308 water temperature gauge gone ?

Dumb question: does the gauge have a calibration feature?

Ken

 

(With the ignition key "on") Touch the wire going to the sender directly to ground (like the nearby bare engine) -- if the gauge doesn't go to the max reading, that's a bad sign for the gauge itself.

 

More than likely it is the gauge, but it is possible that there is a connector in between the gauge and the sender that has high resistance. High reistance means a lower reading. Try grounding directly to the back of the gauge just to be sure before ordering a new one.

 

If the gauge is truly off calibration, I would think you could send it to a specialist to be recalibrated, rather than replace with a new one. There are shops that specialize in this work, they usually have an emphasis in speedometers, but work on all gauges.

Just let them know of the difference in reading to give them a starting point.

 

I replaced my sender but kept my old gauge in an "87 328 after chasing down what might be causing the temp to read too high. Before replacing the gauge I pulled the dash out and cleaned the connectors that attach to the temp gauge. After a thorough cleaning all was well. Try cleaning the connections on the back of the gauge first. It might save you some headaches as well as money.

Dan

 

Hi all,
my car, a 308gtb QV (1984), exhibits exactly the same symptoms described by Martin308GTB and all the test results are identical, including those with the gauge on the bench in my lab.

I understand that this is a quite old thread but, if Martin is here, I would like to ask him a couple of question:
- in your last post you wrote that you were going to replace the gauge: did this solve the problem?
- on the same post you included a table of sender resistance vs temperature: what version of the 308 does it refer to? Do you still remember if the thermistor part number was #104628?

The last question is very important to me because I am using the table values to test the gauge, i.e. if I simulate the sender with a 215 Ω resistance I should read 90 °C on the dial while, on my gauge, I have to use a 150 Ω to read 90 °C.
This means that, if the table is the right one for my sender, the gauge is clearly faulty but if not the problem could be the sender.

Ciao so far and many thanks for any help.

Franco

 

What about a gauge capacitance module? It does take an input ohm value and generate a calibrated output value.

I am using these in order to match generic fuel tank sender to my dashboard gauge. The ohm range of these modules does seem wide enough for temperature and pressure sensors. https://spiyda.com/fuel-gauge-wizard-mk3.html .

Maybe you could drop them an email?

 

https://www.ferrarichat.com/forum/threads/veglia-water-temp-gauge-aftermarket-sender-unit.583781/

my trials and tribulations!

 

@raemin

thank you for the link: it looks interesting and it is worth to be better evaluated.
Only one concern: where to install it in a 308? Behind the dash there is not enough space and near the sender it will get roasted because the engine heat…

@TonyL
The link you provided is really fantastic! Did you inspect the gauge “anatomy” by opening it? In this case what is the best way to proceed? I can’t figure any way other than to cut the chromed bezel.

The “anatomy” that you described goes in accordance to what I thought just measuring the resistances among the gauge terminals.
If we name Gn = car ground, Th = sender terminal and Vc = +12V, the results of my gauge are:
Th to Gn = 170 Ω
Th to Vc = 110 Ω
Vc to Gn = 278 Ω
So I assumed that there is a coil from Vc to Th and another one from Th to Gn: 110 + 170 = about 278 Ω
Just to be double sure I measured again the impedance among the gauge terminals but using a Vector Impedance Analyzer: the value of resistance and inductance are in agreement with my “theory” and only the sum of the inductance Vc to Th + Th to Vc is smaller than the expected series value but I think that this is due the to mutual influence of the coils positioned at 90 degrees.

Following this reasoning I thought to compensate the wrong readings (smaller than expected) by connecting a resistance from Th to Gn, at least to have right values around 90 °C (+/- 20 °C).

If, and I repeat if the Martin table is correct for my sender, at 90 °C the sender resistance should be 215 Ω and I get the right reading with a compensation resistor of 440 Ω.

This seems to work, but on the bench: on the car… who knows...

