GE Kaye

[OlDave]

Do any of you pharmaceutical types have any experience supporting the Kaye LabWatch system? We are in the process of getting a 250 point system up and validated. Unfortunately we have discovered significant drifting of many of our thermal measurement points. Almost all temperature points being monitored are using type T thermocouples and we have seen some channels drift more than 2°C in 3 months. I just recently charted a 0.7°C drift in 4 days on one special test point I set up. System specifications are quite vague but the company claims the system is stable to around 0.25°C per year.

As you can imagine the "blame game" has already begun and is gaining momentum. Do any of you have this system installed at your facilities? Any ideas what's causing the drifting we are seeing? Any and all suggestions or comments would be MORE than appreciated.

[MIRCS]

we don't use it at our plant. We are playing the blame game right now with thier ValProbes as some every few runs no longer read.

Oh yeah also we don't use thermocouples on anything.......all 100 Ohm 385 curve PRT's

[OlDave]

I've been so PO'd at the arrogance of GE Infrastructure Sensing for the last few years that I'm heavily leaning towards going with the Ellab TrackSense Pro loggers instead of the ValProbes. Tell me more about how they are responding to the problems you are seeing with the ValProbes.

Also, what's the maximum wire run you are able to go with for the PRTs, and you are using 4 wire leads I assume.

[MIRCS]

All three wire. JMS makes a hell of a probe.

The longest we have run I would say is 150' Though most runs are from 25-75'

our valProbe problem is with moisture getting inside and just complete failure of them. Start a project with 54 and only get 45-50 to read for a cal verification. They have come up with all kinds of answers which were all just Bovine defecation.

[OlDave]

For the most part we aren't monitoring anything what I call precision work. Just fridges, freezers and an ultra-freezer here and there. I would say our typical wire run is 250 feet though there are some that have upwards of 450 feet of wire to the probe. But wire length doesn't appear to be a factor. I get the same readings when I inject a signal at the end of a 270 foot chunk as I do when I shoot it into the 10 foot test drop on the same channel. And the TCs aren't drifting. I've gone back and tested a sampling of them. That only leaves the POS netpacks, those of course Kaye claims are as stable as a rock.

A 10 to 20% failure rate seems pretty high to me. Are they completely dead or do they come back to life after they dry out? What temp/humidity/pressure conditions are you testing? I could understand it a little if you're testing an autoclave cycle maybe, but anything else doesn't make much sense to me.

[MIRCS]

We have em in our Sterilizer's. So yeah........about 123C for 30-80 minutes, rotating, hitting water.........yeah I can expect failures. The problem is that they completely die and have to get new boards put into them at Kaye. They first told us that it was because we didn't replace the O-rings when battery replacement happened.......................well that has been remedied and we still have th problem.

Kaye has blamed the problem on my temp baths...................................if they think that oil with a cs of 510 is actually getting into the probes.............well they are idiots.

We went to the ValProbes from Data Tracers......due to the fact we could change our own batteries and they were more accurate. I will say that the probes that work are rock solid.......it just sucks when you have made 45 runs on a 90 minute cycle with the sterilizers and I give back the ValProbes to validation and tell em that x number failed...........project failed, make all new runs................that is why we use 54 probes per run when only about 20 are needed.

[OlDave]

We're still trying to make do with an old Kaye KL validator. I'm getting really, really tired of stringing all those thermocouple in everything. That's why I'm pushing for a wireless system.

Have you ever done a site audit of Kaye? That's another area we always seem to have no funding for. The last 3 times I've sent our IRTD probes back for cal I've checked them against our fixed point cells upon return and they were all either just out or so close they should have been. I would love to go check their traceability matrix and processes. But then again, maybe it's better not to know....

[flew-da-coup]

I hope you are not using a Daytron 1281 to check your RTD with. Never use a Daytron 1281 to check RTDs and PRTs due to the current used by the 1281 for 4-wire resistance.

Are you sure you are checking the RTDs with 1mA across the RTD? It really makes a difference. Some meters. like the Daytron 1281, does not use 1mA when measuring 4-wire and when you are calibrating or reference checking an RTD you need 1mA unless it is specified by the MFR to use another known current. I use a Hart Scientific Super thermometer that way I can control my current.

