Mechanical to Electronic Speedometer Conversion
 

I'd like to use this thread to tee up a discussion on the overall topic of electronic speedometer options for our vehicles.

I've run into the problem with my diesel conversion (617 into a 380SL) and ROLLGUY is having a similar problem on his transmission swap of a 722.4 into a 300SD that originally came with an electronic speedo 722.3.

In both cases we thought that a tailcone swap would work but it is now painfully obvious that the two 722 transmissions (.3 vs .4) have different tailcones.

It is still possible that a tailcone from a C220 or a C280 will get the job done, but additional research points to that being a somewhat dubious proposition.

I'll explore some of the more exotic solutions in supplemental posts, but I'd like to set the stage for a discussion and additional options.

First, everything I can find points to the notion that the electronic speedometer pickup on the 722.3 and likely the 722.4 as well is a variable reluctance system consisting of a trigger wheel and a sensor.

Here is a picture of the internals of a 722.3 tailcone (from the 380SL) where the 4 "spoke" trigger wheel is clearly seen mounted to the parking pawl. The sensor reads each revolution through a voltage change from negative to positive.

http://www.peachparts.com/shopforum/...1&d=1329109024

The wave form looks something like this

http://www.megamanual.com/ms2/AC.gif

An in depth discussion of VR devices for triggering spark in Megasquirt using an EDIS device is here- Distributor Pickups

Trigger wheels take many forms and the following image shows some of the kinds that have been used in ignition applications generally as part of a fuel injection system.

http://www.peachparts.com/shopforum/...1&d=1329110238

The first solution I'd like to toss out for discussion and critique is to create a "virtual trigger wheel" using the drive shaft as a base.

Here's a graphic of what I have in mind. The green are the "teeth" of the virtual trigger wheel, the purple is the stock MB sensor and the red is a custom bracket with appropriate adjustments for setting the clearance.

http://www.peachparts.com/shopforum/...1&d=1329110484

This proposed solution would use 4 carefully sized, fabricated and balanced steel "teeth" that would be epoxied to the drive shaft at exactly 90 degree intervals, as close to the front end as feasible. Taken in cross section we would have the equivalent of a 4 tooth trigger wheel, something like this.

http://www.peachparts.com/shopforum/...1&d=1329113089

It would not be prudent to rely on simple epoxy as a bonding agent, so I would use fiberglass, kevlar or carbon fiber cloth ribbon and epoxy to create a composite "band" around the driveshaft. Since these materials are quite light and would be evenly distributed, adverse balance effects should not be introduced. Rather than mount the sensor to the tunnel I would mount it to the rear transmission mount bracket to eliminate all but the tiniest and insignificant relative motion of the transmission and the driveshaft.

It seems as though this strategy would replicate the internal operation of the Mercedes system with an external "universal" alternative using the stock sensor, wiring and electronics which could be fabricated with relatively simple tools and materials.

Here's an article with a lot of great information on the subject of VSS (Vehicle Speed Sensors) - VSS

Where am I wrong and what have I missed?

In the previous post I discussed the idea of creating a virtual trigger wheel and using the stock MB speedometer sensor. In this post I want to take a look at another potential alternative.

This idea comes courtesy of Patrick Schalles in an article titled How To: Installing a Universal Magnetic Pickup Speedometer Instead of using a MB speedometer he is using a universal speedo e.g. the AutoMeter one.

Instead of using a variable reluctance system he is using a Hall-effect system using magnets instead of "teeth" on the driveshaft. This sensor kit is about $25 from Rostra Precision Controls.

http://www.peachparts.com/shopforum/...1&d=1329627187

This system uses a different signal but from his article it appears that we might be able to use this signal by putting it through a GM VSS Buffer to convert the AC signal to a square wave signal. I'm not an electrical engineer but it seems plausible that the signal from the VSS Buffer might be able to be fed into the wiring of the MB sensor and get a signal that way. If this is possible, it means that instead of relatively close sensor-to-trigger-wheel gaps we could theoretically go with the easier and more forgiving tunnel mounting and 1/2" recommended gap.

http://www.peachparts.com/shopforum/...1&d=1329627239

I actually spent about an hour looking for one of these devices at the JY this weekend to no avail.

