Data logging oil temp

[mcerbm]

I was hoping to log oil temp with my vbox. I have a k-series and I noticed that on the coolant side there are two separate senders, one for the ecu and one for the gauge. I was wondering if this was done because if you have two devices reading the same voltage from a sender it affects the result, hence they kept them seperate?

 

I have an oil temp sender in my bellhousing dry sump tank which goes to the temp gauge (coolant / oil temp changed by a switch). If I solder a wire into that gauge oil sender feed would it affect the result?

[revilla]

If you were to attach a high-impedance device like a voltmeter or the input of a device designed to measure voltage, it shouldn't materially affect the result. Those sensors are variable resistance and calibrated to work with a particular gauge. Certainly attaching two such gauges would make them both completely misread. You will need to do some experimentation to calibrate the curve of voltage against temperature which will be non-linear.

[Jonathan Kay]

If you were to attach a high-impedance device like a voltmeter or the input of a device designed to measure voltage, it shouldn't materially affect the result.

Snap!

Would it be better to read voltage or resistance directly?

I was wondering if this was done because if you have two devices reading the same voltage from a sender it affects the result, hence they kept them seperate?

It might reflect history and convenience rather than anything else. (Crossed with #3!)

Jonathan

[revilla]

I guess the reason two sensors are used for water is simply that they require completely different calibrations. The gauge sensor needed to be compatible with the existing gauges and the ECU sensor needed to be to Rover spec.

[revilla]

You can't read resistance directly. You read resistance by passing a known current through and measuring the voltage, or putting a known voltage across and measuring the current. In this case we don't really have a good idea about the voltages and currents due to the attached gauge, which is definitely not a high impedance device. All we know is that the voltage at the sensor will most likely vary with temperature and that reading it with a voltmeter or equivalent shouldn't materially affect it. So what I would do is to prepare a calibration curve by recording the voltage seen against the gauge reading (for want of a more accurate temperature indication) and then log the voltage. You would then need to look up the temperature for any given voltage from the calibration curve.

[Jonathan Kay]

Thanks, Andrew.

I remember the explanation why it was better to read PD rather than resistance in fault-finding.

What I was suggesting here was measuring resistance directly across the sender with the leads attached directly to either side. That would remove the dependence on system voltage... but could it affect the actual resistance of the sender (as seen by the original gauge)? I suppose that depends on the impedance of the added instrument...

You will need to do some experimentation to calibrate the curve of voltage against temperature which will be non-linear.

I think that most of the sender characteristics that I've seen are linear... more, please.

Jonathan

[revilla]

Any "resistance meter" you attach across the gauge in that way will simply be passing a small current through the sender and measuring the voltage, and displaying this as a resistance value. With the gauge attached, the gauge will also be passing a (much larger) current through the sender and so voltage measured by the resistance meter will be much higher than that which would correspond to the current it is supplying, so it would read completely wrongly. Right I'm typing this half awake while eating a hotel breakfast so you'll have to bear with me if I'm not very coherent or make mistakes but by my reckoning, the internals of the gauge are probably a linear circuit (with no diodes etc.) in which case the input terminal to the gauge will be electrically equivalent to some fixed DC voltage source V in series with some resistance G. The sender will appear to be a resistance S to ground. So the voltage at the sender will be a fraction of V given by S/(S+G). At low sensor resistance it will tend to zero and at high sensor resistance it will tend to V but VS/(S+G) is not a linear function of S, so even if S had a linear or inverse-linear relationship to temperature, the sender voltage would not.

[mcerbm]

I’m not sure if this adds anything but I was planning to input the oil temp to my K6 emerald and then use a spare channel on the Canbus which will be connected to the vbox. 

So it’s not a gauge but a pin in on the loom I will be connecting to. 

[Richard Price]

Mcerbm,

are you planning on only reading into the Ecu, or trying to use both the guage and the ECU? If so, you won't be able to use the switch to swap between oil and coolant temperature unless you log whichever the guage is showing.

if you're using only the sender to the ECU and no guage, then you'll need to connect the sender in series with another resistor , then use the ECU to measure the voltage across the sender. You may need to read a raw value from the ECU to the box, then manipulate the numbers in the datalogger  to   Display sensible engineering units.

