01/17/07:  This project was officially started after a few other projects had already begun and after the Major Havoc project had been completed.  I began this project by digging up a copy of a Atari Tempest  manual which depicted a blow up diagram of the spinner.  Already aware of the issues with the controller, all I required the paperwork for was to obtain some basic information on the controller, including the optical board (which, by the way, will be reproduced to spec. as well) and mounting hole spacing/pattern.

01/19/07:  After looking over the controller pictured above, I knew I would have to redesign the controller to accept ball bearings and this was the only major concern I had right from the start.  If I could keep the current design of the cage frame (which, by the way, is about as strong as you can get for a controller frame) and force it to mount real bearings, I'd have something pretty serious.  I'd like to keep the overall design of this controller as close to the original as possible and see if I can keep it within tight enough tolerances to allow the use of as many of the OEM components as I can. 

02/12/07:  After taking several measurements, I knew I wasn't going to have enough room inside of the frame to fit the existing parts AND the new bearings.  So, at this point, the frame will need to be  enlarged ever so slightly.  Although, doing this may negate the use of the original OEM shaft which is a bummer.  I'll have to think this one over a little more.

02/14/07:  After pulling the top plate of the frame off and then reattaching (by TIG welding) it back to the base in a different manner, I was able to score an extra quarter of an inch of room (from top to bottom) to work with.  This was perfect.  Note:  Reassembling the frame in this manner is only a temporary solution so I can verify that my theory of construction will actually work.  I'll clean it up later.

03/26/07: I added two pillow blocks, one on the top of the frame and one on the bottom of the frame.  These custom pillow blocks are designed to each house a type R4 bearing.  I opted to use oil lubed open face ball bearings due to their natural ability to spin so freely.  Note: Each pillow block is attached to the frame via three very small (#2-56) screws which have been countersunk into the frame itself. 

03/29/07:  With the frame modified to accept two R4 ball bearings, I assembled the remainder of the existing parts from the OEM controller to test form, fit and function.  After assembly, I confirmed that the existing shaft would actually work in the new design without any issues.  The same for the flywhee; and the encoder wheel.  I tossed the two bushings and one retaining ring.  Some last minute adjustments required that I enlarge the hole on the top plate of the frame and add two small sleeves to help the shaft ride on the top bearing a little better.  What I ended up with was the prototype controller pictured to the left.  It's an OEM controller which has been modified to accept ball bearings.  Giving it a good spin, it will rotate for quite some time, too!  Ok, now I need to clean up the frame and spec it out for a production run. 

04/11/07:  I've got some fresh ideas on my mind for several controller projects.  The Tempest spinner is going to work perfectly as I had hoped.  I still need to rebuild the frame so it's a little taller than the OEM so it will accept the ball bearings, but this is a non-issue due to the ease of manufacturing this component.  I considered pressing the bearings into the frame while I had some time to think about it, but decided that the pillow block idea was the best way to go, mainly due to the fact that this method lends itself quite well to being serviced by whomever wants to maintain or replace the bearings.  Ever try to pull a pressed-in bearing from a frame?  Without the proper tool, it's a real pain in the neck.  My way, all you need is a standard screwdriver.  Simple and effective, though a bit more costly.  Note the pillow blocks on the inside of the top and bottom of the frame.  The concept works quite well in practice.

04/16/07:  Here's the frame as it should be (minus the black zinc plating).  This is the first article for the frame assembly.  Looks nice, huh?  This frame is nearly identical to the OEM frame.  It's strong and very study.  The only mods are those I had mentioned along the way:  a little taller than the original (by 1/4"), the top hole is a smidge larger in diameter and the pillow blocks and ball bearings have been added.  What a sweet spinner this is going to be.

04/18/07:  I decided to reproduce the knobs as well seeing as Oscar Controls just bit the dust.  We all knew it was coming due to what everyone refers to as "Real Life Syndrome" or "RLF" if you will.  I'm very saddened to see this wonderful resource disappear.  Oh well.  All of the sample knobs I have are broken, so I'm on the search for a NOS OEM spinner knob or an Oscar repro spinner knob.  Either will do.  I'm a little concerned that the added weight of the aluminum knob will add too much additional flywheel effect to the controller.  I suppose this could be a good thing for the MAME crowd?  

06/17/09:  I also made 500 Tempest encoder flywheel caps at the same time as the above encoder hub.  They're made from 6061 aluminum round bar stock, machined, drilled and tapped to original spec.  They have not been plated yet.  The OEM plating was a Chem FIlm (Gold).  I'm unfamiliar with this type of finish but I can see that it doesn't really hold up too well to the test of time.  I'm going to look into having them gold anodized instead - for the "bling" factor!

08/22/09:  I just completed a run of 1000 Tempest encoder wheels.  I recently purchased a photo-chemical etching setup which is somewhat like PCB etching equipment.  Anyhow, the encoders turned out really nice.  I decided to run off extras because I figure this part will be one of the more sought after parts.  So, I'll need enough for my repro controllers AND individual sales.  They're made from 0.010" thick 304 stainless steel and etched to perfection.