Duncan Brown

Bearings are due in the mail today. Then there's a lot of photo editing, web page coding, etc. But soon. Duncan

I don't see why not. Anything 16mm with the appropriate perfs should make good practice film. I've done it in other cameras, though never actually a Bolex - always just went for it with those! (Side question - why the dark? Do you not want to lose even those few feet at the beginning of a daylight spool?) Duncan

200/2.8

85/1.2:

24/2.8:

And some resulting pictures... 24/1.4:

Another little fun/annoying diversion that Redrock sent me down was proving them wrong when they said that the horrific vignetting I was getting with their adapter was due to the coverage of the lens. Seriously? I've been using this thing on a still camera for decades and never saw vignetting but you're telling me they cut it so close I see it on your adapter? "Just zoom in more!" they say - but then why am I using a wide lens, if I'm killing half my field of view? (I knew I shouldn't have revisited this fiasco...grrrrr) ANYWAY, to show them how full of it they were, I came up with a way to show the actual image circle of a Canon FD lens. I cobbled together a recessed lens mount for my 4x5 camera, which would allow me to put the lens at the right FFD from the much larger film... which would then capture everything coming through the lens. Then in PhotoShop I could draw a 24x36mm rectangle on that image, to compare it to the image circle. The rig:

I've been shooting stills with Canon FD glass since I was a teenager (hint: a VERY long time!) and I've always wanted to shoot motion pictures through those lenses I've spent so much of my life peering through. I thought the time had finally come, about 15 years ago, with the Redrock Micro spinning-glass doohickey. So I bought one, and a Panasonic HVX-200, and got to work testing things out. I probably don't need to tell *you* folks what a piece of utter crap that turned out to be! The list of issues was long, and Redrock had no interest in addressing them. I did a lot of testing of everything to figure out what worked and what didn't (I even did a stint as second camera operator in a local indie film!) but I eventually sold the whole rig in disgust. With the advent of full-frame mirrorless DSLRs, it is now finally possible to use that glass to shoot movies! I have a Nikon Z6 and am quite happy with it. It occurs to me therefore, that there might once again be some interest in a massive web page I put together back in those days, with pictures of and through pretty much every interesting Canon FD lens they made. http://backglass.org/duncan/redrock/lenscharts/canon_fd/ I had planned to shoot dollied shots through each of the prime lenses, pulling in on a fixed object, but that never came to pass before I gave up on the thing. Thus all the broken links in the "Fixed Object" column. The concept was that I set up on the 50 yard line of the local high school football field, and shot straight down the field towards the school. Every shot was taken from the exact same position, so you can compare fields of view, etc. I also shot stills there, to compare the field of view that the Redrock rig got vs the field of view normally seen through that lens. Since I am a huge wide angle fan, the hope was that the horizontal field of view would be the same, just narrowed down top and bottom to the aspect ratio of the HVX. Duncan

Man, you just never know with eBay stuff - this 25fps motor's bearings are pretty toast. Grindy, takes a second to get up to speed, makes a hell of a racket once it does. Though bless its heart it was maintaining the advertised 25fps speed once it got there! So this isn't the be-all end-all test of this concept, i.e. I bet the max speed can go higher, on a motor that has good bearings, but just as an early read on the situation, here's what I got: Screwing the screw all the way in, it eventually wouldn't go any farther for reasons that weren't clear. Maybe the spring was completely unsprung and now it was trying to compress it? But it was running at 14.7 fps. With the spring not really being part of the equation, it was a little wavering on that speed. But I think it's safe to say to that guy who was looking for an 18fps motor a while back: I think you could set that and have it be quite stable. Screwing the screw all the way out, it eventually pulls up the unthreaded part so far it run out, so that's definitely a hard limit. In this case it was running 34fps, but I'd be willing to bet with good bearings it would have gone somewhat higher. So there you have it, a much wider range than previously thought, if for some odd reason you wanted something other than 24 or 25fps. And yeah, same motor, just tuned to the labeled speed. (And yes, I admit, the motor was designed to be turned to 24/25fps, and running it near either of its physical limits is liable to be less reliable than sticking to the design specs. But for those who were curious...) Duncan

