BT clutch release, analysis & functional history 36-84:

[Andygears]

“Any failed small parts, springs, pawls etc, with plate type shifter, can drop down into meshing gears.“

And the 5 speed brought back the ratchet lid in a new form but with the large hairspring that the end breaks off and drops right into the high gear mesh.

Andygears

[RooDog]

Add: Ditching of the ratchet lid and change to plate shift was to get rid of ratchet housing, because it was in the way of soon to arrive rear belt and its bigger front pulley, everything went down-hill from there.

Nifty

That may be what the MoCo had in mind, I don't doubt that, but the change to the "Maytag, Cow Pie" top predated the rear belted Sturgis, and FXB models by several years, and the rear belt did not become universal until late inn the 1980s. I'm not arguing the with you, Nifty, just wondering about H-D's thought processes. And the sloppy linkage required for that plate shifter to work is surely another t**d from our beloved MoCo....
....RooDog....

[nifty]

RooDog
Never thought for a moment you were being argumentative
However, the time window between "Turd top" trans and first FXB is narrower than you may think.
I worked in a dealership when these were new, but not 100% trusting memory...
Parts books show mostly -79 transmission part numbers and further state introduction as late 79.
I don't have an FXB parts book, but the first FXB Service Bulletin is H-D TSB M-775 which is dated May 06 1980 and relates to the non adjustable primary belt and trying to get both belts to work. The second FXB TSB was M-818, dated Mar 13 1981 and deals with the Jan 81 intro of new adjustable inner primary cover.
Argh the sloppy linkage, to fix (along with all the normal stuff) I have re-machined OE aluminum main arms and fitted bigger sealed bearings, and in desperation I once carved from solid stainless, a complete new main arm (FL type arms). Can't find a pic of stainless arm.
Cheers
Nifty

[nifty]

More clutch release data:

From 1975 onwards H-D sold 3 release bearing update kits, each with different length pushrod, to retro-fit the new for 1975 Torrington release bearing back to 1941.
37305-41 BEARING & PUSH ROD KIT 41-64
37305-65 BEARING & PUSH ROD KIT 65-69;
37305-70 BEARING & PUSH ROD KIT 70-E84
Simultaneously in 1975 the 39-E75 Clutch Release Bearing 37310-39 and relevant pushrods were deleted from H-D parts books.
37305-41 kit in fact retrofits to at least 39 and possibly 36.

These update kits have been replicated by the AM and at least some are further updated by 24mm Torrington thrust bearing. AM kits use above OE kit part numbers for reference.
The good AM kits include relevant length clutch pushrod, -39B release finger, “C” clip, daisy wheel oil slinger, 2 x 24mm plain thrust races, 24mm Torrington type thrust bearing, and centering bush for bearing.

24mm Heavy-Duty Release Bearing 10mm x 24mm x 2mm, 14 x 4.5mm long rollers, (bearing wafer only, requires bush, plain races etc.) V-Twin Manuf 18-3247; intended to be used with an ID bush .254” x .392” x .070” thick; corresponding plain races must be big enough to fully support thrust rollers. OE 37313-80 Thrust washers/plain races work with 24mm bearing. When the plain races show signs of wear they can be flipped over and re-used.
IKO thrust 10 x 24 x 2 = NTB1024; 16 x 3.6mm long rollers, OD at roller tips .840”, dynamic load 2068 pounds, direct replacement for AM 24 mm wafer.
IKO do not list plain races with bores smaller than 10mm.

Great care should be taken with foot clutch & mousetrap installations (stock heavier clutch springs, greater spring pre-load)
Extreme care should be taken with suicide clutch installations, (especially if no travel stop).
Due to the mechanical leverage through finger, clutch arm, mousetrap or foot pedal, the potential for poor adjustment to exceed design loading of even 24mm bearing is great. A non travel stopped suicide is the worst scenario.
Must never get into a clutch spring coil-bind situation.