A follow on solution using this strategy might be to somehow insert the internals of an AutoMeter speedometer into the MB cluster. Probably not something I'd try, but maybe a path for someone to pinneer.

Anyone have any opinions about whether this kind of a signal might be suitable for a direct feed into the MB system?

So right in the middle of my research and thinking about a conversion from mechanical to electronic speedometer in my diesel conversion of a 380SL comes this tidbit in a Question Posed on FixYa. The Question is

And the response is

Now this seems completely bogus, but I can't confirm or deny based on electrical diagram for my 380SL. The page referenced in the manual conveniently doesn't exist.

If the speedo uses the ABS sensor, what is the purpose of the transmission speed sensor? Bogus?

If on the other hand this is the source of the speedo signal, problem solved. :)

I call BS. What is the purpose of the cable from the tranny to the speedo if the ABS wheels sensors drive the speedo? Conversely, how is the ABS controller connected to the speedo?

Your initial proposal is the most elegant.

Sixto
87 300D

I thought of a more evasive solution, but I probably will not do it because of the risk involved (ruining the parts I have). I will try to describe it, as I have no way of illustrating it otherwise. I thought of making a fixture that would index the center of the tail cone shaft in relation to the mounting surface. I would then have my machinist friend fly cut the nose (the area of the speedometer sensor) of both tail cones so that the electric portion (nose) can be welded on to the housing portion of the cable tail cone. I would then end up with a tail cone that fits the 722.4 trans, but has the fitting for the electric pickup. The one problem I see is that I would need to remove the plastic piece that is inserted in the aluminum casting that the sensor fits into that keeps fluid from pouring out when the sensor is removed. I have a feeling that it is not removable without breaking it. My machinist has the capability to do the cutting and welding, but the whole job may cost more than it is worth. The other problem, is that I am quite sure that the gears in the cluster are bad as well (common with the 126). It has bee suggested that I just find a way to put a 123 (cable operated) speedo in the 126 instrument panel. I am not sure if this is possible, but I might try it BEFORE cutting and welding of the tail cones. I have some 123 speedo's and cables in stock, so if I ruin one it is no big deal. The car is just a beater car anyway, but my friend needs the odometer to work in order to insure it. If the instrument panel looks weird, it is no big deal to him as long as the odo works.

Another thought I had was concerning the 4 spoke trigger wheel. Could it be mounted externally (on the driveshaft maybe)? I thought that maybe it could welded on the nose just behind the flex disc, but there might not be enough space there. Another option would be near the center support where the shaft is thinner. The problem here would be trying to keep the proper distance between the sensor and the trigger (rubber mounted bearing movement). I am just throwing ideas out here, so anyone feel free to post your opinions/suggestions. Thanks, Rich

That's exactly what I thought - but it definitely had me thinking for a bit.

That's exactly what I was trying to suggest with my first solution proposal

http://www.peachparts.com/shopforum/...1&d=1329156036

And the article I linked to had this image of a trigger wheel attached to the driveshaft at the differential.

http://www.peachparts.com/shopforum/...1&d=1329156205


That appears to be a 40 tooth wheel, but it would work just as well with a 4 tooth wheel. I thought it was very elegant the way it piece was designed - no welding and no cutting.

http://www.peachparts.com/shopforum/...1&d=1329156409

When stock hard parts are not available, I always thought this was the best solution ever since Apotheoun had it done to his car when he did a manual conversion on a 300SD.

Although we haven't heard from him in a while, evidently this has worked out well.


.

Well I can now state without doubt that the initial solution to turn the driveshaft into a trigger wheel by gluing "teeth" to it and using the stock sensor with a custom bracket does work.