[mcerbm]

Yes it would be the gauge and the ECU. The switch is just for switching the input to the gauge, I was going to splice into the oil temp wiring pre-switch. So if the switch is in the coolant position and I flick it to oil are you saying that this will affect the result I will be logging? I.E. if I calibrate the software to get the temp it will only work in one switch position due to the way the gauge affects the results?

if so sounds like it’s better to mount a separate oil temp sender

[Richard Price]

The voltage drop across the sender depends on the current that flows which will change if the guage is in circuit or not.

if you want the gauge AND data logging, you need separate senders (maybe with a tee in either pipe to or brim the tank. Or drill and tap the tank for another sender close to the standard sender)

[mcerbm]

Ok that settles it then, a separate temp sender is required. Now where to mount it? I dont feel like drilling into the sump tank. Would it require a flow over it, or would it be ok to be in a static section of oil?

The reason I ask is that I have a mechanical / capillary oil gauge. It is fed via an adapter with 3 ports. One is used for low oil pressure warning light feed and the other is the capillary hose to the oil pressure gauge, the 3rd is the feed from the back of the oil filter housing. maybe I could source a block with 4 ports.

This doesn't work if a flow is needed for a more reliable temp though?

[Richard Price]

No. The sender would need to be positioned actually in the tank, or, with oil actually flowing over it, to give a reliable reading.

[garybee]

Have a look at your oil filter housing.  You should find that there is a spare location for a sender there.    

[revilla]

What Garybee said  You may need to remove a threaded plug with an Allen key:

[mcerbm]

Fantastic, I will remove and add a temp sender next time I do an Oil change. Then route the wiring back to the ECU to be logged as an aux temp and off via canbus to the vbox.

[7 wonders of the world]

Depending which DS set up your using you should look to take a reading from the sump as this will give you the highest temp reading for maximum values.

The Raceline DS has a nice blue plug which secures the filter gauze, this can be tapped to take a suitable sender mounted in the hot flow.

 

[mcerbm]

Its caterhams own dry sump system that I have. If it were to be taken off the oil filter housing would the temps be substancially different?

I must admit that recording readings on the hoses with a temp gun as the oil exits the laminova heat exchanger and the guage readings for the oil temp (from bell housing tank sender) did differ quite significantly.

[mcerbm]

What size is the fitting required in the oil filter housing for the sensor. The Caterham supplied sender for use with their temp gauge may be too low sensitivity to work with the emerald ecu for data logging. Anyone tried this before and have a sender suggestion I could use?

[rj]

I would expect that it's an M10x1 but can't tell for sure. It depends on, if the thread is cut by Rover/Powertrain or it's something added by Caterham - in which case it could be 1/8" NPT.

Best advice it to try to find a thread gauge. 

I have a drawer full of VDO (Stack) sensors similar to the one you have in your bracket, those are M10x1 and if you ask kindly, I can pop one in an envelope for you. (Takes a few days to reach you as I'm in Denmark and compared to our postal service Royal mail is extremely reliable).

If you want one, please send your address in a PM.

[7 wonders of the world]

Needs to be in the sump gallery to get true data, cant remember on the K set up but I'm guessing there will be a similar access cover / bung for the finger filter which can be modified to provide a suitable location for a sender.

I have access to small number of modified filter plates for the Duratec Raceline DS  but none for the K.

[rj]

Regardless where you place it you'll just get the temperature where the sensor is located; it'll always be a relative reading.

My own active sensor is in the lateral wall of (RHS) of the sump pan, where there's just cut an M10x1 thread.

The apollo tank also has a thread for a sensor, but it's not used.

 

[7 wonders of the world]

A relative reading is fine for comparisons however if your going to the trouble of doing this it makes sense to place it in a position to record accurate and usable data otherwise you will never accurately know how much energy the oil is being subjected too.

[mcerbm]

It’s a Caterham bought sender that’s in it just now. But no details on the site for the thread. Link below  

https://caterhamparts.co.uk/senders/17-temperature-sender-96-onwards.html?search_query=sendeR&results=71

It’s already installed so can’t remove it just now without losing oil. I’m just in the process of trying to get it configured with Karl at emerald so it’s calibrated correctly. But he thinks the gauge senders have too low a resistance range. Once I know the thread type he will give a suggestion of what would work better with the ecu software. 

[rj]

I'm afraid you have to:

https://www.lotus7.club/forum/techtalk/k-series-oil-filter-housingsender

If Caterham have tapped it it can be an NPT, otherwise it'll be M10x1.