• Please review our updated Terms and Rules here

Sinclair spectrum +2 gray model video issues (not transistor related, probably)

SiriusHardware said:
What specifically is the cassette mechanism doing or not doing? When you put it into play, does the take-up spindle turn slowly? When in fast-forward, does the take-up spindle spin quickly? When in rewind, does the feed spindle spin quickly?
Click to expand...

Well if I press play, FF or RW when no tape is in there it spins fine. But with a tape in there it played for about 1 second then stopped. I couldn't get it to FF or RW after that. Although it could also be a bad tape. Now that I think about it that tape also stopped in two other tape recorders I tried it in a few months ago.

I really should order some small 10 minute tapes to use on computers, trying to use 90 min tapes isn't the best.

It's funny you mention the Atari ST plug I was just pondering something. One of the things I use the Lynx monitor for is an Atari ST going into the VGA port and this monitor supports the low refresh rate of the ST. So that got me thinking maybe I can build a cable to connect the spectrum to that VGA port as well since VGA is RGB with separate sync pins.
 
That's a rare beast indeed if the VGA mode will run down as far as 60Hz-50Hz so it could be worth a try. Normally you can only get away with running an ST into a VGA monitor in mono mode, where the refresh rate is 70Hz.

You might notice though that the Spectrum + 2 has no Hsync output (why? don't know.)

I have seen suggestions that if you feed Csync (not Composite, but the TTL-Level 'Csync' signal available from another pin on the output connector) to the monitor's Hsync and feed Vsync to Monitor Vsync, then the monitor will extract the Hsync portion of the Csync signal to use as Hsync. So to clarify:

Spectrum Csync-out -> Monitor Hsync in
Spectrum Vsync-out -> Monitor Vsync in.

Of course now the 'health' of the Vsync output generated by TR5 becomes important. I should stress that I have not actually tried this.

The tape you are having trouble with, it sounds as though the auto-stop is activating because it thinks the tape has come to an end. Can you try sticking a pencil through one of the tape spools and winding the tape by hand? It should be very easy to turn, if if it feels very stiff / needs force to turn it then the tape is the problem.
 
Couple of points I forgot to mention: Just try any other cassette tape in there and see if it runs back and forwards and plays. A prerecorded music cassette will do - any one. The Spectrum has very eclectic musical tastes.

Also: Cassettes from back in the day sometimes develop 'bad pad' syndrome, where the dry spring loaded felt pad which holds the tape onto the head turns into a kind of sticky glue. Make sure this has not happened on your suspect tape.
 
I'll have to try that with the monitor but I'll have to order a din cable so won't be anytime soon.
As for the tape it seems to spin freely using a bic pen and the pad seems fine. It also played, recorded, and rewound just fine in the tape deck I have hooked up to my PC which I used to record a spectrum tap onto it.
 
When the tape stops in the Spectrum, is the motor still spinning? If it is, is the belt still moving and the flywheel (the big grey wheel at the other end of the belt) turning?

If you find that the motor is spinning but the belt is not moving that would show that the belt has stretched and become too slack, so the motor pulley just spins round inside it without moving the belt.
 
well I put it on the bench to see what's up with the tape and it seemed to be working fine. So I just put it back together, plugged it into the monitor, and tried to load up the game. It worked. the tape mechanism is kind of loud and could maybe use some lubricant somewhere but otherwise after 5 minutes of loading the game came up.

I'm not sure why it stopped yesterday.
 
All of this has finally pushed me into finding mine and getting it out of the box. First thing I notice is that the belt is looking very weak and stringy - mine is definitely the original belt so that will need to be sorted.

Getting down to the more interesting part, the transistors:-

My PCB is also Issue 3.

All of my TR4, TR5 , TR7 are 2N3904s

If we start with the premise that the majority of the pin outs for the 2N3094 have it correct (e, b, c, with the legs pointing down and the flat face towards us)...

My TR4 is soldered as per the screen print and its collector goes to composite-out via a 75R and LK4. This is INCORRECT, should be the emitter which goes to composite-out so TR4 needs turning around like yours.

My TR5 is soldered as per the screen printed outline - emitter goes to +5V, this is INCORRECT, the collector should go to +5V so it needs turning around.

My TR7 is soldered as per the screen printed outline - emitter goes to +5V, this is INCORRECT, the collector should go to +5V so it needs turning around.

