magnetic tape erase/record circuit question

[AviZiv]

Hi folks - anyone here with knowledge of tape recorder circuitry?  I know that  tape recorders employ an AC bias oscillator that is mixed in with the incoming signal on it's way to the record head. The bias oscillator also goes to the erase head. Now - I would like to modify a tape deck and add a switch that would disconnect the erase head, while recording, for certain applications. This happens to be in a Marantz pmd-222 cassette deck but the question is more general.

Since the bias oscillator goes to both the erase and record heads while recording, will the recording quality be affected if I disconnect the erase head with a switch? I'm looking for a way to cut off the erase function while maintaining quality.  ( yes, I do realize that I will get sound-on-sound when I do that)

If the two heads are designed to always be there as a reqired inductive load, then I may have to think of another way. For example, I could get another erase head and place it in another part of the machine (shielded) and away from the tape and simply switch between the two heads to maintain the same effective circuit. But is this necessary? Any other ideas? Can I switch over to an inductor of the same "value" as what's in the erase head but without the magnet?

Of course I will probably experiment, but thought I'd see if anyone has experience in this already

Thanks!
Avi

[AviZiv]

I'm attaching the portion of the schematic that shows the bias circuit and its connections to the erase and record heads. Note the transformer and its different windings that go to the two heads.

Hope this helps

Thanks,
Avi

[flynwill]

My guess is you will be just fine with the switch.  Do you have a 'scope?  If so you can check to see if the bias level (as measured at CL08) changed significantly with the erase head disconnected.

Not sure what you are expecting as the final result, the "over recording" will likely erase a lot of whatever was on the track prior.

[AviZiv]

Thanks for the response, flynwill! I do have a scope and you're right - it will be interesting to compare the bias signal at the record head with the erase head in and out of the circuit. 

As for the purpose of this - I know it seems odd.... But besides being an engineer, I'm also a sound designer and musician. I'm going to be recording some loops on endless tapes (home made). In normal operation, you cannot complete a seamless loop without a silent gap because of the distance between the erase and record heads. By eliminating the erase function, I can complete the loop. I can do it already by putting masking tape over the erase head but I'd like the convenience of a switch. There are many ways artists want to use machines even if they were not intended. Sound-on-sound is another application and experiments show that you can stack sounds this way. If fidelity degrades it's not necessarily a problem - much of the degradation is embraced if you can believe it or not. There is an entire industry dedicated to distortion We now consider tapes recorders more as creative instruments and not so much as high fidelity sound reproduction machines.

Thanks again

Avi

[DBecker]

I can understand the desire for a gapless loop tape.

But I don't expect that recording over un-erased tape will be satisfactory.  The erase head is broader than the record head in order to erase the entire track.  Without that broadly erased track, any slight misalignment is going to change the bleed-through of the previous recordings.

[schmitt trigger]

The "erase disable" switch, should be a SPDT one, such that when the erase head is disconnected, the switch inserts a dummy load.
That way, bias fluctuations may be lessened when the erase head is not in service.

You may have to find the actual dummy load resistor empirically.

[AviZiv]

The "erase disable" switch, should be a SPDT one, such that when the erase head is disconnected, the switch inserts a dummy load.
That way, bias fluctuations may be lessened when the erase head is not in service.

You may have to find the actual dummy load resistor empirically.

This is exactly why I asked the question - thanks for chiming in. Do you think that a resistor can act as an appropriate dummy load? I thought in the worst case I would put an entire erase head as the dummy load (far from the tape) but maybe that's overkill. I can't seem to visualize in my head the effect of just cutting out the erase head especially considering the transformer that is in the schematic. Maybe some real testing would show me enough.

Thanks!
Avi

[AviZiv]

If you can measure the frequency before and after disconnecting the erase head you'd see if there are loading effects that you need to address, if there are and you have an LCR meter you would be able to calculate the impedance at the bias oscillator frequency and select a suitable resistor.

It occurs to me it might make for some interesting effects if you could modulate the bias oscillator frequency and/or amplitude

You just opened up another porthole into the unknown! I love it. cv-controlled tape bias !?!?! why not.

Thanks for the tip regarding finding the right resistor too. Something to work on. In the very least I should be able to hook up a scope and look at the bias freq with/without the erase head in the circuit.

What a great community - thank you folks!

