs and magnets holding the thing together. Pot metal usually breaks if you try to bend it much so returning the parts to their original shape required extreme care. The nickel path was the most difficult and it is set up to reject them anyway! By the time I had the slug rejecter working I got the manual and was able to put it back and move on to bigger and more exciting things.

Slug rejector before Slug rejector after repair

The Wired Selection Receiver

The first task in getting the electronics working was to make sure all the transformers were in working condition. If any of them were bad, the whole electronics chassis would likely be a junker and there would be no point in doing more work until I could find replacements. Thankfully all the transformers were good. I then went to work on replacing the capacitors in the wired selection receiver. The wired selection receiver (Model WSR5-L6) handles selections from wall boxes (Those Wall-o-matic things you see on tables in diners). The reason I started with it is because it also is the power distribution center for the whole jukebox. I needed to have it working before I could safely power up any of the other electronics. After all the capacitors were replaced, I powered up it using my dim bulb tester. The dim bulb tester allows me to apply power gently in case there are any shorts anywhere or other bad components. For lots of details on dim bulb testers and capacitor replacement (and many other great restoration stories), see Phil’s Old Radios site.

Chassis before from above Chassis before from below

The dim bulb test passed, so I started testing out the various functions of the wired selection receiver. At one point, my “dim” bulb lit up very brightly even though I didn’t change anything. After trying to track down the short, the problem mysteriously went away. Intermittent problems are very difficult and frustrating to try and fix so I decided to continue work, but made sure I always used the dim bulb tester until I was sure I found the source of the fault.

I cleaned up and lubricated the ratchet assemblies for the wall-o-matic decoder and started work on the credit and cancel unit. The credit and cancel unit keeps track of how many credits are left based on the coin that was inserted and the number of selections already made. I didn’t fully understand how this assembly worked, but I knew enough to know it wasn’t working correctly. It seemed like there was always credit available because pin 1 of S10 (the socket that connects to the keyboard) was always grounded no matter how many times I manually operated the cancel solenoid. Time to go to the manual…

Here come lots of details. Feel free to skip over this if you don’t care about how this thing works in detail

Reading the manual I figured out how the whole assembly works. There is a credit wheel with 6 switches on it. All the switches on the credit wheel are wired in parallel, so if any switch is in the on position, a selection can be made. When a coin is put in the jukebox, a coin solenoid flips on a switch on the credit wheel. Nickels, dimes and quarters each have their own coin solenoid. Every time a selection is made, the cancel solenoid rotates the credit wheel one notch. At one point on the wheel, there is a ramp that flips the credit switches off as they go by. This way the number of credits each coin gives is determined by how many positions its coin solenoid is from the ramp. In the case of my jukebox, the nickel solenoid is not used, the dime solenoid is next to the ramp (1 credit) and the quarter solenoid is 3 positions from the ramp (3 credits).

End of lots of details section

After reading the manual, I figured out that someone in the past had crudely bent the reset ramp out of the way so it no longer engaged the credit switches. Because of this, no coins were required to play a song. Unfortunately, this modification has been known to cause the latch solenoid in the keyboard to overheat and burn out after a while. Also, my goal was to restore the jukebox to original condition hence requiring coins. Thankfully the latch solenoid was not burned out and I was able to very carefully bend the ramp back to its original shape. The ramp action has to be smooth because the small credit switches are very fragile.

The Keyboard

The keyboard, or “Electrical Selector” as Seeburg calls it, is how selections are entered into the jukebox. For those that have never used an old Seeburg jukebox, the buttons latch down when they are pushed. After a letter button and a number button have been pushed, both buttons pop back up and your song is queued up to be played. For a very short time after the selection is made, the keyboard is locked while the selection is written into memory. If you push the buttons without putting a coin in first, the buttons don’t latch down. All the button locking and latching is accomplished with two solenoids, some levers and a bunch of springs. As found, the keyboard assembly was filthy with waxy residue and dust. The levers and solenoids barely budged when trying to move them by hand. I knew better than to try and apply power to it in this state. After extensive cleaning, I had all the mechanics of the keyboard working smoothly when operating the solenoids by hand. At this point I should have tested all the electrical contacts in the keyboard. This comes back to bite me later on. I put the keyboard together again and installed it back in the jukebox.

Keyboard before Latch solenoid Keyboard after

First Real Power Up!

