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 12th Dec 2020, 8:37 pm   #61
ortek_service
Octode
 
ortek_service's Avatar
 
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,394
Default Re: Softy 1

An alternative is to try to pick-up a used original 48Pro(+) - The ones I've got access to are now > 10years old and still work OK on modern O/S's, plus there's still regular updates to the (same) free download control software.
However, they seem to want to make their money on protected adaptors, for newer devices - even ones with < 48pins that are just a socket converter
(With not 1:1 mapping on some newer claimed 'high-speed' memory ones)

I doubt there will be many more older devices added in future (Plus not sure Tesla actually still make PROM's, and have heard no-one makes EPROM's anymore, so NOS/used prices could go up)

I did in the past get Dataman to add a couple of new devices to the software for 'free', after agreeing for my works to buy a few of the required adaptors.
And they also fixed bugs I found promptly, giving me beta versions via ELNEC (who produce much more regular updates and it will run on Dataman Pro if you use their drivers, as essentially just has the names changed on the s/w)


I currently use at home the old (Advantech-designed) Dataman 48XP (Parallel Port) one, under XP (And also the original 48 / 48LV ones, that won't run beyond Win9x / Win2k).
But support for these was never that good (we gave up trying to get Advantech support to fix bugs with an existing device, despite buying a specific adaptor from them to use it). Plus had to pay £100's for replacement upgrade boards for LV, then to XP to work with that.

The last Win-XP Only infrequently-updated control s/w for XP/UXP models is now frozen, not being updated to work on Win 7 etc. as they've now made you buy software licences / credits for the new version software that does work on latest O/S, but only supports the 48UXP (Surprisingly still being sold, when Pro-series gives much better support at same / cheaper lower-spec ones). So not worth paying the £100's price to upgrade my 48XP to a 48UXP.
They did at least originally give you a free list of the wiring for many adaptors - including uC ones - for the original 48 model. But more recent Pro models won't work with many older adapters. And if you get a Pro2, it seems it often requires a whole new series and can't use ones from 48Pro.

The original free 'lifetime' s/w updates was an attractive feature of the Dataman's etc - compared to expensive pay per device Data-I/O models.
However, I'm thinking that maybe an Open-source (like with ECAD software) one could be the way to go to supporting old devices, But not seen one yet, Although I discovered the cheap TL866 apparently has open-source Linux s/w (Although firmware in it may not be). I did design my own 2716-27512 one for the Beeb computer > 30yrs ago. But the programming details for many modern devices only seem to be shared with major programmer manufacturers.
There has been an open-source EPROM-Emulator, recently developed:
https://groups.io/g/eprom-emu-ng
- Had found one of few ones still sold, didn't seem to work too well.


Back a bit more back to previous discussions, etc, I did just find that the contents of ETI Oct' 78 issue & Computing Today archive had been discussed on another thread on here last year:
https://www.vintage-radio.net/forum/...d.php?t=156143
ortek_service is offline  
Old 13th Dec 2020, 4:33 pm   #62
ortek_service
Octode
 
ortek_service's Avatar
 
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,394
Default Re: Softy 1

I've just been looking at the Softy-I instructions again, and noted that although it can read (say Bootstrap?) from a 2716, it warns it can't program these single-rail devices.

I presumed the Softy-II - which according to info in my oldest 1985 Maplin catalogue (that unfortunately I only have the last page on this still in there) was like the Softy-I also using the INS8060 / could also be used as a stand-alone control computer - could do both.
But it seems there's even less on the 'net about the Softy-II than the Softy-I - I have found it (along with the MicroDoctor I have) in an old Technomatic advert advert here: https://worldradiohistory.com/hd2/ID...-Page-0044.pdf wjere it says it works with most +5V EPROM's. So a bit unclear whether the 2708 was dropped on this (which definitely had on the rather different S3).
Edit: I've just noticed that Maplin also sold (at £27.95 in '85) a 24pin to 28pin expander adaptor for the Softy-II that allowed reading & programming of the larger 2732A?, 2764, 27128 etc. ones, via a rotary switch that selected which 2K segment of these larger EPROMS was used with the Softy-II's 2K of RAM. So it looks like the Softy-II only support most +5V EPROM's with this extra adaptor, and may have only originally supported the 2758 / 2716 ones (Maybe 2732 24pin with some switching of 2K segment)



