Megajolt Temperature Correction

[RichardB]

If you have MJ, do this! I set it up on the Megajolt this week and it's made a big difference to how smoothly the car runs on choke when warming up. You use a lot less choke, and it's actually close now to how modern EFI cars are where you can't really tell the difference in running between a cold or a hot car, because it runs okay at any temperature. Power seems to be available to your earlier.

 

I used the Classic Mini's 5/8 UNF" 2-pin temperature sensor fitted to multipoint injection models. It directly replaces the old temp gauge sender in the water pump housing without modification, didn't even need the dowty washer I got for it as it seems to have self sealed. You'll need to fit the old temp sender somewhere else if you still want to use it, I fitted mine in a Stanpart stat housing with a 5/8" UNF hole but the gauge behaves very differently after the thermostat now, coming on way later (obviously as the stat is shut) and is a bit less stable.

 

For calibration I used the Ford sensor map that's in the Autosport labs forum, but the figures are way off for the Rover/Mini sensor. Cold water shows up as 89C and running temp is 225C (I'm running an 88C stat). I'll work on getting the sensor calibrated so the readings are more meaningful, as then I can get a nice little datalog of running temperature as well which will be great.

 

Here's the advance correction as it is now:

 

1 - 2 - 3 - 4 - 5 - 6 - 7 - 8 - 9 - 10

Bins 60-80-130-170-200-225-233-240-248-254

Corr 4-3-2-1-0-0-(-1)-(-2)-(-3)-(-4)

 

 

I've attached a printscreen as well of the correction table. There's no science or that much thought behind these values, but they seem to work well with the map I'm currently running.

 

Apologies that this is a bit out of sequence, I should be describing the Megajolt installation first but I still can't find the camera with the in progress build pics I used. Figured its best to get this advance correction discussed now as it's not really covered elsewhere on the net, and for less than a tenner extra it's a huge shame not to take advantage of it if your MJ has the feature.

[RichardB]

Correction map attached.

 

Sensor is part

# GTR185 / ADU7161 / EAC3927, cost is about £6-15 depending where you shop. Mine was £6.99 from ebay in Intermotor packaging.

 

You need a 2 pin junior timer (Durite) connector, these are way too much money for what they are (£3-5 inc terminals and rubber boot).

Edited April 11, 2012 by RichardBaines

[Millstone10]

Many thanks for this info Richard. This was exactly what i was looking for.

 

Nigel

[RichardB]

Here's a photo of the sensor installed into the water pump housing, you can see the bulky durite connector attached to it.

 

The old temp gauge sender has now been moved to the thermostat housing and didn't need a new wire making up.

 

[Nick B.]

Very neat install.

 

I love the fact that even rimmer sell these sensors. Of course there is a Triumph premium on them, so instead of mini-price (7.5 quid) the Triumph price is 21 quid.

 

Anyways - what is the difference between single pin standard sensor and this double pin sensor? Isn't it possible to piggyback it somehow and connect it to the smith-instrument?

 

Cheers

Nick

[RichardB]

Thanks Nick.

 

The main difference between them is that the single pin sensor uses the sensor body as a ground so it earths through the engine block, whereas the 2 pin Rover sensor has a second pin to connect the earth to the same point on the ECU (or in this case, the Megajolt box). Apparently the MJ really wouldn't like being fed a single input signal.

 

As to whether you could piggy-back it, I have no idea - couldn't hurt to try I guess, but I suspect it wouldn't work. Even if it did the gauge readings would probably end up being quite different, whereas on the setup I have now, the optimum temp is at the same point in the gauge (just fluctuates a bit more and the gauge doesn't really come on until the thermostat starts to open).

[RichardB]

...'to the same point on the ECU' - by that I mean same earth location as all the other earths on the ECU. In the case of the Megajolt, that'd be this ground:

 

[Nick B.]

Interesting.

 

Sorry if I keep missing the obvious, but couln't you just as well piggyback the signal from the rear of the tempgauge and go from there to the MJ without getting new sensor etc? Does seem a bit simpler ?

 

Cheers

Nick

[RichardB]

I read somewhere that it wouldn't work on the Autosport labs forums. Can't remember why. Give it a try and let us know the results!

[Millstone10]

I have read that the earth through engine and chassis has to much noise to give a good stable signal for the purpose. Its in the Megajolt forum/guide somewhere but certainly makes sense.

[spitfire6]

I have read that the earth through engine and chassis has to much noise to give a good stable signal for the purpose. Its in the Megajolt forum/guide somewhere but certainly makes sense.

 

To help, I hope!

 

 

Hi,

 

To help understand, consider this.

 

If you added a big wire (more than 2.5mm² )to the body of the grounded sensor and connected it to the negative battery terminal, then you would probably think that you have helped because you now have a better earth.

 

But, this new earth connection you had made might be used by other returning currents from all/some of your car electrics as it presents a lower resistance.. any additional current flowing through your new cable will affect the signal from the transmitter, as what’s in must be what’s out..

 

Now I hope you can see that without adding a cable from sensor ‘earth’ to true earth/ground/negative terminal would be even worse?

 

All small signal returns/grounds should be connected at one point and then returned to (main)ground. This is why the ECU has a common ground point that is floating until you connect it to ground.

 

 

Its quite normal in industrial electrics to have different grounding points that all ultimately are connected to a common ground. Prior to microprocessors and such, grounding a sensor/transmitter via the main earth/body was never a problem. Modern transmitters/sensors always now have two or more wires and neither are grounded at source.

 

Tried to explain, not easy, if there was no such thing as resistance I would not be trying to explain.

 

Cheers,

Iain.

[James]

It isn't a brilliant idea though I believe it can be done. The issue is that it'll effect either the megajolt or your gauge one way or another.

[spitNL]

Won't work, the water temperature sender is used as a voltage divider.

If you add another wire to it (megajolt) this will affect the resistance and as such the voltage, which means both the temperature gauge and the megajolt will be unable to measure it correctly.

[spitfire6]

Won't work, the water temperature sender is used as a voltage divider.

If you add another wire to it (megajolt) this will affect the resistance and as such the voltage, which means both the temperature gauge and the megajolt will be unable to measure it correctly.

Hi,

Not all true, as I doubt the MJ will affect the reading to the gauge, as the gauge is current operated. In any case the reading to the MJ would be wrong. Maybe the MJ has a lookup table you can program so that it reads the non linear signal from the transmitter/gauge?

 

Try find table of volts to temp for the gauge/trx combination and program your MJ accordingly.

 

Cheers,

iain