Hi All,
After my recent battle with cooling I've been off doing what I should, DRIVING :-)
Lots of fun driving, but I'm still not convinced on the cooling.
The new temperature sensor shows her running warm, almost hot, not good in the Desert.
A bit of a search on the MG Experience and it looks like there are two sorts of sensors.
BUT not clear on what which sensor with which engine etc. so out with the kitchen physics kit and time to plot some graphs.
These are the two that I've got.
Original
Part GTR101
Dia : 8.39mm
Sensor length : 13.3mm
Connection : Perpendicular
And then new :
New Sensor
Part GTR104
Dia : 8.59mm
Sensor length : 12.4mm
Connection : Axial
The way the temperature gauge works is all about resistance. The sensor is a thermistor, the hotter it is the lower the resistance.
So out with the sauce pan and multimeter.
Plotting a graph for each sensor it is clear that the original sensor the gives a higher resistance.
Deg C Old New Deg C Old New
26 909 554 70 181 146
28 869 542 72 168 134
30 849 524 74 160 119
32 792 482 76 142 112
34 725 454 78 132 109
36 667 406 80 124 98
38 599 378 82 116 91
40 563 352 84 108 86
42 529 335 86 101 75
44 496 325 88 95 69
46 442 304 90 90 65
48 421 281 92 85 59
50 400 272 94 80 54
52 362 268 96 69 49
54 343 235 98 61 45
56 311 206 99 56 47
58 288 202
60 270 184
62 255 179
64 244 170
66 213 186
68 197 180
Problem is that old sensor shows the engine running Cold (to cold in my mind) and the new one running Hot !
Time to get the laser thermometer out !
ANY IDEAS ANYBODY ?
Cheers,
Richard B.
Find out what resistance is delivered by 90C, then what voltage that allows through to the temperature gauge. If 90C doesn't put the needle in the middle, adjust the voltage with a suitable in-line resistor.
ReplyDeleteOr mark where 90C is on the temperature gauge and drive the car. Of course, this is only sensible if 90C doesn't correspond with the needle up against the stop...
Trying very had not to be dodgy.
ReplyDeleteEasy to fit a small variable resistor and "fix" the problem.
Going to get some real temperature readings (laser thermometer, or maybe borrow a termal camera from work) and see how hot the problem actually is.
I have an idea........GET A TOYOTA!!
ReplyDeleteHa Ha Ha.
ReplyDeleteHow many trips to the garage has your Patrol had to sort out the overheating ?
Think I've got a few more attempts before I reach that lever ;-)
I am currently in contact with an SMP Europe product engineer about the two MGB senders that have been giving false high readings for many years. Repair specialists like myself have to resort to good used senders to get out customers cars to read in the dead center like they were designed to.
ReplyDeleteI have determined the resistive loads necessary to make the gauges read right, and they are nowhere near what is being sold, as you surmised.
To add an elecrtical component such as a resistor skews the reading across the board, so is not a good solution. Until we can convince the manufacturer that there is a problem (and hopefully that will happen someday) and until all the defective units are out of the "pipeline", the only solution is to locate good used units from a salvage operation. There were two different senders. One for all MGB from 1968 to 1976, and one from 1977 and on when the engines had to run hotter for emissions standards, and the normal range of the gauge was shifted back to dead center by the different sender. All the gauges used the same input signals for low, Medium and High.
Hope this helps.
Thansk Glen. I also cheat and have anothed modern digital unit fitted.
ReplyDeletehttp://diary-of-mgb.blogspot.com/2011/03/temperature-alarm-engine-wathcdog-tm2.html
Since the thermostat does not start to open till 80 Deg and will not be fully open till 105 Deg the temperature detected is not represenative of the whole engine temperature as the full flow of water will not be circulating and hot spots will be present.
ReplyDeleteWith a pressurised system, the normal running temp will be above 105 Deg at which point your graphs have converged enough to be insignificant given the accuracy of a bi-metallic gauge
Excellent point.
Delete