I've also just found further references / circuitry description (maybe copy of what's already in the instructions) to the original Softy / some mods to it, are discussed here: https://researchportal.bath.ac.uk/fi...rth_thesis.pdf



It was initially surprising to me that if the Softy-I has the capability of removing the extra +12V -5V supplies, then I thought it probably wouldn't have needed much extra to handle a different Vpp pin. However, from just reading a (NTE)2708 datasheet: https://www.nteinc.com/specs/2700to2799/pdf/nte2708.pdf
it seems programming the 2708 is rather different to 27xx & (25xx) ones, in that on the 2708 you have to put +12V on nCS (PE) pin then apply a +25V 0.1-1ms pulse to program pin, before moving onto next address. Then have to repeated all of this until total pulses length at each address exceeds 100ms.
Compared to 2716's constant +25V on Vpp pin & single 50mS High pulse on nCE/PGM pin.



So can only copy the Softy-I's 2708 EPROM to another and therefore still need another programmer to read it on a PC etc. With very few recent PC-programmers supporting the 2708 - But I have noticed someone recently designed an adaptor for an Arduino to handle 2708's and similar ones I've just read were used on the original IBM PC:
http://www.mattmillman.com/projects/...m68766-eproms/


Alternatively, for just reading 2708's, from this etc., it's probably simplest to just make your own adapter, to add the extra +12V & -5V supplies (but may also want to provide +5V as well, externally, to make sure all 3 supplies switch-on at once to prevent possible damage to 2708 from missing supplies).
And should be able to implement this by just stacking some IC sockets and isolating the relevant pins, as at least the 2708 has the JEDEC standard Data / Address pin-out of the 27xx series etc.

Last edited by ortek_service; 13th Dec 2020 at 4:48 pm. Reason: Extra info on Softy-II Adaptor
ortek_service is offline  
Old 13th Dec 2020, 4:48 pm   #63
TonyDuell
Dekatron
 
Join Date: Jun 2015
Location: Biggin Hill, London, UK.
Posts: 5,191
Default Re: Softy 1

According to the Softy (1) user manual, the '2716' that can be programmed is the TI 3-rail one. It warns not to attempt to program single-rail 2716's such as the Intel ones.

There's a link you have to change to select between 2708 and 2716 programming. In the latter case you program the chip in 2 halves, there's a pin on the PCB that you have to connect to either +5V or ground to select which half will be programmed. So it really is a bit of a kludge.

I feel the Softy 1 is best regarded as a 2708 (only) programmer/emulator.
TonyDuell is online now  
Old 13th Dec 2020, 5:45 pm   #64
ortek_service
Octode
 
ortek_service's Avatar
 
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,394
Default Re: Softy 1

Thanks - I hadn't realised there was a TMS2716 ? that was a 3-rail device
- just that TI only caused confusion by having 27xx devices with different pin-out to others. So had to be careful with TMS ones and led to 25xx ones like the 2532 etc. to distinguish these.
(which I presume the Softy-I didn't support at all, even with link changes, so had to make your own pin adaptors, as I've done for reading out ROM's etc. with these pinouts)

Needing adaptors for more recent devices is a common theme for many programmers, but if the devices are already around it does seem a bit of a kludge - like 24 to 28way adaptor for the S2.

