Technical musings: direct-drive or belt-drive capstan?

sweetbeats

Reel deep thoughts...
These (i.e. the thread title) are the sorts of things I think about...another example: from the time I could drive and owned a car one of my favorite things to read was the factory service manual for any car I owned...read each one cover-to-cover several times and put it to use. And so I take a few moments to share some of these things I think about since I am now, as of the last few years, a renewed tape machine enthusiast. Enjoy, comment or ignore...or...whatever.

I remember wondering why, with its relatively recent manufacture, the Tascam BR-20 had a belt-drive capstan vs. direct-drive (i.e. the capstan motor drives a flywheel with an integral capstan shaft via a belt vs. the capstan shaft being an integral extension of the capstan motor shaft itself). I had asked the question in the past without any conclusive answer. I think I can answer my own question now...

I'd always associated "belt-drive" with being less professional or rugged. This happened because of my observations that many consumer decks would have one or two motors for all three basic transport functions (i.e. supply reel motor, takeup reel motor, and capstan shaft), and more "professional" decks had a separate motor for each...therefore "belt-drive" became, for me, a symbol of compromise.

There are so many things in play here that I've come to know that have helped me to shed my ignorance on this small topic...I remain very ignorant on a great many things. :o

DC servo capstan motors have an innate flutter component as a by-product of their drive mechanism when compared to AC hysteresis capstan motors. IOW a direct-drive DC servo capstan motor has a higher potential to induce flutter into the tape path depending on the motor design and notably the diameter of the capstan shaft (e.g. the smaller the capstan shaft diameter the higher the rotational speed for a given tape speed...the faster the capstan shaft spins the higher the frequency of the flutter component which typically makes the flutter more audible).

A rubber belt that decouples the capstan motor and the capstan shaft acts as a damper of this flutter component.

Mass is also a damper and that's where the capstan flywheel becomes important.

SO...

My point...

Nothing wrong with belt-drive, and in fact it is a great solution to damping the natural flutter component that is produced by the capstan motor. It is not necessarily a compromise.

On a side-note, this question of mine was further irritated some time ago prior to this conclusion when I brought my Ampex MM-1000 home and as I began to dissect the anatomy of this absurdly rugged, if not "crude" transport, I was shocked to find that the capstan shaft is...drum-roll (can you help me out, Otto)...belt drive! One obvious downside of a belt-drive component is that the belt can wear out, slip, break...so it is a maintenance issue albeit a minor one. Ampex addressed this by utilizing a composite mylar belt, not rubber. Supposedly never needs replacing. Also, the capstan shaft is *huge* on the MM-1000 (about 3/4" in diameter), and the flywheel is also huge...best of all worlds? Dunno...makes a racket, that's for sure.

Hope to learn more from anybody's responses to this including corrections to anything about which I'm in error.
 
Great read, Cory and have you ever wondered why some insanely priced turntables (I'm talking many thousands of $$) are belt driven and some truly "budget" ones are direct drive? OK, it's a turntable but there's a lot of similarity to your points, to which I agree.

I feel that both direct drive and belt systems can work wonderfully, especially when engineered correctly but that a belt system is automatically "low quality" is plain wrong, I agree. In many cases, it can be better. The only drawback (which really isn't that bad) is the need to be aware of belts needing to be replaced way before they become a performance issue.
 
Ooo yeah...didn't even think of the turntable world...definitely ignorant to that but know enough to know there are many particulars...seems like "direct drive" became a marketing mantra for every turntable at one point and I'm sure the engineers and marketing folks weren't on the same page. :D

A turntable needle is a sensitive thing...
 
My analog motorcycle is shaft drive. I used to have a one with a digital odometer, but that was chain-drive: what a mess!

In all seriousness, my Otari MX-50 NII is direct drive and it's awesome. I'm not knocking any of the Tascam stuff, because most of my studio is around Tascam machines, but the Otari stuff is just killer.