Here's my "driveshaft-with-trigger-teeth" test setup.

http://www.peachparts.com/shopforum/...1&d=1329160409

I crawled under the car and held my "driveshaft" in close proximity to the sensor and indeed got a speed reading. (Sorry for the poor picture - it's a frame grab from a video as I couldn't be under the car and in the cockpit at the same time to take a picture) This is probably a worst case scenario as my "teeth" in this test were nuts and I'm sure the hole in the middle made the signal appear to be weird to the system, so well proportioned teeth will definitely provide a cleaner signal. And trying to hold a drill and a sensor steady under the car is not a great way to maintain any kind of consistent sensor-tooth gap.

http://www.peachparts.com/shopforum/...1&d=1329160407

As far as I'm concerned, I'll consider this problem solved and move ahead with fabricating a bracket and cutting and mounting my "teeth".

I thought you said something about gluing metal on the driveshaft, and that is why I asked about attaching the stock trigger wheel to the driveshaft somehow. I just got out from under the car putting the driveshaft back on, knowing that I won't need to get back into the rear of the trans to make an electric pickup work. I measured the trigger wheel that I pulled off of the park pawl gear, and it is about 35mm inside diam. The smallest part of the shaft near the flex disc is about 47mm. I am glad you tested your theory out. Could something be welded (instead of glued) on the driveshaft resembling the trigger wheel? That sounds like the direction you are headed, but I will wait and see what you come up with.

A VR sensor interface is set up for a certain waveform amplitude range and to the resistance of the sensor itself. Try to use the original sensor and try to use the same air gap. That way you don't introduce variables that could cause problems that are murder to diagnose.

I like the idea of the tabs on the driveshaft. Very simple.

Yes I did suggest gluing the "teeth" of the virtual trigger wheel directly to the driveshaft. Forget the stock trigger wheel... it's not worth the trouble. Yes you could weld the teeth but now you're into the territory of adding indeterminate amounts of varying weight to a critically balanced piece.

A good epoxy is likely enough, but I wouldn't trust that. My insurance solution is to get some fiberglass, kevlar or carbon fiber tape and using epoxy, layup a ribbon around the driveshaft. They make aircraft wing spars out of carbon fiber so I know it would be more than strong enough. The other reason for going that route is for weight. If your "teeth" are the same weight and the epoxy and cloth is evenly distributed around the driveshaft, any weight discrepancy is likely too small to measure.

If you've never laid up fiberglass (or carbon) here's a little video that will give you a pretty good idea about the process. Hand laid fiberglass body tube - YouTube

For our purposes the analogy is that the cardboard tube is the driveshaft and instead of laying up a 3-4 ft segment for a rocket fuselage, we're laying up a 3-4 inch segment with 4 steel teeth in the middle of the segment. Fiberglass is very much able to go around 3 dimensional curves so it will not be a problem. I'd probably be happy with 3 layers, 4 at the most for fiberglass and 1-2 layers for carbon fiber (just cut the top layer cleanly at the starting point of the first layer to maintain balance) If you wanted to be really trick, you could create some fillets around your teeth using a microballoon-epoxy paste to eliminate the bubbles that will be present between the driveshaft and the fiberglass where the teeth are. In the image below, the purple is the cloth and the green is the microballon fillet. You will also probably want to grind a little concave surface into your teeth to get a good bonding surface.

http://www.peachparts.com/shopforum/...1&d=1329169666

Also note how trigger wheel teeth are generally wider at the base than at the top. This gives a cleaner signal.

http://www.peachparts.com/shopforum/...1&d=1329170381

You could use AeroMarine 300/11 Epoxy ( AeroMarine Products- Epoxy Adhesives) or for the finest materials you could use aircraft grade fiberglass tape and epoxy resin from Aircraft Spruce. (Composites from Aircraft Spruce )

Do you think you need all that? I'll bet the double sided adhesive they use for wheel weights will last the remaining life of the car.

Sixto
87 300D

Probably... I think that a good epoxy by itself would likely be ok too, including even good old JB Weld.

But if one did come loose it could cause some damage as well as throw the driveshaft significantly out of balance. Whereas laying up a composite retaining ring would guarantee long-term integrity.