A 28 to 32w one I believe was available for the S3 may be a bit more forgiving, as 32pin EPRON's weren't nearly as common as 24/28pin ones.
ortek_service is offline  
Old 14th Dec 2020, 12:39 pm   #65
ortek_service
Octode
 
ortek_service's Avatar
 
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,394
Default Re: Softy 1

I looked at the TMS2716 datasheet etc, and can now see why it was said they should be avoided as were incompatible , but I had probably not originally / forgotten the 3-rail supply issue - Which probably led to them not being very widely sold / used (Maybe only for upgrade of existing 2708's).


I also was having a closer look at some of the old Softy adverts at bygonebytes. And noticed that this one doesn't say anything about the devices it supports: http://www.bygonebytes.co.uk/images/Softy_3.jpg

Whilst this one says 'Develop, Copy, Burn, Verify 2708, 2716 and with modifications' http://www.bygonebytes.co.uk/images/Softy_2.jpg
- which doesn't quite seem very clear. And certainly doesn't mention 2716 limitations.
It also said 'For product development with less than 2k of firmware Softy may be the only development tool you will need'
- but if it only had 1KB of RAM that could be rather limiting - particularly if trying to use emulation
(+Not sure 'Now on show in the Science Museum' makes clear it is newish!)


And this later 1983 Dataman ad: http://www.bygonebytes.co.uk/images/Softy_02.jpg
has some interesting information in it

Although odd they show a picture of the 'Softy-II', but just call it the Softy, saying it 'Emulates & Programs 2716, 2516, 2732, 2532 EPROM's'
- So no mention of 3-rail devices like 2708 / TMS2716 and looks like they did drop support for these, despite it also featuring the same INS8060 processor.

And then advertise the 'Softy 1 with PSU', that they still had few left to clear, which they describe as the old faitful 3 rail EPROM Programmers - as seen in Science Museum. If you are still using 2708's get yourself this fine old classic'


Also featured is the rather rare Microdoctor (Only one I've ever seen, is one I bought for a few pounds, not knowing anything about it at first). And the Menta-Z80 Development System - I'd never heard of until I stumbled across a review of it a few days ago, whilst looking for Softy info. It's reduced price wasn't too bad - rather cheaper than the Softy (II) and its INS8060 SC/MP processor address range issues.

And maybe why they eventually moved onto a (CMOS) Z80 (along with built-in LCD) in the S3, advertised: http://www.bygonebytes.co.uk/images/Softy3.jpg
- Where they say it can look like any PROM upto 64K: 25 or 27 Series, but don't say you shouldn't connect emulation cable direct to a 2708 / TMS276 socket (I doubt +12V / -5V would be too good for it!).

Most interesting is that they say they have no secrets and you can change all the programming parameters. Plus were preparing a Developers' Manual which will give source-code, BIOS calls, circuit diagrams etc.
An early example of Open-source, if that was the case, and rather different to their latest programmers with protected adapters and licenses for some devices (I recall you could actually adjust many programming voltages / timings & verify voltages on the original 48).

A truly open-source programmer would be rather handy, in preventing obsolescence but can't see any company making much money out of one.
ortek_service is offline  
Old 16th Dec 2020, 2:19 pm   #66
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,485
Default Re: Softy 1

The main problem with trying to make an open-source programmer is that device manufacturers often did not share their programming algorithms openly and only disclosed them to 'partner' manufacturers of device programmers. I don't think there was anything sinister about that - after all, it wasn't in the manufacturers' interests to make their devices difficult to program. The manufacturers probably just wanted to protect the reputation of their silicon parts by ensuring that they were only programmed with 'authorised' programmers which had been checked and certified by the device manufacturers.

