Gulbransen Theatrum Service Manual - Missing Tones

[organisimo]

Hey guys,
I own a Gulbransen Theatrum and am having issues with some notes not sounding on both manuals. I've gone through and cleaned all the key-switches and bus-bars under each keyboard manual, however some of the tones; for example the 16' & some of the 8' & 2' drawbars and rockers only sound on some of the keys, the missing sounds aren't intermittent they are constant. This configuration seems different to my Hammond TX5 which has bus-bars for each footage tone running under every key.

Does anyone have a service manual or schematics for this model in PDF format that I can download, otherwise, can someone point me in the right direction regarding where I should start looking to get these missing tones back?

Many thanks! :emotion-21:
Peter
Australia

[Jan Girardot]

Hello and welcome, Peter:

It would be helpful to know which Theatrum you own, the 3218 (1971) or the 3318 (1976). I have sold all the Theatrum manuals from collection, so suggest you contact this source:
Dave's Electronics
105A East Penn St.
Hoopeston, IL 60942
USA
Phone: (217) 283-5010
Fax:* (217) 283-5161
Products and Services:*All Pre-1986 Gulbransen models: Parts Service Manuals and Owners Manuals.
email:*daveelec@frontier.com

. . . Jan
the OrganGrinder

[organisimo]

Thanks Jan, it is a 3218 built in 1975.



So far I've got rid of all the dust and crud out of it using a soft brush and compressed air, I have gone through and polished the bus-bars of every key & switch contact and re-seated all of the circuit board cable connectors, but sadly it does not seem to be a mechanical problem as the missing tones don't re-appear intermittently, so I'm guessing I have an electronics problem. Do you or anyone else know which board/s I need to look at that generate the raw tones?

Many thanks :)

[indianajo]

Reseating all the cable connectors is certainly a first treatment, as there seem to be many of them on the bottoms of the boards in the middle of the picture. Brass or tin and <48 v low current signals do not mix very well over time, oxide builds up. One of these sockets could need particular scraping or squeezing or something.
I think establishing the flow of music with a detector would be the next thing, as much as I disrespect those brown and silver electrolytic caps lurking down on the lower left amp chassis.
I use a Simpson 260 -6XLPM VoltOhmmeter with a 2 VAC scale and 20 vac scale to detect music, with a .047 600 v cap attached to one probe to prevent DC signals from reading as AC signals. DVM tend to produce random numbers on AC not at the power line frequency. There is always the oscilloscope, some under $100 at diytools for example. Some people use a "sound probe" which is an amplifier-speaker combination with some sort of limiter on the front to prevent you from blowing up the amp. Powered computer speakers used to be the thing for snooping for sound, although they are all dying of bad power supply caps and dirty connectors like everything else PC related five or more years old. The input voltage limiter could be the cap listed above, followed by two LED's back to back across hot and "ground" of the input plug, to limit the input current to 1.2 VAC. Maybe a 100 ohm resistor before the LED to keep from blowing it up if you hit the amplifier output by mistake and put in a really powerful signal. The advantage of the sound probe is that you hear what pitch it is so you know if you are on the pitch that is missing by ear.
So you look for sound. It starts at those boards with the square cans on them, the oscillator boards. Then it should go somewhere near to divider (by two for octaves) circuits, probably pairs of transistors with resistor-capacitor networks (multi-leaded) for each divide by 2. I would guess those middle boards might be the dividers. However, by 1973 there were medium-reliable RTL divide by two ICs with four or six legs, that could be installed and failed. Especially if there is a pattern of related octave notes being missing. Also check the power supplies on the divider boards that they are about 80% of the bypass capacitor voltage , 8 to 20 vdc anyway. A leaking bypass electrolytic cap can ***** up a whole divider stage. The divider IC's can be substituted by boards from people like OrganService of Marengo, IN, or built up out of discrete transistors capacitors and resistors on perfboard, or with some level shifting, replaced by cd4013b or SN74hc74n flip flop ics on dip project board.
Then the sound probably would go to the key switches, then back to mixer-filter networks.
I could probably make some progress without a schematic diagram, but I know what I am looking for. Buying a schematic diagram is a good way to start. Good luck.

