The Crumar GDS Restoration Blog...

Discussions about anything analog, digital, MIDI, synth technology, techniques, theories and more.
User avatar
HideawayStudio
Supporting Member!
Supporting Member!
Posts: 1387
Joined: Fri Jan 09, 2009 10:41 pm
Real name: Dan Wilson
Gear: 163 tubes in a large wooden box!
Band: Shortwave
Location: UK
Contact:

Re: The Crumar GDS Restoration Blog...

Post by HideawayStudio » Mon Nov 10, 2014 1:43 am

ppg_wavecomputer wrote:How about the Crumar DP electric pianos (or the Baby Grand, perhaps)? They were later but...

Stephen
Thanks for this Stephen.. I'd actually completely forgotten about this - when I first got my Synergy I'd noticed that the keys were very similar in profile to the DS-50 e-piano and indeed there are some other similarities such as the side profile of the instrument and the tactile switches with integral LEDs.

Image

Sure enough the keybed/caps/levers are very similar indeed but extremely frustratingly I think the plastic actuator at the end of each lever is different. You can see the contact assembly behind the levers is very different in the DS-50.

It is possible of course that this aspect of the keybed was custom designed for the GDS and later the Synergy but I'm still wondering if there is a 5 octave intermediate design we've overlooked that the GDS used as a carry over part. Mind you,it is possible there was one developed for volume use but it never saw the light of day.

User avatar
synthroom
Senior Member
Senior Member
Posts: 875
Joined: Sat Sep 08, 2012 3:16 pm
Real name: Kirk
Gear: Fairlight IIx(!), JP-8, D-50, S-50+550, S-760, JX-3P, JD-800, EII, Emax II, Mini, ARP 2600, P-5 Rev.1, Pro-One, Performer, K1m, K5m, few other things.
Location: pdx
Contact:

Re: The Crumar GDS Restoration Blog...

Post by synthroom » Mon Nov 10, 2014 6:44 am

How about 3d printing "new" keys?
Fairlight IIx (Mid-Life Crisis - cheaper than a Corvette!)
Roland JP-8, D-50, S-50, S-550(2x), S-760(2x), JX-3P, JD-800
EII, Emax II, Minimoog, ARP 2600, P-5 Rev.1 (broken...), Pro-One, Crumar Performer, K1m, K5m, MS-2000B, Virus KC, a few other things.

User avatar
HideawayStudio
Supporting Member!
Supporting Member!
Posts: 1387
Joined: Fri Jan 09, 2009 10:41 pm
Real name: Dan Wilson
Gear: 163 tubes in a large wooden box!
Band: Shortwave
Location: UK
Contact:

Re: The Crumar GDS Restoration Blog...

Post by HideawayStudio » Mon Nov 10, 2014 4:43 pm

synthroom wrote:How about 3d printing "new" keys?
This is definitely on the cards - I have indeed been considering using an FDM (Fused Deposition Modelling) machine to perform this in ABS. The issue is that I need to key the parts 3d modeled first.

User avatar
ppg_wavecomputer
Expert Member
Expert Member
Posts: 1007
Joined: Wed Feb 06, 2013 3:37 pm
Gear: more keys than hands (and feet)
Band: ['ramp]
Location: teutoburg forest, eastern westphalia, germany
Contact:

Re: The Crumar GDS Restoration Blog...

Post by ppg_wavecomputer » Tue Nov 11, 2014 2:34 pm

Perhaps you can have some kind of mould made in a dental lab? They might also know what materials to use in terms of longevity, and it might be easier to do as well.

Stephen
"Like the light from distant stars, Stephen Parsick's music has existed for some time, but is only now reaching us on Earth." Chuck van Zyl

https://doombientmusic.bandcamp.com/

https://ramp1.bandcamp.com/

https://stephenparsick.bandcamp.com/

User avatar
HideawayStudio
Supporting Member!
Supporting Member!
Posts: 1387
Joined: Fri Jan 09, 2009 10:41 pm
Real name: Dan Wilson
Gear: 163 tubes in a large wooden box!
Band: Shortwave
Location: UK
Contact:

Re: The Crumar GDS Restoration Blog...

