UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio 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 26th Apr 2020, 5:59 pm   #21
TonyDuell
Dekatron
 
Join Date: Jun 2015
Location: Biggin Hill, London, UK.
Posts: 3,769
Default Re: Maplin Z80 SBC

I am very surprised that any USB-RS232 interface won't talk to a 1489. That IC will work with just about any input signal (even raw TTL!).

That said, could you make a buffer that takes in the signals from your USB adapter and outputs solid RS232 levels? I would be quite a simple project, the worst part would be the PSU.
TonyDuell is online now   Reply With Quote
Old 26th Apr 2020, 6:37 pm   #22
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 4,234
Default Re: Maplin Z80 SBC

If you are convinced that the problem is RS232 levels / voltages then I would consider bypassing the onboard level converters in your gadget with a MAX232, since that has charge pump circuitry onboard to make +/- 10V from 5V as far as I remember. You can get tiny ready made level converters with an SMD MAX232 onboard. Unfortunately all the ones I can readily find are on ebay so I can't point you directly to one particular listing but if you search on

"Mini-Level-Converter-RS232-to-TTL-Adapter-Module-MAX3232-MAX232-Serial-Board"

The ones where the PCB is barely bigger than the SMD chip are the ones I'm talking about. So small, you can fit one inside almost anything. So cheap, it's hardly worth making one yourself.

It's definitely the gadget not receiving, rather than the RS232 source not sending? I'm just wondering if the source is expecting to see hardware or software handshaking before it will send? You may just need to strap two or three of the pins together inside the 'D' plug which is plugged into the USB-RS232 lead.

Last edited by SiriusHardware; 26th Apr 2020 at 6:48 pm.
SiriusHardware is offline   Reply With Quote
Old 26th Apr 2020, 7:01 pm   #23
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 4,234
Default Re: Maplin Z80 SBC

Quote:
Originally Posted by TonyDuell View Post
could you make a buffer that takes in the signals from your USB adapter and outputs solid RS232 levels?
I have seen an ingenious circuit which used two back to back MAX232s, so the input to the circuit was RS232-level and the output from the circuit was also RS232-level, which at first sight seemed absolutely pointless.

The reason for it was that the 'RS232' output on the circuit this interface was used with was the 'right way up' for RS232 but only TTL-Level 0V-5V, which the RS232 inputs on MAX232s are tolerant of.

The input MAX232 was taking in the indifferent 'RS232' signal, passing it on inverted / at TTL level to the second MAX232 which was then re-inverting it and converting it back to full-swing RS232.

The nice thing about this dodge is that it only requires a 5V supply since the MAX232s generate their own higher voltage +/- supplies internally.
SiriusHardware is offline   Reply With Quote
Old 26th Apr 2020, 8:02 pm   #24
IanBland
Hexode
 
Join Date: May 2015
Location: Northampton, Northants, UK.
Posts: 380
Default Re: Maplin Z80 SBC

I bought several of those RS3232 to TTL adapters a while ago, they're as handy as heck. The ones with the connector on in my case.
IanBland is offline   Reply With Quote
Old 27th Apr 2020, 2:10 am   #25
TonyDuell
Dekatron
 
Join Date: Jun 2015
Location: Biggin Hill, London, UK.
Posts: 3,769
Default Re: Maplin Z80 SBC

This reminds me of a very obscure fault I had when at university :

There was an RS232 socket on the wall connected to the departmental computer. It worked fine with a (VT100) terminal but not with my laptop (an old Tandy Model 100). It sort-of worked but lots of data corruption, etc. I thought (without much checking) that the Model 100 with its low-power circuitry (to run off batteries) couldn't supply enough current over the RS232 connector to drive the capacitive load of the cable. So I went to Maplin (shows how long ago it was...), bought a MAX232 and wired it up as an RS232-in, RS232 out buffer (as in Post 23). Powered it off a bench supply, worked fine. Thinking I'd cured the problem, I boxed it up and powered it from a wall wart. Didn't work fine. But the supply voltage to the MAX232 was OK, no noise on the supply lines, etc.

