Adding a 3rd Tape Head to (mono) Reel-to-Reel.

[Folderol]

Are you asking why some tape machines had a third replay head?

If so it was becuse the head construction and winding impedance is not the same for optimum record and replay duties. Replay heads typically had finer gaps and a higher impedance, more turns, giving greater output.
That said, I only recently found out that at least one major tape machine maker (Teac?) used the same head in both positions.

Dave.

Brenell used to do this too.

Ooo! Brenel, not heard that name in a looong time young Skywalker. Built on the same BSH principles as the Ferrographs and IIRC were two speed but had a capstan sleeve so you could get the other two?

Dave.

Yep! I was gifted with a Mk6 stereo (in dire condition) mid '70s. Did a lot of work restoring it, then like an idiot, sold it for peanuts to a friend - no idea what happened to it

P.S. Those two delightfully large VU meters? The were seized solid, but a watchmaker pal, not only cleaned them up thoroughly but also mechanically matched the spring tensions so they tracked perfectly!

[James Perrett]

That said, I only recently found out that at least one major tape machine maker (Teac?) used the same head in both positions.

I seem to remember that some manufacturers said that at high speed the requirements were very similar - narrow head gaps on replay aren't so critical at high speeds as the wavelengths involved are longer. Apparently some engineers only used the replay head on a multitrack for calibration - all recording and mixing used the record head.

Having said that, the Otari MTR90 used a dual winding on the record head with the two windings connected in parallel for recording but in series when used for playback. The performance is also claimed to be slightly better from the playback head (although playback through the record head sounds pretty good to me).

[Tim Gillett]

Ditch the EL3542...

Get something like a UHER if you want to experiment with tape. These have a removable tape carrier. And the heads are available too. New. Not cheap, but hey...

You can pick up a mono UHER for next to nothing. Replace belts. Again, available new. Not even expensive. Clean and lubricate. The chassis and capstan flywheel are built like a tank. And they're all exactly the same. Except for the electronics, but even these are interchangeable.

I'm amazed even the NOS tape carriers show up on evilbay quite often.

Why do they have a dedicated play head that you can use for realtime bouncing and delay?

Are you asking why some tape machines had a third replay head?

If so it was becuse the head construction and winding impedance is not the same for optimum record and replay duties. Replay heads typically had finer gaps and a higher impedance, more turns, giving greater output.
That said, I only recently found out that at least one major tape machine maker (Teac?) used the same head in both positions.

Dave.

Spot on Dave. A repro head's gap length sets the upper frequency limit, or more correctly the upper wavelength limit, whereas from memory the recorded signal isn't so dependent, tending to appear at the trailing edge of the gap. The record head's coil tends to be lower impedance partly to allow practical record amplifiers which don't need ridiculously high voltage rails to avoid clipping. Whereas a repro head needs the somewhat higher impedance to allow a useably high enough signal to feed a practical preamplifier.

Also repro heads usually needed better magnetic shielding than record heads, even to the point of the front face of the head being shielded. A hinged flap often swung over the head's face when record or play was selected.

The third head also enabled "confidence" monitoring of the just recorded signal to confirm by ear that the recording was up to standard.

The third head also made aligning/calibrating of the tape machine to the tape so much faster than the equivalent 2 head machine, especially when aligning a multitrack machine with its many parameters to adjust.

Tape machines still need to be maintained but more of their use these days is playback of old recordings rather than recording new ones, so it's repro (playback) heads that are the more sought after. I haven't bought a record head for many years.

Yes at higher tape speeds, the same head type can be used for record and playback. One reason is that for audio at say 15ips the standard small gap length play head (say 1 micron) can resolve to well above the audible limit, say to 90 kHz, which for audio is of little benefit. So a wider gap, closer to the ideal record head gap, can be used. Very small gap length comes into its own where very short wavelengths need to be resolved such as slow tape speed formats (cassettes), or anywhere where the highest frequency must be recorded and resolved. I think video tape heads could have a gap length of 0.3 microns. Incredibly small.

[Hugh Robjohns]

...from memory the recorded signal isn't so dependent, tending to appear at the trailing edge of the gap.

Technically, the recorded signal emanates from the trailing edge of the magnetic field created by the record head, which extends beyond the gap itself.

H

[Tim Gillett]

Spot on Hugh!

[Folderol]

...from memory the recorded signal isn't so dependent, tending to appear at the trailing edge of the gap.

Technically, the recorded signal emanates from the trailing edge of the magnetic field created by the record head, which extends beyond the gap itself.

H

Dear me. Talk about bringing back memories!

[Hugh Robjohns]

It's an interesting point because it allows tapes recorded in the late 40s and 50s to sound remarkably modern.

Back in the 50s replay head-gaps were relatively wide, and thus the HF response of tape machines back in the day was fairly curtailed. But the electronics in the tape machine usually had a wide-bandwidth and, since the record head-gap doesn't really affect the recorded frequency response, the actual recorded signal stuck on the tape really was a wide-bandwidth signal. That HF just couldn't be heard on replay at the time... But replayed today on modern narrow-gap machines, and all that original brilliance comes back the way the recording engineers heard it directly from the mixing console feeding the tape machines at the time.

An excellent example is Quincy Jones' Birth Of a Band albums from the late 50s, now transferred to CD. Remarkably modern-sounding records with plenty of top end that never made it onto the contemporary vinyl pressings -- and that top end isn't from mastering EQ, it's just because that HF really was recorded at the time, but it needed modern tape machines to replay it!

[Hugh Robjohns]

Dear me. Talk about bringing back memories!

There's lots of lovely tangible science involved in tape recording... and lots of physical and technical imperfections and limitations in the medium that were circumvented in so many ingenious ways.

