by
lmh » Fri Apr 01, 2011 8:53 am
somebody wrote at the beginning: sort out some realistic test samples and a method and get all the manufacturers to do demos. It's got lost in the discussion since, but this is the only sensible way to choose. Even the most well-meaning people have some bias in their opinion or their experience (not many are lucky enough to have worked with the appropriate models from every manufacturer).
There are so many factors to consider: price, software, robustness, sensitivity, price and value of service-contract or other service arrangements, versatility, how it fits physically into the space available, how it fits with other, existing analytical equipment. Every lab and application is different, and the ideal instrument for Dr Schmitt may be far from ideal for Mr Smith or Professor Forgeron.
You can't, unfortunately, trust sales data directly. Each company defines things slightly differently, making the data very hard to compare. For example, most will publish sensitivity using reserpine, but each will report different S/N on different size injections and it takes some ingenuity to put them on a comparable basis. Even S/N can be defined in different ways. If you get your own test data, you can make sure it's comparable. Also, demos put you in contact with demo chemists, who are frequently a little more honest and realistic than sales staff (after all, a sales rep can say "it detects down to 1 attomole", but the demo chemist has to follow this up with a peak!).