So at the beginning of this thread when you were insisting that some of the transistors were probably the wrong way around and I was saying they can't be, mine has never been touched and has worked for 30+ years, you were right and I was wrong. Thanks for bringing this issue to my attention. It really is astonishing that it works as well as it does with three transistors in the wrong way around and I have no idea how it got through QA all those years ago. Thanks for starting this thread. As often happens, I have ended up learning something new from it.

My best de-soldering iron is at work and I won't be back there until Monday so the turning of the transistors will have to wait until then but in the meantime I will see if I can pick up an 8-pin DIN plug tomorrow and make a start on rebuilding the RGB SCART lead for this machine.
 
Yeah it's been a fun project so far and I've learned a lot as well. I made an Atari joystick adapter cable today using an Atari joystick extension cable and rewiring it.

I'm not 100% sure if my TR5 or 7 are in correctly or not so I plan to take a closer look at that. Although I'm not sure what effect I might see one way or the other, maybe slightly cleaner composite video?

Then there's servicing the tape unit, adding external phono jacks for using some sort of external data playback/record. And finally seeing if I can get rgb video through vga.

Somewhere in all that find some fun games to play on it as I've never really used a specy before so it's all new to me.
 
That was a wearisome tactic by Amstrad, trying to force people to buy only the 'official' joysticks by making the wiring slightly non standard. Thanks for reminding me about that. It was only a matter of seconds before someone produced Atari-to-Plus 2 joystick converter leads with a black socket at one end and a grey plug at the other end and I have here a 'Cheetah 125+' - It has two plugs on the end of the lead, one black (conventional Atari wiring) and one grey (Spectrum +2 specific wiring).

I was looking at the 128 +2 last night and thinking that I had never really used it to its full extent although I did play with some of its specific features like the RS232 / MIDI port. As happened with other 'enhanced' versions of mainstream machines there was very little effort by mainstream manufacturers to support the specific extra capabilities of the 128 - they just kept on churning out 48K titles because they could sell those to everyone with any kind of Spectrum including the 128. To be fair there were a a few titles for which '128K enhanced' versions were produced, mainly making use of the three channel sound chip which the standard Spectrum did not have.

After a love affair which started with the Single-Board MK14 'computer' and lasted until about a year after I bought the grey 128 +2 I finally broke up with Sinclair, having been seduced away from them by the comparatively awesome graphics power of the Atari ST.

There is a huge back-catalogue of Sinclair software, most of it available on the excellent World Of Spectrum website. For the 128 in particular, while there was not a huge amount of 128 specific software produced there has always been a healthy 'demo scene' around the Spectrum and you may find some decent 128 demos in TAP or TZX format which will show you what the machine can really do.

As to the transistors, TR5 definitely only affects the Vsync output to the video-out connector. TR7's role is less clear cut, being involved in the IORQ input to the main ULA which among other things generates the video signals. Even if we don't need the Vsync function right now and we aren't entirely sure what TR7 does, it makes sense to ensure that they are oriented as the diagram says they should be.

While I was looking at the tape unit last night I realised that there is a second, small belt in there as well. No doubt somebody sells a 'belt kit' for them.
 
It sounds a bit like the story of the Commodore 128, most of the software being developed during the production/usage of the c128 was written for a c64 and just labeled as "C64/C128 (in 64 mode)" Which is a shame because most of the 128's enhanced abilities went unused.

I was wondering about the spectrum's RS232 port, as it's not a standard 9 pin or 25 pin serial connector I was wondering what I'd need to be able to make use it, maybe plug in a wifi "modem".

I noticed the two belts in the tape unit, in mine they seem to be fine. I might have actually replaced those when I got the computer a year or so ago I don't remember now. It's just noisy when it runs so I think I'll put some silicon grease where appropriate and see if the motor has a hole for dropping some machine oil in.

I noticed that using tape as opposed to floppy disks went on much longer in the UK than in the US. How did people (or did people) use these computers for things like databases? Seems that would be very difficult without random file access.
 
Last edited:
Hmm so today I took a closer look at tr5 and 7. 5 was backwards so reversed it but 7 is actually correct as installed. Strange.
I found out the tape unit runs quietly when not installed in the case but when it is it's quite a loud rumble so now I'm thinking something is rubbing up against the case somewhere.
I was able to load 2 games (only tried 2 so far), "V" and "Turbo the Tortoise"
Both work but "V" doesn't seem to support joystick.
Turbo works with joystick but neither game is outputting any sound.
I do get sound output while the tape is loading so I know the audio jack, cable, and input into the monitor are working.
I'll maybe see how to make a simple basic program to generate sound.
 