[andyg]

I've moved this to the correct section, Home Organs, where it will get more views. Andy-moderator.

[organisimo]

Reseating all the cable connectors is certainly a first treatment, as there seem to be many of them on the bottoms of the boards in the middle of the picture. Brass or tin and <48 v low current signals do not mix very well over time, oxide builds up. One of these sockets could need particular scraping or squeezing or something.
I think establishing the flow of music with a detector would be the next thing, as much as I disrespect those brown and silver electrolytic caps lurking down on the lower left amp chassis.
I use a Simpson 260 -6XLPM VoltOhmmeter with a 2 VAC scale and 20 vac scale to detect music, with a .047 600 v cap attached to one probe to prevent DC signals from reading as AC signals. DVM tend to produce random numbers on AC not at the power line frequency. There is always the oscilloscope, some under $100 at diytools for example. Some people use a "sound probe" which is an amplifier-speaker combination with some sort of limiter on the front to prevent you from blowing up the amp. Powered computer speakers used to be the thing for snooping for sound, although they are all dying of bad power supply caps and dirty connectors like everything else PC related five or more years old. The input voltage limiter could be the cap listed above, followed by two LED's back to back across hot and "ground" of the input plug, to limit the input current to 1.2 VAC. Maybe a 100 ohm resistor before the LED to keep from blowing it up if you hit the amplifier output by mistake and put in a really powerful signal. The advantage of the sound probe is that you hear what pitch it is so you know if you are on the pitch that is missing by ear.
So you look for sound. It starts at those boards with the square cans on them, the oscillator boards. Then it should go somewhere near to divider (by two for octaves) circuits, probably pairs of transistors with resistor-capacitor networks (multi-leaded) for each divide by 2. I would guess those middle boards might be the dividers. However, by 1973 there were medium-reliable RTL divide by two ICs with four or six legs, that could be installed and failed. Especially if there is a pattern of related octave notes being missing. Also check the power supplies on the divider boards that they are about 80% of the bypass capacitor voltage , 8 to 20 vdc anyway. A leaking bypass electrolytic cap can ***** up a whole divider stage. The divider IC's can be substituted by boards from people like OrganService of Marengo, IN, or built up out of discrete transistors capacitors and resistors on perfboard, or with some level shifting, replaced by cd4013b or SN74hc74n flip flop ics on dip project board.
Then the sound probably would go to the key switches, then back to mixer-filter networks.
I could probably make some progress without a schematic diagram, but I know what I am looking for. Buying a schematic diagram is a good way to start. Good luck.

Lots of good stuff there indianajo many thanks!

Thankfully this model has no mysterious LSI chips :-> everything seems to be made up of discreet components, except for these 4 pin ICs in the oscillator section (see pics)


Hopefully this will make component checking and overall fault-finding a lot simpler, especially for my basic electronics training from 35 years ago. I will look at the oscillator boards you mentioned first and make sure I'm getting an output from each tone and then move onto the divider board.


I managed to find someone selling a service manual for this model on eBay in Australia (the planets must have been in alignment LOL) which should arrive next week.

Yes those large cardboard electrolytic caps in the power supply are a bit of a worry, but since the secondary voltages around the power-supply and those delivered around to the circuit boards in the organ are accurate to +/- 1VDC I might leave those last.

Thanks again for your input :) and thank you to Andyg for shifting this to the right forum. :emotion-21:

[indianajo]

Those black IC's with the straddled legs definitely have a Motorola logo. I think the 5xx numbers are RTL IC numbers, that is resistor transistor logic. My supervisor at my first tech job had a book of those, but I never did. I don't think there is anything about them on the internet either. Hope one is not bad, but inside they are just a couple of transistors and some resistors, so if you can figure out what it is, it isn't hard to replace one with a little daughter board with those discrete parts on it. Hopefully you will find sound coming out of each one.
I can't believe you are doing this in the heat of summer. I usually save projects like this for the Ides of January, like today when it is -20C and 0 % humidity or something like that (4 deg F).Go right out and get a sunburn! I'm envious right now.

