BBC Model B Cassette interface problem

[Freya]

Trying to sort out an interesting problem with loading any cassettes on an issue3 board. The incoming signal from the tape can be traced all the way to pin 12 on IC 7.
Just to be sure IC7,IC35 and IC4 have been substituted with ones from a working board, both the BC239 test fine.
Grateful for any ideas from here, it works fine in all other aspects.

I dont think it would make a difference but IC74 and IC75 are not fitted on this board

[TonyDuell]

What exactly is it doing/not doing?

The RS423 buffers IC74, IC75 were not fitted on Model A machines IIRC. So it should work without them.

Does *MOTOR turn the cassette relay (and keyboard LED) on/off correctly. If so, the processor can access the serial ULA register.

Are you getting the correct clock signal at pin 25 of the serial ULA?

When trying to load from cassette, do you get the right clock signals at pins 26 and 18 of the serial ULA. Is there any demodulated data on pin 18

[Freya]

Yes, it turns on the relay and starts the tape and LED. I had possibly incorrectly assumed if it did this pin 25 was good.
Not checked any further so will check the pin 26 and 18 tommorow.
This board did not have its ULA fitted, I did consider if being an early board if it would only work with the Ferranti 2C199E and not the Acorn VC 2026? It also has the full Acorn Econet fitted.
Thanks

[TonyDuell]

I don't think it needs the serial master clock (pin 25) for the relay control. Just the data bus, EClock and CS/. So I'd check that next.

Has the machine ever worked with this tape recorder? It's not something like the cassette speed being way off, is it?

And what exactly does it do or not do?

[Freya]

Hi Tony, not seen this issue 3 ever read from tape, but have a fully working issue 7 machine, I'm not keen on pulling chips to test but did use the above mentioned as substitutions. The iss7 does read perfectly from cassette.

[TonyDuell]

I assume you're using the same cassette recorder and cable on both machines.

[Freya]

Yes, same for both.
I'm usually OK at the basics, it was initially non booting which turned out to be IC45

[TonyDuell]

The other thing that it might be is that the processor can't communicate with the ACIA chip IC4. It could be the slower 1MHzE clock or the ACIA chip select line that's not doing what it should.

Or maybe the processor is not getting the interrupt signal from the ACIA

[Freya]

Thanks for the suggestions, that will give me a good few aspects to check tommorow. Will keep you updated.

[Freya]

Everything checked out with the clocks going as far as the ACIA, decided to go with my earlier hunch and found a Ferranti 2C199 and hey presto it works!
So after putting it all back together decided to try the "non working" Acorn VC2026 in the Issue 7 board and that works !
My conclusion is the issue 3 board definitely needs the Ferranti 2C199E

Thanks for all the pointers helping to solve this.

[TonyDuell]

I'd not be happy leaving it like that. I'd want to know why one type of ULA works and not the other.

Looking in the service manual, there are no obvious changes to the ULA signals although there's at least one change to the cassette read amplifier.

Any ideas?

[Syrinx1]

I'm pretty sure (think it's the service manual) there's a summary of the changes between board revisions - maybe some clues that might explain this?

[TonyDuell]

I skimmed that list of changes, nothing obviously stood out. Unless it's the added capacitor in the read amplifier.

[ortek_service]

A quick glance at the Beeb's schematic does show the Cassette in & out signals (as well as the Cassette Motor Relay drive) going to the 'Serial' ULA. So not too surprising that this is required and was originally fitted to the Model A (Where you just needed to add the two serial level converter buffer IC's - plus the 5pin DIN connector - to add RS423 to it).

I've not encountered any differences between Serial ULA versions before - Although I think I've mainly only got original Ferranti ones, that were fitted to most Beeb's (inc. earlier production of the last Issue 7 board?).
But the Serial ULA is probably not used too much by current Beeb owners - unless they still want to use a tape, or want to use the Beeb as a Serial Terminal etc - So maybe not generally checked that often.

I know the Video ULA changed a bit from Ferranti to VLSI one (Only originally fitted to very-last productions of the last Issue 7 board? / sold as a raplacement part - as my Issue 7 still has the Ferranti one + the large heatsink), which ran much-cooler so no-longer required the heatsink.
And there may have been some external resistor etc value-changes between these.
But no-doubt people on the Acorn-specific Stardot forum would know.