Post by HideawayStudio » Wed Nov 12, 2014 11:29 pm

ppg_wavecomputer wrote:Perhaps you can have some kind of mould made in a dental lab? They might also know what materials to use in terms of longevity, and it might be easier to do as well.

Stephen
One of the broken key actuator spigots is now with a 3d printing company awaiting a quotation for the 3d modelling and printing of at least 30 parts on a Fused Deposition Modelling machine in ABS.
Last edited by HideawayStudio on Thu Nov 13, 2014 12:34 am, edited 1 time in total.

User avatar
HideawayStudio
Supporting Member!
Supporting Member!
Posts: 1387
Joined: Fri Jan 09, 2009 10:41 pm
Real name: Dan Wilson
Gear: 163 tubes in a large wooden box!
Band: Shortwave
Location: UK
Contact:

Re: The Crumar GDS Restoration Blog...

Post by HideawayStudio » Thu Nov 13, 2014 12:06 am

I have spent the last couple of days stripping down Brian's GDS System main CPU unit to a bare chassis.

Image

You don't really get a feel for the scale - the computer cabinet is around 3ft long and over 2ft wide and probably weighs over 50kgs fully laden.

All in all the exterior of the cabinet has weathered the past 33+ years well with the fake wood end cheeks in good condition and a little surface rust developing internally due to exposure to damp conditions in the past.

Having removed all of the S100 cards (10 in total) I set about visually inspecting them.

Image

The second card I inspected was the I/O card which is a stock IMS8000 S100 bus computer card. This card has two main functions in life. It provides the serial communications interface (UART) to drive the serial terminal and is home to the infamous IPL ROM (Initial Program Loader) which is like a BIOS ROM on a PC in that it runs enough code to permit the machine to boot from disk.

Brian had mentioned that there was a distinct burning smell from the right hand side of the card cage and sure enough the I/O card (which is located in the right hand most slot) was clearly the source. This serves as a sobering example as to why it is so important that PTH tantalum caps are replaced in equipment of this age when they are situated directly across power rails (which sadly most are). The black axial caps on the stock GDS computer cards are of a make/type that is well known for failure and usually the failure mode is dead short circuit. This must have made someone jump when it failed!

Image

In the case of the GDS there is so much energy stored in the huge linear PSU reservoir caps that anything in the path of the supply rails is instantly vaporized! If you look closely you can see molten metal has sprayed out of capacitor C13 and has left a solid ball of metal at the point the end of the cap has been blown off. The track feeding the cap has ruptured leaving no trace of several millimeters of track.

The good news is that after recapping the I/O card and repairing the ruptured track the card fired up and tested ok in the KS GDS.

I then set about recapping the Z80 CPU card which also tested ok. It was very satisfying to see the KS GDS boot up under the command of the CPU card from Brian's beast. The floppy disk controller card followed and also successfully booted up the KS GDS.

One of the main differences between the computer system units on the two GDS Systems is that Brian's unit is fitted with four 16K memory cards and the KS GDS has been upgraded circa 1981 with a later single 64K memory card.

I spent the afternoon removing 32 tantalum bead capacitors from the four memory cards and replacing them with modern low ESR electrolytics. At least one of these was looking a little dubious. What I was not totally sure of was whether it was possible to simply remove the single 64K memory card from the KS GDS and fit the four 16K cards without some form reconfiguration or worse still a different system disk. To my relief the KS GDS booted straight up with the four cards fitted and to my delight after a few minutes of running the MEMTEST utility a memory test pass message came up on the terminal.

This is definitely promising as there are 128 35 year old memory chips across these cards!!

The next step is to recap the custom Crumar S100 cards and, like the above, test them in turn in the KS GDS.
Last edited by HideawayStudio on Thu Jun 04, 2015 11:14 pm, edited 3 times in total.

User avatar
madtheory
Supporting Member!
Supporting Member!
Posts: 5156
Joined: Mon May 01, 2006 12:45 pm
Real name: Tomas Mulcahy
Gear: Flangebeast Mk1, Plonkotron, Morovdis Arpeggiator, Maplin My First EQ, Jeff Wayne Thunderchild rack, Thermostat, Buck Owens' Moog.
Band: Minim
Location: Cork, Ireland
Contact:

Re: The Crumar GDS Restoration Blog...

