Weird 915 Race Gearbox Problem: Need the braintrust..
 

Ok Peeps this has me and others totally stumped..... and driving me friggin crazy...

I was at Sebring last week for SVRA event..beforehand i had Kevin Wheeler and the boys from KMW swap my spare gearbox in....( the other was aging a bit and grinding a bit and would not stay in 3rd). Both are short geared , professionally built, plumbed for external cooler , and had a temp sensor logged on my AIM system etc.

The spare was a ZERO hours, never had oil in it, properly stored one from a very well known Porsche gearbox dude. Clutch and everything else were fine including linkage. I changed fluid from Mobil1 synthetic to a non synthetic custom mix that the boys from 901Shop ( see below) have found to be best for 915 gearboxes.

This box shifted fine in the pits .. but as soon as i got it on track, it would completely freeze up .. this happened every single start i tried it. I could not even get one warmup lap, under just warmup conditions, it became stuck in neutral. lever would go back and forth in neutral but would not move remotely forward or back into any gear of 1-4. I could only get it into 5th , and the ooch it back to the pits ( with throttle and clutch..)

Even stranger , when i got it back to pits, after it sat for 10-15 minutes, it would then shift again ( with or without having motor running in neutral during that period). this reversal happened gradually and allowed getting 3-4 back , then 1-2 came back a little later. literally this then felt normal. then when i went back on track, the exact same thing happened ( almost on the exact same section of track). For instance this almost always seemed to occur after the hard rights at T10 and T13 (where there are also alot of bumps). (i know this because i ended up pulling off course each time at the wall safety car gaps at T14 or the new pit entry at T16).
BTW. these were very gentle shifts during warmup : of course proper rev matching heel-toe, and nowhere near the type of shifting i would do in a race.

It seems as something is getting misaligned or jerked out of place inside the gearbox. The linkage was not loose or binding ( In fact worked fine for 5th and revers). The shifter ( standard wevo shortshift). was solidly mounted with no play and there was no play in the shift tube or connector to the trans. I even llfted the shifter unit up from the tunnel mount with a few thick washers to create more clearance for the rod movement in the tunnel... did nothing. All parts of the shifter were fine, including the nylon cup etc. The motor mounts ( both trans ands motor mts are yellow stomski: semi solid) both fore and aft are fine, nothing seems loose. There are no seeable craks in the torsion tube.

THis does not seem to be temperature related. Same thing happend if the trans temp oil was 90 degrees or 150 degrees,. Moreover, I was told categorically by the 901 Shop guys ( Scott and Brady ) that "915 boxes WILL NOT SHIFT if you use synthetic oil... they learned this lesson the hard way" . I was very happy to hear this sunday so i changed the mobil 1 synthetic to their custom mix ... but the same thing happened.

One more thing that is totally strange. I took the car out onto rte 98, (the sebring access rd) and ran it hard up and down the road at the 50-70 mph ( its registered insured and plated). up and downshifting for 35-40 minutes, hot gearbox etc,. AND IT SHIFTED FINE FOR THIS WHOLE TIME!.... So the problem only occurs on course at sebring under pseudo race conditions.

This i cannot explain other than the bumps and hard turns on track vs off. ITs maddening. i did not complete 1 lap in that whole SVRA event... even the announcers on the cool-as-hell live feed on the SVRA App. noted that... "the green RSR from RJR Racing seems to having some mechanical problems this weekend..."..:mad:

Any thoughts?

That's a bummer. Sorry about your race weekend.

It almost sounds like the interlock is keeping you "stuck" in nuetral - you can experience this if you put the lever between gears in any 915 car - place the lever exactly between 1st and 3rd and gently feel around, you'll feel a bit of a notch, try pushing the lever into gear and you'll feel the lever move a small amount and comes up against a hard stop. That's the interlock doing it's job.

I wonder if it's a coupler adjustment problem? Of course that doesn't explain why it happens at random..... maybe a loose coupler clamp? I have one that needs WAY too much torque before it tightens up, and it can slip and throw things off. I have a WEVO billet clamp on order. This could be even more likely if you have a wevo shifter because the spring tension is applied through the coupler at all times when in 1st, 2nd, and 5th/reverse.

