Roller vs carbon throw out bearing

[Rut]

Who uses what and are you satisfied with the product? I'm in the process of a tranny exchange and the carbon throw out bearing is worn down to the metal cup it lives in. Is there a good, cheap alternative to the expensive roller throw out bearings and the carbon type?
Thanks, Rut

 

[Spridget64SC]

Simply put, not really. The carbon types are original and entirely functional. They just are messy and don't take well to extended periods of engagement, such as sitting at the traffic light with the clutch depressed. Don't do that and they will last a fairly long time.

That said, the replacement roller type has gotten a bit pricey. I've made my own setups before using the empty shell from a 948 stock unit and making a steel bobbin on the lathe to accept a "rice-burner" roller bearing. It ends up looking like a homemade version of the "manufactured" part. In the end, I just save the value of my time. A candidate roller bearing part is about $25-30 if I can find the right ones at the right point in time.

Mike Miller

 

[Rut]

Thanks for the info. Does the roller bearing you use fit in the 'cup' of the old carbon TO bearing? If so, what part number or model do you use?
Thanks, Rut

 

[Spridget64SC]

No, unfortunately the roller bearings require an adapter. The adapter looks kind of like a open top hat. I then weld the adapter into the 948 release bearing cup. Notice I say a 948. This is because it does not have the same offset as a 1275 carbon release bearing body. The combination of the 948 bearing body, the adapter and the bearing end up nominally the same length as a carbon type 1275 bearing.

I'm currently using TO-1710 bearings. These are Toyota/Subaru/Mazda type bearings.

I use a similar type arrangement in my race cars with the AP Formula 3 all metal clutch covers. Works very well with these type clutch systems.

HTH,
Mike

 

[aeronca65t]

As Mike says, the carbon unit works fine if you don't hold it down too long at traffic lights etc.

Most of us who are vintage racing are using the carbon units with no problem.

 

[Pythias]

I have no experience with it myself, and have a Datsun so it's not relevant to me.....BUT..

I've heard that the newer Throw Out Bearings (from India? China?) are not of the quality of the original units. Word on the street is that they only last a year or two, instead of years and years and years that they should. If anyone is having problems and having to replace them constantly this could be the reason, and not that it is improperly adjusted or being "abused".

 

[Sarastro]

I wouldn't be surprised. Man, the stories I could tell you about people who tried to reproduce the old stuff but lost the formula. Ended up with something that seemed the same but just wasn't equal. I'm not just talking about cars, too.

It would really be great if someone were to draw up a design for a roller release bearing, complete with part numbers, so the rest of us could copy it. That would be a great service to the community (hint, hint!).

 

[Gerard]

I've always heard mixed results with roller release bearings.I think part of the issue is the the way the fork pivots, the bearing is not always concentric with the contact surface on the clutch. It needs to slide over that surface to some degree as the clutch is depressed. I would think that is why the carbon surface is used. I have seen a few carbon bearings trashed in very bad ways and had 1 or 2 give disintegrate in short notice, so I tend to believe some of the reproductions are as suggested.

I think if you want to do a roller bearing, the solution is to use one of the concentric slave cylinders that go inside the bell housing. I know these are used on Spridgets in racing applications and there are many production cars today.

 

[Spridget64SC]

Gerard in on the right track! All the stock release bearing parts and pieces don't quite line up like you think they would. A flat thrust face on the roller bearing works better than a radiused face. The "Green" company bearing has some kind of plastic film on it that appears to help it center.

When I make the roller release bearing setup for the race clutches, I do it this way;

1) Weld an aluminum tube to the stock front cover. Tube is about 3 inches long. I start with 1.5" OD by 1" ID tubing.
2) Align the OD of the tube 90 degrees to the front cover mating surface. I have a jig that holds the assembly in the lathe for such purposes. I cut the tube OD down to 1.375". I also cut a recess in the inside for a oil seal. This is where the stock scroll would be.
3) Machine a bobbin to hold the release bearing. The bobbin has a 1/2" wide groove in it to accept two pivot pins that get welded onto the stock release arm.
4) The bobbin slides on the aluminum tube welded to the front cover. This aligns the roller bearing radiused surface with the input shaft and the clutch cover assuring correct engagement of the spring fingers in the cover.
5) This setup uses the stock release arm and the slave cylinder. Basically performs the same as an annular release bearing. Similar to later model cars that have the same kind of front cover piece from the factory. Could be used on a stock clutch system as well, but would likely have to adjust tube length, bobbin length and type of bearing used.

