Computer Cassette Recorders

[Eidolon]

or should that be Datacorders?

Yet more issues with things 'repaired' by the place by me

Essentially I had two 80's computer cassette recorders that I used to use with my Sinlair Spectrum supposedly repaired. Blatant drive belt degredation in both. I'd been using one of them as an audio cassette recorder in the attic for music on-and-off for a few years, the other hadn;t been touched. I needed to use them for recording with a microphone a couple of years back. Both were supposedly fixed and fitted with new drive belts. Otherwise they were 'fixed' on account of the new Spectrum Next, which has finally arrived years late after all kinds of production delays. And it's shown a lot of issues.

Boots Recorder - Noticeably plays slightly too slow. Becuase of this, the Spectrum cannot understand the tapes properly, and they can't load as a result. Recurring problem with the play button too, have to hold it down, it's like it's trying to auto-stop or something and I have to hold it until it 'catches' and starts playing properly. Also noticed now that the counter doesn't work anymore, he must have forgotten to connect it back up. Also, quite noticeable distortion to the audio, kind of like it's set higher than 0db, poor quality. I have a feeling this issue may actually be an inherent problem with this machine.

Radio Shack -A lot better. I can occasionally get thisto load if it's something short. Playback speed seems to be ok. This is the one I've used occasionally for music tapes. BUt it keep making faint clicking noises from the mechanism (not on the sound from the speaker) during tape playback. I did notice this before on a few things, but didn't really pay much attention to it. Now I know what it is - something wrong with the auto-stop function. When attemting to load a tape, sometimes the tape speed/playback will be disrupted during one of these clicks, and it'll stop the tape from loading. If i'm lucky i can get away with a short program loading, but on longer ones, a 'bad click' always happens at some point and casues a tape loading error. On the odd tape with a bit more tension than most, it will cause an auto-stop to happen.

Anyway, I need to get them serviced properly. I've been looking at the modern equivalent of these players, and on all account they're poorly made and not reliable.
The main issue is that I need to ensure that the players play the tapes back at the correct speed. Is there some kind of diagnostic device available (that the technician hopefully already might have), that can check for this, and confirm that it's all playing back correctly, so I can hopefully get my tapes to load? Thanks.

[SiriusHardware]

For speed verification, the standard old trick used to be to have a tape which had had a 50Hz sinewave recorded on it by a known good 'master' machine, play that in the unit under test and, using a dual-beam scope, compare the output from the tape machine on scope channel 1 with the low voltage output from a mains transformer on scope channel 2, the premise being that the mains frequency was always extremely accurate.

When the tape machine speed is correctly set the two waveforms should remain almost motionless one above the other, although in reality that sort of precision is hard to adjust for, especially in the more primitive speed control schemes which use a semi-mechanical speed adjusting method.

In these days when most workshops will have a frequency meter, you could do much the same thing by recording an (exactly) 1KHz sine wave onto a tape using a known good master machine, play that back in the unit under test and adjust the speed for 1KHz out.

[TIMTAPE]

I've never used such a system but have much experience with cassettes, cassette players and playing cassettes optimally. I used to see my brother in law using such a system in the late 70's.

I remember reading an online article from someone who went deeply into getting old cassettes to play optimally. Sorry I don't recall the link but there are probably various online articles about this issue.

As you say the speed needs to be accurate. Ideally the same speed at which the cassettes were recorded which may be somewhat different from standard cassette speed of 1.875 IPS. Possibly there is a tone built into the recorded signal which can be used as a speed reference.

Also the sound must be clear and constant with as few "dropouts" as possible, no added distortions or noises etc. I would actually listen to the signal to check its integrity. For example turning the player's volume up too high can add distortion which only masks the needed clean signals on the tape.

To that end, the original units you have may not be the best for the purpose. A more modern machine used for high quality audio would probably better. Preferably one with adjustable fine speed (pitch) control and if possible adjustable tape head azimuth to optimise the capture of the higher frequencies.