A quick Google search on no sound people talk about tr7, so eve though is appears to tone out as the schematic indicates I'm going to try reversing it anyway. If that doesn't help I may have a bad transistor altogether.
 
The tape mechanism may just sound louder in the case because it is using the whole casing as a sounding board. I've ordered a set of belts for mine.

Joysticks: The Spectrum (16K / 48K) didn't originally have or support a joystick so this was left up to a third party, Kempston, to come up with an add on interface and it was quite successful so a lot of the early games which support joysticks ONLY support Kempston. It was a a while later that Sinclair launched the 'Interface 2' which was a twin joystick and cartridge port - of course they didn't make it compatible with the Kempston 'standard' so pretty soon after that games were having to support both the Kempston and Sinclair standard for joysticks and offer that choice in the game's start menu. Fast forward to the 128 +2 and the computers are still using the 'Sinclair' standard for input from the ports so that is what you have to select in-game if you want to use the onboard ports, only on the 128 +2 there is also the additional problem that the joystick ports have been intentionally rewired to be non standard so you then need to use a converter cable to allow standard Atari-wired sticks to be used. If you don't specifically choose to use a joystick in the initial game menus the game will assume you are using keys by default even when there is a joystick connected.

Assuming your game lets you choose 'Sinclair' joysticks, a single player game will usually expect you to have the joystick in joystick port 1, not joystick port 2. Port one is counter-intuitively the right-hand port of the two ports on the side, anyone with a normally wired brain would expect that the leftmost one would be port 1 and the rightmost one would be port 2.

The good (?) thing about the Sinclair joystick standard is that it maps the joystick inputs to the keyboard numeric keys so if you start the machine up and go into (say) 128K mode at the BASIC entry point and move the joystick around, each action up, down, left, right, fire, should 'Type' a numeric character as though you are pressing a numeric key on the keyboard. If even that doesn't work then you may need to go over your joystick converter wiring again.

Not sure I can explain your sound problem at the moment, but bearing in mind that my TR4, TR5, TR7 were the wrong way around the whole time I was using the machine originally, I definitely remember the sound working, I had several games which specifically supported the three channel sound chip of the 128.

A quick sound test: The BASIC syntax for making simple beep sounds is:-

BEEP duration, note

For example,

BEEP 1,0
 
Last edited:
My joystick is working but I get what you're saying that some games may have been designed to work with Kempston wiring.

With your audio out, that is also coming out of the phono jack right? It's not on the rgb port and a part of the scart connector?
 
some games may have been designed to work with Kempston wiring
Click to expand...

It's not just wiring, the Kempston standard arranges for the data to be read from the stick using the IN 31 command. This does not work for Sinclair's ports. If you have a game which only has Kempston joystick support and not Sinclair joystick support the game will not be able to see the Sinclair ports at all. Only if the game has a specific Sinclair Joystick option (and is set to that) will it try to read / use the internal ports of the +2. Kempston defined their standard and then Sinclair decided not to follow it, so the code needed to read a Kempston stick and the code needed to read either of the Sinclair ports is entirely incompatible.

The 'digital' sound coming out of the ULA 'MK' pin and the 'analogue' sound coming out of the General Instruments sound chip and the audio coming from the tape deck are all mixed together shortly before the sound output jack and then they are fed to it and there is no kind of gate or switch involved so sound coming out of the 'Sound Out' socket is always 'on'. So, if any one of these sources is producing sound the sound should come out of the 'sound out' socket. Question: If you have a 3.5mm jack plug plugged into the 'Sound Out' socket, is it a mono plug or a stereo plug? It should be a mono plug.

If the PCB links are set in 'S' (Standard) mode then the combined sound from those sources will NOT come out of any pin on the video-out socket. With the links in the 'S' position, normally the 'Bright' control signal comes out of pin 3 on the output socket but if you remove LK7 and fit LK8 then the sound, instead of the 'bright' signal, will come out on pin 3 of the video port - if that is what you actually want.

The computer uses the same 'MK' pin to both generate the original 'simple' Spectrum sound - (did you try the 'BEEP' command as mentioned above?) and to generate the squealy sound of a program being saved to tape. Write a short BASIC program (maybe just a really long REM statement full of garbage) and save it. Can you hear the sound of the program being saved?
 
Sorry, I missed your question about mass storage and the serial interface.

