Chromatography Forum

We (mid-size biopharma company) are running an Agilent CDS and are far from happy with the performance of the system sofar. The software is slow and there are bugs that will require substantial upgrade work to be fixed with a lot of down time and extra work for us to prepare for. If I had the possibility to go back and redo our CDS purchase, I would definitely evaluate other alternatives.

Based on this and some other experience with Agilent software leads to that I can not recommend Agilent software to anyone working within GMP compliance. However, I consider Agilent hardware really robust and well performing.

Try chromeleon or even chrom perfect

ECM is open lab,
which data base?
at the time we looked into openlab, the "lab servers" did not include an anti-virus, and the agilent people did not want to insert on it our company anti-virus at install for demo,
we simply but it back in the bocks and said thank you, let's stop it while everybody still has a smile

>unmgvar
We have an SQL DB installed if that is what you ask about.

sticktown,
to summirize the end point,
IT support is a very important part of a good C/S setup. the better they now what goes at the start then the better the setup runs in the end

how are the clients in the setting?
are those fat or slim clients?
are you using citrix or TS for them if they are slim?
is it on the same server as that of the data base? do you have more then one server for the TS or citrix?

generally when labs go into C/S settings they overlook a lot of major factors.
for example, is your SQL data base on a dedicated server?
where do you store the data?
is it on the same server or is it on a different server?
how did you set up the users? are you using active directories info of the domain?
I have seen for ezchrom a demonstration of installation of this and it was complex and in the end they did wrong and the entire demo setup was crawling bad.
are you mirroring the setup?

can you check the set up?
I will say now that I am actually very well pro-chromeleon.
one of their great ideas was to create a very usefull DB check in order to find out was could be causing slowing down, and help companies and dionex techs, to find out depending on the result were to look for the slow down.
also from my expereience with ezcrhom, which is the bases of start to openlab/ECM/cyberlab, there are many times in C/S bug of memory leaks of the RAM.
another possible problem in C/S with TS and citrix is client hanging.
see about the clients and see how many time the application is being opened or not closed after client logging off.

but one of the major problems of speed is because IT of the company do not assess correctly how much speed is actually needed, and that it is better to invest more in splitting the different parts of the setup into several servers with enough room of growth to the future.

if the "lab servers" are using windows (which they generally are) then it is important to remember that unlike what we are told that they are there and we can forget them, it is very important to see that they are been restarted every now and then, because windows is built still under the need for restart of the operating system.
there are many other important points to look at and it all depends on the way you are setting up things and what you use, this is why the tech suport from the vendor is the most critical point.

>unmgvar

I agree, active and competent IT-support is probably very important for this kind of installations.

I am not that involved in all the technical details that you ask about and do not have the time to dig into it neither, sorry.

However, I know how the software functions and what support we are getting from Agilent, and I am not completely happy with either of it.

So my bottom line, based on my experience, is; If you do not have an IT-department that has full understanding of this type of software packages beware of Agilent software, as you may find yourself with a piece of software that is not what you thought you were getting and it will cost you a lot of time and money to get it fixed, if it ever will be.

The FDA'er inspecting our lab ripped us to shreds over Chemstation/Chemstore. She took one look at our Chromeleon systems, however, and moved right along. Chromelon is bomb-proof from a data integrity standpoint. So we switched all our ICs and LC's over to Chromeleon, but ended up having to custom-code a proprietary solution to put on top of Chemstation/Chemstore for our ICPMS, ICPOES, GC, GCMS and LCMS systems.

We are looking at OpenLab, but we are a little nervous based on how easily the FDA'er poked holes in the "Part 11 compliance validation" documents Agilent gave us last time. We also ran into an issue with "Part 11 compliant" software that would require an analyst to enter a password 30 times per sample for some of our more complex methods, so productivity is an issue.

Wish there was one "right answer" to the question. For a basic QC lab that does routine compendial testing, the off-the-shelf colutions are readily available. But for a contract lab that does a lot of environmental and R&D work in addition to GMP/GLP work, it has been a very long and difficult road.