In the particular case of Tesla devices, they have their origins behind what was the old iron curtain so they weren't widely known about during the 'golden period' for device programmers, therefore few (if any) 'Western' made programmers supported them. I think the Elnecs were designed in Eastern Europe, which is probably why they (and rebadged versions of their programmers) are one of the few models which properly support the Tesla family devices.
SiriusHardware is online now  
Old 20th Dec 2020, 2:28 am   #67
ortek_service
Octode
 
ortek_service's Avatar
 
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,394
Default Re: Softy 1

Yes, lack of published programming info. can be a problem, although I think that is probably more of a problem for more modern uC's etc. which are often much more involved with multiple-memory areas and protection bits etc. uC's in the past often used to have EPROM-Emulation modes / adaptors-to-EPROM pinout available, so that you could use a standard programmer (but you still needed a special algorithm etc.) and often gave you the circuitry details in the datasheet (but sometimes strangely removed this in later datasheets)

Luckily, for the more vintage EPROM's etc. (that we'd generally want a programmer to support due to many modern ones not doing these), the datasheets were generally quite good at describing the recommended programming method. When I designed my own for 2716-27512's I just read other EPROM Programmaer magazine project details / used info in Maplin catalogue / bought photocopied datasheets from them for these
- This was as a student back in 1988, when everything was generally printed content only, and manufacturer's databooks were harder to obtain by those not working in the Industry.

Device ID's, used by later IC's / more-advanced programmers may have been a bit more difficult to obtain, but not essential and can find this out more easily these days.
I was looking through the manual of my Dataman S(ofty)3 earlier, and noticed how much info it had for these, with all the parameters listed and even explained how to enter your own custom algorithms for other devices etc. Apparently the S3 was £570+VAT, back in 1989, so quite an increase over the S1 / S2 and in relative terms quite a bit more than modern PC-controlled ones - It seems that after the S4 (Possibly short-live S5 one, I once found a reference to on RS), no-one made portable hand-held programmers / EPROM-Emulators, so used S4 are still sought after.

And it seems that after these 'Dataman Designs' ceased doing their own products and re-badged ones from Advantech (the original/LV/(U)XP-48) then Elnec (4x-Pro(+/2)), and who (despite being based in Slovakia) were only formed in 1991, according to: https://en.wikipedia.org/wiki/Elnec
So not sure how long the E.European Tesla PROM's were made for, as I doubt there were many designing with these by then. But may have had large stocks from previous years.
(Strangely, I recently found that N.O.S. Oric ULA's can still be bought very cheap from Bulgaria, even though they were produced by a USA company - via UK sales office near Tangerine - that I doubt had them made in E.Europe)

Luckily, PROM's are fairly basic and so aren't too difficult to program / get info on how to do this (Just need to know the right polarity/ voltage & times to blow the fuses). I found the Dataman 48-Pro+ seemed to struggle with some expensive UK-supplied N.O.S. National ones, taking many retries at several locations, with it having to keep starting again. But with some slightly different N.O.S. Tesla ones, it went straight through first time OK.
ortek_service is offline  
Old 20th Dec 2020, 11:58 am   #68
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,485
Default Re: Softy 1

Just to be clear, you're saying that the Dataman 48Pro+ natively supports the Tesla MH74S571? If so, which other Tesla MH series family BPROM devices does it support?

I have here an ICE TECH Micromaster 1000E which has the capability to emulate EPROMs (that's what the 'E' suffix stands for).

I also have a low cost serial-to-EPROM emulator which was made by a company called 'JPD', it is entirely terminal driven so it can be used with anything which has a serial port (e.g, Raspberry Pi) even now. It looks rather like a typical homebrew build of a magazine project - professionally made PCB, but fitted in a standard and rather oversized ABS project box. That served me very well in the days when the only alternative was to erase and reprogramme EPROMs.

Getting back to the Softy, I could have done with access to one in the late seventies because I had a design idea for a project which needed an EPROM to function as a custom decoder and I went so far as to buy a 2708 thinking that I could lash up a programmer for it using the MK14, the only 'computer' I had at the time, as the 'smart' bit. Then I discovered the peculiar three-rail requirements of the 2708 and it was not practical to have the necessary minus supply on the target system. I had to wait until the single rail 2716 became affordable before I could finally make progress, and by then I had later ZX computers to act as the control host.