With above caveat and provided pushrod & cable/rod correctly adjusted:
IMO the 24mm Torrington type bearing set-up is better than early -39 setup. “Petals” of daisy wheel slinger are more efficient at slinging oil around, especially if kicker installed; the complete Torrington bearing assembly is much lighter than the -39; the captive ball & spring at starter clutch are not required; when clutch engaged, none of the Torrington bearing parts rotate against each other (all bearing parts rotate harmlessly with pushrod, which is driven only by seal/seals in nut); when clutch engaged, there is no connection between finger and any part of bearing. Thus IMO the 24mm Torrington needle roller thrust bearing is technically superior, but undeniably less robust than a good OE -39 bearing.
Quality replacement 24mm bearings & plain races are easy to get & are relatively cheap. It is easier & simpler to install a complete 24mm kit with all parts from pushrod to finger. Which I have done in my own 68 box.

Additionally: Dual seals in clutch nut increases drive torque of pushrod & slinger.

Maintenance: Correct trans oil level, whenever oil checked or changed, have a feel of daisy wheel, if it rotates easily the seals in clutch nut are probably worn and due for replacement to restore full oil slinging capability. When the daisy wheel gets sloppy on its pushrod drive flats, its time to remove cover and replace parts. i.e. before failure.

Conversely, the AM has also produced kits to retrofit repops of the -39 bearing & pushrod to 75-84. internal clearance issues with some covers, Quality & completeness of kits varies, check before you buy.

For 36-64 only, perhaps one could revert to 36-38 pushrod, small ball type release bearing & propeller slinger, but currently these repop kits sell for approx 3 times more than 24mm Torrington kits.

Nifty

[nifty]

Confirmed:
36-38 clutch pushrod extension tip at bearing end is .350" long.
39up 37070-39 clutch release finger has slot .250" wide.
Thanks Paul

Nifty

[nifty]

BT Transmission Mainshaft Bushings.

All years, proper “fit” of clutch pushrod to both ends of transmission mainshaft, is an essential part of clutch release mechanism.
BEWARE some AM mainshafts sold requiring at least one bushing, but bushing not included in sale and no instruction to install a bushing before shaft used.

If pushrod loose/wobbles in either end of mainshaft, have a close look:

Sloppy pushrod at clutch end means the pushrod is supported/centralized only by the pushrod seal in clutch nut, seal secondary function is to drive/rotate pushrod, seal won’t last long if it is also trying to support/centralize the pushrod. When seal gets loose with -39 release bearing, mainshaft & pushrod rotate at different speeds which induces wear at interface of pushrod to bearing. When seal gets loose with Torrington wafer release bearing, the slinger stops slinging oil and all moving parts inside cover suffer.

Sloppy pushrod at kick end means release bearing is not held central, creates lost axial pushrod motion (pushrod forced sideways by finger), clutch plate separation reduced. Operational contact between release bearing and release finger is “off”, which reduces life of both; more likely for Torrington type clip to be displaced by contact with finger.
Some folk insist that when using -39 release bearing. pushrod bush only needed at clutch end. Their belief is that bearing fits nicely over the starter clutch and it supports pushrod. Others opine that when seal is present in clutch nut, no bush required at clutch end. IMO go with close fitting bushes both ends.

37436-36 BUSHING, Transmission mainshaft end, approx .375” x .435” x .500” long. (for clutch pushrod, sometimes just one end or both ends of mainshaft, some bronze, some steel, 36 to Early 80 – El, UL, FL, FX, FLH, FLH•80, Classic
Eastern A-37436-36, steel, from Vital V-Twin Cycles $1.55 ea
Eastern A-37436-36 steel from J&P 620-485; J&P Warehouse 1073971; $2.99 ea
V-Twin number 17-9830 steel, advertised as Eastern, from Dennis Kirk $4.39 ea
IMO fitted clearance .002” both ends, after straightened & polished pushrod lapped to bushes.

Andrews mainshafts come with pre-sized bronze bushes both ends, last one I bought I measured and found clearance to new pushrod to be .005” clutch end .003” kick end.

Pushrod must not bind in mainshaft as clutch arm moves (release finger “tips” travel through an arc against release bearing, similar to rocker arm pad to valve tip ).

Due to seal at clutch nut, pushrod should not spin in mainshaft, but when clutch operated, pushrod is always moving in and out. During assembly, work some moly grease into mainshaft.

More starter clutch data
2143-31 Starter Clutch BT 31-38, 2 keyways; has no captive ball & spring (used V models & 36-38 BT with small release bearing)

Nifty

[RooDog]

Nifty....
Do you have any insights into the later Ball & Ramp release mechanism? Particularly the degree markings on some of the ramp plates. I assume this has to do with the leverage, ramp angles, and amount of push rod travel relative to the cable pull. I know it is a balancing act to get hand lever effort, spring holding power, and disc separation to work together.
Or is there somewhere I can look for this information....
Thanks.....RooDog....