Re the cassettes, often if the sound is weak and muddy, the small felt pressure pad has become dislodged, moth eaten etc and needs to be replaced. Sometimes the actual tape inside the cassette needs to be carefully removed from a faulty or cheap outer shell and transplanted into a better shell.

The heart of it is getting a good, clean, optimal play of each tape fed to the computer in exactly the form it will accept. The computer has only limited ability to adjust to a less than ideal signal. We have to make the adjustments to bring it into the range that the computer will accept.


Hope this helps.

[Eidolon]

The tapes still all.load fine on my Spectrum 128k machine's built in tape deck which is far from top of the line (amstrad).
It's definitely an issue with the recorders entirely. I'm pretty sure the tapes were recorded at normal speed as they can technically be loaded from any tape deck as long as it's mono output.

[TIMTAPE]

Then I would use that built in deck as the reference to align your two other decks.

[DMcMahon]

Of no help to you, but just a comment. In the late 70's onwards, I used to load a lot tapes onto my expanded Nascom 1 microcomputer, using a home made modified version of the Cottis Blandford High Speed Tape Interface design.

I generally found that I got more reliable results with a medium grade Hi-Fi cassette player rather than budget models.

[SiriusHardware]

Originally Posted by TIMTAPE

Then I would use that built in deck as the reference to align your two other decks.

There is a snag, I don't think the output from the internal deck is audible in any way, unless it comes through on the TV (to my shame, I should know this as I do have a +2).

The Spectrum tape audio format does include a long steady tone, perhaps 2-3 seconds long, at the beginning of each section of recorded sound. The original purpose of this was to batter the ALC into submission before the important stuff, the actual data, started to go out, so that the recording level was established before the actual data started to flow out of the machine and onto the tape.

If you knew what the normal frequency of the lead tone was, and if you had a reasonably fast frequency meter, you could measure whether that steady lead tone coming off the tape was higher or lower in frequency than it should be. Unfortunately I didn't find anything, on a brief voyage round the 'net, that states what that frequency should be. You could find that out by capturing it as the Spectrum SAVES the code, I suppose.

[Eidolon]

I've got things to load using an Alba radio-cassette player out of the kitchen. For a number of reasons it's not practical to use on an ongoing basis but it's confirmed that the problems are not with thr tapes in any way, and it's allowed me to transfer some of them onto the Next's storage.

[ben]

Post some photos of the two recorders, and their model numbers so we can see what we're dealing with.

Slow speed may be because a replacement belt has been fitted that's too tight. Check that. There is usually a potentiometer, accessible via a hole on the back of the motor, to adjust the speed.
If speed drops near the end of the tape, and gets irregular, then clean the /replace the pinch roller.

The clicking autostop may be because there is drag on the supply spool or the auto stop 'finger' (a crude design often found in the cheapest mechs like Tanashin that can even scrape the tape in use) is bent. Or some brake is not disengaging (less likely).

To be honest, I would probably cut my losses and look for a slightly better player. Ideally you want something with a tone control/equalizer and a jack output. Cue and review is useful for locating start/end of programmes.
I would ask in the 'wanteds' on here.

[ricard]

Originally Posted by SiriusHardware

[ ... ] frequency should be. You could find that out by capturing it as the Spectrum SAVES the code, I suppose.

Yes, that's a simple yet clever method.

Incidentally , I'm not sure the tape speed has to be that accurate. The part of the firmware in the Spectrum that reads the data off the tape measures the time between edges in the input square wave, in order to determine if it is a '1' or a '0'. In normal operation, there are two frequencies used with a ratio of 2:1, and there's consequently quite a bit of margin in the firmware for detecting which is which, so you'd have to go quite far in terms of speed to get it wrong, I'd say 10% or more.

