An Akai X-200D turned up for auction at a local auction site, and while the price is out of my "that would be fun to have" price range, I decided to see if I could figure out how the auto reverse mechanism worked. After all, most (auto) reversing machines I've seen have either two capstans (such as the TEAC X-2000R) with a double set of heads between them, or a centrally placed capstan with heads on either side (such as the old Grundig TK830). (The Technics RS-1700U comes to mind as essentially belonging to both categories).
Didn't find much written information, but there are a couple of instructive youtube videos, such as this one:
https://www.youtube.com/watch?v=U4RGmXbXBNc
So, yes, it does seem that in reverse mode, the capstan is essentially at the back of the tape path, with the tape tension being maintained by the take-up reel (which at this point is the left reel of course). At first I thought that this would have been tricky on a machine with conventional clutches, but the Akai has three motors so the take-up tension should be smoother. I'd still have a feeling that speed uniformity would be worse in the reverse direction, but I suppose it could be on par with forward mode where tape tension is maintained by braking the supply reel.
The video also confirmed my suspicion that the machine, with its crossfield bias arrangement, uses a combined record/playback head, with separate bias head opposite it. Tandberg also practiced this on a couple of machines, although most of their crossfield machines have separate record and playback heads. I would think that the requirement of a conventional record head of having a wider gap to avoid shorting out the bias field would not hold for a crossfield machine, where the bias is supplied from a separate head anyway, so, lack A-B monitoring capabilities aside, there is no technical reason for having a separate record head in this case.
One minor surprise was the mechanical arrangement of the record/play head. It mechanically shifts position depending on the direction of the tape, apparently using a solenoid. I seemed to recall a similar arrangement on the Akai M-8 for instance, where the channel selector shifted the height of the whole head block.
On X-200D, the channel selector indeed does shift the height of the bias and erase heads. This is definitely the same arrangement as on the M-8. I would have thought that the mechanical complexity of shifting the height of two tape heads would have been outweighed by the electrical complexity of reconfiguring the coil connections, but apparently Akai thought differently.
So, the erase and bias heads are mechanically shifted for track selection (essentially, in mono mode, the head is moved so that one track of the head covers the appropriate area of the tape, with the other track of the head ending up outside the tape), and the record/playback head is shifted to obtain the reverse mode.
One consequence of the record/playback head arrangement is that the coils must be swapped in reverse mode to get the channels right. So there still has to be electrical switching in reverse mode. I would have thought a stationary four-channel head would have been a better option, but again, apparently Akai felt differently.
Obviously, this machine cannot record in reverse mode, or it would need two erase heads which it doesn't have.
A question for the forum: how reliable is this mechanism? Especially for the record/playback head I can imagine two hassles over time:
One, maintaining azimuth. As the mechanism gets work (or even when new) it must be hard to maintain correct azimuth with a movable head.
Two, wear. If the machine is used more in one than in the other direction, the wear groves in the head would cause the tape not to lie flat on on the right-hand track in the reverse direction (assuming the machine is used more in forward than in reverse).
As for the uneven wear on the erase head, I'd think it would be less of an issue, but given enough time, the difference in wear patterns would mean uneven erase in the modes which are not commonly used.
I suppose the machine was quite a wonder in its day, offering auto reverse in a compact machine, with the cross field recording ensuring high fidelity at lower tape speeds, but 40 years down the line I would expect there are many who are working less than optimally, to a greater degree than other machines of the day.