Now what was the real fault?
TonyDuell is online now   Reply With Quote
Old 27th Apr 2020, 9:06 am   #26
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 4,234
Default Re: Maplin Z80 SBC

There aren't quite enough parameters yet. Was the bench supply 0V earthed? If so the real problem may have been a missing or high resistance RS232 GND connection in the Tandy, reinstated via another route via the mains earth of the bench supply. Much depends on what was earthed, and what was not.

The Wall Wart was most likely un-earthed / isolated.

Anyway go on, tell us.
SiriusHardware is offline   Reply With Quote
Old 27th Apr 2020, 11:49 am   #27
TonyDuell
Dekatron
 
Join Date: Jun 2015
Location: Biggin Hill, London, UK.
Posts: 3,769
Default Re: Maplin Z80 SBC

Yes, an earth problem, but not in the Tandy (which was fine). The problem was an open signal ground (pin 7 of the DB25) connection at the wall socket. The terminal server had signal ground earthed to mains earth, as did the VT100, so the circuit was completed that way. The Tandy Model 100 runs off internal batteries so it wasn't earthed. The bench supply did indeed have the -ve side connected to mains earth, so that also completed the circuit when I ran the buffer off that supply. But as you guessed the wall wart was isolated.

Once I'd realised what was going on, I unscrewed the DB25 plate from the wall, spotted the wire hanging off, and resoldered it.
TonyDuell is online now   Reply With Quote
Old 27th Apr 2020, 12:17 pm   #28
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 4,234
Default Re: Maplin Z80 SBC

Well, with hindsight you couldn't really state in precise detail what was earthed and what wasn't when you posed your question, otherwise it would have given too much away.

On a good day, when I have time to find them and I don't have 10 other things to fix as well, I love faults like that - but there are days when I can really do without them as well. I'm not surprised you committed that one firmly to memory.
SiriusHardware is offline   Reply With Quote
Old 27th Apr 2020, 2:21 pm   #29
mikew8760
Triode
 
Join Date: Mar 2020
Location: Bristol, UK.
Posts: 18
Default Re: Maplin Z80 SBC

Just to fill in some of the details of my RS232 problem. In order to establish, as far as possible, the nature of the problem I used my desktop PC which has a real serial comms port. This works fine with the MICE, which sort of confirmed my suspicions about signal levels. Then I also rigged up a real Heath Robinson (younger members, Google him!) contraption with two 12V batteries and a 1488 and 1489 - that also worked fine! I have the schematic for the MICE, and it shows the 1489 Response Control pin connected via a 12K resistor to -12v, which is not the same as in the data sheet, which is at least one reason why I wished to avoid hacking this pcb.
I shall pursue the joule thief idea for a bit, until I think of something better, or just buy one of the expensive (over 6) adapters.
At least it has given me something to do whilst in lockdown!!
mikew8760 is offline   Reply With Quote
Old 27th Apr 2020, 6:59 pm   #30
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 4,234
Default Re: Maplin Z80 SBC

The titchy little MAX232 PCBs without connector are only about 2+.

You need not modify or replace the existing RS232 receive level converter, just break the connection between the RS232 input pin on the socket and the input of the original (unmodified) receive level converter and insert the circuit below.

It will beef up the weedy RS232 being generated by your USB-serial lead and convert it into a full-fat RS232 signal which the input of the original level-converter will then deem acceptable.

This assumes that you are not using hardware handshaking, if you are then you would also have to break and buffer any handshaking input signals as well, using further pairs of MAX232 elements.