Post by madtheory » Thu Nov 13, 2014 12:31 am

So cool that you have a complete working GDS to test with. And of course your know how :)

User avatar
HideawayStudio
Supporting Member!
Supporting Member!
Posts: 1387
Joined: Fri Jan 09, 2009 10:41 pm
Real name: Dan Wilson
Gear: 163 tubes in a large wooden box!
Band: Shortwave
Location: UK
Contact:

Re: The Crumar GDS Restoration Blog...

Post by HideawayStudio » Thu Nov 13, 2014 12:33 am

madtheory wrote:So cool that you have a complete working GDS to test with. And of course your know how :)
I must admit its a real luxury to now have a working system for comparison and test.

I keep thinking it must have been a heck of a long time since two GDS Systems have sat side by side...

I mean.. how often will anyone see a sight like this!!....

Image

And speaking of rare sights...

Image

I'm quite surprised no one has picked up on this yet... if you watch Laurie Spiegel's live improvisation from 1977 at the controls of the Bell Labs Alles Machine aka "Blue Monster" you can just make out the side profile of an ADDS Regent 20 Serial Terminal sitting on the floor behind the desk ie. the same model that was carried over to the GDS.

User avatar
synthroom
Senior Member
Senior Member
Posts: 875
Joined: Sat Sep 08, 2012 3:16 pm
Real name: Kirk
Gear: Fairlight IIx(!), JP-8, D-50, S-50+550, S-760, JX-3P, JD-800, EII, Emax II, Mini, ARP 2600, P-5 Rev.1, Pro-One, Performer, K1m, K5m, few other things.
Location: pdx
Contact:

Re: The Crumar GDS Restoration Blog...

Post by synthroom » Thu Nov 13, 2014 6:59 am

HideawayStudio wrote:
Image
Make sure you don't remove that "warranty voided" sticker! You don't want to get in trouble with the factory service guys...
Fairlight IIx (Mid-Life Crisis - cheaper than a Corvette!)
Roland JP-8, D-50, S-50, S-550(2x), S-760(2x), JX-3P, JD-800
EII, Emax II, Minimoog, ARP 2600, P-5 Rev.1 (broken...), Pro-One, Crumar Performer, K1m, K5m, MS-2000B, Virus KC, a few other things.

User avatar
MiK
Newbie
Newbie
Posts: 10
Joined: Mon Sep 29, 2014 8:33 am
Real name: Michael
Gear: Lots of stuff that once was broken. And some more.
Location: South Germany

Re: The Crumar GDS Restoration Blog...

Post by MiK » Thu Nov 13, 2014 7:55 am

HideawayStudio wrote:Image
Interesting, the left one has no CRUMAR logo. Has it been removed for some reason or did they really have different panel prints with this low number of units made?

If you were a bit closer, i would bring mine to create a shot with 3 of those controllers next to each other :)

User avatar
HideawayStudio
Supporting Member!
Supporting Member!
Posts: 1387
Joined: Fri Jan 09, 2009 10:41 pm
Real name: Dan Wilson
Gear: 163 tubes in a large wooden box!
Band: Shortwave
Location: UK
Contact:

Re: The Crumar GDS Restoration Blog...

Post by HideawayStudio » Thu Nov 13, 2014 10:19 am

MiK wrote:
Interesting, the left one has no CRUMAR logo. Has it been removed for some reason or did they really have different panel prints with this low number of units made?

If you were a bit closer, i would bring mine to create a shot with 3 of those controllers next to each other :)
Yes, I never thought about it like that before... probably the very first time 3 GDS Systems have been in Europe!

Well spotted, there is no large plastic Crumar logo badge on the rear of the keyboard console cabinet either.

