UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Computers

Notices

Vintage Computers Any vintage computer systems, calculators, video games etc., but with an emphasis on 1980s and earlier equipment.

Reply
 
Thread Tools
Old 10th Nov 2020, 9:11 pm   #881
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,386
Default Re: Mk14 vdu

I can't get away with less than 156pF as far as I know (although I could try lowering the values back down in the direction of 100pF). At the moment I'm saying 156pF is the acceptable minimum, at least on my hardware.
SiriusHardware is offline   Reply With Quote
Old 11th Nov 2020, 4:49 pm   #882
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,386
Default Re: Mk14 vdu

A disappointingly small range of low value capacitors here - we only stock what we actually use in products. I've removed the 56pFs and stacked another four SMD 100pFs in parallel with the first set for a total of 200pF on each of A8-A11. I'll see what that does.
SiriusHardware is offline   Reply With Quote
Old 11th Nov 2020, 6:47 pm   #883
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,386
Default Re: Mk14 vdu

Well raising to 200pF from 156pF hasn't killed it or caused any adverse side effects so far. Running charset writing initially to 03xx and looking for corruption to 0Fxx and 0Bxx. I'll try the same with charset writing to every other valid 256-byte block of RAM so this is going to take quite a while.
SiriusHardware is offline   Reply With Quote
Old 11th Nov 2020, 8:46 pm   #884
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,386
Default Re: Mk14 vdu

With firmware FW692 and 200pF capacitors on A8-A11, I have left Charset running for some time, writing continually to each of these memory blocks.

02xx
03xx
04xx
05xx
06xx
07xx
0bxx

In each case the programme was left running for about 15 minutes. At no point were any unexpected writes seen.

I''ll probably now drop the capacitance down to the 'standard' value of 180pF and leave it at that, as I know that the low capacitance value threshold (below which problems appear) is somewhere between 100-150pF for my system.

Tim, if you get a chance can you also try yours with 180pF (or capacitors combined to make up that value) and then remove your RAM upgrade and make sure it still works with the 180pF capacitance?
SiriusHardware is offline   Reply With Quote
Old 11th Nov 2020, 8:49 pm   #885
Timbucus
Heptode
 
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 737
Default Re: Mk14 vdu

Will do having to stay with my Mum for a day or so but, should be able to dig up something in the range to make it I can solder the extras underneath temporarily as it seems to me that 100 on A11 and 47 on the others fixes mine both with and without the RAM

I wonder if someone fancies drawing a PCB with both RAM and OrionView on it - would be nice to have a properly built one... whistles
Timbucus is offline   Reply With Quote
Old 11th Nov 2020, 8:57 pm   #886
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,386
Default Re: Mk14 vdu

I don't know... it would have to be someone who's really good with Kicad... (...strokes beard...)

I think we should ideally not get into the realm of having different value capacitors on each pin, if only because it gravely offends my sense of symmetry. (I was the kind of child who, if I accidentally snapped a propeller blade off a model aeroplane, would snap the rest off to make it look uniform again).

If the circuit diagram has four identical capacitors on A8-A11, people won't question that. If we have one oddball value you can guarantee we'll get questions about that until the end of time, just as you would if you decided to make one pullup resistor of eight a different value to the rest.
SiriusHardware is offline   Reply With Quote
Old 11th Nov 2020, 9:14 pm   #887
Mark1960
Pentode
 
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 202
Default Re: Mk14 vdu

Quote:
Originally Posted by Timbucus View Post
I wonder if someone fancies drawing a PCB with both RAM and OrionView on it - would be nice to have a properly built one... whistles
I was thinking about making a board that could connect to a VI with eprom and ram, possibly with links to select 256 byte holes for components still fitted on the Main board.