-MD
 
For turntables the more operative issue was :::rumble:::, or picking up low frequency noise from outside sources. The playback mechanism ion a turntable s completely different than tape, i.e. the primary content transfer point ( the stylus) is an electro-MECHANICAL one. The number one place where noise is introduced into the signal is at this vinyl/stylus interface. Thats where all those dust and static pops and clicks, scratchs and skips, etc all originate.
What the drive mechanism is supposed to contribute is a steady speed. Higher quality turntables have heavy platters to act as the speed flywheel, with good rubber surface damping to minimize transfer of higher frequencies.

>> Belt drive decouples the motor from the platter, and allow the motor to run at a higher speed wherein any variations in motor speed have less effect.

Intersetingly, I used to work at a radio station, and the turntables used rubber idler wheels which drove the rim of the platter from the underside.
The (huge) motors would run continuously, and there was a mechanism that pushed the roller against the platter internal edge when you went to play.
Advantage: cue up a record and when the rubber hit the road (platter) the record was at speed intstantaneosly. No startup wow. ANd th erubber roller disconnected the platter from the motor.

Terrys Rubber Rollers probably LOVES people who have these old units
 
techno, that's awesome. Thanks for some background on turntables, an analog components that gets much less attention around here.

And, aha! You brought up a capstan drive mechanism I totally forgot about...rim-drive. Usually has the stability advantage of a massive flywheel and some damping via the rubber tire on the drive wheel or on the flywheel.
 
Great read, Cory and have you ever wondered why some insanely priced turntables (I'm talking many thousands of $$) are belt driven and some truly "budget" ones are direct drive? OK, it's a turntable but there's a lot of similarity to your points, to which I agree.
actually ALL insanely priced 'tables are belt drive.
Well ...... a very few are magnetic drive but otherwise it's belts. I can't think of any 10,000+ doloar 'tables that are direct drive.
BUT ..... it's a little different.
You have a needle sitting there that's designed to pick up even the tiniest vibrations and turn that into sound. So motor vibration is a BIG deal and it's not only flutter they try to avoid but also, and more importantly, any vibrations that thye motor gives off.

But tape doesn't pickup vibrations and convert them into sound AFAIK.

I think for tape either system could work well.
But I'm ready to be enlightened about this if anyone has a preference.
 
Cory:

Interesting post. I remember discussing this issue with some 3M experts over the years.

The first version (1965) was the M-23 series which used rim drive. The prototype rim tires worked great, but they had to go with a different source and process in actual production and had some issues with the spot where the ends were joined. My M-23 mix deck works great. There is just a bit of wobble to the phase at 16Khz with a periodicity equal to that of the flywheel. I have an unused flywheel that I could put on there, and it might do better, but I don't seem to be motivated to make the change, since I don't really hear any problem.

The middle series was the M-56 (1968)/M-64 ('67 or '68?) models which kept the all-discrete electronics and AC sync motors, but moved from a rim-drive to a mylar belt-drive system. The belt drive had perhaps the best overall flutter performance, but didn't lend itself to external control.

While the later M-79 generation had direct drive capstan motors that allowed external control and sync, there are those who felt like there was a bit of an issue of induced flutter in comparison to the earlier transports. Personally, I felt they all sound fabulous.

Cheers,

Otto
 
Otto...I was hoping you'd chime in...I remember you discussing the rim drive on your M-23 in your M-23 thread.

So the M-56/64 used a mylar belt? Do you know anything about the wear characteristics or lifespan of one of those?

I've tried to get information about this regarding the mylar belt in my MM-1000 but have come up dry. Replacements are simply unavailable and so I'm trying to get some assurance that I don't need to worry about it...

Anyway, I believe the DC servo capstan kits for the MM-1000 are (as with the MM-1100 and 1200) direct drive. I believe the flutter performance suffered measurably (as opposed to audibly) but the external control capability (though very primitive in the MM-1000) and very rapid start times were a plus. And the flutter component that was introduced was of little consequence thanks to the large capstan shaft diameter. Also, with the DC motor, it shut off when tape was unloaded so the machine could be powered up without sounding like a diesel tractor-trailer idling (in comparison to the AC hysteresis motor which runs all the time when powered up, though defaults to the 7.5ips rotational rate when tape is unloaded).

