Chromatography Forum

I'm currently runnning the original Empower software with an old Waters HPLC system. I've gone through the tutorials and the troubleshooting guides and everything else I can think of, but I can't seem to convince the software to run a full sample set. It runs individual injections just fine, and I've tried everything from adding a shutdown method to removing the "0.0" flow line from my method. I've set it up manually and using the wizard options. No dice. As far as I can tell, I'm giving the correct commands; the software just isn't conveying the "switch to the next injection" command to the pump or autosampler. Does anyone know a fix for this that does not include updating to a new version of Empower?
Thanks!

Is the inj. signal cable running from your autosampler to your detector?

Please provide some more specifics (Waters LC somponent models, standalone, work group or enterprise Empower installation, build number & latest feature release or service pack installed... stuff like that).

We've got pretty much the oldest set-up you can get!

Waters 490E multiwavelength detector
Waters 510 Pumps
Waters 717 Autosampler

This set-up is apparently not compatible with a Windows 2000 computer setup, so the Waters tech installed a Bus/LC card in our computer and we have an exterior module box (Waters bus SAT/IN module) to connect the pumps and such to the computer. We're running the original version of Empower software.

Thanks for the help!

Correct me if I'm wrong...

You're programming all injections on the system controller and the SAT/IN is just collecting the data. Is it possible that your run time on the SAT/IN is a little too long so that by the time it is done sending the data for an injection, the next injection has already happened and has been missed? See that your SAT/IN run time is about 30 seconds shorter than the run time programmed on the WISP (or system controller if that's what you have) to prevent that from happening.

Since you consistently get the first injection, I'm assuming that your start cable is OK.

I'm not entirely sure what you mean by "run time on the SAT/IN" is too long? The only place I program a run time is when I set up the instrument method and the "run time" for each injection via the sample set wizard. Should the run time on the instrument method be different from that of each sample in the set? I don't program the autosampler at all. Can you define these terms a little bit more?

Thanks!

See that your data collection time is a bit shorter than the amount of time that passes between injections. You should be able to do this by adding a "Next Injection Delay" of 0.5 min or 1 min. If you do not see this column in your sample set method editor, right click on a header and bring up table properties. Uncheck "Next injection delay" and this column will appear.

The idea is to give the SAT/IN adequate time between injections to transmit its data and be prepared for the next injection.

I'm a bit confused on how you have instrument control and are using a SAT/IN...

Okay, that makes more sense. I'll give that a try, thanks!

The autosampler is run via an IEEE-488 connection that connects both to the computer and the pump module. The detector is funneled through the SAT/IN module box to the computer. According to the Waters tech that set it up, the detector we're using (490E) isn't compatible with our Windows system unless we funnel it through the SAT/IN relay. I won't pretend to actually understand what that means, but maybe that's why I have instrument control I ought not have?

Right. Your detector is (how do I phrase this delicately?).... OLD.
Since it was not outfitted with anything approaching a modern computer interface (or even IEEE), you are left with the alternative of feeding its analog signal to an external A/D converter (the SAT/IN) that does have a computer interface compatible with your Empower installation.

At the end of each injection, the SAT/IN must ship its data tot he computer and "reset" itself for the next injection. This takes a bit of time, depending on the size of the data file and the quality of the hardware on the receiving end. If the autosampler has made the next injection before the data transmission cycle is complete, you get no data from it.

Okay, definitely starting to make good sense now. Just a couple more clarification questions:

1. Is there a way to tell how long the delay is? I know when I run single injections, there's about a 15s delay between what shows up on the detector and what shows up on the computer screen. Should I assume that is the actual delay or go conservative and give it a minute or two at the end of each injection? Just to reiterate, I allow for this delay by implementing a "next injection delay", correct?

2. I want to be certain that my instrument method is communicating to the autosampler correctly. Currently, my method looks like this (I have two buffers and am running a 0-100% gradient of buffer B over 30 minutes):
0minutes, 1ml/min, 100%A, 0%B, "curve" box blank
30minutes, 1ml/min, 0%A, 100%B, Curve 6
35minutes, 1ml/min, 0%A, 100%B, Curve 6
45minutes, 1ml/min, 100%A, 0%B, Curve 6
50minutes, 0.01ml/min, 100%A, 0%B, Curve 11
Does this method adequately indicate a "stop" signal that tells the autosampler to move to the next injection? If not, what am I missing?

