U+ and U5s etc.

Alvin Bach (BACHA@war.wyeth.com)
Thu, 10 Dec 1998 08:39:21 -0500

Here a Wyeth-Ayerst, I have almost finished replacing the sparc2 computers on four Varian spectrometers (VNMR either 5.1 or 4.3) with either U5 or U10 machines. On two VXR series machines (200 and 300) we installed U5's and on two U+'s (400's) we installed U10's.
Varian does now tell you what boards to get but you have to search through some of their emailed VNMR news issues to find that you need a SUN X1032A 10/100Base T/ W/F SCSI interface. This is also the interface you would need for an INOVA console.

The other important tidbit which you MUST know is that you have to order a special SCSI cable to connect the U5/10 to the Varian DIFF box. This is further complicated by the different generation SCSI connectors on each end of the cable. Varian sends you a cable and an adapter to fit the DB50 connector on the Diff box. I did order this cable from Varian but was extremely disappointed by what I received. There is a great deal of play at the junction between the cable and the adaptor and we have seen times when the HAL board disappears from the computer. I think this is related to the connection between the adapter and the cable. I expected this to be the correct cable without adapters. It also took almost two months to get because of back ordering. The cable is manufactured by SUN but it is not a standard SCSI cable. It appears that the Varian engineers who designed the DIFF box made up their own arrangement for the pins.

I am having a single test cable made with the correct connectors on box ends by a cable vendor. If that works I will order another three. I just don't like the idea of having that adapter on the line. The part description also says something about including an active terminator, I didn't see any in the box, so I purchased some and used them. I guess the cable may have active termination but I didn't want to take a chance.

As for the U+ not working with VNMR6.1A The only problem I've had (So Far!) is with automatically referencing the spectra with the RFP and RFL commands. This changed in version 5.3. There are some fixes for this in the user library which work just fine on the VXR machines. On the one U+400 where I tried this , the referencing seems to be off by .5 ppm. I have not tried it on the other machine.

Good luck,

Al Bach
Wyeth-Ayerst Research
CN-8000
Princeton, NJ 08543-8000

732-274-4092