HEKTOR 1A Restoration

The HEKTOR was a microprocessor training system used by the Open University in the 1980's as the practical element of a few of their courses, TM222, PT502, based on the 8085 processor and programmed in assembler it was 'loaned' to the students for the duration of the course to be returned at the end. I suspect that most were destroyed when the courses ended their run though I have heard that the last students to use them were offered a purchase option. I actually did TM222 in 1983 but of course had to return Hektor.
Specifications: Made by the Open University, UK in collaboration with C.C.Consultants. from 1983-1991
Built in Assembler, Monitor running an 8085 MPU at 3.024MHz with 4K RAM and 8K ROM
Displaying 16 lines of 64 characters monochrome and also having TV/RF output, Cassette port, Serial Port and a 40 pin connector for connection to the Peripheral Board and requiring a 9v AC power supply.
I have now acquired a few of these boards so what follows will be the restoration of one (or more) of them.
Note: the story will be shown in reverse date order so the most recent post will be at the top.
27/06/2022 Update
Picture
So here we are, all finished.
The only thing I forgot to mention was that the 2716 as IC5 was corrupted in places causing some characters to appear strange. As my programmer doesn't support 21v Vpp I used a AT28C16 EEPROM instead. I also mounted a heatsink on the Bridge Rectifier as I expect the current drain to increase when I add the Peripheral Board on (a later project). And lastly, after cleaning the back of the board I gave it several coats of Acrylic Conformal Coating.
Next on the agenda will be a case of some sort, I sense a Lego build coming on :)
26/06/2022 Update
OK, a lot has happened since the last time...
I have 3 Hektor units, all needed work, so I decided to work on all 3 at the same time, making sure that if I carried out a mod on one I did it to them all, keeping them at the same level of restoration throughout. After getting all 3 working at a basic level, mostly this was just poor connections on the 40 year old sockets, either a De-Oxit treatment or changing the socket cured this.
HDMI
​
​The first mod was to add HDMI to the board. Bolted to the top of the UHF modulator case.

The UHF modulators fitted to these boards are next to useless on modern LCD TV's. In all 3 cases it was extremely difficult to get a picture to lock, however, again in all 3 boards the picture on an old CRT TV was OK.

The HDMI board simply takes Video, +5v and ground, the Hektor boards do not use sound. The HDMI board also allowed switching between 720p and 1080p.

​
Essentially the board was complete, however, when I measured the current drain from the power supply it was just over 1.1A, as both the regulator and the Bridge Rectifier are rated at 1A I felt that this was stressing them unnecessarily, this was obvious as both were getting really hot.
The first and easiest fix was to replace the Bridge Rectifier with a higher rated on... 4A capacity, now gets warm but not hot.

I was also interested in replacing the old voltage regulator with a high efficiency one, by Traco.

This was rated at 2A but in actual fact a lower one would have been sufficient.

Once installed the current drain from the power supply dropped to 0.46A, showing just how inefficient the old regulator was.

Now both the Regulator and the Bridge Rectifier run cool.
​
24/05/2022 Interim report
  • Retrobriting the keycaps revealed that the smoke (tar) damage was worse than expected, it removed all the residue but left the keys in strange shades of grey, may try another run when I get the board working.
  • 3 of the 5 RAM chips had broken legs, the remaining legs were also fragile, corrosion had eaten through the pins. All chips were replaced.
  • 74LS373 had a broken pin, IC replaced.
  • Both ROMs were corrupted (invalid checksums), new ROMs burned.
  • Al the keyboard contacts have been cleaned with De-Oxit, and tested.
  • Extra decoupling capacitors were fitted, I am not a fan when components are omitted to save money.
  • All tantalum capacitors changed, c2,c4,c5,c6,c7 with a few of them having really high ESR reading.
  • Electrolytic C21 changed.

At this point it now boots to a blank screen, true fault finding now starts!

Picture
18/05/2022 Initial look at the board.
First job was to power the board up, everything looked to be in place but just to be safe I powered it from my bench supply set to 10v 1.1A current limit, as expected there was no video output. At this voltage it actually drew1A which is normal for this board. The 5v rail was 5.02v so this is OK.
​The first observation was that IC13 (8085) was getting hotter than expected, so it was powered down.
Time for a closer look
Visual inspection only: (updates in red)
  • The bridge rectifier had been replaced at some point with a single diode, now while I agree this may well work it would considerably increase the power supply ripple, so I plan to replace this with a proper 200v @ 6A Bridge Rectifier. (now replaced)
  • As this is going to be a full restoration it may seem that some jobs are done out of order, however, this IS going to work when finished and sometimes it is much better to work on a clean board than one covered in dirt and solder flux. The board already showed signs of heavy rework and it would now be very difficult to differentiate between 'virgin' faults and those introduced by a previous 'fixer'!
  • The video socket was broken, but had been patched out with wire, needs to be better than this. (socket modified with a much stronger connection)
  • I removed all the socketed IC's to allow for cleaning, all the pins were badly tarnished so would need a clean and de-oxit. (done)
  • All keycaps were removed, again to help with the cleaning. Not sure what the significance is of the different coloured pad tops, but they may as well go back like this. The keys will be soaked in warm soapy water in the first instance to see how clean they become. (Although clean they are a brown/grey and the writing is not white, looks like smoke damage, a first test of retrobrighting seems to make a big improvement)
  • IC5 pin 3 (chr ROM) had a wire to IC10 pin 6 (74LS374), IC10 socket had been changed so presumably the pad or track was damaged in the process, (I checked the circuit diagram and this link is valid, and it was tested electrically for shorts/continuity)
  • IC13 (8085)has had the socket changed, this needs investigating to check for shorts and continuity onwards on the board, particularly as this IC gets hot! (socket tests ok)
  • SKT0 (ROM0) has had the socket changed, this needs investigating to check for shorts and continuity onwards on the board. It also has a wire patch (ROM0) pin 9 to SKT1 (ROM1) pin 9, my guess is that the trace was damaged when the socket was replaced. (all tested ok)
  • IC6 (74LS165) has been replaced at some point but it wasn't socketed so this needs more investigation, pin 6 does not have a solder pad on the PCB side, the circuit diagram shows this as going toIC5 pin 17, needs checking, along with the standard check for shorts / continuity, the soldering on this IC is terrible so I may decide to socket it anyway. (all tested ok)
  • The cassette relay has some unusual straps, needs checking.
  • I removed the LAS1605 regulator as I like to re-paste the heat sink compound. (Done)
  • I also removed: UHF modulator, Video socket (Video socket repaired and replaced)
Time for a good clean now.