>>We made this mistake with magnetic heads for credit cards too.
>>The voltage mode circuit is hideous, and has to be tweaked to
>>a range of frequencies.
>
>I've thought of using a current mode circuit for audio tapes--at least
>until I read a Winfield Hill post on the 'net.  He was building a digital
>tape drive (probably the same one in the AoE), and they tried using the
>current from the head.
>
>The only problem was the shorted turn they created from the tape reader
>head slowly erased the tapes, high frequencies first!
>


This seems fantastic.   (Although I give the source considerable weight!)

AFAIK, it is impossible.  Even if the shorted turn were a superconductor,
the resultant field strength could only approach the field strength on the
card, and this is not enough to erase it.

I've not seen any evidence of this, even in trials where we ran cards
through
the reader 10,000 times. They are only mechanically rated for about 2000
passes, and they begin to fall apart. We didn't see any degradation.
At that point, we're examining signal waveforms, looking for indications
that the head is wearing, and at the magstripe itself, for abnormal wear
patterns that would indicate the head being asymmetrically loaded against
the card.  Still, we would have seen it in the waveforms.

It might be an artifact of the media, but even so, this goes against
everything
I've ever read on the subject.  (tab books has an excellent book by Finn
Jorgensen)

There is a phenomena where some domains saturate at a lower field strength
than others, this is very dependent on the media.

I even went over this technique at length with the engineers at Vikron, and
they
saw no problems with it.

I wonder if something else was happening in Win's case?


>
>ps.
>Found the post of DejaNews, in case you're interested:
>>I can relate a relevant story.
>>
>> With a tape head, the voltage output is the derivative of the field
>> on the tape, dB/dt, but the current output (voltage across a small
>> load resistor or better into an opamp S.J. as you suggested) doesn't
>> suffer this "disadvantage."  I became somewhat enamored of the slightly
>> unconventional "measure into a short" approach some years ago, while
>> designing a digital data recording system.   However I discovered there
>> may be a downside.  It appeared that the taped signal became partially
>> erased, especially at high frequencies, by repeated reading with a tape
>> head used in this fashion.
>>
>> To complete my design, I decided instead to use an "advantage" of the
>> traditional dB/dt signals, namely, the peaks of the signal correspond
>> to the point of the original tape flux-changes.  The zero-crossing of
>> the derivative of the dB/dt signal nicely identifies this point.
>>
>>--
>>Winfield Hill    hill@rowland.org
>>The Rowland Institute for Science
>>Cambridge, MA USA 02142-1297