AlfaBMWGuy

I am going to answer my own question here because it's clear I'm diving into fiddly work where few have gone.  Turns out that it is a "friction fit" for the nylon gear on the shaft with the worm gear drive at the other end.  Below in pictures you'll see how I used a small socket (along with a couple pairs of vice grips) to help extract the nylon gear without damaging it and then both my source and destination nylon gears and shafts once pulled apart.  You can see that the rough surface under the nylon gear is designed to grip tight even without glue so it was easy to tap on the better gear to the shaft I needed to have the right odometer calibration.

Astute readers might spot two things:  1. My source/good nylon gear is white where in my picture from yesterday it was yellow.  I actually had a third non-functioning odometer with that white gear I decided to go with.  2. There is a slight bend in the shaft for the source one--I wasn't very careful in removing the nylon gear from this one since I'm not planning on reusing it and this was the result.  Just another reminder that this is fiddly work with delicate parts.  The final pic shows what I intend to run, which I feel much better about in terms of having an odometer that will continue to function.  Note it's the top shaft with the bottom nylon gear (not mangled like the top nylon gear) from the 2nd pic.

-Gary

Thanks!  I was going to mention odometergears.com: https://www.odometergears.com/products/BMW/2002+Series+68-76/23

But, my post was already too long.  From the link above, they only sell the nylon top of the gear and not the whole thing with the shaft.  So that was part of my assumption that it must come off because they are having you put on the new one they provide and, therefore, also take off the old one first.  But, if you are replacing, you don't care about damage to the existing one and that's not the case for one of mine that I want to swap over of course.

I was hoping to get this project done quicker (and saving $25 is never bad) since I do have a decent one, just not on the shaft with the worm gear drive I need.

I am trying to fix a broken speedometer.  I found after removing the speedometer housing that my flywheel had come off the shaft--a not uncommon failure point from my investigation.  There is a nice YouTube video that goes over the repair here: https://www.youtube.com/watch?v=oh4C5a1pSbI

There is also a FAQ thread that mentions this issue, but the video is more comprehensive so I'll just link that in case it proves useful to someone.

I found that my flywheel had not only come off the shaft, but would not spin freely in the housing when reattached, which I suspect is the reason it worked its way off the shaft over time in the first place.  So I proceeded to remove the odometer gear that's pressed into the housing with a ball bearing to see if I could get the flywheel to spin freely.  This piece inside the housing has a nylon gear at the end and is one of the failure points for odometers--the other being the tenths wheel slipping on the shaft between all the number wheels.  I wanted to just clean up this existing gear inside the housing, lubricate it, and hope the flywheel would turn smoothly when reassembled.  However, I found that there was quite a bit of wear on the nylon gear and even though the odometer was working when I removed it, I don't think it's long for the world.

I happen to have a spare speedometer, but it's both metric units (my 2002 is from Canada, but the car is now in the U.S.) and poorly calibrated along with not the correct crosshair face for my '74 tii and it has a needle for the earliest pre-'71 models so someone had messed with it in years gone by, probably leading to the horrible calibration.  So lots of reasons to replace it.  But, the nylon gear on this metric one is in much better shape.  What I'd like to do is use it instead.  But, the shaft it's mounted on, the worm gear part on the opposite end from the nylon gear, has a different thread pitch than one I'm using (the one for miles and not kilometers). 

From previous experimentation I've found that those thread pitches are critical to a metric vs. miles gauge (and I assume that's part of how speedo/odo's are matched to different diff ratios as well).  Below is a picture so you can see what I'm talking about with the thread pitch differences--the top one with the finer threads is for the metric speedo/odo (also has the nicer nylon gear), while the bottom one has the coarser threads (and the more mangled nylon gear as you can see).

With all that, can I remove the nylon gear from the shaft and just swap them over so I've got the worm gear from the bottom one in the picture and the nylon gear from the top one?  Is it just a friction fit?  Maybe push it off with a washer from below since I don't want to put pressure on the nylon gear threads?  Would some superglue make sure it wouldn't slip if I did swap them?  Or is it not possible and I just need to run the more mangled nylon gear and hope for the best so I know I've got the needed thread pitch for a properly calibrated odometer?

