Magnetic Tape recording

Seeing as no one uses magnetic tapes anymore, could I use a few of them as data storage devices? I'm not sure I want to completely rip apart my VCR, but if I solder some wires on to the read/write head and the erase head could I, in theory, use it to store binary integers? Any idea, though, on the voltage required for the write head and for the erase head, and how much voltage I'll get out from the read head? I have a few Lm358 I can use for amplification.

Thanks! baum

FYI The Commodore 64 had a tape device - http://en.wikipedia.org/wiki/Commodore_Datasette

But I don't have a Commodore. I was wondering whether or not I could hack my VCR to use as a tape reader/writer.

As it is possible to record video you should be able to record anything. But VCR is an analog recording technique so you need to do AD and DA conversion, and probably errorcorrection too. The reason why I mentioned the commodore is that it is proven low cost tech, and it may be possible to use a similar recording technique with VCR.

But VCR is an analog recording technique so you need to do AD and DA conversion

Not if I'm just writing ones and zeroes. Activate electromagnet for a 1, turn off for a zero and vice versa. I'm not storing analog data.

Doesn't mag media need a bias? Or is that just for analogue too?

Returned signal levels are pretty low, IIRC.

Low return signals can easily be handled by an op amp.

The old nine-track half-inch tape used to run at maximum density of about 6250 characters per inch (140Mbyte per 2400 foot tape!). Not sure how that translates to VHS helical scan.

I wasn't thinking of having that much information.. just a bit as a side project.

Thanks!

I think the old tone-based stuff on cassette tape managed around 1200bps. Cassette tape moved at 1 7/8th inch per second, so maybe that gives you some idea of what you can do?

Not if I'm just writing ones and zeroes. Activate electromagnet for a 1, turn off for a zero and vice versa. I'm not storing analog data.

  • Better active EM for 1 (field directon up ) and activate reverse for zero (field direction down)
  • In practice a magnetic field on the tape is very analog, a bit is idealy a spot with the strongest field in the middle (sort of) and decreasing to the edges. Reading needs to find the spots (tracking) and the field direction.

long runs: If you have to write 200 zero's after another (can happen) you need some "build in sync" because if the tape runs only 0.5% slower you might read 201 zero's. So at least you have to use a gray-code to prevent long runs of ones or zero's.

I don't say it can't be done, just that it is not easy

If it were me, I would be using a linear tape like AWOL says. Go get a cassette tape deck from a pawn shop or something. Helical scan is just a whole nuther kettle of fish.

robtillaart is right in that you want to encode both 0 and 1 with active magnetization. I think I would recommend using Manchester Phase Encoding at the bottom level, then, of course some bit-stuffing framing protocol and some forward error correction and stuff.

You only have 2K of memory on a 328 and while some of the data massaging can be done on the fly, I think FEC would require keeping whole frames around, so this is going to constrain frame size to the 1K range.

The frequency response of cassette tape is, what, 12KHz at least. You should be able to get 6KBits/sec of raw bits encoded on a tape.

As someone who has designed his share of tape drive magnetic systems... phase encoded NRZI etc... 7, track, 9 track etc... I don't remember it being as easy as you describe -- but I'm willing to learn anew. Carry on I'll follow along and see if I can shove $0.02 worth now and then If I can understand the project.

A helical scan tape drive? (Video Recorder) That is a different kettle of fish though...

There were interfaces to video drives -- you can still buy them -- but mostly, they were abandoned as soon as decent hard drives became a reality...

relevant discussion on the old forum http://www.arduino.cc/cgi-bin/yabb2/YaBB.pl?num=1295774730 encapsulated forever in amber...

OK.

I think if I was going to try this with a (working) VCR, I would do something fairly "unorthodox" - I would purchase this shield (or the chip) - which was recently released:

http://nootropicdesign.com/ve/

I would hook its input to the VCR's output, and its output to the VCR's input. Then I would hook up the play, record, rewind, fast-forward, and eject buttons to the Arduino.

Write the code to use the TVOut library to output checkerboard patterns encoding the data; then use routines (on playback) to read the data (similar to the blob tracking demo). IIRC, this is similar to how various "backup" to VCR systems back in the day worked (although some, I recall, used the audio tracks instead). You could also possibly just record black->white transistions in the video area using the TVOut library, and then read the transitions in some other manner via an analog input...

Ok - maybe its a bit "pie-in-the-sky", but it would probably be the easiest and most reliable way to get something like this working (well, short of just recording audio like the old-school cassette tape systems).

:)

so nobody does mag tape anymore.

But don't tell HP, Tandberg et al.