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.

Closed Thread
 
Thread Tools
Old 14th Jan 2023, 2:47 pm   #1
DavidMS
Pentode
 
Join Date: Nov 2022
Location: Chesham, Buckinghamshire, UK.
Posts: 135
Default SoC Mk14 Pico Uploader

I am currently working on a version of SiriusHardware's Mk14 uploader based around a Pico, which I plan to document in this thread.

In addition to being able to upload from a connected computer it will also work standalone for hex files previously copied to it.

I have a prototype working (see attached pic) and am now working on a basic PCB layout. However as I have a VI replica Mk14 and (in reality) am unlikely to ever own a real Mk14 have opted to solder a single in line connector rather than use the edge connector. That said I would like to make the PCB work for original Mk14s as well. So my question is can I get some advice on how close the final conductor for my board should be to the PCB edge to make alignment work.

Hopefully this make sense
Attached Thumbnails
Click image for larger version

Name:	image14.jpeg
Views:	81
Size:	142.9 KB
ID:	271468   Click image for larger version

Name:	image14a.jpeg
Views:	72
Size:	149.3 KB
ID:	271469   Click image for larger version

Name:	board edge.jpg
Views:	65
Size:	17.5 KB
ID:	271470  
DavidMS is offline  
Old 15th Jan 2023, 2:19 pm   #2
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
Default Re: SoC Mk14 Pico Uploader

Looking very promising. I don't know if your third image already implies it, but it would be best if any prospective PCB has the output to the MK14 on 'edge contact fingers' so that potential users can either use the original edge-type connector or any alternative type of 0.1" pitch connectors such as the SIL IDC connectors you appear to have used.

You can make the reset-out opto connections a direct match for the reset connections on the issue VI PCB and that may also be OK for the 'Oddy' replica if its reset pins are on the same pins as they are on the issue VI. For other original machines and replicas you would need to provide pads / connections for two flying leads with test connection clips on, this would be for machines which don't have reset-in available on the keypad edge connector - in those cases the reset opto outputs need to be connected to the leads of the reset capacitor, hence the flying clip leads.
SiriusHardware is offline  
Old 15th Jan 2023, 2:32 pm   #3
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
Default Re: SoC Mk14 Pico Uploader

As your setup is obviously going to have a user interface it might also be useful to have a way of varying some parameters without having to go into the python script itself to edit them. I'm thinking especially of the keypress delays which can need to be varied by the user for at least two reasons, one, they may be using either a 4.00Mhz /4.43MHz clock and may need to lower the upload speed for reliable loading, or, two, they may have an active VDU connected and that slows down the whole machine, in turn making it necessary for keypad injection uploaders like this to run more slowly otherwise keypresses will be missed.

Also, you might want to make the 'OS mode' (Old OS entry mode / new OS entry mode) a user interface selectable setting as well.
SiriusHardware is offline  
Old 15th Jan 2023, 7:23 pm   #4
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
Default Re: SoC Mk14 Pico Uploader

Another factor to consider: As you will have seen from the original project it turned out to be necessary to provide an alternative opto interface circuit diagram for the JM Precision replica PCB. It's a nice PCB and quite a few people have them but the layout of the connections to the keypad is different to those on all other original and replica PCBs.

To remind us of the difference, image #1 and image #2 show the main opto interface diagram, and the way in which it has to be rearranged for JMP replicas only.

It has been suggested in the past that the pi-uploader opto interface circuits be merged into a single interface with enough extra optos in place to serve either connection scheme, and a new 'Interface = "Normal" or "JMP"' switch setting added to the python script so that you could use the same hardware interface without alteration for either case. The script would then just drive one subset of optos or the other subset of optos depending on that setting.

My feeling at the time was that most people would only have one MK14 or replica and that they would just build whichever interface version their replica needed, but if you are going to commit your design to a PCB then that circuit will be fixed, as defined by the PCB layout. If designed to connect directly one-to-one to the edge connector on issue VI and nearly all other MK14s, then people who happen to have JMP replicas would not be able to use the interface directly.

They could still use an interface made for 'normal' machines but they would have to solder wires to the keypad output connector on your PCB and reroute them to the appropriate connection points for JMP on the way across.

Another way to approach this might be to have not one, but two 16-way output connections on the edge of your PCB, one with the output connections from the 13 optos arranged to match the standard keypad edge connection layout and the other to suit the JMP keypad edge connection layout.

Assuming your PCB is roughly square, put the 16 way connector for 'standard' MK14s on one of the four sides / edges of the PCB and the 16 way connector for 'JMP' MK14s on another side / edge (image #3).