Also, is there any chance you have a Wevo gateshift in the box? Again, coupler misalignment could be the culprit there and given that it's a very exact control over shifting, it needs to have the shifter adjusted perfectly.

Just to add - there are interlock parts in the trans that if it's not reassembled properly, could be on the wrong side of the shift rails in the mid-case. Not sure what the symptoms of this would be, but wonder if the little detent capsules are bouncing around uncontrolled somehow and led to the problem only on the one bumpy corner?

Thanks Jonny . I have the Stomski precision shift coupler. the one pictured with the set screw. The set screw is conical and seems to take up any minute play there is in the whole in the shift shaft.. Its hard to believe there would be a MORE PRECISION coupler out there. but who knows. i will try anything.

have no idea what the internal setup is in terms of gating ... will ask my gearbox builder. but i doubt its wevo.... nothing saids about it on invoice.

thanks very. much for your comments

http://forums.pelicanparts.com/uploa...1614042364.jpg

Is that your actual setup? If so the blue billet clamp is probably a good upgrade. My coupler clamp is the stock one and over time they can stretch and become less effective.

You would know if you had a gateshift, I think - they are not cheap. You can tell by looking at the bottom of the trans, the fork cover is billet so its visually apparent.

Sorry I can't help more..... I know what a letdown that can be. Lots of work and expemse to get to a race weekend only to never take a green.

Bent shift rod from someone in it's past life lowering the gearbox and engine while it was still attached to the linkage?

I would take a real good look at your chassis sounds like the torsionbar tube might be
coming loose and letting the trans mount moving around witch would misalign
your shift linkage . A lot less stress on the road that might explain that.

MATT it did look like the shift rod, the part i could at least see in the tunnel, was warped just a bit because it looked to be slightly bowed as you pushed lever back and forth in neutral. but it does not explain why it shifts fine in the pits,.. .. i am looking for a tower type 915 shifter ( fabcar? hargett?) to get the rod out of the tunnel , so i can exclude that variable and also see that the rod is straight and turning properly. also, raising the shift rod a bit ( in case it was jamming on somnething in the tunnel) using the thick washers did not solve the problem.

i agree blue. something is moving. problem is i cannot tell if its internal or external to the gb.

Matt. its possible that the KMW boys bent the shift shaft while swapping boxes in the paddock.. always a bit awkward to do. ( i did not have this problem with the original box in there...that was a whole different set of problems. i think)

I doubt it. Not only are they professionals, but it would be on removal, not install.

Can I assume that on track you simply could not move the lever fore and aft (except in 5th, and reverse when back in the paddock)?

You would not expect an experienced gear box rebuild shop to the the interlocks (detents) wrong. It is easy to do that if you are working from a parts manual or shop manual in your own back yard - you can get the funny little slugs on the wrong side of things depending on the angle you have the case at on your stand while reinstalling the shift rods. Been there, but whatever issue it caused me wasn't intermittent, that I recall.

You could experiment by removing the two external bolt plugs and removing the springs and whatever else a magnet can fish out (look at a parts diagram to see what these interlock bits look like) and then putting the plugs back to keep the oil in place. If you are careful shifting you won't get stuck in two gears at once. If you have a really bumpy local road, and it replicates the problem Sebring created with things as they are, and removing the springs improved things, that could be a clue - but all that's improbable.

I am having trouble seeing how a bent shift shaft would cause this issue. The 5/R shifting is the closest to the nose cone, and depends on a flat finger on the shaft fitting a rather shallow notch on the shaft the shift for for 5/R. So I would think a bent shaft would be most apt to screw up 5/R. Problems with that are what lead to the "locked in reverse and 1st" problem.

I once forced a shift into 1st when angry, which bent the forged part of one of the internal shift rods where the dongle engages one of two slots. Left me with fewer gears (no 3/4, as I recall) until I took off the bottom plate and fixed things with a crow bar. Silver lining was I then understood how the linkage worked in there. For 5/R the dongle is rotated beyond where the two notches for the two shafts are located, and swings free, letting the finger do the work.

While the rotary action (moving the shift lever right/left in neutral) is unaffected by the dongle, the fore and aft action (which is how the dongle moves the shift shafts) is created by the fulcrum fork which is attached to that 4 stud bottom plate. Hard to imagine the nuts inside holding the fulcrum to that plate are loose after a professional rebuild, but you might want to drain the oil, remove the plate, and see what is going on inside.