My experience from my days as a Spectrum enthusiast was that the azimuth was paramount to getting playback to work properly. I remember having one of the covers off on my recorder at all times so I could adjust the azimuth whenever needed. I found that just listening to the sound in the Spectrum internal speaker was often sufficient to find an optimum setting, turn the azimuth adjustment screw, while listening for maximum 'sharpness' of the sound.

Whereas older 8-bit computers might use frequency-shift-keying (FSK), i.e. a number of cycles of two different frequencies, to represent '0's and '1's, using a PLL to recover the signal, which then needs very little high frequency content, the Spectrum is more brute force and outputs a single square wave to represent a '0' or a '1', and relies on detecting the edges of the square wave on playback. Hence the edges have to be reasonably sharp for this to work reliably.

[TonyDuell]

I can remember being very glad when I could finally afford a disk drive for my (TRS-80 Model 1) computer. Cassette storage was not pleasant for a lot of reasons.

But a few general comments.

The most reliable cassette recorder I had for computer use was the Tandy (Radio Shack) CCR-82. A little battery-only thing (it could run off a 6V wall wart too) with 3 extra computer-related features : A preset volume setting (a click switch at one end of the volume thumbwheel travel) which was just right for computer use, a switch to turn on the internal speaker even if something was plugged into the earphone socket (so you could listen to the recording even with the computer lead in place) and a switch to override the remote control socket (so you could play the tape without unplugging that part of the computer lead). Maybe it worked so well for me because it was a Tandy product and I was using TRS80 computers but it worked well with others too. I still have it (and use it occasionally) along with the service manual. All it has ever needed was new belts.

Secondly, I never had a Spectrum 'back in the day' but I remember there was a Maplin project in an issue of 'Electronics, the Maplin Magazine' called something like 'Spectrum Easyload'. I think it was some kind of analogue filter circuit that connected between the cassette recorder and the Spectrum. I don't know if it would be any help here but it might be worth looking at. I can search my magazine shelves for it if there's interest.

Finally, Azimuth was critical. Many of my cassette recorders have a little hole in the case over the azimuth screw (when the heads are into the cassette if you see what I mean). Either official, or drilled by me. I once borrowed an alignment tape and set a cassette recorder up properly using it. Didn't do a lot of good. Many commerical program tapes were not recorded on well-aligned machines!

[Gulliver]

Hmm...yes a few % slow or fast usually wasn't a problem for the Spectrum. And often 8-bit computers favour simple shoebox machines over anything remotely "hi-fi". Though a tone control can help. In the early 80s I used a WH Smiths branded "computer cassette recorder" which was basically a generic Chinese shoebox machine with a switch for "normal" and "computer" modes....all the latter did was boost the high frequencies a bit on playback.

Generally, any shoebox machine working within acceptable parameters should be good for the Spectrum which is far less finicky than the ZX81. To be honest the most reliable cassette recorders I ever had for my 8-bit computers were a Realistic Minisette 9 which I used from 1986-1991 (it sadly ended up falling into Lake Windemere) and a 1970s very cheap no-name shoebox recorder that had a built in AM radio.

These days I successfully use a pocket cassette recorder from the early 2000s, Radio Shack branded but it was available under other names. Works fine with my Spectrum and ZX81.

I am wondering about your issues. Likely if either machine was working OK they'd be fine. There are still shoebox and pocket sized cassette recorders on the market or try your luck with getting hold of an 80s or 90s model that hasn't been used much.

[McMurdo]

I struggled for a while as a new spectrum owner (orig 48k) with my array of cassette recorders, and eventually splashed out on a Panasonic, this was like night and day and always loaded reliably. No tone control, but it always played music tapes a little 'shrill' or 'toppy' which probably helped.

When I was a lad I used to set up tape speed by recording a long note blown on my school descant recorder on our best music centre, then playing it back on the player under test and adjusting the tone for minimum beat.

[ben]

Originally Posted by McMurdo

I struggled for a while as a new spectrum owner (orig 48k) with my array of cassette recorders, and eventually splashed out on a Panasonic, .

