tape loops (was: Hello and RE: DOD FX-98)

[Haible Juergen <Juergen.Haible@nbgm.siemens.de>]

        >On single-deck looping, I have tried simply mounting a short
loop of tape
        >through both reels and the play/record head, using (in this
case) a bent
        >goose-neck desk-top mic stand to keep the tape from drooping.
It works as
        >a static infinite loop:  basically just a simple looped sample.
But the
        >tape quickly wears out, especially at the splice point.  And
I've wondered
        >if this technique is harmful to the machine because of the very
light load
        >on the transport system (one reel spins very fast, and I wasn't
sure if
        >that was good for the machine over time).

I think I should build a switch into my Roland RE-201 Space to disable
the HF oscillator that feeds the earease head.

The RE-201 is a tape echo with a tape loop of a few meters. It's
intended to
produce short echoes with a distance of a few centimeters between record
and playback heads; the long loop is just to prevent the tape from
wearing 
out too soon.

Which makes me think of several modifications:

(1) disable the earease head. Would add new sound to old sound (after
      the _long_ delay of the whole tape length), but no idea to which
amount,
      compared with previous recordings. (?)

(2) probably it would be an improovement to make the hf current thru the
      erease head _variable_, and fade previously stored signals out by
      partially deleting it. (Again, no idea what this would do to
frequency
      response or distortion of a partly ereased signal ... any clues ?)

(3) Now this would be a tough one (and I think I would not dare to do
      this to my "mint condition vintage space echo" (;->) ): Exchange
      the location of the heads:
      (a) original order: tape comes from storage box,
                                     passes erease head,
                                     passes recording head,
                                     passes several playback heads,
                                     dissapears into storage box.
      (b) suggested new order:
                                     tape comes from storage box,
                                     passes playback head and (plays
what has been rcorded
                                          half a minute before)
                                     passes erease head,
                                     passes recording head,
                                     passes remaining playback heads
(still usable for normal echo
                                          operation, only with one tap
less than before)
                                     dissapears into storage box.

This last one looks *very* tempting ! The feedback would then be
established
electronically, with all its accuracy and possibillities. (I love to put
my latest
toy, a frequency shifter, into a delay's feedback loop !).

Please tell me what you think.
Things like that must have been done before! Especially (3) would be
interesting.
Anybody to share his experience ?

JH.