If you were feeling exceptionally adventurous you could (carefully) hack the 9-way moulded plug on the end of the USB-serial lead to pieces, swap a solderable plug and removable long-bodied plug housing onto the end of the cable and put the mini MAX232 buffer in there, because there will already be +5V available inside that D-plug to power whatever is originally in there.
Attached Thumbnails
Click image for larger version

Name:	RS232_cleaner.jpg
Views:	20
Size:	14.1 KB
ID:	204134  
SiriusHardware is offline   Reply With Quote
Old 27th Apr 2020, 11:23 pm   #31
TonyDuell
Dekatron
 
Join Date: Jun 2015
Location: Biggin Hill, London, UK.
Posts: 3,769
Default Re: Maplin Z80 SBC

Quote:
Originally Posted by mikew8760 View Post
Just to fill in some of the details of my RS232 problem. In order to establish, as far as possible, the nature of the problem I used my desktop PC which has a real serial comms port. This works fine with the MICE, which sort of confirmed my suspicions about signal levels. Then I also rigged up a real Heath Robinson (younger members, Google him!) contraption with two 12V batteries and a 1488 and 1489 - that also worked fine! I have the schematic for the MICE, and it shows the 1489 Response Control pin connected via a 12K resistor to -12v, which is not the same as in the data sheet, which is at least one reason why I wished to avoid hacking this pcb.
I know you want to avoid 'hacking this PCB' but I'd try lifting one end of that 12k resistor and running the 1489 in the MICE with nothing connected to the response control pin. You might find it then works with the USB interface lead. And it's trivial to revese if it doesn't help or if you want to go back to the original circuit.
TonyDuell is online now   Reply With Quote
Old 9th May 2020, 12:55 pm   #32
mikew8760
Triode
 
Join Date: Mar 2020
Location: Bristol, UK.
Posts: 18
Default Re: Maplin Z80 SBC

After some thought, common sense prevailed and I decided the MAX3232 device would be the way to go. I got a little board, with an smd chip and the capacitors, on eBay cheaply. So far I've just stuck all the bits on a breadboard, that is a USB to TTL adapter and a DB9 connector, and this now all works fine. When I have the time I'll put it neatly onto a piece of Veroboard, might even put it all in a nice little box!
Now all I have to do is get to grips with the ancient MICE box of tricks and start tracing code...
Thanks for all the assistance from everyone.
mikew8760 is offline   Reply With Quote
Old 9th May 2020, 1:22 pm   #33
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 4,234
Default Re: Maplin Z80 SBC

If you are still trying to revive the Z80-CPU PCB itself it may be useful to write short test routines (In EEPROM) just to verify whether the CPU PCB is / is not executing code correctly independently of the monitor EEPROM and keypad / display.

On page 5 / Page 6 of the original CPU construction article there is a LED array and short piece of test code respectively which activates the I/O select lines and shows activity on the LEDs, although given the shortness of the select pulses I would not expect those to be very bright.

You could probably get a better visual indication of I/O select line activity using a logic probe with a 'Pulse' indication mode on the I/O select lines, or of course a scope if you have one.
SiriusHardware is offline   Reply With Quote
Old 10th May 2020, 2:11 pm   #34
Slothie
Hexode
 
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 370
Default Re: Maplin Z80 SBC

I've had a lot of success with this 2 transistor circuit that steals the negative supply from the attached RS232 device and you might have the parts in your spares box (transistors are not critical, any small signal ones will do. )
Attached Thumbnails
Click image for larger version

Name:	2tr.jpg
Views:	19
Size:	41.3 KB
ID:	205407  
Slothie is offline   Reply With Quote
Old 19th May 2020, 9:43 pm   #35
mikew8760
Triode
 
Join Date: Mar 2020
Location: Bristol, UK.
Posts: 18
Default Re: Maplin Z80 SBC

Well, I should have known all was too good to last! Having used the MAX3232 to produce the +/- volts which the MICE demanded, all was good. However, while getting to know how to operate it, since I haven't used it in many a long year, it decided to pack up. Seems completely dead, doesn't even produce the usual start-up message. So far all I've done is check the usual voltages and scoped signal lines around the CPU (8085) and there's lots of activity.
I'm just wondering if someone on the forum might have some experience of this ancient Z80 emulator? It's a Microtek Z80B MICE II, vintage 1985. A circuit diagram would be nice, but I'd guess as rare as hens' teeth!
mikew8760 is offline   Reply With Quote
Old 19th May 2020, 9:51 pm   #36
Slothie
Hexode
 