I always detested that design of Panasonics. It was far too easy to press the little record button when you wanted playback, and unwittingly wipe the tape!

[TIMTAPE]

It would be interesting to know what was the maximum recorded frequency they used in the data signal. The higher, the faster it would record or load but the more critical for azimuth and so less reliable. Mono playback is much more susceptible to azimuth error than stereo but has a little less background noise. I'm guessing the max frequency used would be around 5 kHz. Better shoebox decks could do higher but azimuth becomes critical as do errors due to tape dropout. Even when it's manually peaked, azimuth tends to wander the higher the frequency. The maximum frequency chosen would have reflected this tradeoff.

[SiriusHardware]

We have to remember that Sinclair would have planned for these computers to be used with utterly basic, cheap mono cassette recorders because that was always his mantra when it came to his own products - price was everything, performance secondary. They were designed to be used with whatever people already had lying around the house. I don't think the highest tone frequency would have been more than a basic recorder could theoretically cope with.

I can't say I had much trouble using my ZX81 and Spectrum with any recorder that I tried to use them with, but I did usually make the azimuth adjustment 'available' by drilling a discrete hole through the cover so that I could adjust for maximum shrillness / sharpness on any tape which did give me trouble. It goes without saying, probably, that if the units in question have tone controls they should be set for maximum treble, minimum bass.

It does sound to me like Eidolon's recorders are not firing on all cylinders, the one where the autostop keeps clacking and occasionally catching and stopping the tape obviously has a bit of stiffness somewhere which is keeping the tape under too much tension, pushing on the autostop actuator.

[TIMTAPE]

Originally Posted by SiriusHardware

I don't think the highest tone frequency would have been more than a basic recorder could theoretically cope with.

My point. Even with the best recorder in mono, azimuth was and is critical.You yourself made special measures to tune azimuth. The recorder and cassette might be capable of 20 kHz but azimuth error alone will limit it to much lower. Even recording and playing back immediately on the same deck in perfect condition. From machine to machine, all bets are off.

[SiriusHardware]

It was usually the exception to the rule but then I had almost 100% original tapes, I was in one of those 'clubs' where you got a genuine bargain lot of software titles initially and then got sent the 'editor's choice' every month unless you chose something else yourself, so I ended up with quite a large collection of original software. I gave it all away (unfortunately) when I moved on to the Atari ST.

Anyone trying to load tapes which are copies of originals or worse still copies of copies is making things more difficult. You could probably make 'fresh' copies of 90% of the original titles by taking one from the likes of World Of Spectrum as a file and then using one of various utilities to convert it back to a Spectrum tape audio stream. There are only a few classic titles which don't have permission to be on WOS, notably those by Ultimate / Play The Game.

[Richard_FM]

I remember my WH Smiths Data Recorder could normally cope with most tapes for my Acord Electron.

It helped by having a meter so the volume could be set to right level for any tapes that might give trouble.

I don't remember needing to adjust the azimuth to get anything to play.

[Eidolon]

the model numbers are-

Radio Shack TRS-80 and Boots CR255

The playback speed on the Booots one is noticeably slow, and can be heard on audio cassettes too, which also have badly distorted sound. It's not as clear as to the problem on the Radio Shack player. May be more of a head alignment issue.
Both players worked fine during my 48k Spectrum days, and I seem to recall the Radio Shack one was overall a lot better. I used it for years in a car when I had my first car, which didn't have a cassette player built in.
If repaired correctly, they should both technically be fine for loading from again?

My Spectrum +2A has a built in tape deck, and is easy to move the head alignment on thanks to a well placed screwdriver hole. On that, if the heads move out of alignment more than a small amount, then tapes stop loading, that kind of hints as to the limited degree of range for loading sound accuracy.
Is there any kind of documentation on the requirements from cassette playback to get loading to work that I could forward onto the technician? Might make it easier.