Help on Backfire Overload Device

I removed my starter and gear train 35 years ago, now I am putting it back on for my rapidly aging body. I have successfully modified a Harley starter to replace the weak OEM one. Now I am in the primary, reassembling the gear train.
I am especially concerned about a backfire damaging the expensive sprag bearing, so I'm wanting to set up my backfire overload silp torque per the manual.
Old Britts has an excellent article on this, but reading and looking at this picture, I can't determine which way to torque it to test for 50 ft-lbs slip.
Here is the test picture:


Can anyone tell me which way I am supposed to turn the shaft?
And does anyone have a suggestion as to how I can put a wrench or something on the splines to apply the torque without damage?

Thanks,
Jaydee

Not a clue but it's been 30 years back for moi anticipating such a future inevitable fate too. So will watch this thread. Back then I had a Harley -conversion done to 4 brush plate but never installed it as Crazy Combat entered my life then-ish. I hear 50 is important ,weaker being better a tiny bit.

To simulate a kickback situation, the shaft to which the wrench is applied in the photo would be traveling counter clockwise. Although, when you look at the design of the device, it should release at the same torque in either direction.

As Fred points out, the nut will tend to loosen when you turn the wrench, so you will need to mark the nut, check the release torque, make adjustments, mark again, check again, and when you are satisfied tighten the nut to the last mark you made and stake and/or Loctite the nut. This is a case where lighter is better than tighter, as long as it will turn the engine without slipping.

You can do like Fred and weld a thrust washer on an old socket, or you can rig up a socket and appropriate size rods to fit the splines. I forget the sizes, but you can experiment.

There's been some deep threads on this device [google it] and end result was its innately weak and no one has solved it but to keep a spare on hand as a number of tales reported of new ones only lasting a few dozen starts. Any upgrade items may help extend to a few more dozen starts. Jim Comstock gave up on his a few decades ago if that's helps your sense of mechanical skill solutions on thumb commensors. For sure don't set initial spark timing very near its optimal power setting. Might be worth while to try cryogenic tempering. If a compression release were developed that may allow the poor backfire device to survive and maybe make redundant as just step on starts so easy.

You should use a new nut if you remove the existing one. The nut collar will likely be buggered by opening the existing peened section.
Looking at the pic you lifted from OB's, pulling on the wrench will test the drive torque preset and pushing would test the overload against backfire.
I find it significantly easier to mount the overload device vertically and to use a deflecting beam torque wrench with a floating indicator needle for testing existing breakaway. The clicker wrench is a pain in this application.
The overload device is not a precision instrument. Breakaway will vary somewhat by each slip increment since there are multiple balls and driving holes and it will also vary by clock direction, so, one slip in each direction is not an accurate test.
It won't be precisely repeatable either so don't let that be confusing, just keep in mind it isn't needed to survive a moon landing. Record and retain your settings (initial and final) for reference.
A proper fixture with a pin to restrict the gear tooth movement would be better than soft jaws in a mangled vice but is more involved if you lack the facilities.
My opinion is that the overload device is slightly better than nothing at preventing damage to the starter drive train. The intention is good, the execution lacking. If set light enough to preclude any damage it will not reliably drive the crankshaft to start the engine. The promised Alton kit for the MKIII may be the cat's meow if you can wait and original appearance is unimportant.
Wow, sorry for the long winded response, even after deleteing a whole bunch.
A good ignition system properly set and maintained along with a tip top battery will go a long way in preventing backfires.
All the best and enjoy the holiday.

Thanks guys. I took this device off shortly after I got the bike new, so it is OEM pristine. Based on the comments, I think I will just test it as it is and if it is anywhere near
50 I'll leave it alone. AntrimMan made some good points about it not being consistent and precise anyway. I have a torsion beam wrench, so that's a good suggestion. I also like the pins in a socket idea to turn it from RonL.

It will be interesting to see what the factory had it set at. I'll let ya'll know.
Happy Thanksgiving to all.
Jaydee

jaydee75 said:

Thanks guys. I took this device off shortly after I got the bike new, so it is OEM pristine. Based on the comments, I think I will just test it as it is and if it is anywhere near
50 I'll leave it alone. AntrimMan made some good points about it not being consistent and precise anyway. I have a torsion beam wrench, so that's a good suggestion. I also like the pins in a socket idea to turn it from RonL.

It will be interesting to see what the factory had it set at. I'll let ya'll know.
Happy Thanksgiving to all.
Jaydee

Click to expand...

Theres a very good article in Oldbritts. website Technical articles - overload backfire device . In Gearbox section

http://www.oldbritts.com.

Sorry, looks like you've allready found the article

sam

When I check the torque setting of my backfire device, I place the unit in some soft jaws as shown in the first picture to hold the unit, and to turn the shaft to check the torque, I've bought myself a spare thrust washer (part No. 065652 £2.42, the one that fits on the end of the splines on the unit) and ground it into a "hexagonal shape" so that it fits a socket, and then I attach the torque wrench (no welding req'd)......and see when the unit "gives."

Thanks,better than "giving up". Alton where are you ? :idea:

Reggie said:

When I check the torque setting of my backfire device, I place the unit in some soft jaws as shown in the first picture to hold the unit, and to turn the shaft to check the torque, I've bought myself a spare thrust washer (part No. 065652 £2.42, the one that fits on the end of the splines on the unit) and ground it into a "hexagonal shape" so that it fits a socket, and then I attach the torque wrench (no welding req'd)......and see when the unit "gives."

Click to expand...

I did the same thing but left a 1mm undisturbed land to act as a depth stop, as in a flanged nut, so it wouldn't slip out of sight into the depths of the socket. It's buried in the garage and no immediate photo available.
I did a similar alteration with putting opposed flats on the cross shaft coupling for the gear change so I could verify gear change while in the rebuild fixture mounted in the bench vise (the MKIII gearchange shaft exits blindly pointed toward the drive side). It's also buried in the garage and no immediate photo available. Winter weather sucks.
All the best.

OK, I did the test. I cut a 2" aluminum pipe in half, flattened them a little bit, and put the 2 arcs in the vise holding the big gear, shaft vertical. I experimented with sockets and small rods to find a combination that would lock onto the splined shaft. Then I applied torque until the shaft just started to move, it came out to 46 ft-lbs both ways. I did not turn it all the way to the next ball slot, just enough to make the balls climb out of their pockets. It may peak at 50 if I had clunked it all the way. I think it is working fine, so on it goes. Thanks for the help.

Jaydee