Connect 722.6 Transmission to Electronic Speedometer
 

Does any one need any hardware to get there old electronic speedometer working with a 722.6 Transmission swap.

I will be getting quotes on having these wheels made shortly. They install on the front side of the transmission output flange. I will probably need to order a minimum of 5 so the more people interested the cheaper they can be made.

These will be made from 4140 steel heat treated to 1040 MPa, Just like a 10.9 bolt. They will be powder coated black or possibly some other corrosion resistant coating. I will have to see what they have to offer.

This current set is for the 3 web flange only. Unfortunately it is not possible to make one that works with both 3 and 4 webs. So I would need to put up a separate quote for a 4 web design.

These are for use with a hall effect sensor and will require a simple 3 component circuit board to be soldered together for installation. They may also work with a VRS sensor which can be directly wired,but success with that has been limited so far.

JR:

Does the 722.6 trans have any provision for speedo drive, and if not, what is the source of speedo information for cars that use that trans?

Variable reluctance sensors need a clean sharp edge to make a clean sine wave. With your tone ring design I'd expect the " zero crossing " voltage to rise with increased speed and eventually be higher than the trigger point of the speedo. RE: I'm expecting the sides of the open window area to still have an effect on a VR sensor, a sprocket design would be better.

Early Sprinters supposedly used a sensor in the rear of the trans for road speed. The casting hole is there however you would have to dissemble the trans to drill the case and I don't know what the tooth count is.

Making the ring from 4140 is probably overkill, measure the hardness of the output flange and adjust accordingly. I'd expect mild steel to be just fine.

Powder coat is just plastic so over time I'd expect it to cold flow and bolts lose some tension. Depending on the sensing range of you hall effect, a layer of power coat might take you out of range. A VR sensor needs tight clearance so I'd expect that to be even worse.

Regardless, I tend not to like power coat in a corrosive environment and prefer paint because paint tends to have better adhesion. Paint will slowly peel as the base metal rusts where as powder coat comes off in sheets due to rapidly advancing corrosion.

Most cars using the 722.6 get the speed info over the can or possibly a front wheel speed sensor not compatible with the old speedometers.

VR sensors need to be about 3 times the width of the probe to eliminate what you speak of. The widows at 9mm wide and the VR probes I have are I believe 2.5 mm in a 12 mm body. It all worked fine until I bolted it up. These sensors were tested on the lathe and worked very well with the VR sensors, but When I put them on the car I found that the presence of the nuts was enough to piss off the signal and cause issues. I installed it with the installation nuts on the lathe and was able to replicate the problem.

So I tried the Hall Effect sensor and that worked well up to about 80 MPH but using it with the VR wheel was too much the slots need to be shorter. I was able to simulate the Hall effect signal with 50% slot/material and drive the speedometer all the way up to 450Hz (~ 210MPH). Which is the max the speedometer will do.

So the second picture shows the installation with nuts and the old long slots. The first picture shows the latest rev that is threaded to avoid the nut problem. Unfortunately that drove the requirement for the 4140 Steel with heat treat.

I believe this new wheel will work with either VR or hall effect. But I do like hall effect better. Its just a constant 12 square wave output. Compared to the VR's spikes running well over 100V at higher speeds.

As far as the coating I am still up in the air with that. I have considered having the mounting area masked for the coating process, be it paint or powder. But I really would prefer cad plating.

I just use a module that reads CAN messages and sends out a signal. They are not cheap but they work and are programmable. Let me know they are 500.00 pre programmed. I can easily do vehicle speed, RPM, CLt etc. 4 channels total. You are running into the problems I had in my sensor design. But if you want to do what you are doing use magnets instead of steel. I use an omnipolar speed sensor that only reads magnets and will ignore the metal around it. Also you cannot adjust for diff ratio and tire size with a fixed setup like that very easily.

If you put a steel tube over the VR sensors body leaving only the end exposed, that might eliminate the nut interference. I'm thinking the nuts are triggering the side of the coil and not the end.