I'm sure you guys already know this. I just remember having the same problem with not reproducing the same out come as the MFR and low and behold it was my current from the meter not being 1mA. That was 10 years ago and I will never forget the head banging I went through to figure that out.

[OlDave]

The Kaye IRTDs are the so-called "intelligent" probes that have the electronics in the head of the probe and output a digital signal in degrees temperature. I started verifying the probes when they came back from calibration a few years ago against my TP H2O, Ga cell and In cell. I figure I've got sub 1 mK uncertainty with the water and Gallium and about 2 mK on my Indium cell so I get really upset when I see a freshly calibrated probe off the entire limit of 0.025°C at any of those points.

I also have a Hart Superthermometer for my SPRT, nice unit, though I've had a little trouble with the video card in it. And of course all the parts are obsolete so repair is quite expensive. I had to take it apart and mount a heat sink on the video processor and a little fan to pull heat out of the computer "half" of the meter. Seems to be working much better now.

[flew-da-coup]

You must have an older SuperThermometer. I have the 1590 and they still sell them and service them.

I do see your point with the Intel Probes. I never cal'd them, but it does sound like the problem is with the MFR.

I would call their service center and talk to their tech who cals them. I bet you will find that it is in their method.

[OlDave]

I've got a 1590 also. But it's a '97 year model. The LCD screen is obsolete as you can imagine as is the video card. The new video cards aren't compatable with the LCD screen so the only way to really fix the problem is to replace both the LCD and the video card. But of course the current LCDs mount differently so the housing needs to be changed out....hense the excessive (in my opinion) cost.

And I've talked to the service center and it's like talking to the wall. They don't really want to know. It's like dealing with an 800 lb gorilla.

[MIRCS]

Dag on ya'll boys have some super sophisticated equipment.

I'm doing verification on 54 ValProbes in the am. I will let ya'll know how many fail. They made 30 runs with them.

[OlDave]

Don't you dump the data after every run?

Let me know the failure rate. We've got a Kaye rep working for us for the next couple of days so maybe I'll make him squirm a little.

[mjoeng1]

We have an internal customer at my company that's purchasing a Labwatch System. I'm recommending that they don't use thermocouple and opt for Kaye's RTD solution or purchase some compliant RTD from some other source. Thermocouple is wrought with problems if you're going to try and hold a 0.025 C tolerance. It really can only be characterized down at that level and you should expect drift regardless of what a mfg will try to sell you. Kaye makes a lot of money off of their t/c sales.
I wouldn't beat Kaye up too much though. They do make some good products. The X2000 Validator is an excellent product. Very reliable and holds the cal year over year. We've got 12 and haven't had an OOT yet over the last 5 years.
As for the ValProbes, I don't have any experience with them and I don't know which company makes them for Kaye. However, putting a wireless sensor in an autoclave is most likely going to cause failures. It's a pretty harsh environment for anything! If you want to go wireless you might want to check out a company called Mesa Labs. I think the product line is called Data Tracers. I would guess that all of the mfgs of this type of sensor is experiencing problems but Mesa Labs has an advantage over them. They have an additional dock to collect any analog signal (within a certain range) and can be synched with the data from all of the sensors being used. You also have the ability to scale the analog input within the software to be whatever you'd like. This would be a major advantage if you're collecting information from a non-conventional sensor such as a gas sensor, fan, or whatever. Good luck.

[OlDave]

No, I don't expect the TCs on the LabWatch system to maintain 0.025°C. That is the tolerance of the Kaye IRTDs that we have.

For the LabWatch system I'm expecting 0.25 to 0.5°C accuracy for a year. That's more than adequate for monitoring a fridge or freezer. Most of our units have an acceptable range of ±4°C from setpoint so a half a degree is fine in my book.

I do have to admit that our old Kaye X1360 ( I think I remember the number) is solid. Very little drift and is very reliable. It's just a hassle stringing all those probes when you are trying to use it.