Maybe OrtonView could fit on the same board.
Mark1960 is offline   Reply With Quote
Old 11th Nov 2020, 9:23 pm   #888
Mark1960
Pentode
 
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 202
Default Re: Mk14 vdu

Quote:
Originally Posted by SiriusHardware View Post
I don't know... it would have to be someone who's really good with Kicad... (...strokes beard...)
I would use Eagle, i tried kicad a couple of times but didnít like the workflow between schematic and layout.

Quote:
Originally Posted by SiriusHardware View Post
If we have one oddball value you can guarantee we'll get questions about that until the end of time, just as you would if you decided to make one pullup resistor of eight a different value to the rest.
Just think of the opportunities to troll the newbies
Mark1960 is offline   Reply With Quote
Old 11th Nov 2020, 9:44 pm   #889
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,386
Default Re: Mk14 vdu

If you're thinking of making an offboard EPROM / RAM PCB to plug into the issue VI then maybe the really logical thing to do would be to design an issue VI rev 2.0, which would do away with the original onboard 4-bit memories and replace them with 8-bit wide EPROM and SRAM on the main board.

However, don't let us curb your enthusiasm, so by all means make your proposed extension complete with OrtonView. If you do, maybe provide for a dedicated VDU control port / TOP PAGE routing scheme for the VDU and break the 8154 port pins out to a pin row header so they can be used for general I/O as originally intended.
SiriusHardware is offline   Reply With Quote
Old 11th Nov 2020, 11:18 pm   #890
Slothie
Heptode
 
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 666
Default Re: Mk14 vdu

I might be paranoid but I sense a certain amount of (un)subtle hinting here

Well I've been thinking about this, in relation to my memory-expansion-buffer board and I was getting to the position that I would just bring all the INS8154 pins out to a standard 0.1" header, and next to the VDU card connector have another 0.1" header connected to P1-4, TOP, SWAP, INVERSE etc and just use short F-F "dupont" jumper wires to connect the pins as required. With a few spare 0v pins to ground out any connections that would cater for any configuration requirements, and if you wanted to you could make up a connector with a toggle switch on to switch between "favorite"memory configurations (there being probably only 2 you'd want). You could even have pads to solder the toggle switch to the board and pads to connect header pins etc.
As for the expansion memory I'd probably just map a 6116 to the 1,5k hole in lower memory for simplicity. If you wanted to fill the upper memory to replace the need for 2111 memory chips, then using another 6116 with appropriate decoding would be simple enough - Although there is technically enough memory unused (512 bytes) in the first 6116 remapping the address lines to move the 0x000-0x1FF block to 0xFxx and 0Bxx is complicated (although not impossible) but might be possible using a PROM or GAL. I was going to look into this when I get access to my MK14 stuff but thats not happened yet...
Slothie is offline   Reply With Quote
Old 12th Nov 2020, 1:23 am   #891
Mark1960
Pentode
 
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 202
Default Re: Mk14 vdu

Its difficult to decide how far to go in extending the MK14.

I though putting the RAM and EPROM on an extension board would leave it open for me to revert to the standard MK14 if I do ever source a spare set of PROMs.

Another option is to have only RAM, but battery backup and some method of loading the RAM, maybe using an ft245 similar to a previous z80 project.

Adding memory could also include a latch for the high nibble, but then I donít think there is an easy way to enable the onboard IO and memory only in the bottom 4K, and putting holes for the onboard IO in every 4K page would make it difficult to run NIBL which would be the main purpose of adding so much extra memory. It seems to me than an extension to the MK14 should stick to the 4K limit with a separate project to use NIBL.
Mark1960 is offline   Reply With Quote
Old 12th Nov 2020, 2:02 am   #892
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,386
Default Re: Mk14 vdu

Joking aside, little steps before big steps. I think an extension consisting of 1.5K memory + OrtonView would be quite nice with 8154 ports, VDU control lines and some 0V pulldowns and +5 pullups all available on pluggable headers - that's virtually what I have now, scattered across two untidy bits of stripboard.

