Need to find a compatible magnetic speed sensor for speedometer

Thread Starter

MB107

Joined Jul 24, 2016
103
Yes after looking at some of the sizes of these components, I'm starting to realize they can be put into a small package. I still need to fit the board into a housing of some kind.

I'm also leaning away from the relay idea as someone will take it out and replace it with a standard relay of which there would be many of right next to it.
 

Alec_t

Joined Sep 17, 2013
10,698
I still need to fit the board into a housing of some kind.
In days gone by I found 35mm film canisters handy for small projects. The main electronics distributors carry a range of 'proper' project boxes. Other likely candidates are small containers with snap-on lids. Your local supermarket/pharmacy probably has loads.
 

Thread Starter

MB107

Joined Jul 24, 2016
103
I found these. Funny when I do a Google search I get these from Digikey. 1" x 1" bread board 1" x 1" x 0.7" box. When I sear ch within Digikey I cant find anything under 2"

boxes.jpg
 

Thread Starter

MB107

Joined Jul 24, 2016
103
Here you go :
View attachment 179344

Even if the sensor output stays high for a long period the input to the speedometer goes low.
OK results are in at least for the third circuit. Everything works as you have shown up until when I connect the speedometer. The signal changes exactly as you have shown but when I connect it to the speedometer the signal becomes a clipped saw tooth between 10 and 13V . Consequently the speedometer doesn't work. I believe the signal would work if I could get the minimum voltage to < 0.2V

I believe this may have something to do with the speedometer not being a 2 wire display as has been depicted in all diagrams so far. It also needs to be supplied with 12VDC even when using the VRS magnetic pickup.

Hall Effect with transistor and cap disconnected.jpg
Hall Effect with transistor and cap connected.jpg
 

Alec_t

Joined Sep 17, 2013
10,698
Hmm. Looks as though the speedo input needs an active pull-down. Try this :-
SpeedoSensorInvert4.PNG
Note Q1 is an NPN type. R3 could be 47k to reduce standby current. R2 could possibly be omitted altogether (not connected to V1) if the speedo input itself provides a 12V pull-up.
 
Last edited:

Thread Starter

MB107

Joined Jul 24, 2016
103
Hmm. Looks as though the speedo input needs an active pull-down. Try this :-
View attachment 179902
Note Q1 is an NPN type. R3 could be 47k to reduce standby current. R2 could possibly be omitted altogether (not connected to V1) if the speedo input itself provides a 12V pull-up.
Thank You Alec

Fortunately I have all the parts for this so I should be able to get to it this week.
 

Thread Starter

MB107

Joined Jul 24, 2016
103
Hmm. Looks as though the speedo input needs an active pull-down. Try this :-
View attachment 179902
Note Q1 is an NPN type. R3 could be 47k to reduce standby current. R2 could possibly be omitted altogether (not connected to V1) if the speedo input itself provides a 12V pull-up.
With your latest circuite, the signal connected to the speedometer looks absolutely perfect now with the exception of the noise floor. I would think this would be perfect but the speedometer will not budge of zero. Perhaps the noise floor is just too much as I have determined that the minimum threshold for the speedometer was well below 1 volt. My 8 bit scope is limited in resolution but it is reporting +0.4V/-0.2V noise. That would mean that I least have 0.2V and that may be over the threshold that the speedometer reacts to. I will have to look into things a little more tomorrow. Maybe I have some wire crossed.

Noisy Floor.jpg
 

Alec_t

Joined Sep 17, 2013
10,698
Curiouser and curiouser.
Not sure if we discussed this, but can you confirm (1) that the speedo is rated to accept input from both a variable-reluctance type sensor and a Hall effect sensor and (2) what are the conditions under which you have been able to get the speedo to read something?
 

shortbus

Joined Sep 30, 2009
7,330
Are you sure the speeedo doen't send something to a control module? Aren't the modern electronic speedos all stepper motor gauges?
 

Thread Starter

MB107

Joined Jul 24, 2016
103
Curiouser and curiouser.
Not sure if we discussed this, but can you confirm (1) that the speedo is rated to accept input from both a variable-reluctance type sensor and a Hall effect sensor and (2) what are the conditions under which you have been able to get the speedo to read something?
I have no way of telling if it is rated for Hall Effect. The car was delivered and with VRS, so I assume that it was only intended for VRS. I have put sine waves from a signal generator to it, I believe I have tried square and triangular as well. If I recall anything above 0.2V amplitude with the signal generator will get it to go.

