Setting bias on Tandberg 12

[mitajohn]

Well I cannot say but if the head is worn, you can simply look at it. It will normally have a little curve at the centre. If it's worn, the face will have a flat spot on it. Use also a magnified glass.

[ricard]

Usually when the gap has begun to widen it doesn't do it uniformly, so you can see that the gap is wider at one end than at the other. Same thing between tracks.

Hard to see in the picture because of reflections but it does look like there's some residue that could be cleaned off which might account for bad tape-to-head contact. The gap looks ok though.

I wonder if the the weak bias for the tape in question would also account for the 10 kHz treble drop? Or it it could be as John suggests that the equalizer coil is maladjusted, so that the resonant frequency for the equalizer circuitry is lower than it should be. It should be fairly easy to adjust given the procedure in the service manual and given that the OP has access to some form of tone generating equipment. Not sure how easy it is to access the coil though, it would be on the main PCB which is awkward if the coil can only be adjusted from the component side.

The whole purpose of that part of the equalizer is to give a bit of a boost at 18 kHz (at 7 1/2 ips, lower for lower frequencies) to stretch out the high frequency range of the frequency response before it dives down. So if it is adjusted to, say, 10 kHz, that would account for the final peak being at this frequency.

[TIMTAPE]

Originally Posted by mitajohn

... had you set MPX filter OFF?

This is a good point. Perhaps this is the reason for the sharp drop off above 10kHz.

[Eichberger]

The filter was off. I even tested with it switched on just to check that it hadn't been turned round at some point. The only possibility here would be that the switch has malfunctioned in some way. The MPX filter is supposed to take out a notch at 19kHz so I don't imagine it would go down to 10kHz would it?

Originally Posted by ricard

Hard to see in the picture because of reflections but it does look like there's some residue that could be cleaned off which might account for bad tape-to-head contact. The gap looks ok though.

I had cleaned it a couple of weeks ago but I will have another go at it in case I missed a bit.

I wonder if the the weak bias for the tape in question would also account for the 10 kHz treble drop? Or it it could be as John suggests that the equalizer coil is maladjusted, so that the resonant frequency for the equalizer circuitry is lower than it should be. It should be fairly easy to adjust given the procedure in the service manual and given that the OP has access to some form of tone generating equipment. Not sure how easy it is to access the coil though, it would be on the main PCB which is awkward if the coil can only be adjusted from the component side.

I have some LP35 on its way. Certainly will look into those coils.

Thanks again guys for all of your suggestions - I really appreciate you all taking the time to help me out.

[Eichberger]

I just ran white noise through the machine without recording. The signal still goes through many of the same circuits. The plot shows the expected frequency response for a signal is going via the internal amp.

I also tried switching the MPX filter in and could see the little notch appearing in the curve at 19kHz.

So remaining culprits are:

- Wrong tape type.
- The coils in the equalisation circuit. I will do some more testing at different speeds for this as the frequency boost will change at different speed.
- Faulty/worn/dirty record/playback head.

Could magnetisation in the tape path be the cause of this problem?

[mitajohn]

Well done!

So remaining culprits are:

- Wrong tape type.
YES, I read somewhere that the r2r was adjusted via an EMI tape but I don't remember where.

- The coils in the equalization circuit. May be.

- Faulty/worn/dirty record/playback head. YES

Could magnetization in the tape path be the cause of this problem? YES may be a little bit but mainly the head.

For such an old r2r you MUST replace ALL Electrolytic capacitors and check from scratch.

[ricard]

If some part of the tape path is magnetized, a recording will not only exhibit only a noticeable decrease in treble, but also a characteristic rumbling background noise, quite different from the fairly uniform tape hiss which is always present to some degree.

I disagree with John regarding the electrolytic capacitors. My experience has been that electrolytics seldom deteriorate to the point of causing problems in old equipment. (Paper capacitors in valve equipment are a different story, and many makes can essentially be replaced on sight.)

[Eichberger]

This is interesting. These are three plots of the frequency response at the three tape speeds. All show a bump just beyond where the curve drops down and a bit above the maximum frequency response for the respective tape speed. I suspect that this is the HF boost provided by the equalisation circuit. This suggests the coil does not need to be adjusted.

From the manual they should be at:

6.5khz for 1 7/8 ips
12kkhz for 3.75 ips
17khz for 7.5 ips

[TIMTAPE]

Yes as per fig. 43 in the manual. That's if everything else is working properly and we are using a tape for which the machine was designed.

However first step in setting up a tape machine is achieving correct playback response as per 5.3 on page 34 of the manual. Mechanical things like head condition and alignment are really important in this respect.
Playback response must conform to the objective standard. Only then do we set up the record parameters according to the tape machine's own (now correct) playback response.
As per the manual the overall response at 7.5 ips should be 40 - 16000 hz +or- 2db. If all was well with the machine set up for original tape, but we were just recording with a modern tape like the SM911, yes we'd expect a boosted response in the high treble but not the sharp dropoff above 10000Hz. Your plot shows 15000Hz is 30db down which cant be right. The question is why.

