Flat keys, 16K, Velcro and a ZX81.

r0jaws

Mondeo Man
Joined
Jan 23, 2008
Posts
7,237
Country
UK
Region
Lincolnshire
I have been slowly adding to my collection of retro goodies over the last few months, and one of my more recent acquisitions was a cute little ZX81.

ZX81Keyboardreplace0.jpg


My first memory of any interaction with a computer was with one of these back when I was in Primary school. I don't remember being particularly impressed, especially as the keyboard seemed awful to use, and the only programs we had, we had to type in ourselves! (Not particularly engaging for this 7 year old I'm afraid)

Still it was the first computer I had ever seen, and it did have an impact upon me and so is maybe worthy of a second chance. ;)

I picked this up from a work colleague with a big stack of tapes, a spectrum+ and a collection of joysticks and interfaces, no tape deck though. He told me he hadn't fired them up for quite a while, and they had lived in his mum's loft since they disappeared into there back in the 80's.
It was with a small amount of anticipation that I unpacked and fired them up.

The Spectrum+ was absolutely fine and was duly packed back up again until I had time to play with that some more.

The ZX81 was in a bit of a state though. Cosmetically it's not too bad but the RF output is heavily distorted. Bizarrely, the output becomes worse the more I type, this will need some investigating later on.

ZX81Keyboardreplace12.jpg


Unfortunately half the keys were also unresponsive. I knew that it was pretty easy to get hold of replacement membranes so I ordered one and waited for it to flop through my letterbox. It would be something I could be getting on with whilst I tried to find out what was going on with my Video output.

The membrane finally arrived today, so time to get busy. :D
Removing the RAM pack is a doddle. It is only held on with Velcro after all. This was a Sinclair approved and quite necessary Mod in order to hold any expansion in place!

ZX81Keyboardreplace1.jpg


Next, flip over and remove 5 screws, 3 of which are hidden underneath the rubber feet.

ZX81Keyboardreplace2.jpg


The base pops off nice and easily and you are straight at the bottom of the motherboard, with only 2 more screws to remove,

ZX81Keyboardreplace3.jpg


Carefully pulling the two keyboard ribbons free I flipped the motherboard over to have a look. A simple looking design, but then there isn't much to it.

ZX81Keyboardreplace4.jpg


Obviously, I wasn't the first person to open this and have a play. One of the ribbons had quite a substantial tear across it! (n) Oh well, the rest of it seems ok.

ZX81Keyboardreplace5.jpg


Anyhoo, the new membrane/ keyboard replacement is an almost identical match for the original in appearance and feel. I was a little worried that it would be very obviously a replacement, but you would be hard pressed to tell the difference.

ZX81Keyboardreplace6.jpg


Continuing with the rampant professionalism that is the halmark of this little beauty the keyboard is glued to the plastic casing. It can actually be removed quite easily by careful peeling.

ZX81Keyboardreplace7.jpg


One way to tell the difference between the original and replacement keyboards does require you to remove the keyboard in order to see the most distinguishing mark. A handy manufacturing date. This one was knocked up in 1980, the new ones have 2008 on them.

ZX81Keyboardreplace8.jpg


I had always wondered how these worked, and was intrigued by what it would actually look like. If like me this has been a cause of unbearable curiosity, prepare to be gratified.

ZX81Keyboardreplace9.jpg


Lovely isn't it. :D

Threading the ribbons through, pressing the new keyboard into place and reattaching them to the motherboard took literally minutes and I had the whole thing back together in less time it takes to make a cuppa. If only all repairs were as simple.

ZX81Keyboardreplace10.jpg


Firing her up and testing the keyboard out reveals that at least the keyboard function is now fully servicable! The Display, however, is still garbage though. I may investigate a composite hack, this has proven to be a vast improvement on my other speccy's but that's for another day.

ZX81Keyboardreplace11.jpg


I thought that maybe it was down to a weak PSU or maybe some bad caps, but testing the voltages on the PSU everything seems fine. I also swapped out the RF cable with one I know to be absolutely fine and made sure there was no power packs or obvious sources of possible interference in the way.