A wheel weight would be installed on the inside of a wheel (centrifugal force is against the inside of the rim), whereas the centrifugal force on the driveshaft will want to throw off the tabs or whatever was glued on. I think I could make some pieces of steel that would mimic the tabs on the trigger wheel, and weld them on. I doubt the balance would be affected enough to cause a problem. If I did it correctly, the circumference of the finished tabs would be the same as the stock trigger wheel. The speedometer would not know the difference.

Do you think circumference matters? Wouldn't it count 4 triggers as one revolution and call it a day?

If the prop shaft is thick enough, drill, tap and use bolt heads as a triggers.

Sixto
87 300D

Circumference is irrelevant.

Here's a picture of my Ford EDIS trigger wheel with a ruler. I don't think you could go wrong by using the basic dimensions and shape of it's teeth.

http://www.peachparts.com/shopforum/...1&d=1329177939

Hmmm, might even be able to cut up the wheel and use it's teeth. It's made of some kind of sintered metal so I don't know how it will cut and grind.

Don't you think the teeth on the 36 minus 1 wheel are a little tall? I'm thinking the flat pieces of steel used as balancing weights should get the job done, keep a low profile and they'll be pretty tough if spot welded, brazed or epoxied on correctly. I might even try welding 4 beads on the driveshaft, minding heat, of course.

I would not even think about drilling and tapping for screws. That sounds too risky.

Maybe a bit tall, but the taller the teeth the better the signal, to a point. Too flat and you'll have signal problems as well needing to run a gap that is going to be too small.

Drilling and tapping are not a direction I'd go, nor would I just try to run a bead.

For a first go, my working plan is to try to cut the teeth on my Ford EDIS wheel and epoxy them to the drive shaft and then do some form of reinforcement with flox or microballoons and some glass cloth.

During my test the nuts were affixed to my "driveshaft" with two runs of electrical tape and I got the speedometer up to 70mph with no hassle so there is not much force exerted on them. The issue, I think, is more to protect them from getting banged by a sideways force that could pop them loose.

I spent some time cutting and grinding the teeth off my Ford EDIS trigger wheel to see if that concept would work and I must say it looks rather promising.

I cut them in such a way that I had a bit of a lip to provide some additional grip to the composite material I plan to attach them with. It also provided a visual reference for how much to grind to get a matched set. Here's a macro of one of the teeth I cut.

http://www.peachparts.com/shopforum/...1&d=1329346108

It was quite time consuming to try to get 4 that would match, but I managed pretty well.

http://www.peachparts.com/shopforum/...1&d=1329346302

I weighed them and they all turned out the same - unfortunately the most sensitive scale I had only read in grams and all were 4 grams.

Here's a shot of one tooth on the driveshaft. The seam is in a perfect position to provide an accurate point of reference for alignment.

http://www.peachparts.com/shopforum/...1&d=1329346542

I'll need to source some epoxy to be able to mount the teeth permanently, but I'm very optimistic that this strategy will work. I'll also need to make another trip to the JY to be able to pick up a sensor that I can use for fabricating the mount for the pickup. The air-gap between the trigger wheel and pickup needs to ideally be 1mm, so it needs to be both sturdy and have a fine adjustment capability. Here's a sketch of what I'm using as a concept.

http://www.peachparts.com/shopforum/...1&d=1329347827

After pondering the various options for mounting the trigger wheel teeth, I decided to do a test of the welding option. I took a couple of my rejected teeth and welded them to some scrap metal. I tapped on them with a hammer and I'm convinced that they are quite secure. I also don't think that this will introduce any appreciable balance issues, particularly if I attempt to use a consistent trigger time for each weld. Here's a sample...

http://www.peachparts.com/shopforum/...1&d=1329419486

I am glad you tried it this way, because I was planning to weld the tabs on myself. I just need to make sure I get all the paint off the shaft where I am going to weld. I thought about using a strip of sand paper that is used by plumbers for sanding copper pipe.

An angle grinder would do a fine job of getting the paint off and down to bare metal.