Mark, there are a couple of sources over here, one in the Czech republic and one here in the UK (JM Precision) which sell programmed MK14 PROMs, the question being whether they can ship them to you over there and how much would that cost. The JMP ones are currently about 17 GBP per programmed set. I don't imagine it would hurt to enquire about carriage to Canada. RAM I think you can probably find or maybe already have, and I don't think you would have much difficulty finding the 40-pinners over there.

I don't know if you know the Vintage Computer Forums (search VCFED) - mainly USA-centric, but if you put a 'help wanted' post on there you'll almost certainly find someone, possibly even in Canada, who can programme a set of PROMs for you as long as you can provide the code (which I can provide as two .bin files).

I'd be happy to programme them for you of course, as long as you didn't mind them making a very long round trip. (Slothie's original PROMs went on a European road trip to get programmed, as they turned out to be the difficult-to-get-programmed Tesla type).
SiriusHardware is offline   Reply With Quote
Old 12th Nov 2020, 5:51 pm   #893
Mark1960
Pentode
 
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 202
Default Re: Mk14 vdu

Iíll look into getting a set of JMP proms. The local shop had 82s131 at $10 each but listed as refurbished, so a good chance they are already programmed with something. They also had 82s130 with open collector, which could probably be made to work with pull ups, but I might take a couple of tries to program and the effort of building a programmer.

I have RAM, 8154D, 8060N, and TTL so apart from the PROMs just need to order a set of switches and rig up a display until I can retrieve those from my Mumís.
Mark1960 is offline   Reply With Quote
Old 12th Nov 2020, 6:02 pm   #894
Mark1960
Pentode
 
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 202
Default Re: Mk14 vdu

Quote:
Originally Posted by SiriusHardware View Post
I can't get away with less than 156pF as far as I know (although I could try lowering the values back down in the direction of 100pF). At the moment I'm saying 156pF is the acceptable minimum, at least on my hardware.
An alternate for capacitors that might be a bit more healthy for the devices driving the bus might be to add a 74ls08 on the OrtonView. NENIN from PIC on one input, and NWDS from MK14 on second input, then the output to NENIN on the MK14. This would always allow an NWDS cycle to complete, though possibly not complete enough to avoid retry after release of NENIN. It would delay SCMP release of the bus by as much as 500ns, but I think Karen gave enough margin to allow that.
Mark1960 is offline   Reply With Quote
Old 12th Nov 2020, 6:12 pm   #895
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,386
Default Re: Mk14 vdu

Quote:
Originally Posted by Mark1960 View Post
The local shop had 82s131 at $10 each but listed as refurbished, so a good chance they are already programmed with something.
There's just no good way to spin that is there? Either they are new and unused, or already programmed and useless since they can't be reprogrammed.

I had a bad experience with an established little UK seller - not one of the big mainstream suppliers but familiar enough to people on this website - when trying to buy PROMs for another purpose. They sent out three labelled, therefore obviously already programmed 82S123 PROMs and when a complaint was made about that, sent out three more of the batch with the labels (but not the glue) removed and still the same code programmed in them.

That's why buying them already programmed with the MK14 OS is generally less risky than taking pot luck with ones found in the usual places and hoping they will be blank, and then of course you have to programme them somehow.

I would be happy to regard JMP as a trusted seller despite not having used them myself, a number of people including Tim have bought their 'issue 0' MK14 replica PCBs and have been very happy with them, the only drawback being that they don't have the bus connections to the rear edge which the issue VI does, nor do they have the memory hole at 0200-07FF, that only happened on the final issue of the original MK14, issue V, and therefore also on the Czech 'Martin L' clone which is a near exact copy of that PCB.

I actually have 4 * blank DM74S571 here but I won't know if they are OK until I try to programme them.
SiriusHardware is offline   Reply With Quote
Old 12th Nov 2020, 6:34 pm   #896
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,386
Default Re: Mk14 vdu

Regarding capacitors, let's assume that Mark's hypothesis is correct and that the root cause of the corruption problem lies in the difference in the way the SOC VDU and OrtonView load the high address bus when they are not driving it.

