So here we are again.
I will also revisit the ZX81 I/O port that I failed to get working from a 25 year old kit last summer. This time around I am just going to build it myself from the circuit diagram. I will get a bloody light to flash or a voltage to change under my command if it bloody kills me. !*Breadboard*!
Up there, that is my Nascom 1 -It came from a heavily modded third party system, which did not work when I got it. It has now been taken back to its vanilla form. However in doing that i discovered that there are things missing from the board - hence the not working. I will attempt (and fail) to get it working again. !*Solder*!
30th July
And my RC comes to a close with a whimper, a whimper which belies the actual successes.
Firstly tho, let me give a HOOPLA! to everyone else, regardless of whether you were as successful as you planned. I enjoyed a whole bunch of blogs and some of the stuff was wonderfully pointless. Hurrah!
Me? Well ultimately i didn't achieve anything completely. I failed at getting the ZX81 port to work, but i honestly think i am pretty close. In a typical wgoodfian sort of way i count as a success that the ZX81 didn't crash when the port was powered and attached. I do think i need more oomph to get the whole thing working, but i think there are programmable voltages in there, i just need another power brick.
I also got the a free shell account down London way which means that i can cheaply use my Apple //c to dial a whole range of BBS, far cheaper than dialing the US or Canada from Scotland which was the way i had to do it on the rare occasions i did. That was a nice wee bonus.
But mostly tho, the Nascom held my attention for most time. Its a shame that it isn't completely working as the month ends, but the fact that it works at all, is great, a resurrected machine from 1978, 32 years old pretty much. That impresses me. There is still stuff to do with it tho, programming seems to be an issue and i think the cut tracks are the problem. Reluctant as i am, i may have to go in there and try and fix them. But i will buy a better solder iron first. The fact that over the last couple of days i am finding myself (in a very basic way) able to look at hex code in the various old Nascom PDFs and actually make out instructions and what they do, i also find quite impressive (for me). A success without a doubt.
Anything else? Nope not really, it was a busy and hectic month at my end and not nearly as relaxed as it usually is, but then again, thats July for you. There is a reason that i like the WinterWarmUp more; work hasnt got me by the balls for the first few weeks of January!
Anyway, see you next time - and remember, Issue Three of 300baudmagazine should be out come September time.
29th July
I spoke to a Nascom guy online.
He said i should have this- 
Rather that that down there.
I think those cut tracks are to blame. But considering its 10pm, i got in at 9pm and left the house just after 6am, i am doing bugger all. Tomorrow may produce something, but as i off on holiday on the saturday, it cant be promised.
28th July
I think and i wonder if my board is set up in some weird way.
I am pretty sure that i am placing something in the -A- register correctly and i am pretty sure that i am looking at the -A- register correctly afterwards, yet i see nothing. This is either a modified board with added weirdness (two ROMs the cut tracks and who knows what else) or this is considerably harder than i thought it was going to be. Ok, you are right, me being stupid is the more likely option.
27th July
Programing in hex is hard. Z80 assembly is hard. I achieved nothing!
Put it this way:
C80 3E 0C - Loads the accumulator -A- with the clear screen code.
C82 F7 - then calls the ROUT routine in ROM which prints the contents of -A-
This causes the machine to clear the screen, or actually is it resetting -hard or soft?- the machine? Its hard to tell. I can try and load a different character such as erm... 48 and then Tabulate the address C80 to see if it is stored there, but it seems not to be.
I think before i will get anywhere with this i will need to do a fair bit of reading. The other problem is that some commands entered into memory crash the Nascom or reset it. Hex programs are large and without a backspace key, its kinda daunting. I just want a ten line program that works!
I did find a wee 'hello world' program, but i failed to get that to work either. I am scouring the web! Sadly i only have until friday night and i am working late each night. RUSHED!
26th July
Ok i admit it. I am a big pussy! - There i have said it.