At this point I had enough of the jukebox working that I could try out some of its functions. I connected up all the cables to the Wired Selection Receiver with the exception of the 27 pin cable to the pinbank (The jukebox’s memory). I didn’t connect it because I had not had a chance to check it out. My 40 watt bulb on the dim bulb tester came on dimly and the “selection now playing” lamp was on now. When I put in a coin in, the dim bulb lit up brightly and the keyboard solenoid buzzed loudly but didn’t engage fully. Time for a bigger light bulb -- a 150 watt flood lamp… This time the lamp flashed for a second when I put the coin in and the keyboard solenoid engaged solidly. Success! After pushing a letter key and number key, the cancel solenoid did its thing, the “select” light went off and the keyboard solenoid released. So far so good… I decided to take a look at the pin bank. Most of the pins were stuck, but I managed to free a few of them for testing. I plugged in the pinbank and made another selection. This time I heard a very loud buzzing and the 150 watt lamp flood lamp came on VERY brightly. I quickly unplugged and started investigating. Nothing seemed amiss so I powered it on again. No buzz, but the lamp again was blinding. The scourge of the shorting problem had returned. Time to try pulling all the fuses to narrow down the fault… The lamp still beamed brightly with all the fuses pulled - strange. One of the transformers, for powering wallboxes, does not have a fuse on its primary winding so I tried disconnecting that - still no luck. At least now the intermittent problem wasn’t intermittent anymore. I disconnected the hot lead at the power switch and the lamp still lit up brightly! Turns out the fault was in the plug itself. I would have found the problem much more quickly had I simply tried turning off the power switch. Amazing how the most simple problems can sometimes be the most difficult to diagnose. After fixing the plug, I was able to power up the jukebox again.

The story of a cancel solenoid

After putting everything back together, I tried making another selection. The jukebox started making a very loud and angry buzzing sound. I quickly shut it down to investigate and found the cancel solenoid stuck at its end of travel. After barely touching it, it popped back into its normal position. If I moved the solenoid by hand slowly it seemed ok, but if I pushed a little harder at the end of travel it would sometimes stick. Time to take the credit and cancel unit apart… Taking the solenoid off required lots of disassembly, but I found the problem. There were metal shavings inside the solenoid tube. Thankfully, there were no holes in the tube so I was able to clean it up and put it back together. I may end up having to replace the solenoid at some point but it seems to work for now. After putting the jukebox back together, I could make a selection and see the correct pin move in the pinbank.

The Mechanism

Through amazing mechanical wizardry, the whole jukebox mechanism is run by a single electric motor. This motor rotates the turn table, scans the mechanism back and forth and handles transferring records between the record magazine and the play position. Record transfer itself involves lifting the record to the turntable, clamping the record to the turntable, opening and closing a bunch of electrical switches, moving the tone arm to the correct side of the record, locking and unlocking the tone arm, landing the tone arm on the record at the start of play and picking up and returning the tone arm to its resting position at the end of play. This is one busy motor! The motor is run off 115VAC through a relay called the play control relay. This relay is closed any time a selection is queued to be played or when a record is currently playing. It is also possible to turn the motor on through one of the service switches inside the jukebox.

It was exciting to flip this switch the first time and see the motor come to life and watch the mechanism start scanning. Unfortunately when it got to the end of the magazine, it reversed for about 1 inch then went back toward the end where it reversed again. Had I kept the switch on, it would have endlessly “bounced” at the end of the magazine. Clearly the reversing switch was not working right. One of the braces that support the reversing switch was bent. After carefully bending it back, the mechanism scanned correctly.

Normally the pinbank completes the circuit to the play control relay. A washer on each pin contacts a grounding bar when it is pulled (song selected). My pinbank was in desperate need of attention, but I did manage to find a pin that completed the play control circuit. This was cool because I could put in a coin, make a selection, and watch the mechanism pick up that record and start playing it. There was no sound because I had not yet worked on the amplifier, but it was exciting none-the-less. Somewhat less encouraging was that play control relay would sometimes drop out and the motor would stop turning in the middle of the record. This was caused by two problems – a dirty switch in the mechanism and the poorly working pinbank. I then cleaned all the switches on the mechanism. After doing this the mechanism was working nicely so it was time to turn my attention to the pinbank.

The Pinbank

The pinbank, as it is commonly called by people who work on jukeboxes, or “selector assembly” as Seeburg calls it, is the jukebox’s memory. It is made up of 100 levers (pins) that move in or out to indicate whether or not a record side has been selected for play. The levers are moved out by little electric magnets in the pinbank and pushed back in by a solenoid on the mechanism when the record is played. The pinbank is organized in 5 groups of 20 pins. Each group has a relay called the group magnet to ground all 20 electric magnets in its group. A selection is written to the pinbank by grounding one of 5 group wires and one of 20 unit wires. As described in the above section, any pin that is pulled completes the play control circuit through a washer and grounding bar.