I'm wondering if the speedo really requires negative signals. Look at post 36, the two signals have some negative component, Even the signal that was clipped with the diode still has about a -1V spike. Maybe the +/- signal is important or possibly just a negative pulse?
Are you sure the speeedo doen't send something to a control module? Aren't the modern electronic speedos all stepper motor gauges?
 

Thread Starter

MB107

Joined Jul 24, 2016
103
It does appear that the negative component of the signal may very well be important. I connected my wave form generator and attached it to the speedometer I tried a 1 Volt sine wave at 100Hz and moved the offset up and down. In the case of the sine wave, when the wave became slightly all positive (120% offset) the speedo shut off like a switch was turned off. Also when the wave was too far negative (-80% offset) the same occurred. The situation gets worse when you get to saw tooth waves Saw toothed waves were not able to operate at 100% offset. So I believe if this circuit could be modified to pull the voltage down even further the hall effect solution will work.1.jpg 2.jpg 3.jpg 4.jpg
 

Alec_t

Joined Sep 17, 2013
10,698
That's very helpful info.
So here's something to try. It might need tweaking, depending on the input characteristics of the speedo. The output now goes both positive and a volt or so negative, so cuts through the noise floor and provides zero-crossings which it seems the speedo is expecting. D1/D2 clip the negative voltage, but if your speedo input will tolerate (or requires) a larger negative excursion then the diodes could be removed, or replaced with a single 10k (or 47k) resistor to ground.
SpeedoSensorInvert5.PNG
 
Last edited:

Thread Starter

MB107

Joined Jul 24, 2016
103
I will give it a try tonight. I don't quite understand the two Zener Diodes in series. Seems like it would be the same as one
 

Thread Starter

MB107

Joined Jul 24, 2016
103
OK the results are in and we have a winner. The circuit diagram as proposed did work at low speed but not high speed. With diodes replaced by either 10K or 47K resistors the speedometer works beautiful. I like the 47K resister solution as it has the most symmetric wave form more similar to what the original VRS system had. It also has nice crisp vertical lines where the signal crosses zero.

Ultimate test will be to put it on the car. That may take a few weeks, but I am very confident.

Thank You for you patience and sticking with me Alec_t. I couldn't have done this without you. You are a real asset to this forum.
1.jpg 2.jpg 3.jpg
 

Alec_t

Joined Sep 17, 2013
10,698
I don't quite understand the two Zener Diodes in series.
They're ordinary diodes, not Zeners. Each diode drops about 0.6V when the capacitor output swings negative, so the negative voltage spike gets clamped to 2 x -0.6 = -1.2V.
OK the results are in and we have a winner.
Glad we got there in the end (fingers crossed). So it seems the speedo is not designed to work with the raw output from a Hall type sensor, but as we now have a pseudo-VRS it's happier with that.

Edit: Just one thing. Using the 47k option I can't get my simulation waveform to match what you're seeing (it looks more like your 10k waveform); but using 470k it gives a fairly close match. Out of curiosity, can you confirm your resistor value? (Not that it really matters. Hey, if it works, it works :) ).
 
Last edited:

Thread Starter

MB107

Joined Jul 24, 2016
103
Will check the resistors tonight but they are the same 47K resistors I used previously. Maybe I fried it and its open. I have not tried no resistor yet, or maybe I have?
 

Alec_t

Joined Sep 17, 2013
10,698
I have not tried no resistor yet
By all means try without any resistors if you wish, but the 3k3 resistor is a safety feature to limit the current through either the Hall sensor or the speedo input (since their characteristics aren't specified), and the 47k resistor is a pull-down because you didn't want the speedo input staying high (though that is probably unnecessary now that we know the speedo needs negative pulses to operate).
 

Thread Starter

MB107

Joined Jul 24, 2016
103
Resistors checked out OK ~ 45.5KOhm. Re ran it and got the same results. Tried to run without any resister at all and it wouldn't work. It was just a reduced version of the input signal all positive and not crossing zero.

While I don't have data for the speedometer, I do have it for the sensor, see the attached file.
 

Attachments

Top