leakage in the transducer region

[hplctroubleshooting]

I have an Waters autosampler 717 plus and for some reason the inlet and outlet of the transducer has been leaking slightly, presumably due to higher then usual pressures in that region.

time line

-was getting low areas for my analyte peaks
-checked compressibility test, which passed
-checked injection accuracy test, which did not pass
-called Waters technician who replaced the injection valve, he claims the accuracy test passed, but I checked it afterwards and got 47 uL for 50 uL injections.
-opened up the autosampler and found leakages coming from the inlet to the injection valve, so tightened the screw which ceased the leaking.
-a short while after, found leakages from both the inlet and outlet of the transducer.
-peak areas are getting progressively lower, it's gone to about 1/10 of what it should be. Tightened the bolts on the transducer, still a little bit of water is coming out of the inlet and outlets so pressure in that region must be quite high and increasing.


I unscrewed the inlets and outlets starting from the transducer and found that flow was good from the transducer to the sample loop outlet which connects to the needle. My only guess is that the needle is not properly aligned into the stream and that the port may be sealed up perpetually, which is causing pressure to build in the sample loop/transducer region after successive drawing of sample into the loop.

How often does one need to do a seal pack adjustment? It hasn't been long since I did it but then again one of the sensor bars on the needle platform has its plastic broken, exposing the wires inside of it. I asked my Waters technician about this and he said that it shouldn't be a big deal since the wires do the sensing and the plastic is just the casing. So mmy question to you all is - is it essential that I get the sensor bar fixed? I will get it fixed eventually however.

What measures should I take here? Btw, had all three valves replaced just recently, with the most recent one being the injection valve.

Perhaps the broken sensor bar is causing the seal adjustment to go out of whack every once in a while, which is increasing pressure in the sample loop lines, which in turn is damaging the valves?

[Bull76]

Hi,

Any leakage is an issue which needs to be resolved. Sometimes the connections to the transducer can be a little flawed especially if they leak due to the high pressures and people lossing the connections. Replace the fitting's to correct the leak's. The issue with the opto switches for the Injector, well if they are working then not an issue.
The adjust seal only needs to be done once a year or when the seal pack or needle gets replaced. You shouldn't get much difference between the adjust seals values over the course of a year. The needle does need to be in a particular orientation, usually if you have a service contract the Waters FSE will align this.
When the leaks are fixed and the adjust seals completed, compression test passed and then the injection accuracy test passed. Then the system should be good.

[hplctroubleshooting]

Hi,

Any leakage is an issue which needs to be resolved. Sometimes the connections to the transducer can be a little flawed especially if they leak due to the high pressures and people lossing the connections. Replace the fitting's to correct the leak's. The issue with the opto switches for the Injector, well if they are working then not an issue.
The adjust seal only needs to be done once a year or when the seal pack or needle gets replaced. You shouldn't get much difference between the adjust seals values over the course of a year. The needle does need to be in a particular orientation, usually if you have a service contract the Waters FSE will align this.
When the leaks are fixed and the adjust seals completed, compression test passed and then the injection accuracy test passed. Then the system should be good.

yeah, I tightened the bolts on the transducer and that seemed to stop the leakage. But the compression test passed even before that and the injection accuracy test wasn't too far off - 47 uL for 50 uL. So, you don't think that the broken plastic on that sensor bar-basically the sensor bar cracked to reveal the wirings underneath it- would cause any problems in the long run?

[hplctroubleshooting]

Compression test, seal pack test, and injection accuracy test, all passed. But low areas still persist.

I noticed something strange today,

1)unscrewed valve 2 (syringe valve) tubing to the syringe and then opened the valve via the valves motion page while flow was on, nothing was coming out.

2)pulled off the waste tubing on valve 3 (waste valve) and then immediately noticed flow coming out of this outlet on valve 2 that goes to the syringe valve.

Areas increased substantially, from 2 million units to over 3 million units, but not quite the 4 million units that is the normal value we're accustomed to. Could anyone offer up some suggestions as to what might be going on here?

[hplctroubleshooting]

update:

today we took off the metal tubings within the autosampler, one by one, and then filled the syringe , it turns out that the filter seems to cause decreases in the volume withdrawn by the syringe. We observed the amount withdrawn from a vial of methanol by immersing 1)the outlet tubing from the transducer into the vial and then 2)the outlet of the sample loop, which has its inlet connected to the in line filter. There was a significant qualitative difference in the amount drawn between the two. We also disconnected some earlier metal tubing and got the expected volume drawn which was consistent with a hand withdrawl of the syringe - as compared to the amount drawn after the "filled" mode in the valves motions page.

So it seems that the in line filter is an issue, but currently am confused since we had observed the injection accuracy test pass (48 uL for 50 uL injections) a day ago.