Ciao and sorry for my English but in this language I only have 8 bits resolution

Franco

Impedance Vc - Th
Image Unavailable, Please Login

Impedance Th - Gn
Image Unavailable, Please Login


Impedance Vc - Gn
Image Unavailable, Please Login

 

Franco, you certainly are going into this in a very professional way. I wasnt brave enough to strip the gauge down as i wasnt convinced the gauge was at fault. From the data you have supplied it looks like the culprit is the sender, it was on mine. Getting it to read correctly with a aftermarket sender was frustrating.

To get the chrome bezel off is a nightmare and i didnt do that since I found this diagram on the net, it seemed to explain how the gauge worked and what the terminals did.

Installing a fixed resistor to get the temperature to read correctly at 195 was my goal, the problem is that the values are not linear so it varies up the scale. I just got the gauge to read correctly at 195.

Would be interested in your findings.

Veglia changed the temp gauges & senders quite a lot through various cars so some may be slightly different but the concept is the same

Terminals on my car were marked BN, 12v and GN.

BN was on the left, 12v was on the right and GN was ground.

Ciao
Tony
PS-Your English is just fine

Image Unavailable, Please Login

 

Hi Tony,
first of all many thanks for the gift you gave me: the gauge electric diagram that I searched in vain on the net for many hours!

Now what to do next?
First, on the gauge side, I want to better measure, on the bench, the reading non-linearity introduced by the compensation resistor. I already know, that once calibrated for 90 °C, it is acceptable in a small range around this value but I want to record the error in a range of temperatures from 60 to 120 °C: from a very preliminary look I found that the 60 °C reading is a lot to high.
But all of this “experiments” could be completely useless because they will depend from the Martin table and of course if the problem is in the sender: at least I'll have fun...

Then, on the sender side, I plan to have at least one fixed point by reading, with an ohmmeter, the sender resistance during the engine warm up: according to the user manual the radiator fans should start at 84 °C and, if I remember correctly, that is what happened before.
I am also considering to buy a cheap IR thermometer and to read temperature of the block near the sender vs resistance during the warm up: a friend of mine got from Amazon for 32 € one of these chinese “toys” and he says, that once you have found the right emissivity calibration it is rather precise, obviously if you don't expect the performances of a professional one costing more than 300 €.

Finally, if the most likely culprit is the sender, I will replace it but if it is the gauge I will try to open it because it could just be a cold welding.
I'll let you know my progresses.

Ciao, have a nice Sunday and following week.

Franco

 

The biggest problem is nobody knows the manufacturers original resistance readings for the gauge and any measurements taken on a 40 year old sender unit could be way out.

I did map the readings from the Veglia gauge via a small potentiometer but is was rather crude and relied on digital thermomenters, multimeters, IR gun etc.

The values change so quickly it was difficult to get repetitive accurate results unless you can control the temperature by heating the sender very slowly.

It takes a brave man to break into the gauge to see whats going on, i would leave that as a last resort. There is a company over here that repairs gauges (speedy Cables) and are excellent but rather slow.

In the end i settled for getting the gauge to read accurately at 195 F as the temperature below that wasnt of concern, so long as any increase beyond normal operating temperatures was registering properly.

All of the sender units on sale and even those sold as "Veglia compatible" were way off the mark. The Valeo unit was the most accurate but i had to fit an adaptor to the engine (dino 246) block to convert it.

Have fun !

Tony

 

Hi Tony,
I have just ordered the IR gun, it “should” arrive tomorrow and then I’ll carry out the measurements of the thermistor resistance during the engine warm up...
This test should give me an understanding of the sender health.

If it turns out that the gauge is the culprit I have two chances: yust compensate with a resistor to have the right reading at 90 °C (194°F) or to buy a new one. It should be available at Maranello Classic Parts, in Egham, a lovely place that I visited several times in the past and where I even have a good friend. It is priced 245.95 £ + VAT but now, after Brexit, I don’t know how much I’ll pay for the shipping only…
For sure, if I go with the second option, I will open the old gauge trying to understand what was the problem !

Ciao, good evening at to the next.