Avi

[TurboTom]

I tend to agree with CJay. The effect on bias frequency will be the predominant one when disconnecting the erase head. The inductance of this head, in parallel with the tank coil (secondary of LL01) forms a parallel resonance with CL07 which defines the frequency of the oscillator. If the inductance of the erase head is large compared to LL01, the effect may not be that pronounced. Amplitude on the other hand won't be affected much since the circuitry below the oscillator itself uses the rectified oscillator signal via DL01 to control the emitter current of the oscillator circuit to stabilize the amplitude to the desired value (depending on tape type selection and the input via transistor QL31.

If you've got an LCR meter, you may want to check the inductance of the erase head and find a suitable "dummy inductor" to switch into the circuit when disabling the head. Or just disconnect it a see what happens... 

[AviZiv]

I tend to agree with CJay. The effect on bias frequency will be the predominant one when disconnecting the erase head. The inductance of this head, in parallel with the tank coil (secondary of LL01) forms a parallel resonance with CL07 which defines the frequency of the oscillator. If the inductance of the erase head is large compared to LL01, the effect may not be that pronounced. Amplitude on the other hand won't be affected much since the circuitry below the oscillator itself uses the rectified oscillator signal via DL01 to control the emitter current of the oscillator circuit to stabilize the amplitude to the desired value (depending on tape type selection and the input via transistor QL31.

If you've got an LCR meter, you may want to check the inductance of the erase head and find a suitable "dummy inductor" to switch into the circuit when disabling the head. Or just disconnect it a see what happens... 

Looking at catalogues of tape heads, I think their inductance is in the range of a few mH. Funny -  just minutes ago I had this conversation with another guy at work and he suggested to use an inductor of the same value as the erase head as the dummy load. Nice to have options. I'm certainly going to be taking measurements and observing the bias signal to see how its affected.

Thanks for the extra explanations,

Cheers,
Avi

[mikerj]

Looking at catalogues of tape heads, I think their inductance is in the range of a few mH. Funny -  just minutes ago I had this conversation with another guy at work and he suggested to use an inductor of the same value as the erase head as the dummy load. Nice to have options.

You could even use a second erase head as your dummy load if you have room to mount it somewhere.

[cvanc]

It occurs to me it might make for some interesting effects if you could modulate the bias oscillator (...) amplitude

You just described the exact process tape recording uses 

[AviZiv]

It occurs to me it might make for some interesting effects if you could modulate the bias oscillator (...) amplitude

You just described the exact process tape recording uses 

Yes indeed that is one of my options.

So far the possibilities seem to be

1. just disconnect the erase head and perhaps no ill effects (audible) will happen
2. Switch in an appropriate-value dummy load resistor
3. Use an inductor dummy load - the same value as the erase head inductance
4. Use an (remote) second erase head as a dummy load to maintain the circuit exactly as it is now.


Thanks,
Avi

[AviZiv]

It occurs to me it might make for some interesting effects if you could modulate the bias oscillator (...) amplitude

You just described the exact process tape recording uses 

Well, you're right for amplitude at least but you could easily extend the amplitude modulation range outside the normal range to introducing distortion effects by overdriving the head, what I'm not sure about is if there would be any audible effect achieved by frequency modulating it, if it did the effect would vary by tape composition.

I must say that the artist in me loves the experimental aspect of this idea. And I will try it.

Avi

[AviZiv]

I'm back with a quick update: finally was able to get into the machine and take some measurements with an oscilloscope.

The first important confirmation has been that, indeed, disconnecting the erase head does affect the bias signal on the record head. This is indisputable for this machine at least.

The bias signal coming out of the transformer and going to the record head is a sine wave ~82Kz at 172V peak-to-peak. The freq is within the general range I expected. Interesting to see that the part list shows the bias trap tank (to block bias from the audio upstream) is spec'd at 85Kz. So we are in the right range.

After disconnecting the erase head, the same record-head bias signal moved to 27.9Khz (!) but amplitude has not changed.

This confirms to me that I need a dummy load to get the bias back up to 82Khz.  I'll report back, but I'm going to attach a trim pot instead of the erase head and see what that does and if I can make the bias move in a good direction.

(I hear rumors that some multitrack machines have dummy loads built in but this needs confirmation and anyway it's not the case for this unit.)

Thought I'd keep on putting my notes in here, in case someone ever wonders .... haha what are the chances that anyone else will come here looking for this info?

Thanks
Avi

[dmendesf]

Seems the chances are high as I am looking for info about usual current / voltages / frequencies to drive a tape recording head   (but for complete different purposes). So is it usual to drive a 100+V 82KHz modulated sine in these heads? Amplitude modulated, I guess...

[AviZiv]

Seems the chances are high as I am looking for info about usual current / voltages / frequencies to drive a tape recording head   (but for complete different purposes). So is it usual to drive a 100+V 82KHz modulated sine in these heads? Amplitude modulated, I guess...