Last edited by SiriusHardware; 20th Dec 2020 at 12:03 pm.
SiriusHardware is online now  
Old 20th Dec 2020, 1:50 pm   #69
ortek_service
Octode
 
ortek_service's Avatar
 
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,394
Default Re: Softy 1

I can confirm that the Dataman 48Pro (I actually used a 448Pro+, but I believe that's really just a re-designed smaller case version of the original 448Pro that had been supplied at the same price just before release of the 448Pro+ and appeared to support the same IC's) works directly with the similar MH74S287 256 x 4bit PROM - the only Tesla one's I've tried so far on it.

But I do intended to get some MH74S571 512 x5bit PROM's shortly as they are quite reasonably-priced at: https://shop.tvsat.com.pl/en_GB/
And this is listed in the Pg4uW control software, which you can download free & run in demo mode or the pdf supported-devices list:

https://www.dataman.com/dataman-48pro-resources
https://www.dataman.com/dataman-48proplus-resources

Although for some reason, on these resources page the latest version for the 48Pro is v3.32 (2017), whereas 48Pro+ ihas the very latest (Dec 20) v3.63. So it could be that they've deliberately stopped supporting the 48Pro with further updates, and now only do updates for the 48Pro+

But both v3.32 & v3.63 lists these Tesla MH74S ones:
Tesla MH74188
Tesla MH74S287
Tesla MH74S370
Tesla MH74S571



For EPROM-Emulation, even though I've got an S3 I'm tempted to build one of these recent ones: https://groups.io/g/eprom-emu-ng
As being open-source, it makes it much easier to customise & update, with GUI's also being done for it. Unfortunately the designer doesn't ship to the UK the kit / PCB, being sold on the USA version of the online auction website.
But as it's all open-source h/w & s/w, you can just get a few PCB's made cheap in China yourself.


It looks like a Softy-rebuild will be done shortly, as my friend Chris is now currently being sent the 2708 & PROM from one to copy. So could build one of these for doing 3-rail 2708's etc.
But if you really did need to program (but could use it for just reading) a 2708 etc. device, it's probably more convenient to have an adapter to a more standard 2716 etc. for use with more recent programmers. And this one has on-board DC-DC converters to avoid having to have lots of external PSU's / sychonising the switch on of these:

https://buebchen.jimdofree.com/8-bit...gadapter-2708/

It seems there maybe a few (cheaper than a Data I/O) programmers that directly support the 2708 etc.
https://forums.arcade-museum.com/thr...rammer.476950/

But not that cheap, so may be better to build a dedicated programmer, such as open-source one here:
http://www.mattmillman.com/projects/...m68766-eproms/
ortek_service is offline  
Old 20th Dec 2020, 2:30 pm   #70
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,485
Default Re: Softy 1

Thanks for looking up the device list of the Pro+ - it seems there is some hope of obtaining a programmer which can do Tesla devices without having to spend ~£900 or more on a current model which can.

If it's OK with the owner of the source machine being supplied to Chris, perhaps you could attach copies of those SOFTY files in a post here as a preservation measure since there are probably not many working machines left in the wild, and the survivors must be facing imminent danger of EPROM data fade or BPROM failure by now.

I'm sure TonyDuell would appreciate an image of another EPROM to verify his against, assuming they are the same firmware version of course.
SiriusHardware is online now  
Old 20th Dec 2020, 5:45 pm   #71
TonyDuell
Dekatron
 
Join Date: Jun 2015
Location: Biggin Hill, London, UK.
Posts: 5,191
Default Re: Softy 1

The more backups the better!
TonyDuell is online now  
Old 9th Jan 2021, 8:17 am   #72
ortek_service
Octode
 
ortek_service's Avatar
 
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,394
Default Re: Softy 1

Quote:
Originally Posted by SiriusHardware View Post
>>
>>
If it's OK with the owner of the source machine being supplied to Chris, perhaps you could attach copies of those SOFTY files in a post here as a preservation measure since there are probably not many working machines left in the wild, and the survivors must be facing imminent danger of EPROM data fade or BPROM failure by now.

