Old Harley Clutch Wars

This tech will be amazing. I built this 1950 Panhead during the Covid thing, so we called it the Pandemic. We fought with a magneto for a year and finally slipped in a used automatic advance distributor and it started first kick, with a Mike Egan R.I.P. rebuilt M-35 Linkert.

We ran into another issue. The clutch dragged. This bike, a poor basket-case roller came with lots of trashed parts. It did have a belt drive of sorts designed for running in tin primaries. But the clutch internals were made up of stuff I had laying around the shop.

I had issues with the clutch arm pivoting across the top of the transmission and running into the plate under the semi-stock oil tank. Like I said, nothing was exactly stock or correct. I shaved the arm, re-drilled the cable hole and notched it to align with the clutch cable bracket mounted to the frame.

I felt safe with my modified clutch arm since the relationship was working in my favor. The shorter the arm the more pressure plate movement from the handlebar lever. Although, it could add tension to the lever. But lever seemed cool and pulled easily. We didn’t dig into the spring tension discussion much.

I watched a Youtube with Bert Baker. He pointed out that your pressure plate must move .070 for your clutch to disengage. He demonstrated with a completely stock late model Harley, which moved .088. My clutch pressure plate seemed to move .090, so I was golden in that respect—I hoped.

At first, when I dropped it into gear it would kill the engine with a jerk. I took the clutch apart, cleaned and serviced it. Slightly better, and when I adjusted it, it improved, but only a skosh. Now, I could ride it a mile before it attempted to pull me into the intersection at the highway—duck.

Buck Lovell received an assignment by the supreme Cycle Source staff to shoot the Pandemic for a feature. I needed to repair the clutch, pronto. It had to be a running, riding old Panhead or die trying.

Buck said, “Those belt drives heat up and the tension on the clutch increases.” Sounded good to me and we loosened the transmission and backed off the tension on the belt. He even brought over a BDL tool for measuring belt tension.

We also looked into a new set of clutch plates and another solution jumped out at us. BDL makes a complete Belt drive, clutch system for early Harleys, but that would fuck with the Pandemic patina. Barnett makes replacement plates and springs. Buck knew someone at Energy One and we ordered their set of fibers, steels and two tensions of springs.

We were careful to remove all the cable tension on the throw-out bearing before adjusting the pushrod. It must completely release, when the clutch is disengaged.

Although this is a dry clutch situation the directions called for a brief soak in ATF before installation. Okay, the new Energy One clutch system with reduced pressure springs didn’t do the trick, and I spoke to Micah McCloskey who is a master, traditional chopper builder and engine builder. He said, “You must find a Roger Ramjet R.I.P., clutch retainer plate. They still make them but call them clutch retainers.”

I went on the hunt and ordered one from J&P, which my mail person lost, and J&P sent me another one—thanks. I pulled the clutch apart again and took out the bearing support anchor springs and plate. I replaced the plate with the clutch retainer and three small clip rings, which were a bear to install. This plastic plate has three adjustments to prevent the clutch basket from moving with the plates preventing them from disengaging.

J&P directions called for only .018 to .035 clutch basket movement. My first setting was too loose. I turned it two more holes to the tightest setting and installed the clip rings. This immediately made a difference. I could pull on the clutch lever with the bike idling and it wouldn’t move when I popped it into first. But once I rolled down the street shifting into 2nd was still a chore.

I had a date to ride the Pandemic to the photo-shoot location about 15 miles away. I barely got to the top of the hill, and it started to run on one cylinder and died. I goofed with the sparkplug wires, and it came to life.

Low on fuel, I decided to ride into Deadwood and refuel. It died again and I got more aggressive with the sparkplug wires. It fired first kick and I rode it a half mile to the Dinosaur gas station and filled ‘er up. If it acted up again, I was prepared to throw in the towel, but it didn’t, and I started riding out of the winding hills toward the interstate.

The bike handled good and ran like a champ. The clutch was fine as long as I was in 2nd, 3rd or 4th. We took the shots and riding shots. Every time I kicked it, it fired, but I wasn’t done with the clutch.

I attended a 50th Anniversary party, celebrating the Fucking Asshole and the Dream Girl’s marriage with a bunch of AMCA guys riding stock Knuckleheads, Pans and Indian fours. They know their shit when it comes to these models. Charlie told me to drill out the fiber plates with a 25/64s bit to allow for more flexibility.

Mike Kane explained that aftermarket metal plates are stamped out and the rough edge can catch on the clutch basket rails. Cabana Dan, a Hamster who works on early bikes explained that when stamped the metal plates have a smooth edge and sharp edge. The smooth edge should always face out, and he agreed with Mike. The sharp edge should be filed or smoothed slightly.

That turned into my next move, and we tore the clutch apart again. We carefully filed each metal plate and drilled out the fiber plates. It worked like a champ.

Dr. Hamster said seriously after road testing the Pandemic. “So, each clutch adjustment gave you 5 percent improvement.”

Who the fuck knows? After working on old Harley 4-speed clutches for over 50 years, I was blown away to find out so much about the bastards. This tech is currently running is Cycle Source Magazine. Get your copy today…

–Bandit

Sources:

Baker Drivetrain

Barnett Cables

Energy One

American Prime

BDL

J&P Cycles

 
Cycle Source Magazine 

 
 
 
 

Please follow and like us:
Pin Share
Scroll to Top