I do have a very-good 'A Hardware Guide For The BBC Microcomputer' (1983 Wise-Owl Publications) ex- School Library book, I got for a £1, that gives some of the most-detailed info on the hardware (as well as having a glossy A2-size 'poster' of the full-circuit).

A PDF-version is here (But incomplete, only pages upto 47): https://acorn.huininga.nl/pub/docs/manuals/Wise-Owl/A%20Hard...For%20The%20BBC%20Microcomputer.pdf

And a full re-formatted HTML Website version, here:
http://bbc.nvg.org/doc/A%20Hardware%20Guide%20for%20the%20BBC%20Microcomputer/bbc_hw.htm


And this has some useful info in section 6.1 The Cassette Interface: http://bbc.nvg.org/doc/A%20Hardware%20Guide%20for%20the%20BBC%20Microcomputer/bbc_hw_06.htm#6.1

Early versions of the BBC Microcomputer had circuit board errors associated with the cassette interface. Current production boards have, however, been suitably modified in order to make the interface more reliable. Details of these errors and modifications are given below.

i) Due to manufacturing tolerances, there can be an undesirable DC offset voltage on the operational amplifiers inside IC35 (LM324). To check this, measure the voltage with respect to earth at pin 1 of this IC (The case of the UM1233 UHF modulator provides a suitable earthing point). If this voltage is greater than 100 millivolts then the DC offset is excessive, possibly causing the cassette interface to malfunction. Rather than change the IC, which is normally soldered directly to the circuit board, a cure can be effected by connecting a capacitor of value 100 nF in series with resistor R87. The capacitor to be added should be a non-polarised type of small physical size (e.g. RS part number 125-806).

If a particular machine will not load or save data onto a cassette tape which is known to be OK, then it is most likely that the cassette interface lead is faulty. This can be checked by visual inspection or by testing for continuity between the pins of the 7-pin DIN plug at the microcomputer end and the relevant pins of the connector at the cassette recorder end, using a multimeter set to the 'OHMS" range.

If the cassette lead is checked and found to be OK, it is possible that the LM324 (IC35) is faulty. This is usually caused by a cassette recorder that has either an abnormally high output level or high voltage transients superimposed on the output signal.

In most machines, IC35 is not mounted in an IC socket and hence removal. of a faulty device requires very careful desoldering This is best left to a dealer if you do not feel suitably competent. It is advisable, when replacing IC35, to mount a 14-pin IC socket (e.g. RS part number 402-305) in this IC position and fit the new device into it, thus making further replacement much simpler.

ii) The run-in tone detection circuit is incorrect on circuit boards up to and including some issue 4 boards. This can be rectified by changing the value of the resistor (P75 which, along with capacitor C28, forms a CR filter network for the run-in tone detector. R75 may either be a 100 kilohm resistor (early issue PCB) or an 82 kilohm resistor (later issue). Ideally, the value of R75 should be chosen on test as the value that gives the largest pulse on an oscilloscope at pin 15 of the serial ULA (IC7). The value of this resistor should be around 50 kilohms.

iii) Capacitor C32 forms part of the low pass filter of the cassette interface. On some issues of the main circuit board this component was left unconnected at one end, in which case the unconnected end should be connected to the junction of resistors R87 and R88 (see figure 6.1 for details).

iv) When using the microphone input of a cassette recorder, the output (pins 4 and 5 of the 7-pin DIN connector SK5) of the BBC Microcomputer will have to be attenuated. This can be accomplished by connecting a resistor, typically 100 kilohms, in series with the input to the recorder. A convenient point to mount this resistor is inside the 7-pin DIN plug or jack plug of the cassette lead.

v) The record/playback head of the cassette recorder must be aligned correctly. An audio dealer should have an azimuth alignment cassette to set this up. Alternatively, on cassette recorders which have a hole through which the head alignment can be adjusted, a rather cruder but nevertheless effective way to do this is described below:

Obtain a good quality, commercially produced, music cassette tape. Whilst playing back this tape through the recorder whic is to be aligned, adjust the set screw with a fine jeweller type screwdriver until the best high frequency response is obtained. The record/playback head should now be in reasonable alignment for use with the computer and should allow you to load commercially produced software reliably.

vi) Only use cassette tapes of the highest quality for saving computer data. Use of cheap, poor quality, tape is false economy in the long run since it will inevitably break or cause data corruption. Cheap tapes will also shed excessive amount; of their oxide coating onto the heads and rollers of your cassette recorder, requiring you to clean them more regularly.