If you were starting today no one would consider magnetic tape -- or vinyl for that matter -- sensible ways to capture, store, and reproduce audio...

[Martin Walker]

A classic example is Quincy Jones' Birth Of a Band albums from the late 50s, now transferred to CD. Remarkably modern-sounding records with plenty of top end that never made it onto the contemporary vinyl pressings -- and that top end isn't from mastering EQ, it's just because that HF really was recorded at the time, but it needed modern tape machines to replay it!

Wow, fascinating stuff Hugh - it's not often that we can retrieve data from the past with greater fidelity than at the time

Martin

[Tim Gillett]

It's an interesting point because it allows tapes recorded in the late 40s and 50s to sound remarkably modern.

Back in the 50s replay head-gaps were relatively wide, and thus the HF response of tape machines back in the day was fairly curtailed...

That's interesting. I seem to recall that in the late 40's/ early 50's Ampex were mainly concerned to satisfy the FCC broadcast stipulation of an upper 15 kHz limit, as they were working hard to have their new professional tape machines accepted in the broadcast industry. At first Ampex achieved 15 kHz response at 30ips. Soon afterwards they were able to maintain the same 15 kHz response but at 15ips, halved tape speed. This was the late 1940's when tape stocks were inferior to what they would become in the mid to late 50's.

So I'm not sure that even around 1950 they couldn't make a repro head with narrower gap, so much as they didn't see the need to do so, to go above 15 kHz as the upper frequency limit for playback. A 20 kHz upper limit seems to have been a more modern standard or expectation.

You mentioned the full upper range not making it onto the vinyl records of the day. Part of that may have been a deliberate choice to LPF so as not to allow those troublesome highs to be cut onto the disc master. Especially problematic in the inner grooves.

I have in my LP collection a disc named Aretha's Gold. From day one it didn't sound right, and with every play it sounded harsher. A few years ago I decided to look at the audio on the spectrum analyser. It showed a low level but very visible 17 kHz sine wave right through both sides. (It was beyond my hearing range even in my 30's.) Enough to make especially the inner tracks very distorted. Then there were the ill fated 4 channel vinyl experiments of the 1970's where they cut a 30 kHz tone with rear speaker information. So perhaps it was just as well if such upper frequencies on tape masters never made it to vinyl...

But transferring to CD or even higher res formats? A different story perhaps.

[Hugh Robjohns]

So I'm not sure that even around 1950 they couldn't make a repro head with narrower gap, so much as they didn't see the need to do so, to go above 15 kHz as the upper frequency limit for playback.

You may be right; I can see the logic in what you're saying. Unfortunately, all the people I knew who were working in the industry back in the 50s and 60s are all dead now, so I can't quiz them about these things anymore. However, the sleeve notes on those albums I mentioned make a big thing about the high-end not being heard before off tape or vinyl...

H

[Tim Gillett]

So I'm not sure that even around 1950 they couldn't make a repro head with narrower gap, so much as they didn't see the need to do so, to go above 15 kHz as the upper frequency limit for playback.

You may be right; I can see the logic in what you're saying. Unfortunately, all the people I knew who were working in the industry back in the 50s and 60s are all dead now, so I can't quiz them about these things anymore. However, the sleeve notes on those albums I mentioned make a big thing about the high-end not being heard before off tape or vinyl...

H

Is that a total surprise?

[Folderol]

This is almost annoying! I've got a faint memory about something on using a separate head for the bias, opposite the record head, rather than mixing it in the signal.
Ha! Got it
Right here

[James Perrett]

Wow, fascinating stuff Hugh - it's not often that we can retrieve data from the past with greater fidelity than at the time

There are still people working on ways of improving quality from old tapes. I'm currently testing some Dolby A decoding software that claims to retrieve significantly more detail than the old Dolby hardware. There are a huge number of masters that could possibly be improved by using this software.

And of course we have Jamie Howarth's Plangent Process which claims to eliminate wow and flutter using the bias signal or other continuous tones recorded on the tape as a reference.

[Hugh Robjohns]

Yes... I wonder what they'll come up with in the 2060s to improve the reproduction from early digital recordings...

[Elephone]

Yes... I wonder what they'll come up with in the 2060s to improve the reproduction from early digital recordings...

...or really early wax recordings! Could AI actually re-imagine them into pristine renditions?
That's another story.

[Elephone]

"Martin Walker" Wow, fascinating stuff Hugh - it's not often that we can retrieve data from the past with greater fidelity than at the time

Martin

Have you heard what they did with the Robert Johnson recordings? They may have made even further enhancements since then.

[Hugh Robjohns]

I did... and I have some.

[Tim Gillett]

Quoting myself here:

So I'm not sure that even around 1950 they couldn't make a repro head with narrower gap, so much as they didn't see the need to do so, to go above 15 kHz as the upper frequency limit for playback...

I did a few few quick calculations and cross checked another's figures. https://electronics.stackexchange.com/q ... -tape-head
It seems for a tape repro head to reproduce 20 kHz at cassette speed (1.875ips), its gap length must be 1.2 microns or less. But at 15ips, 8 times faster tape speed, that gap length only needs to be 9.5 microns or less.

In the early to mid 60's, Akai was making tape heads for open reel domestic recording with 1 micron gap lengths. I still have an Akai X-IV portable (Germanium transistors!) which claimed 13 kHz upper freq response at 1.875 ips (cassette tape speed). In the late 1950's, or even earlier, that they couldn't construct a repro head with a gap length of 0.374 thou (9.5 microns),
8 times less critical than what Akai achieved in gap lengths the early to mid 60's seems unlikely.

Tim.