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 370
Default Re: Maplin Z80 SBC

I'm not sure there's anything useful here but there's manuals for various versions

http://www.bitsavers.org/pdf/microtek/mice/
Slothie is offline   Reply With Quote
Old 20th May 2020, 8:44 pm   #37
mikew8760
Triode
 
Join Date: Mar 2020
Location: Bristol, UK.
Posts: 18
Default Re: Maplin Z80 SBC

Thanks Slothie for that link. I have the User Manual, but the contents of the 3 EPROMS are very useful. I've checked them against the contents of those on the board and, of course, they are all OK.
I've been checking the data path back from the 1488/89 and get to the UART, 8251. Here I find no activity on the TxD pin. I think it is correct to assume the programming of this device by the 8085 is correct - scoping the CS, WR etc I see what I presume is the data for the sign-on message being loaded into the transmit buffer. This is repeatable for processor resets, but the TxD pin remains high. So, at the moment I'm assuming that a new 8251 is needed?
I welcome any observations anyone cares to make on the above. I suppose 8251's are still available, main hassle will be extracting it from the PCB!
mikew8760 is offline   Reply With Quote
Old 20th May 2020, 9:00 pm   #38
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 4,234
Default Re: Maplin Z80 SBC

You mentioned bus signals, but you didn't mention clocks. Do you have a baud-rate clock signal going into the 8251, pin 9 (TxC) - and while at it also check pin 25 (RxC) and the main clock in from the system on Pin 20 (CLK).

You didn't mention before whether the RS232 connection between the PC and emulator uses hardware handshaking as well as the minimum of TX / RX and GND. If it does, you might also have to buffer any handshaking signals which go FROM the source PC TO the emulator, otherwise the emulator may think it is being informed that it can't send, and is therefore not sending.

That USB-serial lead is likely to be sending out the same rubbish levels on its handshaking output lines as it did on its TX line.
SiriusHardware is offline   Reply With Quote
Old 21st May 2020, 12:19 am   #39
TonyDuell
Dekatron
 
Join Date: Jun 2015
Location: Biggin Hill, London, UK.
Posts: 3,769
Default Re: Maplin Z80 SBC

Also check CTS/ (pin 17) of the 8251 is asserted (low). This pin will prevent the 8251 from sending, it's checked in the 8251's hardware (so however the 8251 is set up, whatever else the firmware checks for, it can't be ignored).
TonyDuell is online now   Reply With Quote
Old 21st May 2020, 12:29 pm   #40
mikew8760
Triode
 
Join Date: Mar 2020
Location: Bristol, UK.
Posts: 18
Default Re: Maplin Z80 SBC

Thanks for the replies all. I skipped over the details you mention; I have already changed the jumpers for RS232 comms on the board, so that no handshaking is required. I should have mentioned that for this 'test' the MICE is not connected to the PC, so no 'interference' from it. ( I previously discovered that a USB - RS232 adapter would work OK on Receive, and the emulator sign-on message would be displayed in PuTTY, but it was the lacking of +/- voltages on Transmit that was the problem.)
Yes, there are requisite clocks, reset is not stuck, but I shall check CTS as suggested.
I have assumed that since this part of the system was working, the 8085 correctly configured the 8251. Not having a schematic, and that this (at least double sided ) board is one of four stacked on one another making circuit tracing awkward. I'm not even sure if the 8251 is I/O or Memory mapped.
Perhaps, when I feel like it, I'll clip on my noddy logic analyser and check the data being fed to the 8251 is what I think it is. I feel that it is, since after the first flurry of activity in sending data all goes 'quiet', presumably as the system is waiting for a ( non-existent ) command input from the terminal.
Watch this space!
mikew8760 is offline   Reply With Quote
Reply

Thread Tools



All times are GMT +1. The time now is 7:44 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 - 2020, vBulletin Solutions, Inc.
Copyright ©2002 - 2020, Paul Stenning.