diy solar

diy solar

SMA KTY Battery Temperature Sensor replacement options

Trukinbear

New Member
Joined
Aug 28, 2020
Messages
259
Hi, last night my SI6048 shut down abruptly with W212 (high battery temperature warning) F206 (battery over-temperature) warnings - restarted it and it did it again (x3). Finally, after using an IR thermometer to make sure my battery wasn't really hot (72.4 degrees F) I disconnected the battery temperature sensor. Now I get W723 (cable break on battery temperature sensor) errors so fast I can't clear them (looks like 6 times a minute perhaps) but at least we're still making power - with thousands of W723 errors.

I had the W723 (cable break on battery temperature sensor) and W722 (short circuit on battery temperature sensor) error last year (that somehow managed to correct itself while I was waiting on a response from SMA) and had requested a new Battery Temp Sensor unit (was willing to pay for it) but they wanted to do the drag-ass warranty thing with me sending in operating files and all that jazz. It started working ok, so I dropped the ball.

I'm trying to find another temp sensor that will work - apparently SMA doesn't sell them in this country, and when I contacted the UK they didn't want to ship one to the US. SMA called it a KTY type sensor.

Ideas?
 
My Sunny Islands don't even have the battery temperature probe hooked up.. never have.

I think the BMS sends all that info to the SI's.

You should be able to find out what kind of thermocouple they use (probably RTD), and just get one on Ebay.
 
Hi, last night my SI6048 shut down abruptly with W212 (high battery temperature warning) F206 (battery over-temperature) warnings - restarted it and it did it again (x3). Finally, after using an IR thermometer to make sure my battery wasn't really hot (72.4 degrees F) I disconnected the battery temperature sensor. Now I get W723 (cable break on battery temperature sensor) errors so fast I can't clear them (looks like 6 times a minute perhaps) but at least we're still making power - with thousands of W723 errors.

I had the W723 (cable break on battery temperature sensor) and W722 (short circuit on battery temperature sensor) error last year (that somehow managed to correct itself while I was waiting on a response from SMA) and had requested a new Battery Temp Sensor unit (was willing to pay for it) but they wanted to do the drag-ass warranty thing with me sending in operating files and all that jazz. It started working ok, so I dropped the ball.

I'm trying to find another temp sensor that will work - apparently SMA doesn't sell them in this country, and when I contacted the UK they didn't want to ship one to the US. SMA called it a KTY type sensor.

Ideas?
Hello,

Did you happen to note the temp reading on the SI while you were dealing with this?

Similar deal at a off grid site that's been working fine for half a year. We've just started getting the F206 but the flooded lead acid batteries are sitting at close to 60F. The onsite person did not mention any other errors. The SI6048 stacked pair randomly shuts down then self clears in anywhere from 10 minutes to 10 hours and restarts The Sunny Island is saying 60 degrees C (~140f) which of course is incorrect. I had the person onsite pull the temp sensor and to my horror the SI now thinks the battery temp is 70C. Had him do a full 10 minute shut down including throwing the battery main to ensure the SI's weren't getting anything but it still read 70C.

Thank you for posting your problems as it's the only mention I can find about the issue I'm having.

Best,
T
 
My Sunny Islands don't even have the battery temperature probe hooked up.. never have.

I think the BMS sends all that info to the SI's.

You should be able to find out what kind of thermocouple they use (probably RTD), and just get one on Ebay.
For LiPo batteries you are correct but for lead acid batteries you would still likely want to have a functioning (BTS) battery temp sensor, particularly if the bulk of your charging was via the SI's (via generator or AC coupling). I do think that if the bulk of your charging is solar with non SMA charge controllers that have their own battery temp sensors then a BTS on the SI's would be less important.
 
Resurrecting an old thread here with some hopefully useful info.

The "solar trailers" that are all over now tend to use Sunny Islands. I got tasked with helping get some going. One big problem we were having was shutting down due to overtemp. Since we had more than one trailer, I swapped out the battery temp sensor for another one. This immediately fixed the problem.

The resistance of the bad sensor was around 50K ohms. The working one around 2K.

Based on the KTA designation mentioned here, I started looking around...

There are a bunch of KTA-XXX sensors. A few that are 2K at 25 deg. C. I homed in on the KTA81/210 as a likely candidate.

I used the wire wheel on my bench grinder to clean the corrosion off the copper block the SMA sensor is embedded in. On one side, the copper had corroded away enough to show what looked to me like a small area of a TO-92 semiconductor device.

Continuing to carefully disassemble the copper, I learned two things:

The corrosion had made it all the way to the inside of the copper block and had eaten through one of the leads. Hence the 50K resistance reading.

The other thing was that the actual sensor was a TO-92 device and the labeling was still readable. It is indeed a KTA81/210. See attached image.


Cost $2.00 USD. Those folks who are charging $100 for a new "SMA sensor" are really screwing us.

I'm gonna order a few and build up my own sensors. Not gonna use any metal block - that's just asking for trouble. Since these are monitoring huge battery banks where the temperature changes are very slow, metal (to decrease the response time) isn't needed. My usual method of waterproofing/acid proofing temp sensors is to seal them inside hot glue lined heat shrink with additional hot glue added at the ends that melts further when the entire assembly is shrunk with the heat gun. Have never had a failure on these.