Again, none of this makes sense as an intermittent issue.

http://forums.pelicanparts.com/uploa...1614053583.JPG
http://forums.pelicanparts.com/uploa...1614054245.jpg
http://forums.pelicanparts.com/uploa...1614054245.jpg
http://forums.pelicanparts.com/uploa...1614054245.jpg

The last I think is the reverse/5th lockout, not the one with bolt plugs. Not usually removed as par of a rebuild. The ones on the side are. For what it is worth.

Shift forks loosening on their shafts? But that doesn't correct itself. Shift shaft alignment? Again, not self correcting or in any obvious way bump dependent.

Good luck

Looks like one just came up for sale.
https://forums.pelicanparts.com/porsche-911-used-parts-sale-wanted/1086739-fs-hargett-shifter-915-a.html
Dean

Sorry for the bad luck geneman. Sucks to go to Sebring only to have this happen.

If you want to distinguish between a problem with your shifter vs. one in the trans, disconnect the linkage from the trans while it's locked up and see if you can shift it by hand using the rod on the nose of the trans. Might be tough I realize since the problem goes away so quickly.

You might be able to trace the issue by disconnecting the linkage and playing around with it while cold.

Also, bad engine or trans mount moving around too much, or one of them came loose?

I was about to make the same suggestion. Only 4 bolts to check, I've had the rear engine mount bolts come loose

Yes. My suspicion is a "compound" issue where the bad part isn't bad enough to cause the problem by itself.

I see you have some good discussion going here. I will think about it and if I come up with something I will certainly chime in further.

Does not sound like a shifter problem. So unless you simply just prefer something different from what you already have for non-related reasons, I might stand pat on the shifter itself.

I will re-read all of your posts. And if you want to discuss off-line, feel free to send me an e-mail. I think you still have my e-mail address.

Thanks Mike and everyone. Really great analyses. Will let you know what turns out. Frank

good advice on the shifter unit... likely not the culprit.

The intermittent portion of the problem leads me to believe it's chassis related. Something is binding or moving during cornering that will not allow it to shift.

We see this a lot with the 80's tube chassis race cars where the engine/trans is a stressed member. When the engine comes loose from the tub, the car won't shift. Nothing wrong with the trans; fix what's broken and it shifts fine!

I used Redline synthetic in my 915 for about 15 years without issue. My application is track driving and racing.

I used their NS version for many of those years.

No need for debate on tranny lube.

From what I have read, the fact that it seemed to shift find on a street run is perhaps most telling. Before opening up the 915 I'd want to check all four mounts carefully, as well as all of the bushings, et cetera. Really everything external, very very carefully.

Longshot... I'd also check the pedal box area very carefully to ensure junk is not under the clutch pedal. Back in 2002 I had such an issue at Laguna Seca. Wife and I had driven the 911 about 6 times by then, coming from a 951. I thought she was having trouble with the on/off RSR clutch. Wrong. It was crap under the clutch pedal.

She ended up #2 in the TT that weekend. The next time out she got TTOD overall.

After that experience I cleaned out that area after every event. Lots of crap under there, especially rubber. And that crap can move about.

I also bought my buddy a small shop vac as a bachelor party gift after I had my experience. He kept it in his trailer and we could all use it at events to get all the rap out from under the pedals.

Again, this is a longshot. Will think more on this situation.

Meanwhile I suggest more street driving to see if it can be reproduced.

Nice to see we have a Pelican in my home town, MKE.

Hey Frank sorry to hear about this. Gordo here on the forum did a nice video that shows how the main shift shaft inside the transmission transfers movement to the shift fork units. It's a really nice complement to the descriptions Walt provided

https://www.youtube.com/watch?v=LLEWjNRjJTI

I would take a look at the bottom cover of the trans and see if the guide fork is at all a potential culprit. That guide fork acts as a fulcrum to transfer the fore-aft movement of the main shift shaft to the shift fork shafts. It would seem that the "dongle" is able to move side-to-side within the "windows" of the shift fork shaft lugs, as you wiggle the shifter in neutral plane. But when you go to move the shifter fore-aft, it seems that you cannot make the shift forks move. Could be the dongle fulcrum, could be the shift fork shafts.