Such fun. Thanks for chiming in everybody!
 
So the M-56/64 used a mylar belt? Do you know anything about the wear characteristics or lifespan of one of those?

I've tried to get information about this regarding the mylar belt in my MM-1000 but have come up dry. Replacements are simply unavailable and so I'm trying to get some assurance that I don't need to worry about it...

My experience suggests that the belts don't deteriorate much and last for a good long time. OTOH, they won't last forever. You might chime in on the Ampex list and see if anyone has some laying about or knows of someone who could make them. Also, I can check with Matt Allen, who helps 3M users find parts, to see if he knows of anyone who could make some more. Maybe there would be several Ampex users who want to do a group purchase? You might also check with John French, as he may know about vendors who could make those belts.

Cheers,

Otto
 
Otto, thanks...all wonderful ideas. Putting that info in the "keepsake".

I did check on the Ampex list but got no response. The mylar belt was only used on the MM-1000 and only 500 or so of those made, and many of them were servo...the MM-1100's and 1200's were all servo...so I don't think a lot of people know or have interest. Now, if there were more people keeping VR-2000 systems going that would be a different story, but not a lot of interest in old quad-video transports. I think those were all AC hysteresis capstan motors.
 
Now, if there were more people keeping VR-2000 systems going that would be a different story, but not a lot of interest in old quad-video transports. I think those were all AC hysteresis capstan motors.

This might be interesting:

"Several methods of motor control are available in servo designs, the most common being:

1. Control of a d.c. supply to an eddy current brake, the motor being supplied with constant power.
2. Frequency control of the a.c. supply to a synchronous motor.
3. Amplitude control of the a.c. supply to a motor running below synchronous speed.
4. Pulse width control of the a.c. supply to a motor running below synchronous speed.
5. Control of the d.c. supply to a d.c. motor.

Of these, methods 1, 2 and 5 are common in CCTV helical scan machines while methods 2,3,4 and 5 can be found in broadcast quadruplex machines."

--Videotape Recording, 3rd Edition, Joseph F Robinson

...it then goes on to point out that while you can use a basic synchronous motor to lay down the recording, playing it back really needs servo control of the capstan and headwheel. While it can be assumed that the headwheel is probably direct-drive, it doesn't describe the usual arrangement for the capstan drive, aside from a selection of circuit designs to drive the motor itself.
 
Since you brought up quad video, it reminded me of yet another drive mechanism. I recalled that a video processor in Chicage I worked for one college summer used these HUGE Ampex 2" quad machines, and those head rotated at a ridiculous high speed...on compressed air! Used air motors. I have no idea how they regulated the speed....some sort of servo mechanism I'm sure. No belts to stretch!!
 
Since you brought up quad video, it reminded me of yet another drive mechanism. I recalled that a video processor in Chicage I worked for one college summer used these HUGE Ampex 2" quad machines, and those head rotated at a ridiculous high speed...on compressed air! Used air motors. I have no idea how they regulated the speed....some sort of servo mechanism I'm sure. No belts to stretch!!
They used air bearings, but AFAIK the headwheel was electrically-driven. It rotated at 15'000 RPM or 14'400 RPM, for NTSC and PAL respectively.

Sadly I've never seen one in the flesh, but it's fascinating to read about. It's worth searching youtube for "quad video" as well.
 
My point...

Nothing wrong with belt-drive, and in fact it is a great solution to damping the natural flutter component that is produced by the capstan motor. It is not necessarily a compromise.

Bingo!

A good direct drive system was generally more costly to make than a good belt-driven system. You can get a poor performing direct drive system if your not careful, so just knowing you have one or the other doesn’t really tell you anything. If direct-drive doesn’t positively impact the specs and performance its just sort of there and you can say you have it. ;)
 
Back
Top