As found, most of the pins were stuck in place and very few, or perhaps only one, of the pins completed the play control circuit - most of the time. After removing the pinbank, I could see a bunch of screws for the cover were missing. This is never a good sign. And, even less encouraging, I noticed one of the retaining clips for one of the group magnets was missing. Apparently a previous “repair” person had messed with it sometime in the past. Taking this thing apart was a serious pain. To say there are hundreds parts is not an exaggeration. There are 100 levers (pins), 100 springs, 100 washers, 100 little electric magnets, etc. Obviously none of these parts are easily replaced so I had to make sure I kept track of everything. All of the contact washers were covered in some combination of dirt, corrosion, and/or grease. The contact rails were also coated with a tarnish/grease like substance. My wife helped me clean some portions of the contact rail and I cleaned all 100 washers. I also cleaned many other parts, although they aren’t as critical as the washers and contact rails. If taking this thing apart was a pain, putting it back together was even more so. I _really_ didn’t want to take this beast apart again so I was careful to test every selection magnet and made certain every pin completed the play control circuit when it was pulled. I also was able to find replacement screws and a replacement e-clip at Tacoma Screw. After putting it all back together it was time for a final test and re-installation in the jukebox. It was great getting the pinbank working. At this point it was possible to select any record side and have the mechanism scan there and pick up the record. A major task was out of the way!

Pinbank before Missing clip Bottom of Pinbank Pinbank contact rails before/after cleaning Putting the Pinbank back together

Having fun with a kind-of-working jukebox

At this point, the jukebox was mostly working with the exception of the amplifier. I had parts on order for fixing the amplifier so there wasn’t much to do except enjoy selecting records and watching the jukebox play them. You could actually hear the sound of the stylus on the record so it was possible to listen to the record by putting your ear up close as it was being played. While doing this my wife and I discovered that selection “E5” (I think) wasn’t working. Through some testing, I determined that no “E” selections could be made through the keyboard. Oops… maybe I should have tested the keyboard more thoroughly when I had it taken apart. The problem was a dirty contact in the “E” switch. This time I was sure to visually inspect and test every switch before putting it back together.

At another point my wife asked why the “now playing” indicator light was sometimes on and sometimes off. Hmmm… It is always supposed to be on. I took it apart and found more dirty contacts. After cleaning the contacts and putting it back together I proudly showed it to my wife only to see the light go out again! #$*@! Turns out the contacts were not making good contact with the bulb sockets. Increasing the contact tension on one bulb socket contact reduced the tension on the other and made one or both go out intermittently. After lots of tweaking I have the now playing indicator working reliably … I think.

The Amplifier

I finally got all the parts for the amplifier and was able to start work on it. This jukebox has the MRA4-L6 model amplifier. This is Seeburg’s first High Fidelity jukebox amplifier. The amplifier has 8 tubes.

NumberTypeUse
V15U4Rectifier
V25879Voltage Amplifier
V36SN7Voltage Amplifier/Cathode Follower
V46SK7AVC Control
V56SL7AVC Amplifier/AVC Rectifier
V612AX7Voltage Amplifier/Phase Splitter
V76L6Power Amplifier
V86L6Power Amplifier

The first task was to test all of the tubes. The 12AX7 and 6L6s were weak so I replaced them. After that I replaced the main filter capacitors. I powered it up briefly at this point to test the new filter capacitors. All went well so I went ahead and replaced almost all of the paper capacitors. The only ones I did not replace at this point were in low voltage areas so they weren’t critical. I also skipped a mica capacitor (the only mica capacitor in the whole jukebox). Usually mica capacitors are pretty reliable and I was feeling lazy. That turned out to be a mistake. At this point I put everything back together, powered up the jukebox, and selected a song. This time I was rewarded with music through the speaker! After all this time, it was very rewarding to have the jukebox working. The only problem was some hum through the speakers that wasn’t affected by the volume control. This was not the harsh sort of hum like you get with bad filter capacitors, but it was just loud enough to be annoying. Also the audio was somewhat distorted and lacking bass.

I went ahead and replaced all the rest of the paper capacitors. Most of these were in the tone control circuits so I was hoping this would improve the fidelity somewhat. After powering it back up, there was still the same hum and the audio didn’t sound any better. It was time to start doing more troubleshooting. All of the power supply voltages were either spot on or off by only a few volts here and there. However, when I went to test the voltage on the plate of the phase splitter it was at 232V instead of 195V. The cathode was at 62V instead of 95V. Apparently this tube was not conducting as much as it was supposed to be. The grid of this tube gets its bias voltage through a resistor off the cathode circuit. I also noticed the mica capacitor I didn’t replace was between the grid and the chassis. I measured the voltage on the grid and it was at 30V. Oops … the expected voltage wasn’t on the chart, but given the way the circuit is designed it should be very close to the cathode voltage. I disconnected the mica capacitor from the circuit and tested it with a capacitor checker. Sure enough, it was leaky. The resistor had also drifted up in value by about 25% so I replaced the resistor along with the capacitor.

After powering up this time, the hum was gone and the amplifier sounded significantly better but still lacking bass. The bass control seemed to adjust the bass level slightly, but not much. As I was fiddling with it the bass suddenly came in loudly and just as quickly went away again - definitely a symptom of a bad connection. I powered down and cleaned all the contacts thoroughly for both the treble and bass control. Upon powering up and playing another selection I was reward with much fuller bass. At this point the audio sounds very good and no other repairs to the amplifier were required.

Chassis after from above Chassis after from below

 

Continued on Page 2

Copyright © 2009 Patrick Nelson