Nascom 1 restoration
|
Arrival image
The wiring started off on from the power supply, then it went.......'random' :) This may need a bit of work! Sold as not working, so I am going to treat it as faulty and rebuild it in phases to simplify testing. A complete strip down will be the first step. |
CASE
Needed a good clean then I blocked most of the 101 holes that had been drilled into it!, but left the obvious ventilation holes.
Also, the case did not have any holes, etc. to fix the top to the bottom, I added some ABS bars across the front and the back then drilled holes through, rather than use nut and bolt or self tapping which can be weak I embedded some M3 pillars in the top with M3 machine screws going in from the bottom.
Needed a good clean then I blocked most of the 101 holes that had been drilled into it!, but left the obvious ventilation holes.
Also, the case did not have any holes, etc. to fix the top to the bottom, I added some ABS bars across the front and the back then drilled holes through, rather than use nut and bolt or self tapping which can be weak I embedded some M3 pillars in the top with M3 machine screws going in from the bottom.
|
MAINS WIRING AND TRANSFORMER Originally the mains was wired directly to the transformer, as the cable had the older mains plug on I just threw the lot away. So, starting with a new IEC cable, correct plug and the IEC connector on one end, I added a fused and filtered IEC socket to the case, from here it went to a rocker switch mounted on the side, then on to a fused connector block, finally from here it went to the transformer. The transformer had a snapped tag on the secondary side so I replaced this. It was interesting that all of the low voltage wiring was routed very close to the mains input terminals of the transformer (see first image above), to fix this I rotated the transformer so that the mains input was as close as possible to the entry point as possible. When it came to the secondary side I tried several methods before settling on the 'over the top' method, routing the wires around the side worked but looked messy. So, over the top they went, to protect the wires I added plastic slides to the transformer edges, I also grouped the 9v ac and the 15-0-15v ac feeds. |
finished rework 
|
POWER SUPPLY
First look, some obvious overheating marks under the R6 resistor and B1 bridge rectifier, and a strange capacitor replacement (c7), presumably when R6 went supernova it melted the original capacitor. I removed R6 to clean the charcoal off the board which although conductive would have made no difference to the low resistance of R6. I also replace the triple capacitor affair. None of the voltage regulators had any heatsink compound on so I removed all of them and added HS compound. The underside was a little messy with the reverse protection diodes just laid across open tracks, etc. so I removed these and installed them from the top. The 4 large electrolytics had started to leak electrolyte so all of these were replaced. I decided to make use of the edge connector facility of the board (originally it was directly wired), so made up a connector so that the board could be easily unplugged for any future maintenance, etc. This released the low voltage points at the other end so I added test point loops. All other components tested fine. Testing... With a 1.7A load the bridge rectifier ran at 147°C (just hot enough to leave a mark on my testing finger!), I added a small heatsink but this was not enough only bringing it down to 97°C. Decided to just replace it with a 800V/25A version, this ran at 69°C and with added heatsinks 48°C. The 2N3055 was running at 78°C (too hot despite the heatsinks), I may come back to this when the machine is loaded with the PCBs as the current draw may be much less and thus run cooler. I added a small self adhesive heatsink to the top of he can just to help a little, I do remember that back in the 80's just about all power supplies based on this cct always ran really hot, most of the time we just added a small fan, this may yet happen. |
|
MAIN BOARD
First job, remove all the IC's, a lot of them were 'crusty' on the legs so all IC's were cleaned with a fibreglass pen, tested with the RCTPro then stored in antistatic foam which had been soaked with contact cleaner. IC18 pin 5 was not in the socket and had been bent under the IC, straightened ready for refitting. |
|
|
The board had previously been directly wired, with some heavy wire and questionable soldering, this can and did lift a number of the pads/tracks, after a good clean up removing all the old solder, etc. I installed Dupont pins in all the connection points. Also removing all the wired links and replacing them with jumper pins to allow easy reconfiguration, some of these I may extend to the case as switches, etc.
To add some structural stability to the pins I added UV resin to the underside of the board where these came through.