I of course checked there first. ( In fact, that is where I ordered some various dowel pins to try using to put the ball drive back on the shaft.) But the cheapest bearing I could find there in that size was a $22 stainless steel one. I'm too cheap to buy a sack of those if I don't have to! McMaster-Carr definitely does not specialize in itty-bitty bearings like this. I'm guessing some RC car or something uses them, though, for them to be so prevalent on ebay. Duncan

Good point - I was trying to keep it all in this thread, where motors had been thoroughly discussed, but it will definitely be worth creating a few new threads after the fact for better future searchability. FYI I also have another constant speed motor coming. Labeled as 25 fps, but I expect to prove to myself that it's the same motor, just adjusted differently. I will then proceed to find out the extent of its adjustability (basically, what speeds it is running at when the adjusting stud runs out of room in each direction) just to settle that question, because I see various references to it not being very adjustable...but if half a turn changed it 1 fps, it seems like it would have a decently wide range. Duncan

That bearing you link to is indeed stainless, but the 634 bearing type has a 4mm shaft hole, not 5mm like the 625. Which is why it is very confusing for them to put the 634 number after the 625 on the bearing. The latest GRW catalog says a stainless steel bearing would have "SS" in the part number, which the ones in the motor did not. Not sure what would have been the case back when these motors were made. Duncan

Took apart a standard variable speed motor and documented it, but I'll wait to post that until I have some new bearings and put it all back together again. It's more complicated than the constant speed motor, and you have to desolder two wires to get it completely apart. This is one I bought used cheap on ebay - and the variable speeds were all over the map as I spun the knob. I'll be interested to see if just cleaning it all up and getting the slidy contacts going on a slightly different path on the variable resistor windings manages to restore it to stable operation. The bearings are indeed the same spec as the constant speed ones - 5mm shaft hole, 16mm outside diameter, 5mm race width. That matches up with the GRW 625 number that's stamped on it, still not sure what the extra 634 number is about. The other motor had bearings that had metal dust shields on both sides. This one had a metal shield on one side, and open on the other, This motor had definitely been worked on before, as there was kind of nasty grease in the bearings, and someone had jacked up the pin that holds the ball drive on the shaft. The original looks to be maybe a brass drift pin, whereas this one had clearly been clamped in a vise before, and someone had tried to drive a roll pin in and broken it off halfway through and then put a nub of another roll pin in from the other side to hold things. I had to drill the whole mess out in a drill press, and will have to come up with some kind of brass drift pin to put it back together. On eBay I ordered some bearings out of Michigan - 10 for $7.99 - with rubber dust shields on both sides. Those have a generic part number of 625-2RS ("2 Rubber Shields"). I also ordered some from China, 10 for $3.99 with free shipping (!!!) that have steel shields on both sides, generic part number 625-ZZ. I'd prefer the steel shields like the originals but A) I'm impatient and B) I'm not sure I trust Chinese bearings that are 39 cents each shipped. But I figured I'll get both kinds and eventually be able to compare them. Stuff from China usually take a month or two to arrive. Duncan

Hey! That's fantastic that someone is doing that! But for a tinkerer like me, where is the fun in that, ha! (I also just picked up a small pile of not-crazy-old NOS Tri-X magazines, which I'll try shooting at a lower EI and pull-processing, and see if I can get usable images, just to test out this camera) Duncan

I could actually use a few of them. Cheap hacked up plastic ones are OK, I just something to keep the dust out when not mounted. Thanks, Duncan

Just going by the other HS cameras I've used (I know nothing at all about the Redlakes) they often have other connectors for remote on/off, timing lights, etc. to help with the insane logistics involved in capturing a brief moment in time without a lot of wasted film. So it's possible you could rig up a remote on/off switch through a connector, to take the place of the one missing in the camera body. All the ones I have make rampant use of "military-style" multi pin twist-lock connectors. I have a giant box full of ones I gathered along with the ones I needed, back when I was getting my cameras going. If you figure out what connectors you need (pictures help, but often pin count is enough) I can look through my stash and see if I have what you need. Duncan