[nifty]

Nifty....
Do you have any insights into the later Ball & Ramp release mechanism? Particularly the degree markings on some of the ramp plates. I assume this has to do with the leverage, ramp angles, and amount of push rod travel relative to the cable pull. I know it is a balancing act to get hand lever effort, spring holding power, and disc separation to work together.
Or is there somewhere I can look for this information....
Thanks.....RooDog....

The markings are supposedly the angles of ramps i.e. 15, 18 & 21 degrees relative to ramp plate, I have never measured to confirm.
15 degrees easy on the hand but not so good for plate separation because there is simply not enough travel. 21 degrees hard on the hand, maybe enough separation when all the stars align.

No specific resource I can recommend
My 2c
Balls & ramps worked well in Triumphs for decades, but apples & oranges if trying to apply to BT.
Rodan tried it on 4 speeds back in the 70's
IMO For hand operation of both the total spring pressure and plate separation required for a Harley clutch, the ramp & ball setup isn’t up to the task.

Not really my area of expertise, but I know there has been lots of dickering with ramp angles, extending the arm of cable ramp, hand levers with different crank angles and bearings at fulcrum, different thickness diaphragm springs, blah blah.
Most of which appear to me to just move the problem around and make adjustment more critical.

There is one area that may make the balls & ramps easier to live with.

Right up front, I’ve never done this on a Harley, but I have on Norton Commandos, and that is to carefully shim the clutch pack. The object of the exercise is to pre-load the diaphragm spring so that ultimately the hand lever will feel quite stiff/hard at initial pull, but as the diaphragm spring goes over-centre/inside-out, the force at hand lever drops off dramatically, so it’s a breeze to hold released against grip.
This does make the clutch a bit more sudden death in operation, but it yields maximum plate clamping force with min holding released force, 2 out of 3 might be as good as it gets. Because you are increasing the spring pressure by packing/pre-loading, you might need to start with one of the lighter diaphragm springs made for easier operation. But on the Nortons all parts were stock except the added shims.

If you can’t get clean separation you might have to play with less plates and more packing.

Once you find the sweet spot of pre-load at diaphragm spring, measure and record it’s as-installed “out of flat” or “dish” dimension.
As plates wear in use, don’t play with cable & pushrod adjustment, instead, re-shim the pack to sweet dimension and then adjust cable etc. I do remember on the Nortons that the shim ended up being pretty well max thickness that would still allow insertion of retainer with diaphragm compressed flat by tool.

For packing you can hand cut sheetmetal shims about same OD as diaphragm spring but only 3/8” or so wide. Shims outside diaphragm spring before retainer, or however your particular clutch works. One thicker shim always beats 2 or more thin ones.

A lot of sucking and seeing, but if you have the time and inclination, why not?
Good Luck

Nifty

[RooDog]

Nifty....
I'm dealing with a Sputhe conversion 5:4 in my 90" '68 Shovelhead, and also a '86 fiver in my Evo with a Rivera Pro clutch. The Evo works well, but the Shovelhead with Barnett Kevlar linings, and 10 springs is messing with me. So I was looking for info on the B&R mechanism. I think you have answered my question: The lower the number, the less effort, but also less pushrod travel/separation. That's all I needed to know....
Now it's just a matter of swapping parts around until I find that elusive "sweet spot".
I also have both a Barnett Scorpion, and a Rivera Primo clutches available for that Shovel....
Thanks....RooDog....

[nifty]

Roodog
Another 2c
Assuming hand clutch
If you can increase your effective hand lever travel/cable pull at lever, it might get you where you want to be (together with ramp angle you can live with)
Perch with fulcrum further from bar
Hand lever with greater distance between cable & fulcrum
Zero slop fulcrum
Zero compression outer cable
Zero stretch inner cable
Smaller OD hand grip
Nifty

[nifty]

BT RELEASE SHAFT (vertical)

All years, proper clearance “fit” of release shaft to both steel bushes in cover, and proper fits of operating arm and finger to release shaft “squares” are essential to good function of clutch release mechanism.
Hammering on clutch arm to remove it from shaft can shear split-pin/dislodge E-clip inside cover, use 2 jaw claw puller or similar to remove operating arm.