This approach would support owners of JMP replicas without requiring any changes to the python script or to the interface hardware (other than routing the additional connections from the optocoupler outputs to the JMP edge connection output).
Attached Thumbnails
Click image for larger version

Name:	MK14_Pizero_Interface_V1.0.jpg
Views:	44
Size:	74.4 KB
ID:	271549   Click image for larger version

Name:	MK14_Pizero_Interface_JMP_V1.0.jpg
Views:	37
Size:	74.0 KB
ID:	271550   Click image for larger version

Name:	interface_suggestion.jpg
Views:	44
Size:	19.4 KB
ID:	271553  
SiriusHardware is offline  
Old 15th Jan 2023, 7:40 pm   #5
DavidMS
Pentode
 
Join Date: Nov 2022
Location: Chesham, Buckinghamshire, UK.
Posts: 135
Default Re: SoC Mk14 Pico Uploader

Quote:
Originally Posted by SiriusHardware View Post
Another factor to consider: As you will have seen from the original project it turned out to be necessary to provide an alternative opto interface circuit diagram for the JM Precision replica PCB. It's a nice PCB and quite a few people have them but the layout of the connections to the keypad is different to those on all other original and replica PCBs.
Thanks for the suggestions, where options can be catered for in s/w it should be relatively easy to accommodate. The level of flexibility h/w may be harder as there us very little room on the PCB already - I am limited as I use the Eagle freeware version and don't have time currently to get to grips with an alternative package as much as I know I should !

That said I will look at the 2 two interfaces and see if I can do something with jumpers.

The info I still need is where relative to the corner of the PCB I need to place the first pin of the edge connector to make physical interface to an original Mk14 as easy as possible - this would be the Row 0 pad. If the location of the edge connector on the VI replica is as per the original PCB - I can work the rest out ?
DavidMS is offline  
Old 15th Jan 2023, 8:35 pm   #6
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
Default Re: SoC Mk14 Pico Uploader

Well, it just so happens I have my issue II beside me here - I don't have one of those fancy digital micrometer things here but I make the distance from the CENTRE of the row-0 connector finger to the lower right hand corner of the PCB = 2.50mm.

The same measurement on my issue VI replica looks exactly the same, 2.50mm. I can't vouch for the JMP replica, Czech issue V replica or the 'Oddy' because I don't have any of those available to make measurements from.

If you are thinking of making the output connection link selectable for normal / JMP modes then I think there are a total of four 'column' optos whose emitters will need to be disconnected from their current destinations and redirected to different places on the edge connector for 'JMP mode'.

As well as providing pads for flying clip leads from the reset output optocoupler you will need to allow disconnection of the reset opto output from the top two edge connector fingers which on most MK14s are shorted together and connected to 0V, the exceptions being the Issue VI and the 'Oddy'.

Last edited by SiriusHardware; 15th Jan 2023 at 8:52 pm.
SiriusHardware is offline  
Old 15th Jan 2023, 9:01 pm   #7
Mark1960
Octode
 
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 1,264
Default Re: SoC Mk14 Pico Uploader

On my issue IV its slightly more than 2.50 mm. Just measured with a steel rule. I would guess 2.54mm as its most likely to have been laid out on a 0.1 inch grid.
Mark1960 is offline  
Old 15th Jan 2023, 9:14 pm   #8
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
Default Re: SoC Mk14 Pico Uploader

That seems reasonable, I guess we now need a casting vote for the distance on the issue VI.

I think the main thing to watch for is that other replicas -may- differ greatly from the 2.50mm / 2.54mm already stated, so it may be best not to depend on that being an exact dimension on all versions (for whatever reason).

Last edited by SiriusHardware; 15th Jan 2023 at 9:19 pm.
SiriusHardware is offline  
Old 15th Jan 2023, 9:22 pm   #9
Mark1960
Octode
 
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 1,264
Default Re: SoC Mk14 Pico Uploader

Quote:
Originally Posted by DavidMS View Post
Thanks for the suggestions, where options can be catered for in s/w it should be relatively easy to accommodate. The level of flexibility h/w may be harder as there us very little room on the PCB already - I am limited as I use the Eagle freeware version and don't have time currently to get to grips with an alternative package as much as I know I should !
I use Eagle 7.6, I have a modified script for the panelization step that makes all the characters on the silk screen the same size. If its any use to you I can forward a copy. I also have a python script that can be used to copy and merge pcb files with rotation and shift for panelizing smaller boards. I downloaded it from somewhere, I’ll see if I can find the link.

Searching for it again I found this one, which seems to run inside eagle.
https://maker.pro/pcb/tutorial/how-t...sing-v-grooves
Mark1960 is offline  
Old 15th Jan 2023, 11:42 pm   #10
ortek_service
Octode
 
ortek_service's Avatar
 
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,392
Default Re: SoC Mk14 Pico Uploader

I'll try and measure the ChrisOddy replica, tomorrow.