I think the key here is to remember that it is on the user responsability to see and adapt the products to perform in the needed validated environment
most CDS provide one or more solutions to the needed requirements
the user have to decide how to make the CDS function correctly based upon a well defined URS.
never pass over that responsabilty to the vendor
their work is to get the CDS functionning as you want. and if you feel that the functionnality sucks, like for Cerity cases then do not buy the product
and yes validation is your sole responsability as well. the user must check all the vendors paper and see what they are what they mean and what they cover and if needed to require further tests and procedures

we as well very much like Dionex because of the dionex support and as well the local rep support that were capable of supplying solutons and information to all our sometimes crazy demands

The other reality:
I work with several CDS
Varian Star: HPLC Varian fully controlled, HPLC Waters where only signal treatment.
Varian Galaxie for Varian GC.
Chromilion for other 3 HPLC systems: Dionex and Waters
Chemstation for an Agilent 6890.
In two weeks Empower for another Waters HPLC.
I slip up all the days, several times each day.
Sometimes I feel I should go t the toilette to cry....
Any advice?

between you and me

get rid of the varian hplc and you can control all the rest HPLC by chromeleon
the 6890 is controled by chromeleon
for the varian GC tell us the details and it is most probale you can control it or get a signal from it

why did you buy the waters with empower if you already now how to control waters with chromeleon?

I didn't buy anything. At my Lab there is a Pesticides residues in vegetables Department that is the Jewel of the Crown. They discard equipments and I receive them to do my work.
The three Waters I have are composed by 600 pumps, two RI detectors (410) and one Variable UV plus a DAD. All of them older than the people who assist me.
If I connect all of them by Chromeleon I can't stop the pump at the end of the work automatically, for instance. I should calculate how long will the work take.
I have no problems with Varian Star. Given that I've worked with it for a lot of years, I have a good knowledge over it.
The Varian GC is a 430 model that we (this time) bought new, As I saw the Galaxie I understood I sliped off. We tryed to obtain a signal to use with other CDS but we couldn't (The Varian Technical assistance and me).
About Chromeleon, I like it but I hate the ICs-3000 I have. The Dionex Technical service installed it to control a ds-600 pump, a Wisp 717 Autosampler and a Waters DAD where I lost a lot of its abilities. I also had communication problems that blocked the equipment.
I finally found the solution that was closing and re-opening the server.
About Chemstation I agree completely Karen 01, I hate it.

Well, I talk too much.
Is it possible handling the 6890 by Chromeleon?, Does it handles autosampler and GC conditions?

which chromeleon version do you have?
the software is capable of showing up how much solvents will bw consumed well before the other CDSs did it.
shutting flow and other things is done simply by using PGMs with the specified commands there is no problem with that. again what version do you have for wich modules?
waters has 2996 and 996 pda that can both be controlled, check firmware on the instruments
for the GC 430 you can for sure get the data sent to chromeleon to be integrated and calcualated
the problem with varian is that it is not a serious player that could warrant either HPLC or GC control by the others

IC-3000 have great sensitivity and are right now top IC systems, but i always like to remind, sensitivity is a shrp double edge swords
there are application of IC that you should not breath next to the system or it will pick it up

CM 6.8 handles all gc commands, EPC as well
for the sampler you need to say which one?
for the standard one the control exits

I think the key here is to remember that it is on the user responsability to see and adapt the products to perform in the needed validated environment
never pass over that responsabilty to the vendor

I fully back this statement. However, many chromatographers are unfamiliar with IT terms, computer system validation and so on and use their system as a 'black box' as supplied from the vendor.

the user have to decide how to make the CDS function correctly based upon a well defined URS.