I'm sure TonyDuell would appreciate an image of another EPROM to verify his against, assuming they are the same firmware version of course.

Yes, the owner of the Softy who loaned these to Chris, seems to be happy for these to be made available online - as now also updated his webpage: http://www.bygonebytes.co.uk/Softy.html

with a link to Chris's new Softy-1 webpage: http://81.174.146.201/acorn/PROMs%20...1/softy_1.html

- Which now has finally released online (for the first time?) copies of the Character Generator PROM AND Crucially, the 2708 EPROM IC's, that he managed to readout from the ones he recently received


This website will soon have a domain name, similar to his main site, that's no longer updated in readiness to switching to hosting it all himself on an RPi web server. But this IP-address should still work, as he has a fixed IP-address


Also on this webpage, is a Python script he's written to convert all of the Character Generator PROM's content into an ASCII-symbol representation of all of its Character-set in a text output (also on there) to verify it readout OK

And a disassembly of the firmware -Where he's discovered an invalid-address Jump instruction. So either there was an original bug, or a byte has got corrupted in the EPROM he had, so would be useful for TonyDuell to verify against his one sometime...)


Both of these websites now also have a newer Hi-res copy of the instruction manual - originally on bygonebytes, who has an original copy for his Softy-1 and has now re-scanned it to make the schematic a bit more readable for Chris etc. to make his replica from (Another Webpage documenting all of this is under construction (with CAD re-drawn PDF etc. like for his Mk14 replica), and should be released soon, once he's got his replica PCB ordered.

I've also OCR'd this latest manual, but can't attach it here due to size (8MB) - It loses some un-OCRable text at the bottom if I reduce file size to around 2MB, with Acrobat Full-Standard's option for this I normally use)
So I will probably have to upload it somewhere & share.

I couldn't find any reference in the Manual, to the 'Conversion Card' that was apparently available and enabled programming of 2508, 2758, 2516 (=Intel 2716) and 2532 EPROM's. Or from looking through all of the adverts in Wireless World magazines from back then

The original Softy was only advertised a few times between Jan 1980 - from advertiser 'Softy' (With 'Dataman' in address), then briefly GP Industrial Electronics - sometimes separately from Softy in the same Issue!.

But by the end of 1980, neither were advertising it anymore, with GP moving onto much-more expensive Professional (E)P4000 Development / Production programmers, doing lots of both 3-rail & Single-rail supply types.
And GP seemed to have handed the Softy selling over to Dataman, who eventually dropped advertising under the 'Softy' name, becoming 'Dataman Designs' late in 1981, just after advertising new S2 version with full-page ad.

Most original-Softy adverts were short on details - often not even a picture!

So if anyone has any details on the 'Conversion Card', that would be of interest (Chris thought he'd seen it mentioned on here)


In the process of designing a similar replica board, Chris has also discovered the original schematic in the manual has a few errors:
The most significant are IC12 no clock and IC6 PA0,1 connections missing.

Also, on his replica, he's added links allowing NENIN and NCARD to be tied to 0V obviating the need to solder wires to the edge-connector, that you needed to do on the original to make it work! (A bit like SENSEB link on MK14, with later SCIOS version)

Last edited by ortek_service; 9th Jan 2021 at 8:25 am.
ortek_service is offline  
Old 9th Jan 2021, 9:06 am   #73
TonyDuell
Dekatron
 
Join Date: Jun 2015
Location: Biggin Hill, London, UK.
Posts: 5,191
Default Re: Softy 1

I'll get my Softy out sometime when my bench is a little less full...