I'm new to this circuitry so I'm learning as I investigate but if you look at the schematic I attached early in the tread, you will see a cap attached to the erase head and they made a point of noting that it is 100V rated. I didn't understand the meaning of it until I saw the oscilloscope reading. I guess the transformer produces the high voltage as part of its function, but the current is low. I looked up cassette-recorder erase heads in old catalogues, and they are rated in the 10's of mA.  85Khz bias signal seems one of the known practices in the industry, although higher freq also exist. I'm guessing that my 82Khz measurement reflects old components or maybe just tolerance.

Avi

[AviZiv]

Another update - I was able to measure the inductance of the erase head and it showed to be about 0.34mH. I decided to order a small kit of inductors and see if I can switch in a dummy load. While waiting for the delivery I thought I might play with resistors and see if I could use a resistive load to give me an equivalent dummy load AT the target 85Khz frequency. It did not work. A resistive load killed the bias signal. I really don't understand it but I guess I'll table this for now.

The inductors came and, indeed, switching in a 0.3mH (more or less - within tolerance) inductor, brought the bias signal back to the target 85kHz and the same levels. I think I'll accept this solution and built the switchable mod into the machine.

Leaving this here just in case someone ever wonders about these matters...

Thanks,
Avi

p.s 1 but whhhyyyy didn't the resistive dummy load work?.......hmmm....
p.s 2 It's very very hard to find specs for these tape heads. I thought my measurement must be off because the one and only tape head manufacturer catalog I found showed their cassette erase heads (not necessary mine) to be about 2mH... but I had to trust my measurements.

[chuber]

Thanks for these updates! I'm intrigued by this thread and the possibilities of achieving varying forms of sound degradation by changing the behaviour of the erase head. Did you by any chance conduct any more experiments that might be of interest? Like connecting a poti switch to the erase head or anything like that?

Electronically a lot of this stuff is over my head, I'm coming from the musician side of things. But I'd love to get some ideas to work towards and maybe get some help from more electronically talented friends when needed.

By the way I assume this applies equally to reel to reel decks? I'm currently setting up a Sony TC-570 as a looping/delay/sounds device for my saxophone solo set.

[david77]

I can understand the desire for a gapless loop tape.

But I don't expect that recording over un-erased tape will be satisfactory.  The erase head is broader than the record head in order to erase the entire track.  Without that broadly erased track, any slight misalignment is going to change the bleed-through of the previous recordings.

It might just work fine. In the days of open reel tape recorders it was common to have a switchable erase head on many decks. I have never seen it on a cassette deck, though. The narrower tracks might be problematic and lead to sound degradation, who knows?

[Circlotron]

I seem to remember that the bias frequency and bias level for best s/n ratio and lowest distortion were two different figures. It would be interesting to have a circuit that varied the bias frequency and level according to the average audio signal level being recorded at any given moment to see how much of an improvement could be made.

[Le_Bassiste]

it also heavily influences audio high frequency range, so you would adjust the bias level in a way that would get you optimum flat frequency response. less bias enhances high frequencies at the expense of overall THD, more bias dampens high frequencies but improves THD. the automatic bias level adoption to audio recording levels that you are suggesting is called "DOLBY HX Pro".
https://en.wikipedia.org/wiki/Dolby_noise-reduction_system#Dolby_HX_Pro

[chuber]

In my case getting a flat frequency response is not always desired. Having the option to change these settings from song to song would be awesome, allowing you to switch from loops that repeat without changing colour to loops that degrade into a dark mess over time. Since I'm planning to use it more on the musical/experimental side, mainly with endless loops, these kind of imperfections are intriguing to me.

[Maniac303]

Am I the only one reading this that thinks you might be re-inventing the wheel here? There are plenty of multi-head tape echo machines out there, I am currently working on a Dynacord with 2 play heads and 3 record heads and one erase head, for example (tube based). I applaud the DIY approach and repurposing a mechanism in an older deck, but I think you might get plenty of ideas for your project by studying the schematics of the existing machines, none of which have been mentioned (so far) in this thread, and examining the approaches that each of the designers used in the classic machines. Roland, Korg, Dynacord Echocord, Klimt Ecollette are some of the brands that come to mind. Just MHO and again, I applaud your creative energy! Good luck with the project!

[David Hess]

Feedback through DL01 controls the oscillator level so disconnecting the erase head should not matter, however if it does, instead of disconnecting the load presented by the erase head, get a second erase head and transfer the load to it.