I'll relate my experiments with a reel-to-reel tape deck...
I have tried putting a guard over the erase head, or disconnecting the erase head. Then I recorded as usual. I found that this partially erased the previous recording.
You may find it is 30-50 percent attenuation, which you state would be satisfactory.
I'd really appreciate any help finding examples of an AC bias circuit specifically made for recording onto
I believe the bias is a 19 kHz sine wave. Its voltage amplitude needs to be strong enough to coerce the particles to move. But its amplitude cannot be so strong that it causes all the particles to align all the way (which would be saturation).
By keeping the bias within that middle range, it deposits your audio on tape, with high fidelity.
I was going to say the bias works like amplitude modulation. As I think about it, I don't know that this is correct. It would need to be a much higher frequency than 19 kHz.
The bias should have no DC component. If it did, then it would magnetize the head. The idea is to run only an AC bias to the head. Even so, heads require demagnetizing every so often. However cassette decks are not so prone to this, I've read.