The shift fork shafts simply move through bores in the transmission case. As long as the forks are properly secured on the shafts, they should move the selector sleeves. Of course the shift fork movement is dictated by the allowance of movement between selector sleeves and tripod hubs for 1-2 and 3-4 gear selection. Those pretty much are what they are and there's no adjustment to them. The 1st-2nd and 3rd-4th sleeves and hubs are not interchangeable (noticeably different in size) so they must be paired together properly.

The above info from me is somewhat meaningless because you were able to shift it fine for a long time on the open public road.

So I have to jump on board with the other guys and say it's external. I would go over the mounting points:

• mounts in the trans crossmember- two main M12 bolts that secure the trans to the torsion bar tube
  
• the four M8 bolts and nuts that secure the mounts to the crossmember
  
• two M12 engine mount bolts in engine bay
  
• the four M8 bolts and nuts that secure the mounts to the engine bay
  
• four studs in the tail case to mount the crossmember to the trans
  
• check four nuts that secure the trans to the engine
  
• check the torsion bar tube to make sure it's not come free from the chassis. I've seen it crack at the outboard mounting points in the wheel wells. This was on one car that had coilovers and the spring plates were replaced with hard-connected 935 spring plates

Don't want to get too far off in the weeds...

Here are some things I do (my application is racing):

I use a WEVO bearing retainer plate. Although one of my expert friends believes that the best benefit of such a plate is with alu boxes-- he believes that the steel inset with mag boxes like mine render aftermarket plates a little redundant.

I used metric j nuts and threadlocker for the bearing plate or alternatively socket head bolts and safety wire to hold the retainer in place.

I use socket head bolts and safety wire for the transmission mount to chassis locations. And I paint dot them.

I use a WEVO shift rod clamp.

After seeing a friend's race ruined by a loose coupler set screw a few years ago, I started to use thread locker on my coupler set screw. I never had mine come loose in about 15 years, but figured why not take a precaution.

Again, the biggest clue I think I see so far is that the thing shifted just fine driving up and down the street.

Chassis cracks... look carefully. The nutserts in the t-bar bearing cover area may need close inspection including a wire brush to see the bare metal.

I think the chassis cracks would have to be substantial to be a cause here... but again, everything probably oughtta be checked carefully

Surgeons leave sponges in patients. They count them and use products with RFID tags. I have stories about that...

I have to thank Matt Monson, Jerry Woods, Hayden Burvill (WEVO) and Mat Lowrance for helping me through the years... plus a few others. And yes I have had at least one tranny mystery in that time-- none of those guys had ever seen it before. That one cost a lot before I located the culprit. But it did not have the symptoms you describe.

Gordo's video is great - much easier to understand than writing about it. What he doesn't show is 5/R: there the dongle tip moves past the 3/4 slot and is out in open space, kept from flopping around by the fulcrum fork, but able to move fore or aft without hitting anything in that part of the transmission.

I still don't know exactly what the "feel" of this not being able to get into 1-4, though. Obviously it isn't just grinding of engagement teeth, or he'd have been able to get into 3d to get back to the paddock. It isn't popping out of gear (he'd have been able to hold it in a gear with one hand on the shift lever to get back off the track). I suppose it was more like the trouble, especially on the older 915s, of shifting into 1st when still moving along? You get the lever in the right position (and it is easy to get 1st into the right position - just hold all the way over) and push hard. The lever moves forward a bit, but that's it - no grinding, but doesn't engage either.

We know nothing obstructs rotation of the lengthwise parts - the dongle moves freely through its three positions, and 5/R work. 5/R working on the bumpy track sort of suggests nothing odd is loose inside the center tunnel of the chassis blocking fore and aft motion somehow.

If you position the shift lever between where 1/2 should be, and 3/4 should be, you can't move the lever fore or aft more than the clearance of the dongle head and its slots, plus the detents, because the detents prevent moving both internal shift rods simultaneously. However, since (without a Seine or other spring loading aftermarket shifter) you are free to move the shift lever right and left while trying to move it fore or aft, and I bet he tried that, this doesn't quite fit the symptoms. Even if the transmission was rotated you'd think he would have been able to get 3d and 4th if it rotated the way it did, and for that matter not 5th - if the dongle couldn't rotate beyond its two slots, for instance, or the 5th/R finger couldn't move far enough to engage that rather shallow slot on its associated shift rod.