That said, it's been a good start and a simple little machine to work on. I mean 'little' quite literally, these things are much smaller than I remember. But then again, I was quite a bit smaller myself back in 1981.

The only thing I was disappointed with really, is finding out about the tear in the ribbon cable. I had believed my colleague when he said it had been untouched. I could have save myself a tenner by having a look and trimming the ribbon cable instead! According to some of the forums I have visited, this is quite a normal repair option. There is plenty of spare cable after all! Ho Hum, It's quite possible that the old keyboard contacts could have been on their way out anyway. :)
 
Good thread!

My first computer was the NEZ8000, which is a ZX81 clone: 1kb RAM, no "slow" function. Programs(?) loaded from tape, up to 48kb of maximum memory.

Piggybacking another 1k or 2k memory chips is dead easy, btw. But on its time memory chips was insanely expensive.

Ah... Those where the days...
 
I still have my original ZX81, bought way back in 1982.

Coincidentally, the RAM pack I bought to go with mine was the same as yours (although 64K rather than 16K). The anodised aluminium somehow felt more professional than the plastic of other packs. (The Velcro was a better substitute for a lump of Blu-Tack, too!)

I still have 3 original game tapes which I also bought from WH Smiths around the time. 3D Monster Maze still loads even now. My first attempt at "modding" had me soldering a push-to-make switch across 2 of the connectors on the back of the unit to act as a reset button.

(In truth, it was time wasted. All I needed to do was move the ZX81 about a quarter inch across the table and the unavoidable "RAM-Pack Wobble" would do the rest, all on its own!)

I recently picked up a ZX81 composite kit which came with the required instructions, components and a high quality composite to SCART lead. It's a tad more involved than a Spectrum composite mod but relatively easy all the same and the difference in quality is spectacular!

Well... about as spectacular as you'll get from a monochrome display! :Doh:
 
The ribbon tears are a known problem with all Sinclair keyboards.
I have several ZX81 & Spectrum units, they all came with dead keyboards which turned out to be due to tears in the ribbons.
Those ribbons just get dry & break by themselves over the years, nothing can be done to save them, and I tried :(

And BTW, as you can see, the RF modulator has a wire coming in in its side, thats a standard composite signal!
I added composite-out to one of my ZX81 machines in a matter of minutes :)
 
And BTW, as you can see, the RF modulator has a wire coming in in its side, thats a standard composite signal!
I added composite-out to one of my ZX81 machines in a matter of minutes :)


He's right you know Phil, that's how I've modded specy's in the past :nod:

TC :)
 
Found myself with a bit of time over the weekend, and with my PPC away and my Amiga lying unusable at the moment I thought I'd crack on with the composite mod for this little puppy.

I had a look around and found a schematic that shows what I need to do to hook up the Mod. Found HERE, under "Hardware projects", "simple video output for ZX81".

This uses the existing RF socket as it's output, so will leave the case looking as if nothing has been altered. (My favorite kind of mod :D)

Firstly, taking apart the case again as before to leave me access to the mobo is a simple and quick process.

Prying off the top of the RF case comes next.

CompositeHack0.jpg


On the left hand side of the RF case are 2 inputs. The one on the left is the video (composite) and the one on the right is +5v.

Desoldering the existing connection to the RF output takes another few seconds.

CompositeHack1.jpg


Next I solder the NPN Transistor into place, I used a crocodile clip between my solder joint and the transistor itself, as these are very heat sensitive.
Base went to the video input and the +5 connects to the Collector.

CompositeHack2.jpg


I did snip the existing connections to the RF.

Next I connected the Emitter to the RF socket.

CompositeHack3.jpg


Flipped the board over and with the aid of a generous blob of flux and a large amount of heat finally got a weld onto the RF case to act as Ground.

CompositeHack4.jpg


Final piece of the puzzle was to connect a 100 Ohm resistor between the Emitter and Ground. According to big Z this is to 'Pull down' the signal emitted.