Well it's almost certainly something simple, but not necessarily simple to fix! Like, if the contacts deep inside on the other end of that main button were tarnished, say. Maybe some gentle raps on various parts of the camera while holding down the release switch? (Twist-lock the power switch into the always-on position to make it easier to do this.) I deal with old electromechanical pinball machines a lot, and they get angry if they aren't used for years...the fix being to just use them! But sometimes getting that first spark of life out of them is the hard part. Duncan

I would fiddle with the off/battery-test/forward/reverse switch first. Maybe it's not making proper contact in some of the positions (have you tried reverse?). Then maybe the speed knob and fixed/variable speed switch next. Anywhere a contact or variable resistor could have gone open or high resistance with age and tarnish. Maybe fiddle with the takeup and feed spools to see if you can get the movement inside to alter its position even slightly. The R16 has that cool feature where it always stops with the shutter closed, but maybe that also means if something gets old and stops making connection, it won't start up again from that position? Duncan

I bought Sean's kit to do the replacement. I'm pretty sure he's morally obligated to discourage you from doing it this way, as it's insane. He also won't help you do it that "right" way on newer cameras because that is also insane and should only be done by a real experienced competent Arri tech. As far as he's concerned, his kit is only for the older cameras with the two bolts holding the transmission. But, here you go. Next tech who disassembles this camera will probably curse the light tool marks I no doubt left on the transmission shaft, but frankly the nut simply wasn't that hard to spin off. the 11" handles allow a pretty tight grip at the tip of the pliers, and there was no slipping. So any tool marks would be insignificant. Duncan

Oh just another trick I learned along the way. I have a later camera, for which replacing the drive coupler "requires" pretty deep disassembly to be able to hold the transmission in your hands to replace the coupler. That seemed a bit much to ask just to replace a rubber tube, so I figured out a vastly easier way. Just work down through the motor opening! Now, you need to be especially careful not to drop little parts down into the bowels of the camera, so this is not for the faint of heart. But with the right tools, it just took me literally a couple of minutes. At which point I wasn't sure I had gotten the little pinch rings in properly so I took it apart again... discovered I HAD, indeed, done it correctly, then put it back together again. I bet Arri had a custom tool that did it this way, perhaps with some grippy and/or curved surface on the tips to minimize tool marks, but this definitely gets the job done. These two pictures should explain it all, let me know if they don't. (I'll put the second one in the next post, because it says the two of them are too big for one post.) Duncan

As promised. http://backglass.org/duncan/arri/constant_speed_motor_24fps_teardown/ When I get my spare variable speed motor I'll give it the same treatment. With luck it will use the same bearings, so I can go the extra mile and pull them off to parameterize them better. Still wondering about the seemingly spurious number on there. Duncan

OK, full pictures and details coming later, but I got the motor all apart and back together again, and adjusted to about as dead-on to 24fps as you can get with an electro-mechanical governor like that. (I can see why people eventually went to crystal synced motors) For the technological era involved, it really couldn't have been any simpler to adjust the speed, if that's all you need to do. And in the end that was all I needed to do. My bearings were fine, my main brushes were fine, one of the governor brushes is getting on down there but given the level of use I'll give it, will probably never get to "worn out." Even replacing the bearings would only require one wire to be desoldered, and one roll pin or something to be pressed out (where the drive ball is held to the shaft.) Kudos to Arri, this is a good design for the time. I do have a spare stock variable-speed motor coming, which I'll pull apart and document too, but guessing it's pretty similar. Duncan

Answered my own question by just doing it - tapped the pack at 6 cells, gives around 8.1V unloaded, still runs at the exact same slow speed. Time for some motor surgery. Duncan

Interesting about the mirror stripes - that actually makes sense! Before tearing down the motor, I actually did a little more testing on the camera. With the battery hooked up and the switch on and the motor out, I can measure the voltage at the motor contact. It's the exact same as at the battery, so I'm probably not losing anything through the camera (yes, it's possible somewhere there's a high resistance that would affect the motor but not my meter, but again I would think that effect would not be so precise as I'm seeing.) BUT!! The voltage is 9.4V. My fully charged nominally-1.2-V cells are putting out more like 1.35. Could that affect the motor governing? I guess I could modify my pack to tap off of 6 cells instead of 7, which gives me like 8.1V fully charged, but it would drop to 7.2V after some use I would think. Hmmmmmm Duncan