We think Brian's unit is at least in part made up of a curious mixture of prototype custom GDS parts (including the front panel) from 1979 and a stock IMS8000 computer from 1981. It was originally a sales demonstrator so this is very much a possibility and is therefore an interesting piece of history in its own right. The older style 4x16 memory cards are dated 1979 so my theory is that a customer received the computer's newer 64K card (the KS GDS has had this upgrade too) and the customer returned 4x16s have been installed in Brian's unit.

There is a running number on the main oscillator boards - the KS GDS is marked /6 and Brian's /3. My guess is that 10 or so oscillator boards were made in one batch in 1979. Question is, what happened to /1 ?

User avatar
HideawayStudio
Supporting Member!
Supporting Member!
Posts: 1387
Joined: Fri Jan 09, 2009 10:41 pm
Real name: Dan Wilson
Gear: 163 tubes in a large wooden box!
Band: Shortwave
Location: UK
Contact:

Re: The Crumar GDS Restoration Blog...

Post by HideawayStudio » Thu Nov 13, 2014 10:30 am

synthroom wrote:
Make sure you don't remove that "warranty voided" sticker! You don't want to get in trouble with the factory service guys...
...trouble is, I seem to have become the factory service guy! :D

User avatar
HideawayStudio
Supporting Member!
Supporting Member!
Posts: 1387
Joined: Fri Jan 09, 2009 10:41 pm
Real name: Dan Wilson
Gear: 163 tubes in a large wooden box!
Band: Shortwave
Location: UK
Contact:

Re: The Crumar GDS Restoration Blog...

Post by HideawayStudio » Thu Nov 13, 2014 11:35 pm

I'm pretty knackered this evening - today has been another pretty intense recapping session with over 70 axial and radial PTH tantalum capacitors removed and replaced on Brian's GDS board set...

Image

This has included the caps mounted under the huge heatsink for the 5 volt regulator for the oscillator board.

The lab currently feels like its a shrine to MTI/Digital Keyboard's wonders all in pieces awaiting attention at the moment...

Here is Brian's GDS DAC board recapped and awaiting testing complete with its gorgeous output transformer and 16-bit Burr Brown DAC. Note the two devices that look like enormous resistors are in fact filter inductors.

Image

The output DAC and main High Speed Digital Oscillator Board were reassembled onto the top lid of the computer system.

Image

The hope is that the PSU can be recapped early next week and the whole GDS computer system reassembled for its first full power up under controlled conditions shortly after the remaining three custom Crumar S100 cards have been recapped and a fresh IPL ROM programmed.

User avatar
HideawayStudio
Supporting Member!
Supporting Member!
Posts: 1387
Joined: Fri Jan 09, 2009 10:41 pm
Real name: Dan Wilson
Gear: 163 tubes in a large wooden box!
Band: Shortwave
Location: UK
Contact:

Re: The Crumar GDS Restoration Blog...

Post by HideawayStudio » Fri Nov 14, 2014 10:44 pm

After another hard day recapping the remaining three boards in Brian's GDS S100 board set I spent a couple of hours this evening rubbing copious amounts of metal polish into the empty computer chassis metalwork. After a fair amount of elbow grease the nickel plating started to come up quite nicely and the exposed parts should look nice on reassembly.

Image

Earlier today I finished recapping the remaining three S100 boards - the three custom Crumar boards which act as the interface to the main digital oscillator board in the top lid, the interface and regulated supply rails to the keyboard console and the memory mapped I/O and ADC for the sliders and switches.

I was only replacing the tantalum caps but on closer inspection I found the top of one of the disk capacitors was missing leaving a blackened hole where it looks like the cap has failed short circuit. This is actually quite rare in PTH ceramic disk caps (much more common in SMT) and it may well have been the result of some kind of physical damage probably scuffing from the adjacent card on removal at some point in the past.

Somewhat unnerved by this I decided to replace these 100NF parts on the affected board with high quality metal film caps.

Speaking of bad caps...

Whilst working on something entirely different I found an almost comical ECN (engineering change note) detailed in the back of a service manual from 1983. I won't say who the manufacturer was but it was a very major name.

Image

This is fascinating because its strong evidence that the infamous dipped blue caps in many pieces of kit (ARPs!) of the era were known to be troublesome even back then!

