Electronic speedo drive removal - help needed.

[hagarep]

Not sure if this is the right section but most speedo post are in here.

Has anyone got any tips on how to remove the electronic sender for the speedo, from the gearbox?

Mine appears to be welded in with 44 years of dirt. My only option at present is to use a small screwdriver as a pry bar or chisel.

Was hoping not to destroy it as they are a tad expensive.

Jason.

[proporsche]

i assume the one on the side of the transmission...soak it with a lot of wd40

 

Ivan

[hagarep]

Been hitting it with rost off and even rost of ice but it is well and truly in there.

[Jonny Hart]

I had to lever mine out with a screw driver.    The sensor separated from the plastic ‘cup’ but I managed to repair by filling with  Araldite.   Been fine for 10 years.  

[SilverWT]

The three I have removed from three different gboxes had all been stuck/sealed in with some sort of compound. The only option was to gently pry them out. They all came out OK, but had to have the compound scraped off before reuse.

Mark

[hagarep]

Thanks guys. I guess if there is sealant, I can try the heat gun too.

Did you find anything on the sensor that could interfere with the signal? My speedo reads  30mph when the car is doing 40 by gps. Quite a significant error. Ground to the speedo is good and the speedo itself was recondition by a friend of Julian Reap. I opened it up and it had new gears and looked new inside, so I am assuming its the sender unit.

[SilverWT]

Senders will tend to either work, be intermittent/erratic, or dead. If the speedo is steady at a fixed speed and goes up and down smoothly, I would suspect speedo calibration.

The other possibility is some of the permanent magnets have become detached from the carrier on the side of the diff, giving a false reading. Not sure that is very common though.

If the same degree of error is present at higher speeds it could just be the speedo needle needs repositioning. There was a post about this last week.

 

Mark

 

[angusb]

Are your tyres the same size as originals?

[hagarep]

Normal 225/50 16's but my car was supplied on 15 cookie's, now that you mention it.

I have tried the adjustable POT on the back of the circuit board in the speedo, but that has a range of around +/- 5 mph and my error is much higher.

Try to test with a signal generator next. 60 mph is 106.66 Hz according to my math. There are plenty of threads on Pelican and videos where people have done this.

Even considering a GPS signal conversion.

Jason

[Leicestershire]

On 08/06/2020 at 09:32, hagarep said:

Thanks guys. I guess if there is sealant, I can try the heat gun too.

Did you find anything on the sensor that could interfere with the signal? My speedo reads  30mph when the car is doing 40 by gps. Quite a significant error. Ground to the speedo is good and the speedo itself was recondition by a friend of Julian Reap. I opened it up and it had new gears and looked new inside, so I am assuming its the sender unit.

Is the speedo error consistent at -10 mph regardless of speed?
If the unit has been rebuilt there is a good chance that the needle has been re-attached in the wrong position. They are just a push fit.
On most new cars the speedo is calibrated to display 2 - 3 mph more than true speed.

[SilverWT]

 

9 minutes ago, Leicestershire said:

If the unit has been rebuilt there is a good chance that the needle has been re-attached in the wrong position. They are just a push fit

This^^^, and this -

On 08/06/2020 at 12:31, SilverWT said:

If the same degree of error is present at higher speeds it could just be the speedo needle needs repositioning. There was a post about this last week.

Which is here -

Mark

[hagarep]

Successful test using iPhone soundwave app. Speedo worked correctly, but is seriously inaccurate. I will add the wiring diagram for the test along with a table or graph of the results. Basically, using the potentiometer, you can set the speedo spot on for a single speed and it will read high above that and low below it!

This means that my sender unit is not working correctly, but if I spend £170 or more on a replacement, the speedo will still be inaccurate.

if I convert to a GPS sender, it still sends pulses to the speedo so would it be any more accurate?

[Leicestershire]

The potentiometer is exponential gain??

Can you set the pot to display a linear reading - not focusing on a true value at stage but a consistent +/-  number?

 

[hagarep]

[hagarep]

3 hours ago, Leicestershire said:

The potentiometer is exponential gain??

Can you set the pot to display a linear reading - not focusing on a true value at stage but a consistent +/-  number?

 

No, linear.

It appears that the rate of change on the speedo output is too high and never parallel the calculated linear increase (reference). The pot only increases or decreases the output across the whole range, so, you can adjust it to be correct at only one point. I set mine to 60 but the problem is the inaccuracy is worse low down, where we need 30mph to be accurate for the speed cameras. But set it correct at 30 and it is massively out higher up. I will refit the speedo and drive to work this week, as the weather is nice. This should confirm my sender unit is not working properly, leaving me with three options:

1) Order and fit a new sender unit, live with the inaccuracy.

2) Fit new sender and something like a Yellow Box (www.yellr.com) which has some positive feedback on Pelican

3) Fit an Auto Meter gps speedo converstion

4) Realise I have currently run out of jobs on the car and am stuck in work with way too much time to kill. Remember that I can't see the speedo because the Prototipo wheel is in the way.

[angusb]

As you know , the Speed sender is a Hall switch, which closes when faced with the rotating magnets in the gearbox. There is a easy method to test , by hooking a multi meter to the sender cables,then spin the rear wheel. Your multi meter should beep at regular angular intervals as you spin the wheel.From memory my 88 Carrera has about 9 beeps per 360 degrees wheel rotation,. Your car may have a different number.

 If you do  need to replace the sender , make sure that you get a geniune Porsche one, the aftermarket ones do not work.

Personally would question the accuracy of a GPS  to measure speed,all cars speedometers are inaccurate to a degree,the fact the you are running larger tyre than original has compounded this inaccuracy. 

[hagarep]

Thanks Angus

I have previously checked the sender for output, but did not count or measure it. A new sender is £108 with club discount from Porsche.

I believe the gears affect only the milometer and are dependent on the gearing/final drive ratio. The wheel and tyre change will make no difference as the critical dimensions are the same.

I need to road test the car once I have fitted a new battery and still need to go on Pelican and research the gps and yellow box conversions.

[Leicestershire]

The standard C3 speedo is calibrated to a 215/60-15 tyre (2007 circumference).

225/50-16 is a later factory fit alternative size.

[hagarep]

Well  I'm sure tyre size isn't the main issue. Got the car to 80 yesterday, but the speedo went to only 50. At lower speeds it was barely moving. Time to count the pulses and maybe pull the sender unit out by any means.

[hagarep]

So last update on this subject.

Got a new sender from Porsche, got the old one out with heat gun, hammer spike and pry bars.

Fitted the new one and took the opportunity to connect the multi-meter and check the pulses.

5 instead of the 8 I was expecting. Speedo reads 50 when the car is doing 80, so I guess this makes sense.

Nothing to do now but use Waze and maybe something with just large numbers for speed, that better suits my eyesight these days, until I have enough saved for the gearbox rebuild!

Jason.