But I endorse the check a whole lot of stuff before pulling the transmission approach.

This is really weird - Sebring is bumpy, but not all that bumpy everywhere - why would mount/chassis cracks etc move things around exiting the paddock onto the track when building up speed? But then sag back to where the shifting worked? I can't see a fore or aft up or down motion of the transmission causing trouble - that's why the shift coupler has some degrees of freedom to it. Which makes a rotational movement more likely, if not obviously so.

By now enough guys who have had enough problems or otherwise know a lot about how transmissions work have been stumped to make it especially interesting.

A very sincere thank you to everyone for responding so quickly and thoroughly. At this point i do not have any more hard info. to share. Your suggestions are superb. i sent my backup to the backup gearbox, just gone thru by P. Dawe down to Zotx Racing today. Hopefully they wiil swap early next week and do a thorough check of chassis, mounts, torsion tube, pedals etc. Stay tuned. Frank

Interesting thread … leaving us hanging … what was the culprit??

Should start a poll with our guesss...

I'd missed the "shifts fine in the paddock, not on the track" part.
Club Racers with these now quite old cars are seeing increasing cracking of the tub. One place this seems to happen is at the rear firewall where various parts - the floor pan, the firewall, the tube for the shift rod, and the center tunnel - come together. The operational symptom reported is inability to get gears on track.

I haven't actually looked at a car with this issue, and don't know specifically how it would misalign the shift mechanism (maybe it could cause shift linkage parts to rub or hit chassis parts?), but I believe the shop guys who have mentioned this - it is a real chassis issue. So one could check for cracks.

I saw Frank at the track in September and it seemed fine then, he must’ve gotten it sorted.

Frank, Frank … come in Frank …

HI Everyone. Rosco. Thanks for jogging my amyloid addled mind to respond. and anyones interest in this saga. Well. ... JeezusFC.... what a saga. I think 915s and me are cursed. Anyway. I sent the supposedly pristine newly, totally refurbished , never used box ( THAT WOULD NOT SHIFT AFTER 1/2 lap at sebring) back to the guy who rebuilt it . TOTAL PROFESSIONAL. He told me it was fockin armageddon inside the box. everything. dogs, bands, shift rods. etc , SIMPLY MAKES no SENSE at all. The only real issue was that I had used synthetic lube in the very beginning and during that weekend at Sebring.) and not the recommended Swepco 210. Plus, not sure my trans cooler was working well, but temp never got over 160 which I log on my AIM system. even when I was hammering it outside the track. Matt tells me that the wrong lube in a highly stressed racing 915 box can cause quite rapid deterioration. BTW I had the 901 shop at track at sebring exchange the synthetic for their "custom. magic 915" gear oil.. bluish stuff .... that still did not solve the problem) . So. to be clear. I ran synthetic ( stuff you are supposed to use for G50 boxes) for the first warmups and starts at this one event,. in reasonable heat. BUT still I cannot get over the fact that I could take the car outside the track EVEN WITH THE SYNTHETIC GEAR OIL. and it worked fine. I am talking pounding on it for 30-40 minutes up and down Rte 98. in high heat of day. from ": Spray and Pray " gun shop to Lordida and back. IT WORKED FINE!!!. Then I take it to the next start at Sebring in the afternoon and it totally locks up just after that right at tower turn...... riddle me that Batman. thanks again for everyones input. FRank. OH. BTW. Bill Rader did the build initially and then was kind enough to redo it when I sent it back to him ., literally for close to free,. That is truly amazing customer service ....

Bill Rader is the best!! I love that guy....I won't let anyone else work on my 915 (well, except for one guy but he is retired).

Stick with the Swepco 210....you will be happy...but actually filling the 915 with the gear oil helps with longevity!!

Weird 915 Race Gearbox Problem: Need the braintrust..
 

My guess is you have a Wevo gate shift in there … and the engine and trans was able to move enough (lateral G’s) on rough sections on the mounts to upset the selector .. which would then relax again once cooling down back in the pits. Reports from way back of hard driven race cars unable to find gears due to engine / trans movement on the mounts .. Wevo gateshift exacerbates this ..

Good call Roscoe., indeed it is a Wevo.... will check it