The hope is to start reassembling the recapped pcb set into the chassis over the next few days and recap the PSU early next week.

Image

User avatar
HideawayStudio
Supporting Member!
Supporting Member!
Posts: 1387
Joined: Fri Jan 09, 2009 10:41 pm
Real name: Dan Wilson
Gear: 163 tubes in a large wooden box!
Band: Shortwave
Location: UK
Contact:

Re: The Crumar GDS Restoration Blog...

Post by HideawayStudio » Mon Nov 17, 2014 10:56 am


SUCCESS... GDS runs under an HxC disk emulator for the first time... ever!...


After two particularly intense days not only have I managed to hook up the HxC disk emulator as a second disk drive to the KS GDS but I’ve successfully created a suitable blank disk image from scratch and format the first disk image under CP/M 2.2 in place of one of the 8" disk drives.

Image

This involves building a cable to map the later 34 pin Shugart floppy disk data cable standard found on 5.25" and 3.5" drives to the original 50 pin Shugart SA800 pinout. I was initially going to build this cable myself but JB Emond, the Frenchman behind the wonderful flash kit disk emulator for the Fairlight Series II, supplied me with the data cable ribbon adaptor from the kit which is perfectly suitable for the IMS8000 GDS computer too.

I temporarily set the HxC up as drive B and one of the 8" disks as drive A.

This initially permitted me to boot up off the 8" and copy the entire contents of all of Brian's Synergy factory patch disks in GDS format to several new disk images. The virtual copies loaded up flawlessly.

The absolute beauty of this system is that it stores several thousand virtual disk images as files in FAT32 format on an SD Card which can simply be pulled out, plugged into your PC, and all the files copied off in seconds. This means when I setup the second system in the same manner literally hundreds of little CP/M files can be copied over in workable form in seconds having already done the hard work of imaging the disks.

I have now fitted this wonder to several DK Synergies (Kaypro II), Emulator II, Prophet 2000/2002 and AKAI S900/S950 to similarly great effect.

After some further tinkering the following day I managed to reformat another raw image and successfully transfer the IMS8000 CP/M OS boot sectors over to the image thus making a bootable virtual system.

After some further fiddling in order to reconfigure the HxC as drive A and the floppy as drive B (and remembering to fit the termination resistors!) I managed to make the KS GDS boot and run from the HxC!

Image

So I managed to run up the entire KS system under twin disk emulation (no physical floppies at all).

Fuelled by this I spent this afternoon reassembling Brian's fully recapped (bar the PSU) GDS system unit along with the HxC system in place of both 120V 60Hz drives (I’ve got some kit on order in order to test these and the HxC is capable of emulating a twin disk system) and to my complete and utter delight his system BOOTED AND RAN AND SUCCESSFULLY LOADED UP THE VOICE APPLICATION AND PLAYED!!!!!

This is literally the first audio out of Brian's machine for quite some time....



And a picture of this wonderful event.

If you look carefully you can see that the US version sports two mains receptacles on the rear panel which are not present in the EU version (all two of them built!)..

Image

This was under the control of the KS GDS keyboard console – I still need to recap Brian's and the PSU early this week.

There is still a lot of work to do but this is extremely promising progress...

Some of the tasks that still lay before me...

KS GDS:

To dismantle remove and polish the keyboard action bus bars (very badly tarnished with age)
To refit the keybed and front panel assemblies into the newly re-veneered keyboard console cabinet
To strip and powder coat the top panel of the CPU cabinet and the new front trim strip for the keyboard console
To source and fit 6 replacement foil and foam pad key contacts to the ADDS terminal
To mount the HxC drive on a panel in place of the A-Drive

BK GDS:

To overhaul the serial terminal and rebuild the keyboard (most of the keys are sticking badly)
To prove the 3d model for the 30 or so broken key actuators and test the first 5 rapid prototype replacements
To rebuild the keyboard action with the replacement actuators
To strip down the front panel assembly and recap the front panel board
To reassemble the keyboard console
To test and run the 120V 60Hz floppy drives using a pure sine invertor
To mount the HxC drive on a panel in place of the A-Drive

Post Reply