From the beginning Sinclair Spectrum owners had the option of buying the 'official' mass storage solution which comprised the 'Interface 1' and at least one 'Microdrive'. The medium used in the Microdrive looked like (and was) exactly like a miniature 8-track audio cartridge and it had the same continuous tape loop. From the user point of view it worked exactly like a rather slow floppy drive and you could save a named file and reload it in exactly the same way. Sinclair themselves never provided a real floppy drive for the Spectrum but a couple of third party suppliers did, they were super expensive but that is what you would have bought if you were a 'power user' using your Spectrum for serious purposes. Amstrad eventually provided an actual floppy drive in the Spectrum +3 but it was the less common 3" type as used in the Amstrad CPC 6128, Tatung Einstein and others.

Floppy drives were in the main just too expensive for UK users to be able to afford them and they were not provided as standard with any of that generation of low end machines but they were available, the C64 had its official 1541 option and the BBC model B had an option to use a more or less standard 5.25" drive, athough you had to retrofit the FDD controller IC and the additional EPROM which provided the disk commands before you could attach a drive. Not until the arrival of the 16-bit machines (Atari ST, Commodore Amiga) did we have a situation where everybody's computer had a built-in FDD.

Serial interface: Although the 'RS232 connector' is non standard the signal levels in-and-out are standard RS232 levels thanks to the 1488 and 1489 level converter ICs - but - the signal lines don't go to a dedicated UART IC, they just go to standard I/O port lines on the AY-8910 sound / parallel I/O chip and the serial RS232 comms are bit-banged by software routines in the Spectrum ROM. This means that the RS232 port basically has no buffering. It's fine for output-only applications like printing to an Epson or similar printer with a serial connection port, but for anything which required fast exchanges of data both in and out, it would probably fall over. If you wrote your own handling routines in assembly language you might be able to run it fast enough to handle proper comms. If you really wanted to do proper comms to and from a modem you would have been better off building yourself a real serial interface to attach to the rear slot on the computer, based on a 6402 or Z80-SIO or Z80-DART or similar UART chip.
 
SiriusHardware said:
(did you try the 'BEEP' command as mentioned above?) and to generate the squealy sound of a program being saved to tape. Write a short BASIC program (maybe just a really long REM statement full of garbage) and save it. Can you hear the sound of the program being saved?
Click to expand...
Yeah I did and it sounds fine. Then I loaded up Turbo the Tortise again and cranked up the volume. There is sound there but it's extremely low.

I am using a stereo 3.5mm plug though so I'll try a mono one. Strange that the tape audio is loud as well as the BEEP command. Maybe it's just how this game is. I'll have to experiment with more software.
 
I don't know the TTT game so I wonder if it is using the AY- sound chip? I have seen a mod somewhere which aims to equalise the sound output level from the AY chip with the sound output level from the ULA because by default the 'Beep' sounds generated by the digital output on the ULA are apparently significantly louder than the ones from the Analogue AY chip. It is possible to rebalance that by changing a component value or two. I have to admit I never noticed this problem back in the day but I probably just turned the volume up if a game was using AY sound and back down if it was using digital sound (original Spectrum sound is virtually unlistenable, to be honest).

As you have an ST you no doubt know that the ST demo crews have taken the very bland YM Audio chip in there and tortured it to the point where it can make sounds which hardly seem possible, often approaching what could be done with the C64's famous SID chip. The AY chip in the Spectrum 128 is virtually the same chip as the ST's YM chip and in the past few days I have been looking through a lot of Spectrum demos on Youtube and some of them have fantastic soundtracks. These are techniques which have been developed long after the original heyday of the Spectrum so the sound in commercially produced game software dating back to the 1980s is not in the same league as some of the more recent work done by demo crews.

The reason I queried stereo / mono is that the 3.5mm mono sockets used by Sinclair were a little bit odd, the contact closest to the entry hole on the socket is set quite a long way in - on a mono plug it will still land on the long GND portion of the plug, but if you plug a stereo plug in there then the GND contact of the socket will often land on the middle ring of the plug rather than the GND section closest to the body of the plug.

If you only have a stereo plug you could try connecting the middle section of the plug to the GND section of the plug, so that the section near the body of the plug and the section in the middle of the plug are both GND. Or for an easier life, just swap a mono plug in instead.
 
Well I did find one thing out, I don't need to mod the computer in order to load wav files in from a PC (using a toshiba libretto). I just needed to use one of those cheap cassette adapters they make for cars. I've had one sitting in my closet for years because the sound quality is terrible but apparently the speccy is happy with it. That opens up being able to test out lots of games without having to use up tapes or taking time to record them.
 
I've heard of this being done but assumed you would not have one of those tape gadgets handy. Good that you have found a use for it.
 
You must log in or register to reply here.
Back
Top