I'll see what I can do to come up with some pictures and maybe some drawings for a stock type setup.

Mike

 

[Rut]

Having no experience with putting together anything like this I've made a few observations regarding TO bearings in general. Most appear to mount on the transmission input shaft and the clutch fork fits in the grove of the bobbin. The bearing rotates around the shaft, loosly it seems, but enough to locate it. This in turn keeps the face of the TO bearing square with the clutch and allows the fork to move it in and out to actuate the clutch. On most designs the fork also locates the TO bearing and it would appear that it's a simple arrangement. I've obtained a sealed roller thrust bearing that fits the cup of an old 1275 carbon TO bearing to within 0.051" and has a diameter suitable to clear the shaft. I'm making a dummy mount with a rear plate, flywheel and clutch assembly to test what I build on the transmission complete with slave and hydraulics. I haven't decided how to mount the thrust bearing to the old cup yet, but I'm leaning toward set screws thru the cup and into the band of the thrust bearing. If this works it would be very simple and cheap to repeat and little would have to be changed. If it doesn't, I want to investigate a standard type of roller TO bearing and change to a yoke type fork or modify the existing fork to fit the bobbin. If anyone has ideas I'm open. My goal is a good, durable, roller TO bearing at a very low cost.
Thanks, Rut

 

[Sarastro]

Interesting discussion! When I've thought about this, it seemed to me that keeping the bearing concentric with the axis of the clutch and transmission input shaft would be necessary, and what Mike describes, or something with the same idea, seemed to be the way to do it. It's neat to see that someone has done this already.

I haven't seen the inside of a modern clutch in many years (which, I guess, is the same as saying I never have). How do they maintain concentricity?

 

[Gerard]

When I make the roller release bearing setup for the race clutches, I do it this way;

1) Weld an aluminum tube to the stock front cover. Tube is about 3 inches long. I start with 1.5" OD by 1" ID tubing.
2) Align the OD of the tube 90 degrees to the front cover mating surface. ....

Mike

Exactly the idea I had in mind. This is how it's done on the Datsun 5 speed covers. I'm glad to hear you've actually done it. Now if you added the lip seal to the back of the cover, you'd have the absolute perfect solution. I'd like to see pix if you can post them sometime.

 

[Rut]

Steve,
It looks like the sleeve or bobbin fits the transmission input shaft and that keeps everything lined up. The clutch disk fits on the shaft as well and the pressure plate is bolted to the flywheel. I know this is pretty basic stuff, but it's interesting to me to try and figure out how to do it with minimum effort and expense. Timken had a TO bearing, AU1715, that's been discontinued and I've searched for NOS. The aluminum sleeve that Mike welds to the front cover keeps things centered and I wonder if you could do the same thing by threading the cover and screwing the aluminum tube into the cover, or will that mess up the input shaft scroll 'seal'? The more I dig into this the more questions I have. More thoughts?
Thanks, Rut

 

[Gerard]

The one thing that has been missed is that in order for the bearing to remain concentric, it can't be affixed to the clutch fork. If the pivot point is fixed, it will always want to move in an arc, not a straight line. The fork would need a sliding pivot or be allowed to slide freely when pushing the bearing.

 

[Rut]

Gerard,
That's why I was leaning toward the bobbin style TO bearing and modifying the clutch fork to fit. I guess there are several roller TO bearings out there that would meet the specs needed IF I could make an adapter yoke for the clutch fork. I wonder if I could cut the top out of an old carbon TO bearing cup and use it for a yoke? That coupled with a sleeve over the input shaft might keep everything lined up.
Thanks, Rut

 

[Gerard]

Well, I'm sure I could come up with a good solution with the help of my machinist.

 

[Rut]

I thought about going to a machinist as well, but that would kind of defeat the whole purpose of this exercise. I would like to get all the ideas I can and figure out how to do this at the most basic level (mine)and share it with the forum. Cheap, easy, and effective.
Rut

 

[Rut]

Mike, I've got a few parts and pieces and I'm going to start assembling this thing. Any pictures or drawings will be very helpful. Also, have you ever tried an internal hydraulic TO bearing?
Thanks, Rut