Oxide coating deposited onto the components of the cassette recorder, particularly the heads, will load to poor playback and recording performance. It is a good idea to remove periodically, any oxide deposits -from your machine. This can be done using a proprietary head cleaning kit or, alternatively, using cotton buds moistened with isopropyl alcohol. It is important that the cleaning solvent is only used sparingly and is removed Using a clean, dry cotton bud immediately after application.

vii) The internal loudspeaker of the recorder can usually be disconnected by plugging a suitable jack plug into the external loudspeaker or earphone socket. with some types of recorder a dummy plug is provided in the accessories pack supplied with the machine for this purpose.

FIGURE 6.1 CASSETTE (ANALOGUE) CIRCUIT

[ortek_service]

Originally Posted by TonyDuell

I'd not be happy leaving it like that. I'd want to know why one type of ULA works and not the other.
Looking in the service manual, there are no obvious changes to the ULA signals although there's at least one change to the cassette read amplifier.
Any ideas?

Originally Posted by TonyDuell

I skimmed that list of changes, nothing obviously stood out. Unless it's the added capacitor in the read amplifier.

Originally Posted by ortek_service

>>
I've not encountered any differences between Serial ULA versions before - Although I think I've mainly only got original Ferranti ones, that were fitted to most Beeb's (inc. earlier production of the last Issue 7 board?).
>>
>>

Well what's probably the very-last version of the Acorn Service Manual that covers the original Beeb Model A/B
- as it also now covers (in an additional sect 2) the later model B+ (with apparently Acorn putting B+ boards in ones sold as original Model B, once they ran out of original boards),
they still only list the original Ferranti type numbers for the ULA's
(Which they mostly now refer to as the Serial Processor & Video Processor).

- A copy of the first section of this, covering the Model A/B only from Issue 1 to (last() Issue 7 boards, is here:
http://chrisacorns.computinghistory.org.uk/docs/Acorn/Manuals/Acorn_BBCSMOct85_Sec1.pdf

And there's no mention of any later Serial (or video) ULA versions, that I could see.

Although it does, usefully, cover all the circuit changes from Issue 1 to the last Issue 7 boards.


However, I have just noticed in the (original, at least - not the 'New' prefixed one which also covered the later models but maybe lost some original content) Advanced User Guide for the BBC Micro:
https://stardot.org.uk/mirrors/www.bbcdocs.com/filebase/esse...puter%20Advanced%20User%20Guide.pdf

that they do detail a 'bug' in the original Serial ULA, that was kind of resolved by changes in OS1.0 onwards (but you needed to be aware-of if accessing the IC directly):

20.10 The ‘BUG’ fix required when using the cassette system0

On early versions of the operating system, there was a bug
which led to erroneous recording of programs on cassettes. This
has been corrected on later versions of the operating system (1.0
onwards) but it is necessary to know how to fix the bug if the
cassette hardware is being used directly from a user program.

The problem occurred because it was possible for the serial
ULA to get out of sync for a few bits when the 6850 divide bits
were changed. This tended to corrupt the first character of the
first block in a SAVE, or the first character of any block during
sequential access (since the 6850 is reset for each block during
putbytes).
The cure is to write a dummy byte to tape at the start
of a SAVE and the start of every block during putbytes. If the
leader of a prerecorded program is played back. a run in tone
followed by a blip (the dummy byte) followed by more run in
tone will be heard.
It is necessary to have a run in period of high
tone after the dummy byte. Preferably this should be done by
polling the 6850 to check if the TDR is empty, since it is difficult
to accomplish if the 6850 is continually interrupting. The 6850
can then be turned on to interrupt just before starting the block
write operation.

So maybe this was fixed in the later one
- Although rather-late to do it by then, towards end of Model B production (But maybe the B+ also used the same Serial & Video ULA's, until the Master).