37074-38 Release shaft 38-69 retained by cotter pin (split-pin) at bottom, used with washer adjacent to pin. Can be used to 84+

37074-38 with hole for pin repop.jpg (407.29 KiB) Viewed 723 times

37074-38A Release shaft, BT, 70-84+, retained by E clip, will fit back to 36. Through 1978 at least, OE versions of -38A shaft have no undercut where square sections abut major OD; later production OE -38A shafts have deep undercuts. IMO undercut shaft more likely to break at upper undercut, especially if used with a pre ‘68 long clutch arm. Modern day MoCo parts aren't always better.

no nasty undercuts in earlier clip type shaft

37074-38A Release shaft 78 OE.jpg.jpg (7.26 KiB) Viewed 723 times

vipers nest of stress raisers at undercuts

2008 undercut version -38A release shaft OE.jpg.jpg (225.55 KiB) Viewed 723 times

Most AM release shafts are copies of 37074-38A; AM supply both undercut and non-undercut styles, AM shafts use same E clip as OE.

Release Shaft upper & lower male squares, nominal ½”, measured example .495”;
All shaft squares must be a good fit in fingers and arms i.e. arm should be firm to tight and finger should slide on without slop.
Release Shaft upper OD .605”-.606”
Release Shaft lower stub OD .408”-.409”
Length of lower stub (E clip type shaft) approx .975” to shoulder, some repops are significantly less and thus useless unless lower face of lower cover boss & bush ground away.
5 speed release shaft has 2 flats at finger end, shorter

2 flats for finger

80-85 5 speed release shaft.jpg (8.52 KiB) Viewed 723 times

11000 Clutch Release Finger Shaft Ring 70-E84 (E-clip for 5/16”- 0.312” shaft, 0.022” – 0.025”thick) also fits most AM shafts. Manuf: 96-309, J&P 630-243, J&P Warehouse 1026944; $0.99

6443H W, WASHER, Finger Shaft, (upper hardened thrust washer nominal 13/32” x 11/16” x 1/16” thick, 36 to E84 BT, goes between bottom of finger & top of lower steel bushing, previously 2436-36A / 6320 / 6442); actual measured dimensions 0.420” x 0.683” x 0.064” thick. With many OE covers the 6443H W washer sits in a machined recess. Ensure thrust washer sits flat either on cover, or on top of lower bush. Thrust washer must be dead square to release shaft. V-Twin Manuf 12-1158, J&P : 620-741, J&P Warehouse 1029563, $1.99.
Can replace 6443H W with a nominal 3/8” grade 8 flatwasher, (they are hard and usually fit over Release Shaft lower stub OD .408”-.409”. USA made Brighton Best gr 8 washers are approx .415” ID x .817” OD x .064” thick) Ensure gr8 washer OD does not foul cover, or release bearing, might have to grind gr8 washer OD.

6702 WASHER, 5/16”x 11/16”x 1/16”. all split pin type shafts, goes on below bottom bush, before the split-pin. However, many times a thinner washer is found in this location (probably an unrecorded factory floor work-around).

Keep release shaft washers separate/tagged, the washer against finger should be hardened, specific OD so it clears -39 release bearing & cover, specific thickness for finger to be centralized to mainshaft.

528 Release Shaft cotter-pin/split-pin, must be secure, “stick-out” must not foul cover
In “E” clip release shaft applications, there is usually no room for lower washer and none is listed in parts books (plus gravity keeps finger/fork against thrust washer).

RELEASE SHAFT BUSHINGS
33290-38 Release lever shaft bushing LOWER steel approx .410” x .548” x .740”

33291-36 Release lever shaft bushing UPPER steel approx .606” x .862” x 2.253”, big outside chamfer* at bottom for finger & -39 bearing clearance, zinc plated. *AM covers often foul -39 release bearing at this chamfer, relieve as required.

Release Shaft fit in bushes, no OE spec, should be free without excessive slop, some are found worn to hell (.040”+) and thus a source of lost motion at clutch plate separation. With new parts I lightly lap shaft to both installed bushes, just enough to knock down high spots & smooth operation.
I use measured clearance 0.010” max, prefer .005” or less after light lapping.