I had a look at the tracks in question that seemed damaged; in fact they were cut rather than damaged. Extend that out, it means that they were cut by the original modder and likely have nothing to do with the keyboard problem. I looked at the tracks around the plug on the motherboard, i found many areas across the board where things looked dodgy, heck i nearly lifted a section of track trying to remove what i thought was stray solder. A big blob (not mine) just sitting there, benign. I thought and considered and then came to a conclusion.
When i started i didn't have a working Nascom, now i pretty much do. The Nascom, has a special place in UK home computing and i have pretty poor soldering skills. Simply put, i am calling it quits! - I may later attach a reset switch, but frankly i would be mad to do anything else. It works and i would likely kill it. So i suppose i will hover over worldwide ebay looking for a LICON keyboard. 16pin ribbon plug. Anyone, anyone?
So that gives me a few days until the end of the month. I guess that either leaves me the ZX81 port. (too bitty at this point) or trying to get something to run on the Nascom except arithmetic and the screen clear routine. So that i think, will be a fine end point. Printing my name will be sufficient i think.
Well its going to be hard! I mean its assembly and i have to program in hex. I think 'Hello World' will be more than sufficient. I barely manage in BASIC for gods sake!
Take care - love you!
25th July
Howdy!
The mountains were nice, although i think i saw more bars and beer taps than mountains. Lets say that i am slightly jaded today, 39 year olds shouldn't go out to clubs, its unseemly. Anyway, i am back home now and i had a look at the Nascom again. Inspired by Adrian Mac Medics efforts at fixing his Mac Portables i thought i should look at the tracks under the keyboard plug section of the motherboard. From that inspection i think there are two areas which should have continuity but dont. I am too sleepy to do a damn thing about it and i certainly don't think holding a soldering iron would be good, so a task for tomorrow.
And it has to be said, tomorrow as the lead to the last week. Golly, didn't we get here fast.
22nd July
Well i am stumped - I simply cant think of a place to go with the keyboard problem. I have found the IC that deals with the key-press and i have found the IC that deals with the sense function for that row of keys, but i will be buggered if i know what to do from there. Everything seems ok. I have gave the board a wee scrub, but no change. The best i can do is re-solder the joints on the underside of the motherboard at the keyboard socket, but thats about it. The IC that deals with the key-press is really just three transistors and frankly i dont know how to check a transistor, either the input and by extension the output. And besides, if we assume that either of the two ICs are goosed then i dont have replacements just now. They are common i believe, one certainly so. But that would take time. Hummm-de-Haw! - Anyway as it stands, i am away for a few days so i will give it some thought, at the very least i might try to write a wee program, it all has to be in hex as the basic, basic system that i have doesn't do mnemonics; which is just as well as part of the keyboard is buggered. Silver lining and all that. Any ideas on the keyboard, then let me know. See you all on Sunday or Monday. I will keep twittering the best i can on everyone else.
21st July
Well i found some schematics for the Nascom Keyboard, but its a whole lot of tracing to find out where things are and tonight i cant be bothered. I did quickly check the continuity of a dodgy part of the cable, which was fine and i checked the bank of resistors that at this point seem to be associated with the line of keys not working and they were fine. What i need to do is trace it all and find out what each line to the to keyboard ICs should be doing and from that find out if pin 4 of keyboard IC one (the culprit so far) is doing what it should be. I have at this point no idea how to do that, but thats what needs done.
I did look around for some little programs to try and failed completely to get them to do anything of note. The programs are small and my entry looks fine and indeed they let me execute but two different programs both resulted in the screen filling with the number 1 and scrolling. I tried various little programs, but no luck with any. I then swapped to the second ROM, well i tried the best i could. I cant J-ump to the second ROM address because of the keyboard problem, so i tried another way, which only resulted in the same program giving me a screen of 'fuzzy squares'. I think i may try and write a very small program and see how i get on that way.
Anyway, i have tomorrow night for a wee play, then its away for a few days, so nothing until sunday or monday.
20th July
Ok!
