>
Do I need to pass a current through the head or is it basically passive???
Playback is easy.
Works the same as a phono cartridge or an electric guitar pickup, except the magnet and moving iron stuff left off, it feels the magnetic wiggle on the tape.
Any phono preamp can be hacked into a tape preamp. Change the RIAA equalization network (usually 2 R and 2 C in the feedback) to NAB/IEC (usually 2 R and 1 C) and fudge the gain to suit your heads. If you have very odd impedance heads, a phono preamp may not give best noise; "Walkman" implies cassette and most of these were similar to phono-cart impedance, anyway tape noise usually exceeds amp noise. Ampex and other snazzy tape machines sometimes used very low impedance heads and a step-up transformer (not unlike moving-coil phono carts).
LM381 is/was a better-than-junk tape playback preamp. http://cache.national.com/ds/LM/LM381.pdf
LM389 was a junk speaker amp with extra transistors: http://cache.national.com/ds/LM/LM389.pdf page 6 shows a complete record/play circuit typical of $13 cassette recorders.
>
Can I use a record head as a play head and vis versa???
If you are talking surplus or junk-box heads, just try it.
As you see on LM389 pg 6, low-performance tape machines used the SAME head for record or playback. The differences are that a play head never has to handle huge signal so can use a soft core without saturating; HF response in play is related to the gap-width but in record it is related only to the straightness of the gap trailing edge (a record head can use a large gap), record heads were sometimes lower impedance so they could be driven from 12V rail, play heads generally higher-Z to reduce preamp requirements. But both improved materials and general cheapness converged to not-very-different. If you have to meet a 17KHz playback spec, you will need a very good head optimized for that work; the majority of ~12KHz decks seemed to use all the same head.
RECORD is tough. You need considerable EQ, and you need "bias" to get off the flat-spot of low signal level. The LM389 pg 6 plan shows an awful but awful-cheap bias scheme: they run DC from the 39K through the head! If the DC flux is about 1/4 of the total peak-peak saturation flux, you can use almost half the tape's total flux swing without real gross distortion. What the Germans discovered in WWII is that a supersonic AC bias will give the whole range of the tape with quite low distortion, but you are really better off stealing a pre-made record-amp and head than rolling your own.
Pete has posted
an excellent book from 1952 that tells how to make a tape recorder using only common household scrap like dead radios and roofing tin. While 1952 scrap like 78RPM record player motors is now hard to find, the basic explanation and detail is worth reading.