NOTHING KILLS A TRANSMISSION LIKE WATER.

Top of release upper bush is a water entry point, more clearance = more potential for water entry.

not the worst I've seen

probable pressure washer water damage 5 into 4 33353-78 long bumper and cast deflector.jpg (216.36 KiB) Viewed 723 times

For a reasonably effective home-made water seal, an approx ¼” wide neatly cut slice of 5/8” ID oil resistant rubber hose, installed between operating arm & upper bush (with a little water-proof grease), helps keep some washing and rain water out.
Adjust width/height of new “seal” so just slightly compressed when operating arm installed. Will not withstand determined assault by pressure washer, but neither will anything else.

just a little crush & some waterproof grease

home-made water seal at top bush.jpg (127.93 KiB) Viewed 723 times

present to myself

one of a kind stainless upper bush with real seal & grease nipple rsz.JPG (174.38 KiB) Viewed 723 times

Nifty

[RooDog]

Correction
Delete: "Then there are the inconsistencies, like keeping the ratchet lid trans on FXWG, long after everything else had changed to plate shift lid."

Add: Ditching of the ratchet lid and change to plate shift was to get rid of ratchet housing, because it was in the way of soon to arrive rear belt and its bigger front pulley, everything went down-hill from there.

Any failed small parts, springs, pawls etc, with plate type shifter, can drop down into meshing gears.

Nifty

It's funny that the first few years of the rubber mounted, 5-speed FXR, FLT, and the 4-speed kicker FXST still retained the rear chain drive. I just suppose the MoCo was using up parts they had in stock, as they are wont to do.... Ya' gotta love 'em.....
....RooDog....

[panhead_kicker]

I've seen this idea of the MOCO using "using up" previous model year parts on many threads here....
it's new model year 1966, a customer needs to warranty a cracked head on his 65 Electra-Glide he bought last year, but the MOCO tells him "SORRY!, We used all those up!!"
I don't buy into that line of thinking.
I don't have a different explanation, but this one just doesn't seem to hold water in my opinion.

[nifty]

[/quote]

It's funny that the first few years of the rubber mounted, 5-speed FXR, FLT, and the 4-speed kicker FXST still retained the rear chain drive. I just suppose the MoCo was using up parts they had in stock, as they are wont to do.... Ya' gotta love 'em.....
....RooDog....
[/quote]

IMO In these kind of instances, the MoCo were hedging their bets against buyer resistance by offering both chain & belt and while preferring to sell the "newer" design, didn't really care so long as the public bought one or the other.
Also back then and earlier, the MoCo didn't spend much on exhaustive testing, instead they relied heavily on dealer feedback from fixing failures. Regulations were different, no major threats of enforced recalls.

Panhead-Kicker
Your example is a little extreme, but if it were real, the MoCo of the day would offer Shovel top end as being backward compatible to repair the motorcycle.

As for "using up parts", consider this, every part costs money to design & manufacture, its more frugal/cost effective to use a part that is either gathering dust in the warehouse, or which was used previously and blueprints-tooling were still on hand. The MoCo have always done this.
Just off the top of my head:
Topper carburetor studs re-purposed for 65up BT electric start
K/XL cylinder studs re-purposed for 65up BT trans to inner primary studs
Topper dowel pin re-purposed as travel limiter in 80's BT oil pump
Etc etc ad infinitum

Nifty

[nifty]

More release data:
Interesting that July 1968 TSB 573 makes no mention of ball & spring in starter clutch being a possible contributing cause of wear at pushrod & adj screw. Perhaps MoCo knew there was a problem with some particular bearings?

Except as an illustration of what 37310-39 bearing looks like, the drawing appears to have little relevance to the text.

However, the drawing of 37310-39 bearing is itself interesting in that without any text reference, it shows the 1/64" clearance required between tin shield and outer race, this clearance is a lubrication pathway. Some repop bearings come with the shield actually rubbing where this clearance should be. Perhaps MoCo were hinting that in 68 some OE bearings had insufficient clearance as indicated?

Curiously, the drawing pushrod appears to be 36-38 with extension tip. Perhaps the drawing was extracted from a much older MoCo drawing? Which may indicate original intent for 37310-39 bearing was to use pushrod with extension tip located in .250" wide slot of -39 finger?

Nifty