I have a working Nascom 1 board. The random character thing? It seems was because the serial link was loose. That was firmed up and now no problems whatsoever. Well except the keyboard. It works except the keys G H J K L @ ; : and backspace. Yeah, backspace is a bummer. These keys did work, but now dont. Sadly the Nascom doesnt use a normal pressure contact matrix, is all wonderfully high tech (for 1978) and digital. Clearly its a martix problem, but its not a nice wee pressure sensitive grid with contacts that you can clean. The keyboard in many ways is actually sealed, even tho it seems open to the world. Rivets hold the metal top plate to the PCB. Hmmm! - From that it looks to be a component problem (well thats all i can get at!). I will test the 3 ICs that are available, to the best of my ability, but it feel i may get nowhere. This of course is very very annoying as the computer is working perfectly, i just cant use it properly.
I do have some sample programs that wonderfully dont use the keys in question, but thats luck. I mean i cant L-oad from any where, and each line has to be correct, no corrections allowed.
Hum-de-hum!
Actually as i typed that i discover that the actual programs are all hex. Numerical plus ABCDEF - Oooer this might not be so bad, at least for fun play. I am sure i found a wee version of breakout the other night. Hmmm!
19th July UPDATE
I swapped the ROMS but no change to the fuzzy square problem. I re-seated the character generator ROM and now the screen fills up with random stuff and ultimately the machine doesn't work again!
YAY! COCK!
19th July
Well, however you look at it, i have a working Nascom 1 from 1978. Its mostly working, in that i have this repeating character problem that comes and goes and is not keybpard based in the traditional notion of it. But i typed in a small program this evening and it worked, in that it accepted the code and didnt fall over. The program didn't actually work. But tats a different thing!
So, what can i say - I wanted to have a working Nascom 1 by the end of this and i now do. YAY me! - I honestly didn't think i would manage it,especially after the 3 months solid on the Acorn Atom board last year, but it seems i am learning stuff. Useless stuff, stuff that won't get me a better job or make me more attractive to women, but stuff.
I am KING!
18th July
So the ZX81 was put away, i thought and i thought and i couldn't get away from the power issue. I don't have a pack, so off it goes to storage; well to be placed on top of the PET anyway. The port lies adjacent, looking forlorn.
So back the the Nascom! - I soldered on the last four resisters and switched it on and no real change was visible. I tried various key-combos, but again nowt happened. SO i then decided to see if the CPU was working, were we at least at that stage, so i reset the Z80 by grounding pin 26 and bingo - !!! -
Man!!! This is so unfair!! - I had made progress!!!
FUCKFUCKFUCKFUCK!!!
Sigh - We shall see i suppose...sniff... - Its a busy week coming up with at least one late night and i am off to Aviemore for a few nights at the weekend, so i only really have a couple play nights this week. This was not how it was supposed to start. FUCKFUCKFUCK!!!
UPDATE - its working again and i am back to where i was with the system prompt! - Thank fuck!!
17th July
Well i didn't really do anything today as i frankly couldn't be arsed. The wife is 'oot on the lash' so i have been busy with domestic stuff. I did look at the ZX81 and the more i look and the more i consider the problem that presents, then it really does seem that i need more power, not volts, but amps. Looking at the ZX81 info 600milliamps is the minimum with it reaching as high as 1200milliamps when a Sinclair printer is used. I have used the Sinclair printer for the ZX81 and it worked from the 700mamp supply i have, but it does further explain why the Alphacom printer for the ZX81 i have has its own power supply. Anyway, i don't have a brick that puts out near an amp, well i have one, but its wired to the Epsom HX-20 i was playing with at xmas; its also been hacked a wee bit. So really i need to buy another, which i cant do for a day or so, perhaps not until the end of the coming week. I guess i will see. Anyway, thats what i need to progress. I did try and wire the voltage output from the ZX81 and the external power together, prior to hitting the breadboard, but it didn't work. I pause.....for a while.
I did get a rom cart thingy for the DS today in the post, sadly i don't have a micro SD card to use with it, so i couldn't try that Apple II emulator that Gavo mentioned, way way back. Still, Ebay sold me a 512mb one for £1.99 including the postage.