I'm starting to get a good idea about how I'm going to fabricate my sensor mount. Here's a sketch of my current thinking.

http://www.peachparts.com/shopforum/...1&d=1329462058


The transmission mount bolt provides a very stout place to tie into. It's a 19mm nut and there is a channel that will provide support against any lateral movement. There is not a lot of room for anything over about 1/4" so it makes sense to make the attach point steel. I want to be able to take the mount apart in segments, so that is why I'm using aluminum and allen screws to form the sensor support. The sensor is a metal cylinder 18mm in diameter, so an 18mm hole with a clamp feature will provide the fine adjustment needed vertically to maintain 1mm of clearance between the trigger wheel and the sensor.

The sensor is quite similar to this in design. (Don't have a picture of the one on the car yet)
http://www.peachparts.com/shopforum/...1&d=1329462684

WOW! You are just getting too fancy LOL!

If it doesn't cost an arm and a leg, I would request that you make a bunch of these! I could sure use one, and I am sure you could sell a few more. I was thinking I would not get that fancy, but just bend something up out of flat CRS stock.

Don't plan on doing that, but as I go through the fabrication process, I'll add dimensions, photos and notes on how I built mine.

Not even sure this is the final design yet, just part of the solution process.

Having the parts bolted together will definitely help, as the trans mount bolt is hard enough to tighten as it is.

That was exactly the reason for a design that could have the vertical part of the bracket installed after the transmission mount bolt was installed and tight. Since the "base" is relatively thin (1/4") steel, the aluminum part of the bracket needed to be fastened from the bottom. I don't have the capability to weld aluminum so attaching the final piece had to be with allen screws.

Barring any modifications as a result of the process of finishing the fabrication and installation, here is the design and the dimensions of the bracket. For best results cut the piece for the transmission mount fairly rough, and then grind to a snug fit.

http://www.peachparts.com/shopforum/...1&d=1329634648

I made a trip to the junk yard today and grabbed a speed sensor to use in finishing the fabrication of the bracket. I didn't want to cut the original and then have to re-splice the wires under the car and worry about polarity. It's easier working on the bench rather than under the car. Here's the actual sensor

http://www.peachparts.com/shopforum/...1&d=1329635188

I'll most likely cut the tab with the mounting hole and just use a cinch bolt in the bracket to hold it in place with a 1mm air gap.

By the way, we can now verify that the sensor is a VR (variable reluctor) rather than a Hall effect sensor because it is a two wire device.

Well after all this R&D (mostly on your part mach4), it looks like the car that needs the electric speedo will become a parts car. Another (much nicer) '81 300SD became available that has a bad engine. This thread: http://www.peachparts.com/shopforum/diesel-discussion/312704-broken-camshaft-617-81-sd-how-could-happen.html
is about the car. I tried to feed in a new timing chain, but it seems there is more damage somewhere else inside the engine. This car has a good trans, decent interior, and a pretty good looking exterior. The engine from the other car ('81 SD with mechanical speedo trans) will go in this one, and the trans will go in a friends 300D that needs it. The rest will be for parts. It is still nice to know that the mod can be done if needed in the future.

Got the fabrication done and ready to be installed in the car. In this picture I'm getting the center-line for the teeth of the "trigger wheel".

http://www.peachparts.com/shopforum/...1&d=1330136046

Stripped off the paint, marked the centerline and the position for the teeth and prepared the surface for welding.

http://www.peachparts.com/shopforum/...1&d=1330136263

And here's the finished "trigger wheel".

http://www.peachparts.com/shopforum/...1&d=1330136452

The final test of this strategy will unfortunately need to wait for a bit until I get the engine back in the car and get all the other components hooked up.

I never finished off this thread, so I thought I'd update it with some pertinent information.

First, the solution has been working really well. Absolutely spot on speedometer and rock solid. No balance issues with welding the "teeth" directly to the drive shaft and the pickup has been working great. With one annoying exception...

When stopped, I would periodically get a "bouncing" speedometer and the odometer would slowly increment as well. Moving the car just a couple of inches would cause it to stop. I suspected that the sensor was mounted too close to the "trigger wheel" teeth such that when the sensor and trigger wheel were aligned, the slightest vibration would induce enough movement to trigger a false signal voltage change that would cause it to register with the computer as movement.