FWIW, the 81/210 I recovered is still functional, as it changes value with temperature. However, it reads half of what it should (1K). I suspect the corrosion made it to the "guts" of the TO-92 and has affected the chip inside.

Something for all you SMA users to consider - the KTY series sensors are being EOL'd by all the manufacturers:

"End of Life: Scheduled for obsolescence and will be discontinued by the manufacturer."


"Last Time Buy"


As cheap as they are, you could buy a bunch "just in case". Mouser and DigiKey don't have any right now but others do.
 

Attachments

  • kty81-210.png
    kty81-210.png
    548.8 KB · Views: 4
Update:

There are (as can be seen in the datasheet) actually 3 leads on the 81/210. The third one was corroded off flush with the case. I used an X-Acto to scrape away enough of the TO-92 epoxy so I could get a pin-style probe stabbed into it and checked again to the #1 lead (reading from #1 to #3 instead of #1 to #2). Saw 2K. So it is actually still functional. Gonna be a challenge to get a small piece of magnet wire soldered to it, though. I might just wait for the replacements to show up:

 
Resurrecting an old thread here with some hopefully useful info.

The "solar trailers" that are all over now tend to use Sunny Islands. I got tasked with helping get some going. One big problem we were having was shutting down due to overtemp. Since we had more than one trailer, I swapped out the battery temp sensor for another one. This immediately fixed the problem.

The resistance of the bad sensor was around 50K ohms. The working one around 2K.

Based on the KTA designation mentioned here, I started looking around...

There are a bunch of KTA-XXX sensors. A few that are 2K at 25 deg. C. I homed in on the KTA81/210 as a likely candidate.

I used the wire wheel on my bench grinder to clean the corrosion off the copper block the SMA sensor is embedded in. On one side, the copper had corroded away enough to show what looked to me like a small area of a TO-92 semiconductor device.

Continuing to carefully disassemble the copper, I learned two things:

The corrosion had made it all the way to the inside of the copper block and had eaten through one of the leads. Hence the 50K resistance reading.

The other thing was that the actual sensor was a TO-92 device and the labeling was still readable. It is indeed a KTA81/210. See attached image.


Cost $2.00 USD. Those folks who are charging $100 for a new "SMA sensor" are really screwing us.

I'm gonna order a few and build up my own sensors. Not gonna use any metal block - that's just asking for trouble. Since these are monitoring huge battery banks where the temperature changes are very slow, metal (to decrease the response time) isn't needed. My usual method of waterproofing/acid proofing temp sensors is to seal them inside hot glue lined heat shrink with additional hot glue added at the ends that melts further when the entire assembly is shrunk with the heat gun. Have never had a failure on these.

FWIW, the 81/210 I recovered is still functional, as it changes value with temperature. However, it reads half of what it should (1K). I suspect the corrosion made it to the "guts" of the TO-92 and has affected the chip inside.

Something for all you SMA users to consider - the KTY series sensors are being EOL'd by all the manufacturers:

"End of Life: Scheduled for obsolescence and will be discontinued by the manufacturer."


"Last Time Buy"


As cheap as they are, you could buy a bunch "just in case". Mouser and DigiKey don't have any right now but others do.
Thank you very much for this information and links. I had been searching for this information to replace the BTS.
 
The battery sensor on my "solar trailer" has been intermittent, and unfortunately, causes a complete shutdown of the Sunny Islands when it malfunctions.

SMA wants about $150 for a new assembly (including cable), if memory serves. I ordered and received several KTY sensors, but hadn't yet figured out how I want to package it.

My usual method of waterproofing/acid proofing temp sensors is to seal them inside hot glue lined heat shrink with additional hot glue added at the ends that melts further when the entire assembly is shrunk with the heat gun. Have never had a failure on these.

Appreciate you sharing this approach, I'll probably do it that way. I know you said the metal block isn't necessary, since the battery temp doesn't change that quick. I guess that's probably true, but I wonder if there's even a point to hooking it up, aside from getting temperature compensation from ambient. I think I remember reading that without a sensor connected, the SI will run as though it's circa 100º, compensating with lower charge voltage parameters, in an effort to preserve battery longevity. It'd be easier on the battery, and one less point of failure...
 
The battery sensor on my "solar trailer" has been intermittent, and unfortunately, causes a complete shutdown of the Sunny Islands when it malfunctions.

SMA wants about $150 for a new assembly (including cable), if memory serves. I ordered and received several KTY sensors, but hadn't yet figured out how I want to package it.



Appreciate you sharing this approach, I'll probably do it that way. I know you said the metal block isn't necessary, since the battery temp doesn't change that quick. I guess that's probably true, but I wonder if there's even a point to hooking it up, aside from getting temperature compensation from ambient. I think I remember reading that without a sensor connected, the SI will run as though it's circa 100º, compensating with lower charge voltage parameters, in an effort to preserve battery longevity. It'd be easier on the battery, and one less point of failure...

Ambient isn't what you're interested in. You want the temperature of the battery(ies), which can be significantly different from ambient due to their thermal mass.
 
Back
Top