I am now going to have a Duval and watch some Johnny Cash bio-pic.
Later homies!
16th July
Well its been an interesting head scratcher of a night.
Wire fixed on the A3 line. Then testing. Firstly the C port of the 8255 is latched. This means (as far as i can figure) that it stays at whatever state it is in unless programmed out of it. So if i put a line high and i don't bring it low later it will stay high. I still dont have the ability to go high and then go low, but i can reset the 8255 by opening and then grounding the reset line and then setting the lines high with a poke. Sounds good huh?
Well its a progress of sorts. Clearly the port works to some degree. What doesn't work fully is my understanding of how the port works. I have spent recent times pokeing the memory address of the port, but i only discovered tonight that i need to set the mode of the port first by pokeing a separate location. Yes, yes, i know - Any novice programmer guy would have known that, well i am even less than a novice it seems. Anyway, from that i need to figure out how to properly output on the port and to be honest figure out if it happened. I need to reappraise my understanding as to what is going to happen. Should it be a single voltage change on one line and can i measure that, or do i need to amalgamate all lines and take my output reading from that?
Other problems? Well power being one, the best way to turn on the setup is by externally powering the port from a 5v supply and then switching on the ZX81, which also feeds the board with 5v and GND. Now when i do this i get approximately 0.68-0.74 volts off of any given line (D0-D7) measured. Now if you divide 5v by the 8 lines, each would be around 0.62v. So less than i am getting. So i cant trust those readings can I?
Now the power up is just that, a power up, once running i can turn off the external supply. When i do that and i read the lines i get 2.25v or thereabouts (importantly, under 2.5v) or 3.7v. I don't know about you, but it seems more like the values for a digital high and low. Now then, the problem with this, is after switching off the external supply to the board, the keyboard on the ZX81 stops working so i cant experiment further. Yes, really.
I can only assume that the ZX81 is having difficulty powering the board and itself. Would using the ZX81s 9v line be a better option with some sort of capacitor or 5v regulator in the way?
This is getting bloody complicated! - But, i am sure that the board is functional in some significant way. All i need to be able to do is set -high- and -low- and i will be sure of it. But the power issue seems to be the barrier now.
Another week at most then back to the Nascom.
15th July
So the keyboard problem!
Oh in case you are wondering, tonights tipple is a bottle of Aventinus double brew beer, German i think. 12%; tastes a lot smoother than that would suggest. Cant see anyone wanting more than one mind you. Its a dark hoppy thing, but crisp, not heavy and treacly as you would think. Anyway back to the port.
The ZX81 port that is, not the sherry-esq drink!
Ahem... ok, i will get to it. The port, i have to assume works in someway and it has an error of some sort, thats all i can say at this point. When everything is first switched on, the keyboard works fine. After about 10 seconds or thereabouts the keyboard stops working. Simply put when the board is fully powered a short or something is effecting the keyboard. But i have no idea what. All the connections look ok to me, there is noting loose and nothing seems 'wrong' or 'wronger' than it was when i wasn't getting any where.
UPDATE - even tho i hadn't posted this yet, i had half finished but went back to play again, and discovered one of the address lines has came loose A3 to be exact. Sigh. I guess i will fix that tomorrow. I don't fancy a soldering iron or a glue gun and scalpel tonight. Did i mention my beer is 12%?
13th July
Golly gosh was i tired yesterday. But today i am as fresh as Stilton, so at least had a little bit of a play, but golly gosh big dogs cock isn't this retro computers thing a trial! -I have no bananas thats for sure.
Anyway, lets start at the beginning. Port A wasnt working out for me, so i decided to try another port on the 8255. B requires buffers to work and i dont have a 7407 handy, so port C it is then. Once hooked up the first thing i noticed when running the test program is that the numbers i was getting were different. Previously port A registered 16 and port C registered 0. I assumed that 16 meant something, but i wasnt sure what, because of other testing i certainly didnt think it meant that the port was working. Now i am not so sure. I hooked up port C, ran the program and this time Port A gave 0 and Port C gave 36. Ummm, doing the same tests as with port A gave the same zilch result, but the C port works differently, in that it can work as two 4bit ports as well as a single 8bit port, so i decided to check each output data line separately. I poked to turn data line 7 high, and what should i find but 1.28v - Golly i thought! I wonder where that came from. I was then going to alter the pokes to see if i could change it, or even if a reboot would zero it or or or...