Thanks,

Gary

I measured the set I sourced from a FAQ member a few months ago and got 37-3/8" for that measurement between the top outside mounting holes.  Not sure the exact origin of the ones I have, but I know they are reproductions from a few years ago.  Apparently no shrinkage in my case.

Thanks,

Gary

4 minutes ago, jp5Touring said:

I have zero photoshop skills, I guess I need to buy some software for my MacBook.

35 7/8 is the Liner. The door is 37 3/8.  This is off the top door fastener hole.

 

 

Thanks for the verification with pics!

2 minutes ago, bergie33 said:

Center to center on the top outer holes of the door is 37-3/8".

Perfect.  Thanks Karl!

Thanks Karl! I was actually trying to figure it out without taking off my door panels.  So if you could give me that center point to center point measurement between the two outer "upper most" mounting holes on your '74 tii (I have the same), I would greatly appreciate it!  I can see from the pictures provided by @jp5Touring that where those holes are corresponds to the widest spot of the liners.

-Gary

On 5/30/2022 at 10:12 AM, jp5Touring said:

I tried to center the liner and short on both sides.

The top door fastener hole is actual 37 1/4 inches and the liner is 35 7/8 wide. So approx.1 3/8 short.

 

Can you clarify what is the "top door fastener hole"?  I'm just trying to figure out where you measured 37-1/4" and 35-7/8"?  I've got some reproduction liners in a tube still and would like to measure them to know if I should even attempt an install since what I'm running now might be better than something that's too short by that much.  Alternatively, if you could measure the entire edge-to-edge length of your liner along the bottom through the center of those evenly spaced holes and give me that number, that would do the trick for me and would be greatly appreciated!

Thanks,

Gary

That advice is greatly appreciated cda! I was backing off my vacuum advance to 6 degrees max because of the concern on the total advance with more being around 40. Now I can still get some extra vacuum advance while having the bit more aggressive centrifugal curve where the tii really performs, as you stated.

Also to answer my own question in my original post, that source for the vacuum line to the 123 vacuum port is a decent one.  I just need to swap out the blue silicone line for something appropriate under the hood of a classic BMW.

Since I was still confused, but neither Mark nor Dave indicated I was doing anything too crazy, I decided to take a more detailed look at the help provided on the 123 USA website.  There I discovered in one of their YouTube videos something I hadn't seen documented on the FAQ, although it probably is and I just overlooked it because I wasn't trying to solve my current problem when I saw it.  I also found it discussed here on the 123 website: https://123ignition.com/support/create-curve-tutorial/

The magic is that the values in the MAP curve you create must be offset by atmospheric pressure, which at sea level is 100 kPa.  This offset converts from relative pressure (as shown on the app dashboard gauge) to absolute pressures that are independent of elevation (and thus allows uploaded 123 vacuum curves to still be good as you change elevation).  Therefore when I see -0.6 bar on the app dashboard gauge at 1000 rpm idling, that really corresponds to a data point in the MAP curve of 100 - 0.6*100 = 40 kPa!  That's assuming I'm at sea level, which is not quite the case, but not too far off.  I did find that trying to upload a curve file with negative kPa values gave me an error so that's not even feasible in terms of directly inputting the app dashboard MAP curve values into a 123 vacuum curve.  But, the offset for atmospheric pressure means it all makes sense and 0 kPa in a MAP curve corresponds to -1.0 Bar on the 123 app dashboard gauge.

With that revelation, for me at least, I was able to tweak my curves with confidence and get the total advance down lower than the 40 or so degrees I was seeing before.  One other issue with that reference curve (which is also uploaded on the 123 USA website as a tii curve, not just in other FAQ threads) is that with only vacuum advance at low rpms, starting is much more difficult.  Need to add in some centrifugal advance to help with that.  I'm still trying to pick the curve I'm going to run with so I won't provide one now, but maybe my trials and tribulations along with starting to come out the other end into the light will help someone else out.