[organisimo]

Those black IC's with the straddled legs definitely have a Motorola logo. I think the 5xx numbers are RTL IC numbers, that is resistor transistor logic. My supervisor at my first tech job had a book of those, but I never did. I don't think there is anything about them on the internet either. Hope one is not bad, but inside they are just a couple of transistors and some resistors, so if you can figure out what it is, it isn't hard to replace one with a little daughter board with those discrete parts on it. Hopefully you will find sound coming out of each one.
I can't believe you are doing this in the heat of summer. I usually save projects like this for the Ides of January, like today when it is -20C and 0 % humidity or something like that (4 deg F).Go right out and get a sunburn! I'm envious right now.

I have a bunch of semiconductor reference books from the early 80's when I was studying electronics hopefully there's something in there about these ICs, but then it is unlikely for low-power semiconductors to fail under normal working conditions, unlike the components in power amps and switching power supplies. I'd be more inclined to think it's a dried out capacitor or dry joint somewhere among the hundreds of pin headers and daughter-boards, but I won't really know until I can see a system block diagram so I know what each board does - hopefully my service manual arrives soon.

Yes Australia is going through a heatwave at the moment, though we've had a bit of a reprieve over the last 7 days and the temperatures have dropped down to the low 30's (about 80 deg F) B-) plus it's given me a good chance to leave the organ outside in the sun so I can get rid of that old furniture smell before I put it in the studio - came from deceased estate ;)

[organisimo]

Okay so I've done some more fault tracking this morning and have discovered that all the footage tones sound however not on every key, so for example: The 16' tone sounds on mostly the higher octaves of both upper & lower manuals while the 4' sounds on all keys. The upper & lower manuals fail on exactly the same tones, this brings me to the following conclusions:
1) All the key contacts work however not all tones sound on every single key
2) The tone generator is working on all tones
3) As an aside, only the C key is able to sound every footage tone on every octave

The keyboard upper & lower manuals run directly into the left-most rear board which are then routed to the right-most multiplexing board, before going on-route to the tone generator board via the blue multi-core cable - see picture. Would I be right to assume that the fault lies somewhere in those 3 rear boards?

[indianajo]

Unless a public source of schematic diagram becomes available, or someone that has done this repair answers, only you can determine the flow of Ac signals in your organ. If signals go into a section on one board, and come out, but other signals go in and fail to come out, that is your sign that something is wrong in that area. As the boards on the pulldown shelf appear to be soldered, I wouldn't suspect connectors in that area, is the best advice I have. Again the "sound" probe or protected amplifier-speaker combination can tell you if you are tracing the same pitch in and out.
If you discover a correlation between boards that don't have go-outa but do have go-inta, and DC power supply variations, then perhaps you will have some data on what could be wrong. You may develop other theories based on your test data.

[organisimo]

I won't get my service manual till Tuesday so I'm just doing a little detective work till then. My hope is to use this forum topic as a running dialogue to help me get through this repair and hopefully with the aid of my detailed pictures be able to assist someone else who might have a similar problem in the future ;)

[organisimo]

Okay so the good news first my Gulbransen service manuals arrived today complete with Service Bulletins from 1976. In the package the seller also included service manuals for; President, Paragon, Palace & Music Computer organ models as well!

Anyway, after spending an hour studying the block diagrams, I dusted off my old CRO, Frequency Counter and AF Probe and started checking each oscillator to see that they were generating a tone... all produced a fundamental tone.

But then my joy turned to sadness as I began to check the octave dividers - yes those little 4 pin ICs - and discover that eleven (11) of them are dead :'( damn it! Of all the components that could have failed it is these little bastards! According to the manual they are Bistable Multivibrators and marked as: Motorola 818 504. Can I even buy these anymore? I did a quick search on the web but found nothing!!