But! - In the finest tradition of the retrochallenge, as i was about to do this the numbers 1-5 and letters y-p stopped registering key presses. Give me a canine and i will show you a big dogs cock! - i mean really!! I was about to verify if i had got it to work and the keyboard borks!!
So i switch off and its the same. I play around with the power and its the same. 10 keys no longer work. I then remove the port and the keys start to work. I reattach the port and they stop working again.
I honestly have no idea - I didnt do anything. I am away to drink tea and watch TV.
FFS!!!
12th July
Well back to work today so tired- tired- tired
No ZX81 port work tonight for that reason, although i have a couple of ideas. The 8255 is a three port chip after all and i have only been working from port A, i have also noticed that the port is set to a particular mode from two specific lines from the ZX81 to the 8255. Some more head scratching needed.
Anyway - through a random email that hit my inbox, someone offered me access to their dial up Alphaserver 800. Now in the UK a free dial-up shell account is unheard off - well i didnt manage to fnd one a couple of years ago when i was trying to get net access for my Apple //c. So this was/is a treat. I dont have the time for a proper play, i have at least managed to connect to RetroNet with my Apple //c, although it borked at the password. Wish me luck!
11th July
Well i fixed the cable and i took apart the port and rebuilt it from scratch but still no joy. Nada, nowt!
I need some time for some head scratching. Anyway, today is my last day of leave so i am back to work tomorrow and with that the early starts and late finishes that get in the way of such pointless exercises as making ports for the ZX81. I wonder i wonder.
As an aside to that i may be a little tardy with the twitter thing too. Bloody work.
10th July
Well many many hours have passed and i am no further forward.
What do i know? - I am partial to a blueberry muffin and a mug of tea at midnight.
What else?
Well the port can now be attached and powered and the ZX81 doesn't crash. However it doesn't seem to work. I have powered the board from the ZX81's 5v line, a separate 5v line and running both together. The machine stays up but, as conceitedjerk may say, i have no bananas. I have measured the voltages across one data line and all data lines during various output POKEs and no change. I have PEEKed at one data line and all data lines when running 3v and 5v across the lines, and again no change. It simply seem not to work, but its addition to the ZX81 when powered does not crash the ZX81.
What else, well power across the three IC, the 7402, 7410 and the 7430 were as you would expect, in that using pin 7 for GND and measuring pin 14 gave the expected 5v, regardless of how the board was being powered. However, the 8255 which is the IC that really is the 'port' doesn't register in the way i think it should. If i try and measure 5v at the appropriate pin i get less than 1v. If i ground away from the 8255 IC then i can indeed get the 5v reading. Its curious. I am not experienced in such things but i cant figure that one out. I swapped out the 8255 with another one, but got the same results.
As far as the edge connector cable, well its mostly good. Mostly? Well, Address Line 2, which is an input into one of the NOR gates of the 7402 is a bit dodgy. But would that line if not working bring down the entire port? The output of that gate is the input to an eight input / one output NAND gate. Hmm, looking at it like that, if that failure with A2 feeds through, then actually i can see why the entire port would fail.
Ok, that was a useful discussion. Thanks for that. The question is how to fix it, the port is getting dodgy the more i much about with it. And what about the weird voltage on the 5v pin on the 8255?
Hmmm - Onward!
9th July
Well my bedroom is now decorated. Yes i know you dont care, but it was the thing standing in my way and pissing about here. So i only had a short time to do anything. What i did do was remake the edge connector cable thingy. This time i thought i would try glue rather than solder. So far it seems to have worked. Everything seems very snug compared to the solder effort the last time. I cant try things out however as i still need to label each line so i know what is what. Tomorrow that will be done and all going well the post will be attached. Time permitting of course.