I'm thinking it is less of a slot ( window ) length issue and more of a metal length issue. With a short metal length you are probably running into a hall effect reaction time issue where a VR sensor reacts faster.


Got it, however you will be losing self locking threads that the nut provides.

Will the speedo accept a square wave? I'd be minorly concerned that the rapid on / off could affect charge time of a capacitor / inductor and lead to an inaccurate speedo or even component damage.

As for the 100 V,was that an unloaded or loaded voltage? RE: a current transformer not connected to a meter can reach very high output voltages but when loaded the voltage drops dramatically. Increasing sensor to wheel gap is a way to reduce voltage.

I'd go for a thin coat of plain black paint in a spray can.

Rather than going to all the effort of another wheel, why not mount a second sensor on the diffs axle flange and use the wheel you already have then use a speedo fooler to generate the needed pulses? Even with your proposed new wheel, will you have the proper pulse ratio to give an accurate speedo reading?

We did talk about that about a year or two ago I was all set to buy it, but you seemed to loose interest so it kind of made me loose confidence in it.

10-4 on the adjustment but I'm working on that as well but once I have this a simple frequency adjuster is all that's required.

It would but there is only about 1mm clearance between the side of the VR sensor and the bolt. And the bolt only sticks out about 2mm past the nut.

Not sure but simulating the hall effect sensor output and running it through my circuit show it to work best with more metal. Actually 60% seems optimum but anywhere from about 30% to 90 metal will run the speedo to full speed. The output of the hall effect goes from zero when the sensor is over the metal to + 14V when the sensor is over air. Its extremely clean.

Absolutely Loktite is your friend.

I have tried all the waves on my function generator. The speedo loves them all, Sine, square, triangle L and R saw tooth. The sine wave requires the smallest voltage threshold to get it to respond, followed closely by the square wave.

Those were peek to peek voltages. I measured both loaded and unloaded. Loaded drops it about 25%. But those numbers were at very limited speeds. My lathe only goes to about 45 MPH. I also got some data at about 70MPH withe the wheels jacked up from both a stock MB system and mine. But what ever I got the speedo would need to accept twice that. You can increase the gap but you need to watch you minimum voltage threshold at low speed. That's the beauty of the hall effect although the speedo doesn't seem to care. I tried a circuit with zener diodes to keep the voltage to never exceed 12V. The circuit worked great but it didn't change a the way the speedometer worked at all.

Its all a matter of perception, of what you feel is going to be the shortest effective path to getting the job done. And the path you don't choose may be the path you end up on after all is done.

I didn't want to use my diff wheels because they are too far removed from MB's original design 48, tooth verses 4 teeth. I know those frequency changers can shift a frequency +/- 50% but can they shift it 600%???. And then there is sure I have 48 tooth wheels but I would need to make a mount for the sensor. Both those tasks have about the same level of complexity. And then theirs why didn't MB do it that way. Why would they put both a 36 tooth ABS wheel in the diff and a 4 point wheel in the trans??? Bottom line is rarely is something ever as easy as it seems and the pitfalls to each path will only be uncovered by trying it out.

I am sorry if I flaked out. I vaguely remember talking to you about it. I think the problem was getting a car of your vintage in to look at to confirm the values.
I was thinking that my speed sensor kit may work to drive a speedometer as well. Problem though is pulses per rev and correction to the differential ratio and tire size still. It would still be easier just to do it over CAN. What year is your donor again?

I'm pretty much convinced I have the solution. Its been a long way getting here but its done. The hall effect seems to alleviate any of those interference problems. The VR sensor is really a nice easy solution for fixed speed applications but if its got to work from 10 to 210 MPH your sensor really needs to be no where near anything else. After doing what I did I'm glade I didn't wast my time trying to pick up the head of the flange bolts. Yeah they would work but be very difficult to get a smooth enough and consistent pulse over the full speed range.