While I was under the car recently to just do a visual check and to do a tightness check on all the bolts since the swap was completed and I've got several thousand miles on it, I added some clearance to the sensor/trigger wheel.

That adjustment fixed the problem and now that "bouncing" needle syndrome has been completely eliminated.

So I can say with a great degree of certainty that if you need to add an electronic speedometer to a car without it, and you can't find a transmission or tailcone with the pickup internal, this external solution works. The bracket can be fabricated without welding or any special tools.

nice work paul and thanks for finishing the report!

Question...

What parts are required to make an electronic spedo work in a car that previously had a mechanical spedo? I grabbed a 400E spedo to use in my 606 swap (into my '87 300D). I might have to go back to the junk yard and pull the dash and trace out wires next weekend.

That is a really, really good question. One that I tried to figure out when I was setting up my "driveshaft trigger wheel". Fortunately my car already had the electronic speedo, so all I had to do originally was hook up a trigger wheel and bracket. (Subsequently, I've swapped the 722.4 transmission for a 722.3 that I was able to put the 380SL tailcone onto and get a completely stock electronic speedometer)

Here's the section of the wiring diagram on the 380SL speedometer. It's obviously a "solid state black box" but this diagram only shows power and ground and an output to the cruise control. I have no idea where the VR pulses get to the speedo.

http://www.peachparts.com/shopforum/...1&d=1362267403

The wires on the VR sensor are solid brown and solid white (at least on the 380) so you might be able to trace them. It definitely appears that the cruise control is not a source, since the speedo provides only output to the CC.

Please post anything you find, because I, for one, am definitely interested in knowing how this works.

Looks like I will go back to the junk yard next Saturday and spend some time pulling the dash and steering wheel to trace out the wiring.

I'm doing a 722.6 swap and will need something other than the small display that shows speed. I did grab the 400E spedo so I will get back asap to figure out the w124 system so I can retrofit it into my car. I do not have any welding capability so I might have to farm that out.

You might want to start at the other end and work back towards the speedometer. It might be as simple as determining that the VR sensor hooks directly to the speedometer black box. On the otherhand you may find that there is some kind of a "speedometer amp" or some other electrical device sitting between the sensor and the actual speedometer module, perhaps analogous to the tach amp (or EGR controller) for the 123s tachometer. Those devices also use a VR sensor as the input for the gauge.

Interesting. I will have to do that. I clipped three plugs to keep for the gauge. I'm going to make a new thread and reference this one just to keep this one organized

A lot of work has been done to adapt. Has anyone ever considered a bicycle speedometer. A magnet could be applied to the rear axle or wheel assembly and the sensor put in the dash.

There are units made for motorcycles like the trail tech vapor that includes a tach and temperature gauge that is configurable.

Just saying.

Some people want the quick and easy solution and others want it to be done correctly. I prefer it to be correct.

Hi Don,
If we were showing these cars, the only "correct" way would be completely stock. How boring would that be.

I pointed out a solution that will give correct speed an mileage. It is easier. It uses a reed switch that is tripped by a magnet. Appropriate technology for the rpm involved (600).

Really nice.

Wondering what speedometer, sensor and any other electronics/wiring you did as i'd like to replicate this on my 5 speed without mechanical speedo.

If you have a mechanical speedometer in place you might want to look into a GPS speedometer cable drive box. Once calibrated to how many cable revs/mile, you can change tire sizes, transmissions, rear end ratios, etc and it stays accurate because it's getting your speed from GPS.

I've never used one but this seems to be the "in thing" for transmission swaps and rear end ratio changes.

That's true, however, i would really rather one driven from the wheels turning rather than GPS.


GPS just tells me what i was doing a few seconds earlier where as a speedometer driven from the wheels is " now"


I have a gps on my 5 speed 717.422 transmission swap :/

I'm driving a Mercedes diesel, so the speed I was going a few seconds ago never differs a whole lot from what I am going now, even with the pedal on the floor. :)

Thanks for the info though, that was an potential issue I had never considered with those.