Software crashing, return of the noise, broken injection

Hi All

We have a few persistent problems that i am hoping someone can give comments or suggestions.

1. robocyte software crashing (version 2.0) We have noticed that if we double click on a single well in the recroding screen plate graphic in order to look in detail at a recording, that, on occasion (1 in 5 or so), the software will freeze and we have to reboot the system (or kill the roboocyte with the task manger). On these occasions we also loose ALL the data for that run so needless to say we no longer check our data in detail during a run. Has any one else had this experience, it is not a great situation as it limits our ability to interact with the experiment. Are there any suggestions as to how XP should be configured, ie is XP SP2 ok with robocyte ver 2? Should anti virus/anti-spam be turned off etc?

2. We earlier reported a problem with electrical noise on the signal that occured only when the perfusion was on (we use the gilson system). With MCS's help we grounded the waste solution ad for a time seemed that this problem went away. However, it seems as though this problem has returned and has also changed a little. In addition to the baseline eletrical noise during perfusion we also some see regular variation in the amplitude so it looks like it is modulated by another source of noise. Unlike last time we havent been able to get rid of this by grounding of either the waste or the main perfusing solution. Any suggestions?

3. Broken injection needles. We have been buying the 20 packs of injection needles and have noticed that around 4-5 in every shipment have broken ends - (the back ends not the pointy ends). Whilst not a terminal problem it is a pain as it leaves bits of broken glass in the injection needle holder that must be cleaned out. Is this a general problem or is it due to shipment to Australia where they may get more rough handling? We are considering pulling our own - anyone have any experience with pulling their own and what glass would the recommend?

4. Software suggestion box. One frequent need is to repeat a certain protocol several times over in order to increase our confidence in the data. It is of course possible to code the script each time to count and index successful attempts but the relatively simple scripting language makes this task more complicated than we would like, especially if one needs to be running different protocols for different compounds sets and different screening needs. Would it be possible to have a higher level interface the plate level where you could instruct the system to run script X until i get Y successful completions and then load and run script A until i get B successful completions etc? This could be done by a graphical user interface at the time you select which oocytes to record and would remove the neccissity to code this into every script and deal with all the debugging and limitations of the scripting language?

thanks

steve

Dear Steve

1.) Two of our customers reported exactly the problems that you describe, but at both sites, the problem vanished suddenly, and unfortunately, we cannot reproduce the problems here at all. Very important: You may not use any firewall or antivirus software on the data acquisition computer at all. Those programs can interfere with the data acquisition and/or the Roboocyte communication in an unpredictable way. So I would ask you to deinstall those programs completely. (Deactivation alone does not always work as some programs load parts of themselves into the memory at computer startup.) Microsoft service packs are ok. If you have deinstalled all those programs and the problem does still persist, it would be nice if you could send us your database so that we can try to reproduce the problem again. If this is not possible, there is a chance that the next software version to be released in about two months will solve this problem, because this release will improve the program performance.

2. Could you please send us a data file or a screen shot that shows this problem?

3. This should not be the case, of course. Generally, all injection needles are fire polished at the ends, and so far, we had no complains about broken ends, so maybe it is due to the shipment. Could you perhaps make a digital photo of such a broken end or send us back some so that we can have a look at it?

4. Thank you for this suggestion. This would be of course quite useful, but it is not so easy to implement, because each user has different requirements for his protocols, and you can set up quite complex protocols with controls before and after the main test, and so on. We will keep this in mind, but in the moment, we are working on the implementation of the new amplifier and some nice stuff like fitted dose-response curves.

Best regards, Christine

Dear Christine

Thanks for your suggestions.

1. I will remove any anti-virus software from the PCs and see what happens.

2. Yes we can arrange to send some images of the screen showing the problem - as soon as the machines are offline

3. no problem

4. thnaks for your response , although i am not sure i clearly expressed the feature i was wishing for. maybe we can clear it up at a later date as it is not a major concern at this time.

Dear Steve

I spoke again with our software engineers. It is ok to have a virus scanner installed on the Roboocyte computer, but it should be deactivated manually before starting to operate the Roboocyte. If all files are scanned during data acquisition, unpredictable problems may occur.
Our support team is working on the other problems.

Best regards, Christine

Hi Steve,

1. I haven't noticed this problem here, but normally I did not you this way (doubleclick on a single well). I open the Recording Display with it's button from the toolbar and leave it open. This maybe helps you to check your data during a run?
I'm not allowed here to run a computer WITHOUT a virusscanner. We are using "TREND MICRO OfficeScan Corporate Edition". So the software could run together with a virusscanner, maybe not with all.

2. Which frequency has your noise? With which frequency do you collect your data?

3. No problem with broken needles here. (Shipping within Germany!)

4. If your interested, open a new thread and give me some more details!

Ciao,
Mike

http://webkatalog.xantiva.de/ | http://blog.xantiva.de/ | http://flat.xantiva.de/ | http://singles.xantiva.de/

Dear Steve,

I was just wondering about the issue with the software crash that you described. I have also observed this sporadically. I think that it may occur less frequently when I set the acquisition frequency lower. However, I have never lost data because of this. I was wondering if you use the "save" command to save data after each recording.

I hope this is helpful.

Ian

Hi Ian,

>data because of this. I was wondering if you use the
>"save" command to save data after each recording.

But with a big database it takes several seconds up to minutes to save the data with the Version 2.0! I save the data after each "good" oocyte. (Think about IV measurings and several seconds between 100ms recordings ;-) )

In the next version the overall performance will be increased!
I'm just testing it :-)

Ciao,
Mike

http://webkatalog.xantiva.de/ | http://blog.xantiva.de/ | http://flat.xantiva.de/ | http://singles.xantiva.de/

Hi steve,

>2. We earlier reported a problem with electrical noise on
>the signal that occured only when the perfusion was on (we
>use the gilson system). With MCS's help we grounded the
>waste solution ad for a time seemed that this problem went
>away. However, it seems as though this problem has returned
>and has also changed a little. In addition to the baseline
>eletrical noise during perfusion we also some see regular
>variation in the amplitude so it looks like it is modulated
>by another source of noise. Unlike last time we havent been
>able to get rid of this by grounding of either the waste or
>the main perfusing solution. Any suggestions?

Try the following: Ground the needle from the GILSON!
I connect the needle (using the GILSON-cable for the fluid-level-detection and a little bit soldering) with the metal holder of the "z-axis".


Ciao,
Mike

http://webkatalog.xantiva.de/ | http://blog.xantiva.de/ | http://flat.xantiva.de/ | http://singles.xantiva.de/