I do know that the Keyboard one on his was different to the (shorter distance) Exp.Bus distance. As I found that an old 'RS' green-colour SIL card edge connector, of the same ways as the Keyboard contacts, was about right if one of its supplied L-shape end-stop brackets was left slid in. Only having to remove the other end, as edge-connector pads area didn't protrude like it did on some other computers.
But would probably have to remove both / fit a filed-down one on the Exp. bus connector.
ortek_service is offline  
Old 16th Jan 2023, 9:20 am   #11
DavidMS
Pentode
 
Join Date: Nov 2022
Location: Chesham, Buckinghamshire, UK.
Posts: 135
Default Re: SoC Mk14 Pico Uploader

Quote:
Originally Posted by SiriusHardware View Post
If you are thinking of making the output connection link selectable for normal / JMP modes then I think there are a total of four 'column' optos whose emitters will need to be disconnected from their current destinations and redirected to different places on the edge connector for 'JMP mode'.
Quick question, on the JMP board do you know if pins 3 & 4 are just left 'open' i.e I can leave the col 6 & 7 connected as per the original - [as well jumping them to 6 & 7 obviously].
DavidMS is offline  
Old 16th Jan 2023, 11:51 am   #12
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
Default Re: SoC Mk14 Pico Uploader

Unfortunately I don't know, one or two people here have JMPs so hopefully they may be able to answer that question.

I can however tell you that on real MK14s and correctly wired clones of same, pins 6 and 8 of the keypad edge connector ARE connected to internal circuitry on the MK14 so there shouldn't be anything connected to those when the interface is in 'normal' mode. The 'NC' designation on those pins is a bit misleading, it should be taken to mean 'connect nothing'.

The best approach might be three solder pads, closely spaced in a straight line, with the opto emitters going to the centre pads and the outer pads connected to the two possible destinations for that opto emitter. To set the mode, you solder-blob the centre pad to one of the two outer pads (exactly as per the RAM select link on the underside of the issue VI).
SiriusHardware is offline  
Old 16th Jan 2023, 2:18 pm   #13
DavidMS
Pentode
 
Join Date: Nov 2022
Location: Chesham, Buckinghamshire, UK.
Posts: 135
Default Re: SoC Mk14 Pico Uploader

Quote:
Originally Posted by SiriusHardware View Post
Unfortunately I don't know, one or two people here have JMPs so hopefully they may be able to answer that question.

I can however tell you that on real MK14s and correctly wired clones of same, pins 6 and 8 of the keypad edge connector ARE connected to internal circuitry on the MK14 so there shouldn't be anything connected to those when the interface is in 'normal' mode. The 'NC' designation on those pins is a bit misleading, it should be taken to mean 'connect nothing'.

The best approach might be three solder pads, closely spaced in a straight line, with the opto emitters going to the centre pads and the outer pads connected to the two possible destinations for that opto emitter. To set the mode, you solder-blob the centre pad to one of the two outer pads (exactly as per the RAM select link on the underside of the issue VI).
Thanks that is fine will leave them isolated when not in 'use'
DavidMS is offline  
Old 16th Jan 2023, 9:04 pm   #14
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
Default Re: SoC Mk14 Pico Uploader

To also answer your earlier question re: JMP version connections, the attached crop from an online image of the PCB clearly shows that pins 3 and 4 of the keypad connector -are- connected to two pins of the 7445 column driver IC.

The same image also appears to show the tracks from edge connector pins 5, 6, 7, 8 coming to a sudden halt and going nowhere but this is only due to the poor quality of the image after being digitally 'magnified'. In reality those tracks will pass through vias at the points where they appear to stop and continue on their way to other pins of the 7445 on the underside of the PCB.
Attached Thumbnails
Click image for larger version

Name:	JMP_Keypad_Connections.jpg
Views:	33
Size:	56.8 KB
ID:	271646  
SiriusHardware is offline  
Old 16th Jan 2023, 10:58 pm   #15
ortek_service
Octode
 
ortek_service's Avatar
 
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,392
Default Re: SoC Mk14 Pico Uploader

Quote:
Originally Posted by ortek_service View Post
I'll try and measure the ChrisOddy replica, tomorrow.

I do know that the Keyboard one on his was different to the (shorter distance) Exp.Bus distance. As I found that an old 'RS' green-colour SIL card edge connector, of the same ways as the Keyboard contacts, was about right if one of its supplied L-shape end-stop brackets was left slid in. Only having to remove the other end, as edge-connector pads area didn't protrude like it did on some other computers.
But would probably have to remove both / fit a filed-down one on the Exp. bus connector.
I've now physically-measured one of Chris Oddy's MK14 replicas
- It may have been easier to just ask him to check the CAD files, but IIRC he based all measurements by scaling photos of other ones, as he never had access to an MK14 (original or otherwise).