Now I'm at a crossroad; do I just salvage the Leslie & 35 watt amp and scrap the rest of the organ (it owes me $100 plus the hours spent cleaning it) or, do I persevere and try to hunt down the dozen or so of these elusive ICs? :-\

[geoelectro]

It is highly unlikely you have a dozen of these dividers bad. They work in series. If one fails, all the rest after it won't run. Simply find the first divider not working in the chain, (high to low) and replace it. Rather than ask how many total notes are dead, how many notes in a one octave scale are affected? A, B, C etc.

Geo

[organisimo]

Hi Geo, yes you are correct, the square wave output of each divider feeds the next one all the way down to the lowest octave in each key, so if one fails, the rest in that chain will not sound either because they need that square wave trigger at their input - a bit like the old-school xmas tree lights where the bulbs were wired in series.

To give myself more room to work, I removed the two toneboards from the organ and hooked them up to a lab power supply set to 17VDC. I then proceeded to test each divider individually, by injecting a clean square wave into the input pin of each IC and monitor their output; the ones that did not produce a sound were marked as bad ones. I know how unlikely it seems to have so many dead divider ICs, but I triple checked each one and sadly there are almost a dozen bad ones - this is just incredible bad luck I guess :( based on the research I did on these RTL chips on the web, they have been a redundant technology for some decades and replaced by TTL technology.

Does anyone know of where I could get a dozen of these divider ICs from or someone that is parting out a Theatrum or similar vintage Gulbransen model that I could retrofit the tone boards out of?

Thanks :)

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

EDIT: I looked into how Bistable Multivibrators work and they don't seem to be very complicated circuits.

They are basically a couple of transistors wired in a flip flop configuration. What makes them special in this application however, is that they require two state changes (pulses) to output a single state change hence why they are ideal as an f/2 frequency divider. The ones that Gulbransen use in this model operate on fairly high voltages i.e the oscillator buffer outputs a 9vpp square wave signal and as far as I know the maximum operating B+ for TTL chips is 5VDC or around 12VDC for CMOS gates.

Now, if I can't find these chips or an equivalent modern day replacement somewhere, and nobody is parting out a similar model organ, I may have to look at creating something from scratch using some sort of TTL chip that contains a gang of flip-flops (as indianajo suggested in an earlier post) but geez that would have to be an absolute last resort...

[indianajo]

RTL was so notoriously unreliable that they were a dead issue at the company I worked at by 1973. They still used vacuum tubes for some things, but the RTL had been purged out of the system. Probably a packaging issue, IMHO Motorola was the first to use plastic injection molding on IC's and TI was doing it better with ceramics on the TTL.
If you feed the not Q output into the D input, and tie the resets to ground, a CD4013 IC will divide by two up to 20 V. They only put out about 1.5 ma though, you may need more than that produced by a transistor or a interface IC 75xxx. You could put them on 21-4580 dip project board from mcmelectronics. Use sockets, you don't want to solder directly to an IC.
With the drive requirements maybe needing a transistor, I think I would just build up some of the 2 transistor flip flop circuits. Organservice of marengo IN sells them specifically for divider replacment, or you can get circuit boards made up to your pattern by various shops listed in the construcion threads of diyaudio.com. Lubuntu 11 op system has some schematic layout programs built in or look up the windows versions of CAD on the same forum.
If you want to just replace them all, a CD40103 has eight divide by twos in the same package. Low current output again, you'll need to investigate buffer IC's for that.
Or you could buy a cheapo import organ from the late seventies or early eighties, use the top octave and divider board, throw the cheap tin keyswitches away, throw away the cheap filters and plastic case, and use the Gulbransen filters. Some voltage and current level interfacing probably required. You don't want one with 40 pin IC's, probably the keying functions are built into the master IC or something. You want just a TOS (top octave synthesizer), followed by a divider chain IC with maybe 24 pins.