The bottom of the board then treated with a silicone conformal coating. 
|
|
|
I was not keen on the way that the extension board was connected to the main board, so, I installed a 43way Dupont socket mounted in from the edge enough so that an edge connector could still be used if necessary, more on this later. |
|
This uses the same 'inductive' coupling method I previously rebuilt on my Nascom 2, back then I had to 'rebuild' a lot of the key internals as the steel wires used had corroded to dust. Unfortunately without dismantling every switch it is not possible to check for corrosion, etc. as they may still test ok but be very close to failure, it really depends on how the unit was stored, anything stored for 40 years in a garage or loft is pretty much guaranteed to be faulty.
|
|
|
So, keytops removed for cleaning, all key switches desoldered for checking, they all turned out to have minimal corrosion but were still passed through an ultrasonic bath, then they were all relubricated with silicone spray. |
|
|
Key switches were reinstalled along with the tested IC's. The bottom of the board then treated with a silicone conformal coating. A new keyboard cable was made up using ribbon cable. |
|
|
Time for a first powerup, without the keyboard to begin with, I had to manually force a RESET. with the original ROM fitted it reset to the prompt. I dumped this ROM and compared it with the stored images I had for all the variants, there were no matches but it behaved like the early NAS-SYS ROMs and was the closest match. I programmed 2 replacement 2708 ROMS with NASBUG4 and swapped the original out. And finally with the keyboard fitted. All seems well at this point. |
|
|
Started working towards the expansion board so made up a bridge board, the original main board connection used 43 pins which included a gap for the keyway and a +5v feed which I would not be using, the SMART-1 board originally had a 40 wire ribbon cable, however if you included the 2 physical gaps and the use of only one of the two ground leads this became 43 way. The SMART board also had to be shifted about 5cm towards the back of the case to avoid fowling the keyboard.
|
|
|
SMART-1 memory board The extension board was a mess, components hanging off the underside, wires off the top, etc. so the plan here was a complete strip down, reposition/solder some of the components, testing them where appropriate. |
|
|
Edge connector pins fitted first.
IC's all removed, cleaned and tested then reinstalled. I swapped the hard wired links out for DIP switches and then using the schematic I identified each IC number as this was not marked on the board. There was a shortage of decoupling capacitors in the logic area, these are not as important as the memory area ones but I added 4 in the appropriate areas. I didn't really want to drill the board but the +5v and ground rails were accessible topside. I also added the new power supply wiring. |
|
Additional notes:
There was a capacitor 221J (220pf) between IC12 pin 4 and ground, this is the MREQ line, on the scope this made no difference to the signal quality so was removed.
There was a capacitor 102k (1nf) between IC12 pin 9 and ground, this is the RD line, when fitted the capacitor caused -ve ringing on this line so was removed
These may have been necessary when the board was connected using the 'original' 6 inch ribbon cable but were not required when connected via the new edge connector.
There was a capacitor 221J (220pf) between IC12 pin 4 and ground, this is the MREQ line, on the scope this made no difference to the signal quality so was removed.
There was a capacitor 102k (1nf) between IC12 pin 9 and ground, this is the RD line, when fitted the capacitor caused -ve ringing on this line so was removed
These may have been necessary when the board was connected using the 'original' 6 inch ribbon cable but were not required when connected via the new edge connector.
|
I had to reverse engineer the add on board to identify where the loose wires came from.
I also suspect that the board never worked as the builder had omitted a ground connection to the 74LS154. Chips removed and tested. I wanted to replace this board with a new one without the wired connector. I used a long pin IC socket which could be plugged into the main board. The socket on the main board was not going to accept the slightly larger pins so I replaced it with a normal dual wipe one. I was also not a fan on the original board where one of the IC's was inverted. so the redesigned one would address this. Also, the unused gates on the 7404 were left floating, these were now tied to ground. |
|
|
SMART-1 board fitted and wired in, lets call it a 'snug' fit! Powered up and tested OK, did some simple memory tests first to make sure that all 4k boundaries were respected. Tested the cassette ports by saving and loading data, all OK. Put the top case cover on and left it on soak for 8 hours with no problems. |
|
|
I was in two minds as to whether to add LEDS and switches on the case, but in the end decided against it, back in the day this would have been quite normal as the 'owner' of the machine, however in this case I see myself as the custodian and would leave these mods for the next owner to make.