CompositeHack5.jpg


Continuity tests showed that all the connections were good. Happy Days. :)

Putting her all back together again, it was time to test out the new mod. It didn't really go as I expected unfortunately. My LCD TV output a pure white screen, my son's CRT TV put out a completely black screen, and my plasma didn't react at all! :Doh:

Now I know there are problem with modern TV's and the signals sent out by old Computers, so I obviously have some research to do to find out what the problem is. Ho Hum. :coffee:

Expect some results in the near future.
 
@r0jaws

Hmmies indeed -

1. did you disconnect (cut) the resitor to the phono out ?

I dont think you have in looking at the photo, that needs to be cut -

have a lookie at this picture

speccompmod.JPG



Both RF units are identical with the Zx81 and Spectrum, the only difference is the strength of the Composite single - I am unsure what the Zx81 has to offer here as I dont have schematics - but it would be a save bet you probably wont need the transistor to boost the signal.

However saying that, Lokking at the schematic using the above in-line with the transistor should work.


speccompmod2.JPG
 
hm..

I had the same problems with my zx81, I had it inside a cupboard for years..
I got a new 14' tv (was working perfect with an old color tv) then the RF signal was
like s****.. I thought it was problem with the RF modulator, then I did the AV mod
but the signal is so weak! (sorry no pics the zx is on my homecountry now...)

chris
 
@r0jaws

sorry chap, I missed one of your earlier photo's it probably didn't load in as I skimmed the hardware pr0n -

I would seem you have certainly seperated the original in-series resistor from the phono-socket.

this indeed is perplexing - let me dig up some schematics my friend, and see if i can help =)


heres a Zx81 schematic - to be honest the mod should work sir..



Zetr0
Ponders......


I am just being a sausage - are you sure you got the Transistor the right way round my friend?
 
opps.. as I see the video (UK) is provided directly from the IC1..
thats why its a bit weak?
@r0jaws can you measure the signal?

Chris
 
Okay,
I finally got a new battery for my multimeter (damn thing uses a23's!!)
First thing I tested was the voltages. At the power input I got +11.81v, according to the spec sheet, this should be +7v to +11v DC, so a little over.
At the +5v for the RF, I got +4.98v.
At the Emitter I was getting +2.93v, and at the Base +3.60v.

I tested a spare BC 108 transistor, making sure that I was getting only one way signal between B and E, B and C and zero signal between C and E.
I tested the transistor in situ and found that I was getting signal BOTH ways between B and E and B and C. Aha! I think, Faulty transistor!

Smugly I set about carefully replacing the "duff" transistor with the known good one and as soon as it's in place it acts the same! Bugger I think, I've toasted this transistor too.
Just to be on the safe side I test the transistor I've just removed, and it acts as it should. :Doh:

So, I then de-solder the base of the transistor in-situ and placing my positive probe on the base I get a signal from the C and E respectively and when I swap the positive to the C and E and test against the Base, I get nothing.
I also get a signal when I test the positive to the C and negative to the E, but nothing when reversed.

To be absolutely, doubly sure I removed the transistor again, and using the info on The Electronics Club website connected the Base to ULA pin 16, The Collector to the +5v and the Emitter to the RF output and Ground with a 100 ohm pull down resistor.
I also tested that the resistor was functioning correctly, just to be doubly safe.

I'm still getting the same results on my Tv's. Black screen on one and White on the other. :(
Now, I am quite prepared to admit that I am a biff with electronics, and my understanding isn't terribly good, but does this all seem right?

After having a read through THIS thread, it appears someone else had a similar problem, and built a circuit to compensate. According to Andy_rea the ZX81 doesn't produce a Back Porch that the tv uses to set its internal reference to the black level in the video signal. This could explain why I'm getting a signal that is picked up by both TVs but is displayed differently (and incorrectly).

