Revox A77 recording problem

[pentoad]

I don't think it is the input selector because input monitoring is fine on both channels of course which I forgot to mention before
1)The other thing I have noticed that this fault exists on all inputs
2) With the input levels at zero in record with tape monitor on normal listening level, channel I is very noisy - hiss and and intermittent rustling sound.

Any help would be most appreciated

[TIMTAPE]

You now say subsequent cleaning of the head has fixed channel 2. So when you said that the tape head was clean, you were wrong? Channel 1 on the head is higher up. It is both harder to see and harder to clean than channel 2. Is it possible you still haven't cleaned channel 1 enough?

Again, do you have a suitable inspection mirror to see what you are doing?

What sort of tape are you using? Certain tapes can dirty up the tape heads in seconds.

[pentoad]

Tape: BASF studio master 911

I cleaned the heads thoroughly with meths and cotton buds and inspected with a mirror and lens
I monitored a new recording, both channels I and II in tape monitoring appeared to compare very well with the input monitoring at first and I thought the problem was solved but gradually both channels tape levels dropped as the recording proceeded and the high end decreased compared to the source input. I had a look at the heads again there was a tiny bit of oxide which i cleaned off and tried another recording but this time both channels tape tevel were low and muffled from the start.

[DMcMahon]

What was the outcome of your open circuit oscillator slow start relay coil measurement, where Lawrence had commented that looked like that would result in no erase/bias signals to the erase/record heads ?

[jamesperrett]

While BASF tape doesn't usually shed much, certain tapes shed more than others and it sounds like you have a bad tape (or maybe someone has wound a different type of tape onto a BASF reel). It would be worth cleaning the heads really thoroughly (as all your problems could be caused by dirty heads) and trying a different tape.

[Ted Kendall]

Towards the end of production, most European brands used the same basic brew for coating and they all suffered some problems, although these tended to affect isolated batches. Get hold of somethng that doesn't shed to prove the point - BASF LGS35 is pretty much immune to this, and still performs well.

[pentoad]

Originally Posted by DMcMahon

What was the outcome of your open circuit oscillator slow start relay coil measurement, where Lawrence had commented that looked like that would result in no erase/bias signals to the erase/record heads ?

There was nothing wrong - I realised that I was taking a reading from the wrong points on the pcb, my mistake.

[pentoad]

Originally Posted by jamesperrett

While BASF tape doesn't usually shed much, certain tapes shed more than others and it sounds like you have a bad tape (or maybe someone has wound a different type of tape onto a BASF reel). It would be worth cleaning the heads really thoroughly (as all your problems could be caused by dirty heads) and trying a different tape.

The fault exists with the heads absolutely spotless, although tape residue did compound the probem on channel I earlier on.

[pentoad]

Originally Posted by Ted Kendall

Towards the end of production, most European brands used the same basic brew for coating and they all suffered some problems, although these tended to affect isolated batches. Get hold of somethng that doesn't shed to prove the point - BASF LGS35 is pretty much immune to this, and still performs well.

I have tried Agfa 478 and Afa PEM 469

[jamesperrett]

Originally Posted by pentoad

I have tried Agfa 478 and Afa PEM 469

I don't know 478 but my experience of PEM469 is not good. The tapes that I've attempted to transfer have left a residue on the heads after only a few seconds and the only way to get them to play has been to wrap some interface fabric around the erase head to catch as much of the resdue as possible before it reaches the play head. They do not respond to baking.

On the other hand, most of the PEM468 tapes that I've transferred have been very good.

[Chris55000]

Hi!

Faults of this nature really need an oscilloscope and a signal generator to sort out, plus (ideally) a known good machine!

If you're not got much T & M a "audio tone" signal generator app on a mobile will do, the first thing to do is to feed an audio tone of 1kHz into each radio or line input of your Studer together, adjust the record–level controls for "Zero dB" (or whichever point is indicated in your user book–work as max!), then you can use a DVM, set on a.c. volts, with a 1uF polyester cap in series with the positive lead to work thro' the recording amplifier chain against the circuits in the s.m., I think nominal signal voltages as well as d.c. operating voltages are marked on the diagrams!

If you do have a good oscilloscope available, you should get an idea of the peak–to–peak voltage at the take–off points to each record head – this will be sitting "on top and bottom of" the h.f. bias, but you can make a simple demodulator probe (one germanium diode, two 100pF 250V capacitors and a 47k resistor as per a standard a.m. demod.) to get rid of the h.f. bias so it doesn't overload your 'scope's vertical amplifier – the bias may be 100V r.m.s. or more but the audio 1V or less!