7th July
Bonjour mon ami - I hope the evening finds you well?
Anyway, last night i did not in fact work as was planned, as it was cancelled shortly after i turned up. I 'worked' for 20mins, but the whole travel thing meant that i lost 3.5hrs. Cock!
Still when i got home i was tired, i meant to do something but instead i picked up a book that i had bought a few years ago and due to a shuffle of books due to redecorating the bedroom, it was prominent so i read it. A history of the UK fascist music scene. Yes really. From the beginnings of the Skin movement and the National Front up to the late 90s and the development of the BNP as the main driving force. Very interesting. Ahh i remember idiots at school sporting National Front graffiti on their school books. I remember being bullied by them too. Dumb fucks. Anyway, it was a good read. Its called 'White Noise' and its a Searchlight publication, which is the UKs main anti-fascist organisation.
Anyway - Politics aside!
Ahh but the ZX81 was upside down so i could not type a bloody program in to test it. Och! I rewired it the correct way around, but after much shoogling it still didnt work. But it did at least once. Now i knew i was going to have to rebuild the edge connector thing, so no real problems, but i at least get to do it with the idea that i may actually have a working port at the end of it. Wouldn't that be surprising!?!
6th July
Nowt from me this evening as i have worky work stuff to do. Alas!
5th JULY
4th July
Just for a laugh!
3rd July
Firstly, good stuff out there huh? I am quite liking some of the starts so far. Keep it up - (said the actress to the bishop)
Anyway tonight i continued the effort for the ZX81 port. Its a mighty fiddley job i must say. Lots of wires soldered on to an edge connector plug. At one point it looked as tho it was all going to be fine, in that all the wires were attached and continuity was fine so i covered it all in hot glue. Bingo, i thought.
I then set about making 30 little labels so i would know where each line was. During this, two of the address lines fell out of the edge connector plug. Bastard!
And with everything covered in hot glue, its a tricky one to fix up.
And i am a bit shit of course.
2nd July
BAH!!
Well i started the soldering tonight and on the one hand it went pretty well in that there were no fuck ups that were obvious, it all looked fine, well at my side at least.
So, up until this point i am used to four band resistors and the fifth being the tolernace; the ones i was supplied with were five band, but as they were all labeled on the bags it was still fine. Anyway i am installing some 2.2k resistors and i notice that they look very similar to the 220 ohm ones. I check, they are in fact the same, my 220ohm are banded the same as my 2.2k ones. Hmmm. I check, both bags, online and multimeter and both are in fact 2.2k.
Bollox! - Somebody at the store has buggered up my order, i check the rest and i have the same error in another bag. 150ohms asked for, 1.5k delivered.
Sigh...
I did add everything else to the board but i am still down four resistors. I was unable to buy them locally the last time and bought them online, for that reason. Oh well, no great surprise that it still didnt work when i tried it, for blind luck and no other reason.
So sourcing some parts is still on the agenda. I guess i will switch to the ZX81 for a wee bit then.
Or if i am really lucky i might be able to salvage them from something.
1st JULY
Well i am too sleepy to be messing around with a soldering iron so i thought i would just update to give a fuller idea as to my plans.
Up above there you can see my Nascom 1 as it currently is. As you can see it works, but only sort off. As i have said above, the board came from a heavily modded third party system, upon investigation i eventually noticed that the board is incomplete when removed from the bigger system that it was attached too. In essence the problem is the component gaps as can be seen in the photo below.
My other task for this month is to make an I/O port for the ZX81. The hardware on the ZX81 is quite basic so to make a port you need to make the hardware. Now, last year i made up a kit, and it failed to work and i couldn't figure out what the problem was, this time around i am just going to follow the schematic and build it from scratch on a breadboard.
Saying that, i dont need to build the entire thing. The full plan gave three ports to the ZX81 on completion, i only really need one, so i have cut things down and it should be easier to figure out - i hope. Anyway, the photo below is the build in its current form.