-Gary

Thanks Mark.  The curve I gave was definitely more of a performance oriented one, but apparently was created to mimic Curve 4 (0 231 151 008) for the switched 123 distributor for BMW 4-cylinders.  Mr. 123, who created it, is a Frank Botenbal and this curve was in that 26 page FAQ thread for the 123 going back many years and covering a variety of 123 issues.  I do greatly appreciate the performance bump it offers over a more standard curve like you provided.  That extra advance bump at 3500 is right where you want it in a tii since mid-rpms are where they are lacking something esp. compared to a dual carb '02.

Throughout the FAQ threads that talk about 123 curves there are disclaimers about the need to customize to your engine, modifications, and preferences.  I just was figuring it wouldn't be something this drastic in terms of me seeing kPa values 100 less than expected.  I will do some more experimenting with what the MAP gauge in the 123 app is giving me in different conditions since really I just have those two reference points now, idle and steady state at 50mph in 4th gear.

-Gary

Hello,

I have been trying different advance curves for a 123 on my stock '74 U.S. spec tii.  I am currently using this combination of centrifugal and vacuum curves and loving the performance while also getting good idle and off throttle characteristics:

Redline 6000

500    0

1000  0

1400  12

1600  16

2400  24

3500 35

8000 30

MAP Start   0

0    10

60  10

87   0

100 0

200 0

I found this in multiple other FAQ threads and it was referred to as a reference standard that reproduced one of the hardware curves in the older "switched" 123 and was very appropriate for a tii and created by someone referred to as Mr. 123.  The problem is that with this loaded I'm seeing up to 40 degrees of advance in the 123 app "dashboard" gauge just at steady throttle at 50mph around 3K rpms in 4th gear.  While I'm not experiencing any pinging that I can detect, I know I shouldn't be running this much advance as damage can be done.

I think I know what's happening based on the MAP gauge readings also in the 123 app dashboard.  I am getting much lower vacuum/MAP readings with that gauge than I was thinking I should be seeing based on the vacuum curve above.  For idle, I see -0.6 Bar.  For that steady state 50mph, I see -0.3 Bar.  For WOT, I don't think I get more than 0.0.  So what I think is happening from those values is that the 10 degrees of vacuum advance only specified for 60 kPa and less is instead being applied 100% of the time on top of the centrifugal advance curve values (since I never see anything approaching 60 kPa).  At idle I do see 10 degrees of advance, meaning nothing added to that vacuum curve value of 10, which is consistent with the centrifugal curve, but also seeming to confirm that anything less than 0 kPA also gets the 10 degrees of advance (my idle would correspond to -60 kPa if I have the Bar to kPa conversion factor of 100 correct).

So the clear indication is that I must be taking the source for the 123 vacuum port from an incorrect location to be getting pressures that far off expected.  Below are two pics showing how it's setup now.  That blue hose (yes, I know an incorrect color choice and it will be changed out for black one way or another) is the one going to the 123 vacuum port.  You can see how a tee was added to the line from the intake manifold to the '74 tii specific bit of Bosch emissions equipment on top of the air box--decel valve or whatever it is.

So is my solution to instead run this line from somewhere else?  If so, from where?  Or, would it be feasible, if a bit kludgy, to just create a new 123 vacuum curve that offset the pressures values with what I'm seeing?  Since idle at 1000 rpm should be 40 kPa (from what I've been told) where I'm seeing -60 kPa, would it be legitimate to just subtract 100 kPa from the values in the curve above, meaning -40 and lower would get 10 degrees, but at -13 kPa (87 - 100) it turns to 0 degrees?  Does it even work for the 123 with negative pressures specified in the curves?  My concern with just hacking the curves is that if my source for the vacuum line is wrong, it's then not giving me expected/consistent pressures in all conditions (partial throttle, decel in gear, WOT) to simply do that offset by 100 kPa.  Therefore, the hack might work for idle, but not consistently across the board in other conditions as is the whole reason for wanting to add in the vacuum curve at all.

Thanks,

Gary

Emailed.

Thanks!

-Gary

NT

NT

...and gone.