It all gets back to the last paragraph in my previous post to 97 SL320. You were unsure about programming it correctly, I was unsure if I could get it to work on my own, just as I was unsure weather or not I could take signal off my 48 tooth wheels and convert them. So I did what I thought was the simplest thing, doing what Mercedes did. Turned out to be not so simple but I do believe its done and will be a robust solution.

In reality they are all pretty simple solutions once you know which screw to turn.

In ref to powder coat / plating / paint:

Paint seems to hold up pretty well on the driveshaft so it would be fine on the tone ring. What type of finish is on your axle tone rings? I'd say these are more exposed to weather than the trans flange.



I don't know. I also don't know that type of circuit they use.

If I was making something from scratch because the fooler could not make that great of a change, a counter circuit that outputs a pulse to the speedo fooler every 12 pulses would be a good start. The speedo fooler would be for scaling the signal to your diff / tire combo. Sort of like a 2 stage pressure regulator.


It may be possible to duplicate the axle sensor bracket with just a clock position change. I'd still call this , from a mechanical stand point, less work than the trans wheel + bracket.


ABS needs greater resolution than a speedo because ABS needs to tightly control degrees of wheel rotation where as the speedo needs just enough pulses so the needle does not bounce at low speed. You could add damping but then the speedo would be too slow to react. ( As a side note, I wonder if some performance cars reduce needle damping during hard acceleration to make the car appear " faster " ? It would only take a few lines of code to see > 80 % throttle = less damping. )

That 4 point wheel on the trans becomes a ~ 12 point wheel at the diff when counting wheel revolutions so the gap isn't as large as it sounds.

There is also the issue where , in the pre 722.6 era , various systems evolved more or less independently. RE: speedo cable to electronic speedo. ABS needed greater resolution and 4 wheel signaling so the speedo 4 point was never considered.

Also, given ABS systems are made by others and relatively new at that point, it would have been impractical for Bosch / Tevis / ATE to integrate other systems for everyone's brand of car.


True. Too bad many people think because a car makes it's way down the road it took no effort to do so and if there is a problem at some point in the cars lifespan, why this was not foreseen.

The ABS wheel speed rings are painted. They don't hold up very well but I believe the POR 15 paint will be alot more durable. The next time I paint them that is what I will use. But its really just cosmetic the original 129 rings were located on the outside of the axles and I don't believe they were painted. They were just rusty sheet metal rings pressed to the outer axle.

It is definitely possible to clock another location onto my current bracket but that bracket is not easy to make. For me to fabricate either a tone wheel or the abs bracket is about the same . For people with the right machines to fabricate these same parts the tone wheel is definitely faster and cheaper. That's because the brackets really require tooling. The tone wheel on a machining center with live tooling is all programming then 10 min in the machine. For me on a 1HP 1960's Bridgeport and 1.5HP 1940's Logan lathe its a 20 hour job.

Hey I'm not saying it wont work, I just chose a path that appeared to have the least amount of unanswered questions. And even though my first attempt didn't work that doesn't mean it cant work. Each attempt required finding and flushing out all the pitfalls. I had the same issue when I first installed the ABS sensors. My first brackets weren't stiff enough and vibration caused them to have problems.

The speed sensor is in the trans itself,amongst the clutch packs . there are two magnets which collect the signal from a toothed wheel passing over them. The magnets are in the plastic contact plate,so if you ever have to change a contact plate you can see the them. On one of mine, a needle race has come apart in the trans and the speed sensor magnets are covered in needle rollers. Yes it drove until the number of needles prevented a signal reaching the magnets and the drive was lost. :rolleyes:
i cleaned the needles off,replaced the plate and lower pan and proceeded to use the car for a few more weeks until the trans gave up completely. :P
So.. If you want to be able to get a speedo to read the signal from the trans, the loom coming out the side is where you start. the Mercedes WSM has all of that info (nothing much on line regarding which color wire but suspect it will be typical DIN and white. You only need to come up with a device which will power a digital read out when activated by the ignition key .

None of those sensors can be used since none of them are on the output shaft. Your speedometer would only read right in 4th gear.