Fun Fun Fun. ;)
 
@r0jaws

I read up on the schematic and we can build it here (*have protoboard will travel*) although I dont have any LM1881N V Sync Seperators -

I am a little concerned with those voltages though - the Composite Video should only output 1.0 vpp (give / take 15% tollerance) yor emitter is thumping out 3 times that - I know TheCorfiot would be able to confirm this my friend -

One thing we can checkup is the ULA, the thread you mention discoveres the difference in ULA's is important in the circuit the newer ULA - 2C210E (date code 8303) found on the rev 3 boards and above wont need the circuit, perhaps some extra cleaning but thats about it.

The previous version of the
ULA - 2C184E (date code 8142) that is found on the rev 2 motherboards and before however have known issues with TV's

At the moment I think we need to establish

1. what rev mobo and ULA you have
2. is composite amplifaction nessessary ?

Looking at the schematic rev 3 by Andy_Rae its pretty simple, it is however one you build in situ - that is once built you fine tune it to the video display set its plugged into.

still LOTS of fun, alas this is where I should save my pocket money and buy a scope =D
 
@r0jaws

After reviwing your pics, I have just confirmed that you have a Rv 3 (or higher) motherboard

you have a ULA - 2C210E - DATE 8310 - one of the *better onese* so the ciruit mentioned is not for you, as this ULA already has the approrpiate timings for the "Back Porch"

What I think we need it get these output voltages under controll
 
I had an idea...

this might be a dirty - power rail issue -

Disconnect the 16K ram pack and try again - if you get a picture, then we may need to look at modifying the power rail to the ram pac (i.e. adding some radial *thats is if its 9vDC* condensers)

good luck =D
 
I have been doing all my tests without the RAM pack fitted just to ensure there was no "external" issues with the board.

Thanks for having a look at this buddy, it's appreciated.

Would changing the resistor have the effect of reducing the voltage?
 
indeed, looking at that voltage I would say somthing in the region of 150 ohm (in series) to bring it inline

this means an In series resistor between output of the transistor and the connection to the phono socket

at the moment try out 150ohm is an edumacated guess - I will work it out properly later but think it will need to be higher -

you could always test this by placing a resistor on the probe end and check your reading when connected to the C-video output

try and get this in the 1vpp range (1volt)
 
As Zeets has mentioned the Output needs to be 1.0V p-p. you are simply driving the Video Amplifiers within the TV's into saturation, even LCD's will attempt to buffer the composhite signal prior to A-D conversion.

An In line resistor is a good idea to reduce the output level, I'm pretty certain this is the issue here.
I'll have a look for an old Basic Video Amplifier circuit for you.
Out of curiosity have you tried driving the set directly without the "buffer", If it was man enough for the modulator to use, the set maybe OK too... but be careful not to short this "unbuffered signal".

A further tip which just occured is.........Do the ZX Spectrum schematics
show a buffer from it's ULA output prior to being fed to the PAL encoder.

Don't worry about the lack of Back porch too much, Since our friend the VCR came to life in the late 70's, TV Frame circuits were des-sensitised due to the abysmal timing output from VCR's. Just a remnant from the original spec for luminance video that old valve TV's required to synchronise when flyback periods occoured.

Hope this Helps
TC :)
 
tis my experience with the zx81 the old edge connector gets dirty just clean the edge connector and push on and pull off the rampack a couple of times it cures the problem completely

just had to do it with 2 zx81's
 
Hi Yorkie, long time no see! :D I haven't even reconnected the RAM pack yet, as I'm trying to get a basic video output and am having a bit of trouble getting that going!!

I tried a 100 ohm resistor inbetween the output and the RF connection and it made no difference to the voltage. :huh: I even tried a couple more resistors upto a 1k one, and it made no difference.

I also desoldered the whole shebang and connected the Composite output directly to the output. This gave a 3v reading and when connected to my LCD TV just produced a grey screen. I also went up and down the brightness and contrast ranges just to be safe and got nothing.

I had a look at the ULA output, and on the motherboard it appears there is no buffering between pin 16 and the output.

It's a tricky one to be sure.
 
Back
Top Bottom