Again before any recording is made it's important to go back to basics and make sure playback response is correct right out to the highest required frequencies which in this case is around 16000Hz. Ideally we use a calibrated test tape.
In the absence of a calibrated test tape I sometimes make a basic test of playback response by playing back a blank (bulk demagnetised) tape. On the FFT analyzer we should notice (above the playback preamp's own noise) a healthy spectrum of noise extending out to 15000Hz or more at 7.5ips, which of course drops when we pause the tape. If not, and the response only extends to around 10000 Hz or less, that needs attention and it's probably to do with the head and its contact with the tape.

[mitajohn]

I have a question to make: Are these measurements presented correct?

I remember that in audio applications, one of the best ways is to use pink noise as a reference tone to check frequency responses and becomes particularly useful when coupled with a 1/3 octave spectrum analyzer.

In the absence of a sweep audio generator why don't we use pink noise?

[TIMTAPE]

I take it he's already doing that.

[TIMTAPE]

Just in case it wasnt clear, the test procedure I suggested uses tape hiss as a signal source. To confirm, yesterday I got out one of my Tandbergs and fed the output into an audio programme with a FFT Spectrum Analyzer.
The tape noise/hiss was quite apparent on the screen, and extended up to 20kHz at least.

[Eichberger]

Thanks for the suggestion TIMTAPE. Hopefully will have a chance to try that later.

[mitajohn]

May be this is helpful:

https://www.youtube.com/watch?v=G8t4zi_uEL0

[Eichberger]

I couldn't find any erased sections of tape but I do have a couple of old tapes recorded by someone else on a different machine. They mostly have classical music on with some speech and one section of (backwards) pop. All recorded at 3.75 ips.

These are plots from playing them at 7.5 ips to shift the frequency range up. All three and in particular the 'modern' music show considerable energy in the 10kHz - 15kHz region. Where my recordings plummeted 30db in that range these only drop 10db or so and this drop could well be caused by the limitations of the source material.

So it looks like my problem is with recording.

[TIMTAPE]

We cant expect "as new" performance especially in the extreme high frequencies unless all is in good condition and set up well. I wouldnt even attempt to optimise record bias unless I first checked/rectified the mechanical head and tape contact issues. Again my guess is there is poor tape to head contact. Possibly the head face needs relapping.

The beauty of using random tape hiss as a high frequency source is that it's independent of head azimuth angle. And you're almost guaranteed extended high frequencies. But it still requires excellent tape-to-head contact...

It used to be said the tape head is the "heart" of a tape machine. If we want to extract the maximum performance out of the machine and tape, that's where we need to focus as the first step at least.

[ricard]

Originally Posted by TIMTAPE

Just in case it wasnt clear, the test procedure I suggested uses tape hiss as a signal source. To confirm, yesterday I got out one of my Tandbergs and fed the output into an audio programme with a FFT Spectrum Analyzer.
The tape noise/hiss was quite apparent on the screen, and extended up to 20kHz at least.

I'd assume that you could make a tape on the recorder you're going to test it on, by putting a piece of paper in front of the record/playback head and setting the machine to record mode. That way you'll erase the tape but there will nothing at all actually recorded on it, not even the hiss and noise from the record amplifier.

I must say I like the tape hiss method now you've explained it, it's clever and simple.

That said, the plots above show that the playback seems fine up to 15 kHz, as it should be, so as the OP has concluded, there must be a recording issue of some sort, and possibly not tape-to-head contact either.

[TIMTAPE]

Thanks Ricard. Yes making a "recording" with just with the erase head sounds like a good compromise in the absence of a bulk demagnetiser.


I find the last spectrum shots ambiguous. There might be "some" recorded energy at 15kHz but it still might be a long way below what it should be. Before anything else need some way of confirming the playback response is essentially flat up to the highest frequencies. As we know, even a speck of dirt on the head face is enough to drastically reduce the highs.

[ricard]

I did think though that the plots show a more uniform response in the 10-20 kHz area than the previous record-play plots which show a marked peak at 10 kHz, which hints at what John was getting at with the EQ adjustment. But observing the hiss of a (bulk) erased tape should provide a more definitive answer. In the absence of calibrated test tapes it's hard verify accurately. One way would be to compare the frequency response with a known good recorder.

Otherwise, verifying the equalization adjustment in the recorder should be a fairly easy task; the adjustment affects both playback and record (but the latter more than the former as seen by the plots in the service manual), and could be done by setting up a (slow) sweep of 10 to 20 kHz in a signal generator (of whatever sort), setting the machine set to 7 1/2 ips in record mode, observing at which frequency (the target being 18 kHz) the peak occurs by looking at the deflection on the record indicators.

[TIMTAPE]

Yes and I suspect even easier given he has a digital spectrum analyzer. When I played back the bulk erased tape the different peaks at the three speeds were visible, and were even visible when it was just pre amp noise with no tape in motion. Spectrum analyzers are a very handy tool.