I used my favourite easy-measuring using Digital Calipers - Invaluable for measurement things when designing PCB's and even the cheap < £10 ones seem as accurate as the original Mitutoyo ones I bought at work back in the 90's (just the cheap ones never really switch-off, drawing about same current when off (10x the off-current on Mitutoyo) as when, so battery only lasts 6months if left in.
And I tried not to measure at too-much of an angle, for best accuracy.

I measured 3.0mm from board edge to closest-edge of the first pad of the keyboard connector. And I measured his pads as approx 1.70mm wide (With the gaps between each pad being 2.54-1.70=0.84mm).
So the distance from board edge to centre of first pad will be 3.0 + (1.70/2)= 3.85mm

For the Expansion-bus connector, I measured 1.6mm from board edge to closest-edge of the first pad of the Exp. bus connector (which has the same-size approx 1.70mm wide pads (and so also 0.84mm gaps between each pad)
So the distance from board edge to centre of first pad on Exp. bus connector is 1.6 + (1.70/2)= 2.45mm


BTW, Chris has been wondering if anyone has ever seen a QWERTY keyboard for the MK14 / circuit details of one? It seems there are enough potential keys (40, as used on ZX80/81 & Spectrum) in total, on the 4 x 10 matrix. But the original keypad connections don't lend themselves to just wiring the 4 rows and 10 columns to a normal QWERTY / 1234567890 40-key keyboard layout. And unfortunately the 8x5 matrix Sinclair later used on the ZX80/81 & Spectrum, means you can't just connect one of those either.
- Not using the standard SCIOS at least. I presume ELBUG etc used the same Keypad layout as NS KitBug / Introkit Ref. designs, although it seems if you expand the memory too-much, with fuller-decoding, it breaks SCIOS that relies on memory wrapping-round.

Last edited by ortek_service; 16th Jan 2023 at 11:10 pm.
ortek_service is offline  
Old 17th Jan 2023, 12:10 pm   #16
DavidMS
Pentode
 
Join Date: Nov 2022
Location: Chesham, Buckinghamshire, UK.
Posts: 135
Default Re: SoC Mk14 Pico Uploader

This is basically what the corner of the PCB will look like (with some extra clues as to what each jumper does). Essentially right short for original and VI, left for JM Precision. Note the edge connector is duplicated on the underside


I have spent far too much time fiddling with the layout so once I have done a final check will send off for a few to be made.
Attached Thumbnails
Click image for larger version

Name:	Screenshot 2023-01-17 110206.jpg
Views:	35
Size:	45.7 KB
ID:	271685  
DavidMS is offline  
Old 17th Jan 2023, 12:39 pm   #17
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
Default Re: SoC Mk14 Pico Uploader

It's good that you have managed to squeeze in actual jumpers because there are one or two people who have a JMP and a 'normal' replica so they will be pleased you have made it so easy to swap between the two options.

I take it the square 'J3' group of pads allows configuration of the connections from the reset-out opto, ie, they can be linked to connect the reset opto directly to edge connector pins 1/2 for issue VI / Oddy, or flying test clip leads can be soldered to two of the four pads for boards which don't have reset-in available on the edge connector?
SiriusHardware is offline  
Old 17th Jan 2023, 12:57 pm   #18
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
Default Re: SoC Mk14 Pico Uploader

Owen, there was never an official 40-key keyboard circuit. We talked about establishing a 'standard' key matrix for one ourselves somewhere in one of these threads but really an alphanumeric keyboard would be of little use without a matching output device (ie, VDU) and then a new OS written to use both.

Something like this was actually in the MK14 pipeline but instead, the decision was made to develop an all-new machine with integral alphanumeric keyboard, VDU and high level language - the ZX80, in other words.
SiriusHardware is offline  
Old 17th Jan 2023, 4:07 pm   #19
DavidMS
Pentode
 
Join Date: Nov 2022
Location: Chesham, Buckinghamshire, UK.
Posts: 135
Default Re: SoC Mk14 Pico Uploader

Quote:
Originally Posted by SiriusHardware View Post
I take it the square 'J3' group of pads allows configuration of the connections from the reset-out opto, ie, they can be linked to connect the reset opto directly to edge connector pins 1/2 for issue VI / Oddy, or flying test clip leads can be soldered to two of the four pads for boards which don't have reset-in available on the edge connector?
Yes that was my assumption on what people would do
DavidMS is offline  
Old 17th Jan 2023, 4:30 pm   #20
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
Default Re: SoC Mk14 Pico Uploader

Looking very nice so far. What sort of optocouplers have you gone with?

I chose the individual surface-mount TLP185 optocouplers for the original project because we had them by the bucketload at work, but it makes sense to reduce the cost and component count by using multi-opto packages as you seem to have.
SiriusHardware is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 10:39 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 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.