As an aside, I have both the EP4000 and P4000 programmers. The former is more like the Softy, it lets you display the data, edit it, etc and it's also an emulator. The latter is an 8-way EPROM copier. I've reverse-engineered both and dumped all the EPROMs if anyone is interested. I think I passed that information on to the chap who runs a website for such units too.
TonyDuell is online now  
Old 9th Jan 2021, 11:03 am   #74
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,485
Default Re: Softy 1

Well done to Ortek... (whose name we still don't know), and friend Chris. Another giant leap for micro preservation. Very thorough investigation of the EPROM contents too, by the sound of it. It will be interesting to see if Tony's EPROM has the same apparent error, but be aware of the possibility that the machines might not have the same firmware version (Tony's may have a 'version+1' where that error has been found and fixed).

In an ideal world the EPROM's label would always have the 16-bit checksum of the contents written on it, that was quite a widely adopted convention at one point. It made it very easy to check whether the EPROM contents were intact or not. I would still suggest doing that even now, when programming and labelling EPROMs.
SiriusHardware is online now  
Old 9th Jan 2021, 11:53 am   #75
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,485
Default Re: Softy 1

Quote:
Originally Posted by ortek_service View Post
(A bit like SENSEB link on MK14, with later SCIOS version)
A minor correction: Sense-A is the line which has to be held at 0V on MK14 with 'new' OS, rather than Sense-B.
SiriusHardware is online now  
Old 9th Jan 2021, 3:02 pm   #76
ortek_service
Octode
 
ortek_service's Avatar
 
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,394
Default Re: Softy 1

Quote:
Originally Posted by TonyDuell View Post
I'll get my Softy out sometime when my bench is a little less full...

As an aside, I have both the EP4000 and P4000 programmers. The former is more like the Softy, it lets you display the data, edit it, etc and it's also an emulator. The latter is an 8-way EPROM copier. I've reverse-engineered both and dumped all the EPROMs if anyone is interested. I think I passed that information on to the chap who runs a website for such units too.
Yes, it would be interesting to see what's inside either of these, as I imagine they've got quite a lot of circuitry crammed in there (especially the P4000 8-way copier) to handle switching between so many listed EPROM types inc. 3-rail and single-rail types without adapters, compared to the much more limited Softy. And strangely, it seems the P4000 could also handle the larger EPROM's without an adapter (that was needed for these on the EP4000). Although needing 8x adaptors would be a bit messy, and the EP4000 did have other capabilities (TV output of hex looks similar to the Softy) / was a slightly-earlier design.
GP Ind. Electronics, then moved onto the quite similar-looking EP8000 / P8000 not too long after the 4000 ones, then the rather different XP-640 with built-in LCD, that looked a bit more like Elan / Stag etc ones of the time, in their move to the higher-end of the market (From starting-out just advertising EPROM Eraser units, it seems)

I hadn't found a particular website specialising in info on the (E)P4000-series, So ut would be interesting to hear what that one is
If not: http://matthieu.benoit.free.fr/ - which has quite a broad range, but mostly referred to frequently for DataI/O info by this group: https://groups.io/g/DataioEPROM (which also has some from other high-end manufacturers on there, but need to join to access their files)


I presume GP moved away from the SC/MP CPU for the 4000-series / later ones, with its max. 4K page size limitations - Whereas Dataman (Designs) kept with it for the S2 they developed themselves, according to info on their adverts

Strangely, there seems to be a lot less info on the S2, than the original Softy - Maybe because it was then only available ready-built, rather than a kit so maybe no published circuit. But I've not found a copy of the manual anywhere on the 'net (or even a photo of the insides for the S2!).
Dataman did make as big thing about how open they would be with the S3, and its hardware and a developers / technical manual with full circuit (rather than just block diagram in the provided manual) and source-code.
But as they later advertised you had to pay £150 to get hold of the 'developers manual', I doubt too many copies were printed. The circuity isn't too complex to rev-eng (especially with quite detailed block diagram), but their use of a Toshiba Mask-programmed for them (with SOFTY marked on it) low-power Z80-core micro and custom? Texas-Instruments made Keypad etc. would have made building you own from these details a bit more involved.
ortek_service is offline  
Old 9th Jan 2021, 3:04 pm   #77
ortek_service
Octode
 