Franco

P.S.
From your profile I saw that you are located in UK.
We, me and my zoo (wife and daughter) lived there for three years, in the eighty, while I was working for Unilever Research and we used to return there almost once a year but not the last and you can easily guess why :-(

 

Franco
I very much doubt Maranello would have that part in stock unfortunately, if it is defective then i can recommend you try Speedy Cables, they can repair most gauges and may even be able to match it with a new sender. My speedo drive broke some years ago and they machined up new parts to restore the old one and only cost £80GBP, a new one was over £390 !!

https://speedycables.com/

Keep us posted on your findings with pictures, these posts are invaluable to others in the future who are brave enough. I do remember a post sometime ago where a Fchat member did this successfully. As you say the chrome ring is the difficult part. Speedy do the chrome rings i beleive!

 

https://www.ferrarichat.com/forum/threads/instrument-gauge-dis-assembly.370853/#post-141502838

is this what you were referring to?

Ciao.

 

Looks like one of the old posts i recall seeing, there are others from a quick search.
tony

 

Last week I couldn’t carry on the programmed sender tests because I was too much involved babysitting my two granddaughters …

Yesterday the first test gave me temperature vs resistance values too high, both during the engine warm up and cooling down, so I suspected that it was due to a bad choice of the ground point.

This was the test setup during this trial:

Image Unavailable, Please Login


Today I repeated the test using a different ground point and…
the results during the warm up were so bizarre as to be completely useless (jumping measures and in some cases with extremely low values) such as to make me suspect that there was a strong influence from the nearby high voltage cables and other eddy currents.
For the measures I used an IR chinese toy (that seems accurate within +/- 3°C) and a reliable Tektronix TX1multimeter.

During the cooling down the results were definitively more stable although all lower than expected, at least compared with the Martin table above and the one that I found at:
https://www.ferrarichat.com/forum/threads/water-temperature-sensor-curve-chart-for-308-gt4.409456/
and, in range 75 – 83 °C, the curve seems to flatten too much.

Image Unavailable, Please Login

Using the found resistor values I have simulated, with a trimmer, the sender on the gauge still on the bench: with 446 Ω I read the correct 60 °C, with 250 Ω the gauge shows a little less than 75 °C but with 215 Ω the reading is only a little bit over the 75 °C mark; the sender resistances, to make my gauge happy, should have been 228 Ω for 75 °C and 169 Ω for 83 °C.

This is the reading with the trimmer set at 215 Ω (I should read 83 °C on the gauge):

Image Unavailable, Please Login


So the question remains: which is the culprit, the gauge or the sender?
Now the next step is to remove the sender from the engine and to test it with hot water in a pan; removing the sender doesn’t seems to be a painless operation (get rid of the oil filter ?) and, BTW, is it possible to do it without emptying the coolant circuit ?

Franco

 

FWIW: I helped a friend with this issue on his 308. Connected all sorts of electronic temp measuring devices adjacent to sender on manifold, checked with IR temp gage etc etc. After much time spent, problem was with gage.

I just had a thought: ~~~~~
Gage responds / sender calibration 'should' be fairly robust ~~~ Why does gage suddenly start reading too high?

What if return spring holder for galvanometer movement (gage itself) is adjustable and it slipped its position allowing needle swing to be exaggerated?? Not certain of gage construct, but this is a typical component on D'arsenval movements. On "spring for control action", there's typically a little brass spring holder that adjusts the springs fixed position. Could holder have slipped??

image here:
https://www.rfcafe.com/references/electrical/NEETS-Modules/NEETS-Module-03-1-11-1-20.htm

If I run into this again, I'll dive deeper. We just replaced gage this time.

 

I am with you if a gauge starts reading too high ...
but mine is reading too low, mainly in the “high” temperature range 75 – 83 °C making me thinking that the sender, when the temperature increase, doesn't decrease its resistance as it should but rather flattens out to a value greater than the right one.

Now I'll try, as I said, to test the thermistor alone and, at least, I'll have some fun …

Ciao and thanks.

Franco