Thorn DC43 deck valve mono tape recorders - bias setting and head markings queries

[dazzlevision]

Circuit extract and adjustment procedures.

[DMcMahon]

Originally Posted by dazzlevision

Can anyone shed any light on the meaning of the coloured dots and stripes shown on the photo of the rear of the record/play head fitted in these Thorn group DC43 (three speed, four track) tape deck, mono valved tape recorders?

I see on my Grundig TK 141 schematic/service sheet under the HF Bias Adjustment procedure, it states to adjust the trimmer for various voltages depending upon the head colour coding, i.e. Red 32V, White 36V, Black 40V and Yellow 44V, I guess this is to cover variations in heads due to manufacturing tolerances.

David

[TIMTAPE]

Originally Posted by DMcMahon

I see on my Grundig TK 141 schematic/service sheet under the HF Bias Adjustment procedure, it states to adjust the trimmer for various voltages depending upon the head colour coding, i.e. Red 32V, White 36V, Black 40V and Yellow 44V, I guess this is to cover variations in heads due to manufacturing tolerances.

David

Yes since oscillators werent true constant current sources the amount of metal in the laminations on the head face affected head inductance, which affected bias current. It was difficult to maintain a precise inductance value in head manufacture, hence the grading of heads sometimes.

This is why we measure actual bias current (via a series resistor) rather than mere bias voltage across the head winding.

It's also why as the head face laminations wear down and inductance decreases, the bias current can increase, leading to a tape overbias situation. I suspect that not understanding this phenomenon, some assumed the worn head (which was now making muddy sounding recordings) was useless and replaced it with a new one. Whereas assuming tape to head gap contact was still acceptable, a mere restoring of the bias current to correct value would have often fixed the muddy recording problem due to overbias.

My old Ferrograph Series 5 carried a label on the PS stating the correct voltage across the R/P head as (I think) 11 volts, but as the head wore and I relapped it a couple of times, further removing some metal, that 11 volts figure had to go progressively downwards to represent correct bias.

[DMcMahon]

That is very interesting and clear.

So the tape recorder procedures (like the TK 141 example) that only measure the bias voltage across the record head is not the best/optimum way. The TK 141 does have procedure for the head current (by measuring voltage drop across 100 Ohm in parallel with head) but this is for the record signal (with signal generator input) with the bias disabled.

David

[ms660]

1kHz input with the input and record levels set as per the manual and reduce the bias level somewhat, then increase the bias to peak at 1kHz when playing back the recorded signal, then continue to increase the bias until the recorded signal when played back is 0.5dB below that peak value, that was at Sony, in most cases that resulted in the frequency response being within spec. if it didn't then a final tweak to the adjustment would do the trick if everything else was in order.

Lawrence.

[TIMTAPE]

Yes that was a pretty standard method. A more precise method used the "overbias" level of say a 10kHz tone. Correct bias for a given tape might be when increasing bias increased 10 kHz to its maximum and then decreased it by maybe 3 or 4db.

More pro tape decks had the extra flexibility of a high EQ (pre emphasis) gain trimmer, one for each tape speed, plus adjustable bias for each tape speed. So you could optimise bias and high frequency response somewhat independently. The OP's Thorn seems limited to adjusting only the bias current. Going further seems to involve swapping components.

Bias level was always a trade off between good high and low frequency response consistent with low distortion. The discovery of AC bias in the early 1940's was a crucial moment in the development of high fidelity magnetic tape recording. Before that, magnetic recording wasnt taken very seriously as its distortions were far worse than the current disc recordings.

[dazzlevision]

Thank you Timtape, for your insight - most informative.

I have now returned the original (Ultra 6212) chassis to the workbench and made some more frequency response recordings. The 3.75 ips response is now just about in spec, although one of the "select on test" capacitors that adjust the bias feed has been removed in order to achieve this (as you say, likely due to head wear effects).

The 7.5 ips response is still a bit excessive at high frequencies, but consistent with the results obtained on my spare recorder chassis.

Test recordings from a CD at the three tape speeds sound as expected, improving with tape speed.

I think that is about as much as can be done, apart from fitting a new R/P head.

All I need to do now is try to reduce the visible flutter that I could see when playing back spot frequencies! I'll start by cleaning the motor pulley, flywheel and intermediate idler wheel.