3. In addition to the SAT/IN module, we have a pump module, which I presume does the same for the pumps that the SAT/IN does for the detector. There is a cable connected to the "relay" slot on the terminal strip on the back of the box. Currently, the other end of the cable is not connected to anything. Should it be and, if so, where?

Thanks for your help!

1. Is there a way to tell how long the delay is? I know when I run single injections, there's about a 15s delay between what shows up on the detector and what shows up on the computer screen. Should I assume that is the actual delay or go conservative and give it a minute or two at the end of each injection? Just to reiterate, I allow for this delay by implementing a "next injection delay", correct?
I'm looking for a minute or two here.

2. I want to be certain that my instrument method is communicating to the autosampler correctly. Currently, my method looks like this (I have two buffers and am running a 0-100% gradient of buffer B over 30 minutes):
0minutes, 1ml/min, 100%A, 0%B, "curve" box blank
30minutes, 1ml/min, 0%A, 100%B, Curve 6
35minutes, 1ml/min, 0%A, 100%B, Curve 6
45minutes, 1ml/min, 100%A, 0%B, Curve 6
50minutes, 0.01ml/min, 100%A, 0%B, Curve 11
Does this method adequately indicate a "stop" signal that tells the autosampler to move to the next injection? If not, what am I missing?
This method is setup for gradient, re-equilibrate, go to low flow rate if no more injections are made. If your 45 minute line is to reset initial conditions, I'd change that time to 36 minutes, or even 35.5 minutes. That will get it back to starting conditions faster. So after making that change, what I'd do is tell Empower to collect data for about 40 minutes and add a next injection delay of 2 minutes. Your total run time would then be 42 minutes between injections and you would have 40 minutes of data for each injection. If this does not help, then we need to look at the cables you mention below.
3. In addition to the SAT/IN module, we have a pump module, which I presume does the same for the pumps that the SAT/IN does for the detector. There is a cable connected to the "relay" slot on the terminal strip on the back of the box. Currently, the other end of the cable is not connected to anything. Should it be and, if so, where?
There should be something going to the SAT/IN to tell it when an injection has been made. If there is a 2 lead wire coming from either the autosampler or controller from a spot that says "Inject/Start" (or something similar), it should run to the SAT/IN. This is what triggers the SAT/IN to start collecting data for the run time specified in your sample set method.

I'll alter my method and see if that corrects the issue.
In regards to the cables, there is a 2 lead wire running from the autosampler "Inject/Start" to the SAT/IN, so we're good there. Should there also be a cable running for Inject/Stop to the SAT/IN?

Thanks!

No need for a "stop" wire (assuming SAT/INs work the same way as eSAT/INs do except for the data connector).

Okay, if there's no need for a stop wire, then I would say our cables are set up correctly, based on what you indicate they should look like.

I adjusted my method somewhat like you suggested. As I've been collecting data through all this, I didn't want to change the gradient too much. So, ultimately this is what I programmed:

30mins, 1ml/min, 0%A, 100%B, Curve 6
35mins, 1ml/min, 0%A, 100%B, Curve 6
45mins, 1ml/min, 100%A, 0%B, Curve 6
60mins, 0.01ml/min, 100%A, 0%B, Curve 11

Then, when I set up my sample set, I set the "Run time" for each sample to 50 minutes and the "Injection Delay" to 2 minutes.

Based on your earlier post, I think this is what you said to do (?). However, this did not help, and I still needed to manually switch to the next injection. Is there a chance there's something wrong with the autosampler itself, not recognizing the signal to "move to the next injection"?

Thanks!

Could be...

If you have any sort of service contract (Empower software or Waters hardware), I'd either get on the phone or call a service rep respectively - especially if the thing was working properly after the SAT/IN was hooked up.

Eureka!
Apparently that cable connected to the "relay" spot on the pump module also needs to be hooked up at the "inject start" site on the autosampler. We randomly hooked it up just to see, and voila.

Thanks so much for your help and advice!
Cheers!