(Some service manuals suggest temporarily stopping the bias oscillator for setting–up record amplifiers or fault–tracing – this is worth while if Studer provides a convenient link, etc., for this!)

The idea with using a second known good machine is that you can repeat the test to record a length of tape as a "test tape" with a steady amplitude known signal on it, which you then play back on your Studer with the recording fault to make sure you've not getting some unexpected fault in the playback amplifier chain as well!

These machines have a great deal of interconnecting wiring, switch contacts and electronically controlled switching, and only by working thro' them meticulously against the diagrams in the service manual is the method to bring you success!

I agree about the presets, if you're limited in test gear you'll need to measure the originals (if possible! – there might be a mark on the track showing where the slider was if the thing's dropped to bits), if so, you can measure from the "cleaner" mark to each end to determine the approximate original position of the wiper!

Chris Williams

[Chris55000]

PS!

A77s have completely separate R/P heads with their own amplifier chains, so normal playback does not imply the entire recording amplifier chain from the I/P sockets, thro' the amplifier stages and all the multitudinous interconnections, plugs, switch contacts, wiring, etc., etc., as well as the Record head itself, is free from defects, so you have to go through both R & P channels as I suggest above separately, to determine exactly where your fault lies!

Chris Williams

[pentoad]

Originally Posted by Chris55000

Hi!

Faults of this nature really need an oscilloscope and a signal generator to sort out, plus (ideally) a known good machine!

If you're not got much T & M a "audio tone" signal generator app on a mobile will do, the first thing to do is to feed an audio tone of 1kHz into each radio or line input of your Studer together, adjust the record–level controls for "Zero dB" (or whichever point is indicated in your user book–work as max!), then you can use a DVM, set on a.c. volts, with a 1uF polyester cap in series with the positive lead to work thro' the recording amplifier chain against the circuits in the s.m., I think nominal signal voltages as well as d.c. operating voltages are marked on the diagrams!

If you do have a good oscilloscope available, you should get an idea of the peak–to–peak voltage at the take–off points to each record head – this will be sitting "on top and bottom of" the h.f. bias, but you can make a simple demodulator probe (one germanium diode, two 100pF 250V capacitors and a 47k resistor as per a standard a.m. demod.) to get rid of the h.f. bias so it doesn't overload your 'scope's vertical amplifier – the bias may be 100V r.m.s. or more but the audio 1V or less!

(Some service manuals suggest temporarily stopping the bias oscillator for setting–up record amplifiers or fault–tracing – this is worth while if Studer provides a convenient link, etc., for this!)

The idea with using a second known good machine is that you can repeat the test to record a length of tape as a "test tape" with a steady amplitude known signal on it, which you then play back on your Studer with the recording fault to make sure you've not getting some unexpected fault in the playback amplifier chain as well!

These machines have a great deal of interconnecting wiring, switch contacts and electronically controlled switching, and only by working thro' them meticulously against the diagrams in the service manual is the method to bring you success!

I agree about the presets, if you're limited in test gear you'll need to measure the originals (if possible! – there might be a mark on the track showing where the slider was if the thing's dropped to bits), if so, you can measure from the "cleaner" mark to each end to determine the approximate original position of the wiper!

Chris Williams

Sound advice - thanks for that

[pentoad]

Originally Posted by jamesperrett

I don't know 478 but my experience of PEM469 is not good. The tapes that I've attempted to transfer have left a residue on the heads after only a few seconds and the only way to get them to play has been to wrap some interface fabric around the erase head to catch as much of the resdue as possible before it reaches the play head. They do not respond to baking.

On the other hand, most of the PEM468 tapes that I've transferred have been very good.

What kind of material is "interface fabric" ?

[Ted Kendall]

469 does generally respond to baking, but most Agfa doesn't need it - a couple of batches were duff, but it wasn't on the Ampex scale. Note that baking times have been drifting longer over the thirty-odd years the issue has been known - initially overnight would have sufficed, but now three or four days is the minimum I use. Reels of Scotch 207 which had actually got wet needed several weeks to prevent the matt backing ripping up the mag coat.

[jamesperrett]

Originally Posted by pentoad

What kind of material is "interface fabric" ?

It is the fabric that goes inside shirt collars apparently. In the US it is known as Pelon but UK haberdashers seem to know it as interface fabric. I picked up the idea from Richard Hess who has a good document on playing different problem tapes on his website.

I must admit my immediate reaction was to bake these tapes but the ones I tried didn't respond at all to baking. There may have been an issue with storage of these tapes as I had a reel of PEM468 from the same source (a US tape library) which was also bad - the only bad reel of 468 that I've ever encountered.