Writing a "well defined URS" is a special and highly sophisticated art. Users usually want a system that
a) is easy to use
b) fast and efficient
c) compliant to 21CFR Part11 and all sorts of regulations
d) good field support
Nothing more and nothing less. Non of these points is specific and traceble.
Many people in this forum think e.g. chemstation is a easy to use software. As long as you think in files this is true. Also many big pharma companies use Chemstation, so it must be efficient and compliant.
To be more specific, one could even add the macro-capability to the URS to be more flexible, even though many of us think, macros are just an excuse for poorly designed software.
From my point of view a database based CDS is much easier to operate. However, do I need MS SQL, Oracle, MySQL? As a chemist I can't even tell the difference. Picasa and iTunes are also database based and can be installed without an engineering degree!
Why not use a simple file based system and copy the area values into (validated) excel sheets? Or even in a LIMS that calculates everything?
It also depends on your procedures. Can u change them? In one of labs we used a structured file system with paths like:
$stability protocol\$time point\$climate or
$customer\$product\$date
in an other lab, owing to chemstations and empowers (1) ability to deal with just one folder or project level they created an external database to create analysis numbers for each of the subtasks and labeled the folders or projects with this number.
Without cleanup the number of analysis number folders gets exessive even though there is not so much data.

we as well very much like Dionex because of the dionex support and as well the local rep support that were capable of supplying solutons and information to all our sometimes crazy demands

I like the chromeleon concept because of the connection of a database and a spreadsheet reporting tool with the ability to submit, review and approve each sequence. Many of our stuff don't get the powerful query capabilities (average age much closer to 60 than to 40). However, I like the possibility to view the column history (column S/Ns are entered in a custom colum) and the efforts for each project and employee (projet code entered as custom column). I like the possibility to collect data from validation projects just with a simple query.

If there was some scheduling ability (when do I have run this test) and some workflow capability (which sequences do I have to review), chromeleon could be used as a full-blown LIMS. Nothing impossible with queries, especially as one can the whole thing even without chromatography (e.g. just enter the sample weights and get the distribution, rsd and mean in a report).
Looking for new buisiness models, I think that could be promising. however I really like my job as analytical chemist.

Alex

PS: I just read in a paper that Excel gains more space in buisness calculations, as it it is much easier to 'program' than traditional ERP packages (like SAP).
People just develop Excel sheets (without any controls) to calculate special cases (development, buisness outlooks, special queries...) as doing the whole thing in SAP would require prepare a proposal, get it singned by your boss, get it singed by IT, write a URS, get it programmed, get it validated, get it signed by the IT boss, by your boss, by all other importand people...... get the roll-out.
That would be a chance for Chromeleon, as it combines database capabilities (security, differrnt sign levels....) with ease of use (spreradsheet instead of code progranning).
Maybe I should rethink my creeer desicitions.

To unmgvar:
Sorry the delay to answer you. I needed to talk with my supervisor (my wife, imagine...) about the solution you propose.
The agreement I obtained is go to Chromeleon step by step.
The first one is control the Agilent 6890 through Chromeleon.
Here the data you asked for
Agilent 6890 Serial US10226093 Firmware Rev. N.06.07
Autosampler G4513A Firmware Rev N.06.07
Can you send me the know-how?

WillyOne
the right thing to do is to talk to a Dionex rep and give him this information

pay attention to the LAN cable of the GC, you need to see if it is crossed or not and decide to use one crossed or go for a HUB connection with the PC.
normally you should not need any additionnal hardware but that only a dionex rep on site can tell you for sure.
we have a 6890, 7890 and 5890 all on CM, but the samplers are CTC

regarding the waters DAD, do you have a PDA option in the license?
a lot of functionnality with CM is present in the command options, and you simply need to know how to find it and use it.
you should post here what functionnality you lost and maybe I or another person can help you out

To get a Varian (now Bruker) GC to talk to analogue data systems you will require the right boards in the GC they were not always fitted as standard. Bruker are working with Dionex for Chromeleon control of the Bruker 430/450 and this will include 3800, 4900 and newer.