ortek_service's Avatar
 
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,394
Default Re: Softy 1

Quote:
Originally Posted by SiriusHardware View Post
Quote:
Originally Posted by ortek_service View Post
(A bit like SENSEB link on MK14, with later SCIOS version)
A minor correction: Sense-A is the line which has to be held at 0V on MK14 with 'new' OS, rather than Sense-B.
Thanks - I knew recalled it was one of those
(So had a 50% chance of getting it right, without digging out the circuit etc.!)
ortek_service is offline  
Old 9th Jan 2021, 3:12 pm   #78
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,485
Default Re: Softy 1

Wasn't being pedantic, but someone might possibly have surfed onto that post at some point in the future and spent many happy hours thereafter trying to get their new-build MK14 to run programs with Sense-B tied to 0V.
SiriusHardware is online now  
Old 9th Jan 2021, 3:26 pm   #79
ortek_service
Octode
 
ortek_service's Avatar
 
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,394
Default Re: Softy 1

Quote:
Originally Posted by SiriusHardware View Post
Well done to Ortek... (whose name we still don't know), and friend Chris. Another giant leap for micro preservation. Very thorough investigation of the EPROM contents too, by the sound of it. It will be interesting to see if Tony's EPROM has the same apparent error, but be aware of the possibility that the machines might not have the same firmware version (Tony's may have a 'version+1' where that error has been found and fixed).

In an ideal world the EPROM's label would always have the 16-bit checksum of the contents written on it, that was quite a widely adopted convention at one point. It made it very easy to check whether the EPROM contents were intact or not. I would still suggest doing that even now, when programming and labelling EPROMs.

Hello, My name's Owen (I would have put it as username, but I think you needed more characters, even for my email name so had to add a bit too it. And it didn't seem like too many others had got their name in the user name).

The photo on Bygonebytes shows the EPROM labelled '304'. Which I guess was version 3.04, and they didn't use a . in the numbering as not all computer systems allowed this without it meaning an extension.
And Chris has also put this 304 in the filename at: http://81.174.146.201/acorn/PROMs%20...softy1_304.bin

Yes, I've seen checksums on many printed EPROM etc. labels - produced by 'programming house' subcontractors?
But there were still many manufacturers who just used a small hand-written version number on their EPROM's - even on fairly mass-produced products like the BBC Computer and the DFS 0.90 8K EPROM (before it was superseded by the 16K mask-ROM DNFS 1.2, with Econet & 2nd Processor Tube Host code support added)

Maybe the issue with checksums, is knowing which 'standard' to use. A straightforward additive one isn't always too good, so a more complex CRC is often used. And I recall the later Dataman programmer software / some hex editors gives you several different ones (Early version just give 32bit additive? checksum, so typically not much more than 16bits for small code)

Owen
ortek_service is offline  
Old 9th Jan 2021, 3:39 pm   #80
ortek_service
Octode
 
ortek_service's Avatar
 
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,394
Default Re: Softy 1

Quote:
Originally Posted by SiriusHardware View Post
Wasn't being pedantic, but someone might possibly have surfed onto that post at some point in the future and spent many happy hours thereafter trying to get their new-build MK14 to run programs with Sense-B tied to 0V.
That's OK - It's good to have things correct as facts like that can not always be too widely-known (especially when it was buried in the manual, and you had to add your own link, so it must have caused a lot of problems with those who'd missed this)

Plus why I also commented that the original Softy schematic isn't quite right (Chris's Replica Softy webpage will have all these issues found on it, and I have suggested that it may be useful to put a warning not wherever a download link is, that there are original errors)
- As I've found many original schematics had omissions / errors in the days they were drawn by hand. Although when Acorn re-drew their Acorn System ones in CAD, they also kept all the errors in!

And is something this is trying to resolve: https://www.baldengineer.com/bit-pre...ith-kicad.html
ortek_service is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 11:26 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.