Battery temp sensor testing

[booster]

We recently had a discussion in another thread about how accurate the battery post mounted temperature sensor might be. As usual, there were a lot of different theories put out , but that is always a good thing.

When we did our 80% discharge/recharge test a couple of days ago, I tried to get some idea of if the battery sensor reading seemed plausible, over the range of conditions.

By watching how long the temp stayed at max, once the van wasn't moving or running, I think I got a fairly good feel for how much of the temperature rise was because of the ambient air around the sensor and cable getting warmer from the exhaust, differential, and other underbody heat sources. Our batteries sit right behind the rear axle, but are somewhat heat baffled. The exhaust is also on one side of them, but baffled and insulated.

Bottom line is that we saw a max of about 30*F temperature gain at the battery temp sensors. When we stopped, we quickly lost about 7-9 degrees of that gain, and then the drop was very slow after that. It took almost 48 hours inside the shop, once we got home, for the batteries to get to ambient. My guess from this is that the first quick loss is getting rid of the ambient heat issues, and then the high thermal mass batteries slowly lose heat after that and would be a decent indication of actual battery internal temperature. That would leave us with about 22* of temp gain from putting in 350ah very quickly. It does point out that we really could benefit from temperature compensation on the alternator, also.

Since this is a really easy test to do during a recharge, I think it would be very interesting to see how other battery locations and charge rates react, temperature wise. I know many of the AGM installations are not in high ventilation or really cool areas, so they may have less of the ambient heat, but also be slower to cool, or they may not.

If those that have the capability to run a similar test feel like it, we might learn something, and help others with future installs if we see something that works better than other ways.

[phantomjock]

Now, what is my immediate interest in a somewhat esoteric topic? I have misplaced/lost one.

My research has yielded the following, and I thought I’d share.

The sensors work on the principle that as the temperature changes, the resistance changes. This difference in resistance is what is what the measurement device, display, regulator, etc. uses. With a well-defined resistance vs. temperature characteristics, it might be important to “get the right one.”

Platinum is the most common and accurate material used to make RTDs. They are often available with a 100 Ω and 1000 Ω resistance at 0°C. They are referred to as PT100 and PT1000 respectively. But which one is commonly used for the Alternator, and which for the Battery – or are they both the same? How can I tell which one I should have? Or, macht nichts?

A Platinum RTDs is near linear response to temperature changes, they are stable and accurate, they provide repeatable responses, and they have a wide temperature range. RTDs are often used in precision applications because of their accuracy and repeatability.



The configurations most often seen is a 2 wire. I understand these can be limited in length to prevent accuracy becoming an issue. Here is a handy table I dug up:



Any ideas which of these is commonly used? I bet the K is what is used with a multi-meter. The Balmar Battery Sensor is a MC-TS-B, and the cable is 20 feet long It is described as; “…Cable includes solid state sensors sealed in heavy duty lugs for easy installation.” Hmm, that suggests perhaps something is going on inside there. From internet images, many of these look the same, physically, but functionally? The Battery Sensors are silver and the Alternator appear copper color and have a maximum length of 80 inches. The Alternator Sensor is model MC-TS-A.
There is some info available that suggests the length has little relevance, other than convenience.
In fact;

“The alternator temperature sensor is a semiconductor device and is designed to bolt on to one of the case mounting bolts of the alternator. …”

“…The TS-A and TS-B sensors are identical except for the wire length one is 52" long and one is 20 feet. They are interchangeable. The TS-A is supplied with a washer and so if you get the TS-B you will have to provide your own washer.”

First, it is good to know they are interchangeable. But Is there a way to measure the sensitivity (and therefore accuracy), short of doing a full scale test?

An inquiring mind...

Cheers - Jim