SOC VDU uses proper tristate buffers to gate itself on and off the bus, whereas OrtonView just puts its address driving pins into input mode, so the theory is that when NENIN strikes in the middle of an SC/MP NWDS pulse, the state of the high address lines changes state during the NWDS pulse causing at least one other address to be written to.

The capacitors are fixing this by slowing down the rate of change of state of the un-driven high address lines so that they remain in the same states for long enough for the NWDS pulse to end, but deliberately rounding off digital signals to achieve an end is not really an ideal fix.

I'm wondering if we can take out the capacitors and replace them with series resistors in the A8-A11 lines between the OrtonView and the MK14. The PIC pins have very strong pull up and pull down action when in output mode (around 50R from output to 0V or +5V, I think Karen said) so they drive from rail to rail, therefore should not have any trouble driving the system address lines into valid logic states even with resistors in series. When the PIC pins are in input mode those same resistors may lessen the 'pulling' effect of the PIC pins on the high address lines.

Incidentally the idea for using a 74LS08 with NWDS and NENIN - only one problem with that, historically speaking the NWDS signal is not available to the VDU although in practice there is no reason why it can not be.
SiriusHardware is offline   Reply With Quote
Old 12th Nov 2020, 7:39 pm   #897
Timbucus
Heptode
 
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 737
Default Re: Mk14 vdu

I think the 74LS08 would be only for the OrtonView so would live on that card and require an extra wire to the standard wiring of the SOC VDU - but, as you say original owners who have retained the SOC extra wiring on the Issue 4 and 5 might be wary of even adding that extra wire - considering they are likely to pay close to £600 for a replacement unit at current market prices and the fact that changes to what is now a historical artifact may be considered unethical unless they are only what was already part of the history of the device like the guys here are doing restoring their setups.
Timbucus is offline   Reply With Quote
Old 12th Nov 2020, 7:45 pm   #898
Timbucus
Heptode
 
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 737
Default Re: Mk14 vdu

As regards JMP that is Jerry Walker (I think he is a member on this board) and I can vouch for the reliability and quality of the products I am sure his PROMS will be tested. You only have to watch his dedication to restoration on the youtube channel to see that - watch the latest series on the MADAS calculator and how he has to fix the motor- https://www.youtube.com/watch?v=HkWtbaMCPnM

Do watch out though anyone using the MK14 board as it has a different KB connector layout if using the PI/Arduino programmers elsewhere, they need a tweak in software and some extra opto isolators if you want to use them on both boards interchangeably.
Timbucus is offline   Reply With Quote
Old 12th Nov 2020, 7:48 pm   #899
Timbucus
Heptode
 
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 737
Default Re: Mk14 vdu

Finally on the interface lark I really think a 1.5K RAM and Orton View PCB with the headers for the IO ports and Flags - so the Audio, PROM programmer and Cassette interfaces can be easily hooked up is a great idea.

We have discussed in other threads that the >4K RAM should be on a board on the CPU socket following BITD cannon suggestions for adding NIBL to the machine - probably in the memory expansion thread,
Timbucus is offline   Reply With Quote
Old 12th Nov 2020, 11:06 pm   #900
Slothie
Heptode
 
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 666
Default Re: Mk14 vdu

Quote:
Originally Posted by Timbucus View Post
Finally on the interface lark I really think a 1.5K RAM and Orton View PCB with the headers for the IO ports and Flags - so the Audio, PROM programmer and Cassette interfaces can be easily hooked up is a great idea.

We have discussed in other threads that the >4K RAM should be on a board on the CPU socket following BITD cannon suggestions for adding NIBL to the machine - probably in the memory expansion thread,
Yes I think I posted a rough draft design...
Slothie is offline   Reply With Quote
Reply

Thread Tools



All times are GMT. The time now is 7:57 am.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2021, vBulletin Solutions, Inc.
Copyright ©2002 - 2020, Paul Stenning.