On 5/7/2022 at 2:45 PM, m-tuner said:

alfa guy I have been swamped with tii resto, I will put together all the ones I have in one shot maybe 6-10 if you see one you like PM me,   Bruce

Thanks Bruce.  Shoot me some pics when you are able.

-Gary

Thanks Andrew.  Are these Hella VR's known to be better quality than the Beru ones that BMW is now calling OEM and are really quite the pieces of junk right out of the box?  The Hella does look identical to the Beru, which both being different (not as tall) as the older OEM VRs.  I'd prefer an older one that had both the correct look and was better made so I'll just keep looking, but keep this in mind as a fallback since it does look more correct than my current NAPA VR.

-Gary

Still looking 5 months later if someone has a nice looking OEM VR available.  I've got a scruffy OEM one so just looking for one that looks great as well as working properly.

Thanks,

Gary

Still looking 6 months later.  Either a new OEM (right-hand side) passenger door seal or a used one in nearly new condition is acceptable.  I know it's a longshot since BMW hasn't been shipping these for a long time, but I thought I'd give it another try.  I ended up installing a URO seal, but it doesn't fit nearly as well as my new OEM driver door seal.

Thanks,

Gary

Interesting and thanks for the extra info.  On my last 2002, an early '71, the OEM door seals on both sides went on easy with no interference.  With my current '74 tii, it's more of a struggle on both sides, but the OEM driver door seal wasn't too much of a pain (just barely avoids door panel interference) compared to the URO passenger door seal--mix and match was never the desired solution, but at least for the coming year or so, it's reality given that BMW hasn't been shipping passenger door seals and won't for many more months.  The seals for the '71 were bought back in 2017 so maybe that was just a better designed batch or those and the newer OEM seals both fit better on earlier '02s, not sure.  I bought my current OEM driver door seal last September, while the URO I bought late last year when the delay on the OEM passenger seal became intolerable.

Thanks Jim.  Actually, the URO passenger seal stuck out an extra 3/16" beyond a new OEM driver side seal.  On the driver side, the seal didn't tear itself off based on how it was (improperly) designed.

As an update on the situation with the vertical section of my passenger door URO seal sticking out too far and the interior door panel ripping it off when the door is opened/closed, my mechanic ended up taking off the seal again, taking a power sander to the inside of it where it adheres to the frame for this vertical section, and regluing it.  He took off about 1/8" of the rubber material and says this was just enough to keep the door panel from tearing it off.  So we are going to declare victory on this for now even though I'll be patiently waiting for the OEM RH seals to become available again in the months (years?) to come.  That level of customization should never be required, but it was preferable to going back to my worn original seals--the whole reason for buying new seals in the first place.

-Gary

I recently got word back from my mechanic on the difficulty he's had in installing the URO passenger door seal.  This is after using 3M adhesive remover, Wurth rubber glue (after the 3M weather strip adhesive wasn't up to the job), and sanding the URO seals after removing the mold powder with brake cleaner and alcohol--all based on advice I found from searching the FAQ.  The issue is that the URO door seal pulls loose around the door striker even with the strongest glue and fastidious preparation as the seal stands out an extra 3/16" on the passenger side vs. a brand new OEM seal on the driver side.

So, he sees two possible paths forward.  Either he'll cut out the vertical section from my ancient OEM passenger door seal (the one I was replacing because it no longer makes a tight weather seal) and splice that in instead of the URO seal for that length or he'll try cutting down the new URO seal specifically for that vertical section.  The hope with the latter approach would be to eliminate enough of that 3/16" that it stands out so that it no longer pulls the glued seal from the door frame when the door is closed/opened, but at least have fresher rubber than what could be original seals from 1974.

Both of those are interim solutions and the long-term solution is to wait for BMW Classic to start shipping new OEM passenger door seals again or for URO to redesign their passenger RH door seal.  From what I see on this thread and heard from BMW parts suppliers, both of those may happen in the same timeframe early in 2023.

@johnny02 your issue definitely is exacerbated by not having the plastic door sills.  Sourcing those seems like your best path forward, but cutting the lip off the URO seal seems like your best contingency plan if they just can't be found.

-Gary