[ms660]

With a three head machine doing a comparison at various frequencies was quick and easy, lost count of the times I did that adjustment and the rest of the adjustments, almost every machine had an "alignment/adjustment" check along with a de-mag. It was easy for us because we had all the manuals, test tapes, gauges, jigs etc and most of the common spares for all models, any spares we didn't have in stock could be gotten quickly from Japan, usually within a few days or so.

EDIT: post crossed.

Lawrence.

[dazzlevision]

Out of interest, I have made record-playback frequency response test recordings for several different types of tape, including the later "high output/low noise" formulations that appeared in the early 1970s (IIRC!).

I used the same head bias current setting (23mV across a 100 Ohm resistor) throughout and tape speed was 3.75 ips. I also cleaned the tape heads after each tape was used. The input signal level from my generator was kept constant. I used a top frequency of 13kHz, rather than 14kHz, due to the 14kHz tone causing intermodulation distortion during the recording process. So, on playback, there was an audio tone of around 1.5kHz present in addition to the 14kHz, making accurate measurement difficult.

The differences are significant and underline the Ultra 6212 instruction booklet's advice: "For the best results always use the type of tape that was supplied with the recorder".

I know that there were changes made to the treble boost (record) and bass boost (playback) circuitry during the course of production of these Thorn group DC43 deck recorders (1965 to 1969). This was because at least three different brands of tape were supplied with new recorders over this period (Scotch Boy, EMI and BASF).

Of course, it is the case that some brands of tape were not always made by the company itself. I have some Philips tapes that are marked "Made in Holland" or "Made in UK".

[DMcMahon]

Interesting results, quite a spread.

For the Bias set-up how do you determine peak recording level, to be able to set to the 20dB level below peak frequency response recording ?

David

[ms660]

Originally Posted by DMcMahon

For the Bias set-up how do you determine peak recording level, to be able to set to the 20dB level below peak frequency response recording ?

8 volts RMS between tag 21 and chassis so far as I can make out from the instructions posted, -20dB is 800mV with reference to that.

Lawrence.

[DMcMahon]

Thank you Laurence, I can see it now and calculate the 20dB reduction, previously either had missed the reference to "peak modulation" or had not associated it to be peak recording level at the record level meter calibration.

David

[dazzlevision]

Yes, that is correct, the preset pot in the meter drive circuit is adjusted so that the record level meter pointer is at the boundary of the blue and red sections (or at 0dB in the later type of meter), with a signal amplitude of 8V rms (at 1kHz) at solder tag 21 on the PCB.

When making frequency response test recordings, by the time 14kHz is reached, due to the rising HF boost characteristic of the record amplifier, the meter registers nearly 0dB.

So, if the record level was set to 0dB at 1kHz, there would be gross over modulation of the tape at 14kHz.

[ms660]

On most of the machines I worked on the feed for the record level meter circuit was taken before HF boost.

Lawrence.

[dazzlevision]

Originally Posted by ms660

On most of the machines I worked on the feed for the record level meter circuit was taken before HF boost.

Lawrence.

Yes, that is preferable, but it's not the case with this recorder (see the circuit extract in post #41).

Another related disadvantage of this recorder's circuitry is that there is leakage of HF bias into the meter drive stage. That's why the instruction booklet states the record level is set with only the record key depressed (as the bias oscillator is only powered up when both record and play keys are depressed).

This particular Ultra 6212 recorder has twice the maximum value of C11 fitted (100pF vice 50pF), which makes the effect on the record level meter more noticeable (with both record and play keys depressed). It's sometimes necessary to adjust the record level after a recording has started.

[TIMTAPE]

Actually I think meters which read the pre emphasised signal was a good idea, at least with slower speed recordings where HF saturation/distortion could be a real problem, and where the lack of real time 3rd head monitoring (via headphones in a live recording) meant you had no idea if the signal was saturating or not. At least the meter gave you a better idea of what was going on in the highs department.

It sounds like in the Thorn the metering circuit wasnt well implemented, possibly due to cost considerations. I think the Tandberg TD20a was one model which had this metering of the pre emphasised signal.

Of course the poorer/older tape stocks exacerbated the HF saturation problem which may help to explain the huge listed discrepancies in the HF playback response from older to newer stocks. It's not always understood that high frequencies effectively increased the record bias level, creating a runaway situation where the more you increased the HF record level, the weaker the HF playback. This limit would be reached more quickly with an older, lower energy tape.

To get the best out of the old original tapes dating back to the 40's the record bias needed would probably seem ridiculously low compared to the record bias even a typical 1960's tape was happy with.