MAX (win2000) cannot see devices in my MXI-2 HP E1302B system.

[gsiemens]

I must say to start that I am a newbie to programming these devices, but I have extensive computer hardware skills and am a VB programmer. I have set up two HP E1302B systems with VXLink.

This is my system (1-17-2002):

- a recent Dell Optiplex GX150 w/ WIN2k
- latest Max (2.1?)
- latest NI-VXI (3.2)
- PCI-MXI-2 to B size MXI-2 interface
- HP 1302B 20 slot mainframe
- two HP E1326B Multimeter cards LA 24 and 32
- 2 HP E1353A Multiplex cards LA 25 and 26 on E1326B LA 24
- 2 HP E1353A Multiplex cards LA 33 and 34 on E1326B LA 32

This is a test system to prove out the MXI-2 interface for our two existing VXI chassis'. These existing systems have HP VXLink PC interfaces which are ISA based and are being obsoleted by the lack of ISA in new Dell PCs.

I am trying to get the interface to the point our programmer can do his work on modifying our existing in-house VB DAQ application to work with the MXI-2.

I install all of NI's software; NIDaq first and then NI-VXI. Install the card and device manager and Max sees it. I power down and connect the cable PC-MXI-2 to MXI-2 in the E1302B. Reboot. Run Max and then System Resource Manager. A small window pops up saying it is configuring the mainframe. The 5 second wait for the SYSFAIL to clear ends and a message says "Configuring Mainframe 255" and the system completely locks up (mouse won't even move - hard reboot a win2000 machine?!).

I have waited up to several hours for this app to finish but it never does. If you power of the E1302B after the lock up the app finishes instantly and returns to Max.

Next I pulled all of the cards but the B size MXI-2 card and reran resource manager. This time the configuration process completes fine. After trying various permutations it turns out that of the four E1253A cards, two cause lockups.

With those two cards set aside, Max lists only the MXI-2 interfaces, one in the PC and one in the mainframe (as near as I can tell). It does not list any of the 4 other cards.

Am I correct in thinking that these cards should show up? Using HP stuff their resource manager would always list all of the cards in the chassis.

I then changed the LA of the two E1353A cards to that of the cards causing lockups. Resource Manager still runs fine (not an LA problem).

I checked all of the jumper settings on the two groups of E1353A cards and the ones that work match the ones that don't as far as i can tell.

After looking all over the NI support site I can only find some information on checking simple things like bent pins and the like. I will check these but what else could be the problem? What else could be the problem? Thank You Much. GReg...

gsiemens@trane.com

[Julianna]

Hello Greg!

I have a few questions about your system configuration. First, are the various components of your system VXI or VME (the MXI-2 card, the chassis, the instruments)? I would definitely recommend checking the logical address you have set for the PCI-MXI-2 and VXI/VME-MXI-2 cards. It concerns me that the chassis is identified as Mainframe 255. The mainframe is identified by the logical address set for the MXI-2 module located in Slot 0 of the chassis. If the VXI/VME-MXI-2 is set to LA 255, then cards with lower LAs in that chassis will not be recognized. The PCI-MXI-2 should be LA 0, and I would recommend LA 1 (for single chassis configurations) or LA 16 (for multiple chassis configurations) for the VXI-MXI-2. Also, are all 4 cards stil
l recognized and functional in the VXLink systems?

[gsiemens]

Thanx for replying.

First off, I goofed when I said that the mainframe was 255. I should have said 225.

All of tha cards are VXI cards as far as I know. The 4 cards are all working cards that have been in the VXlink systems at one time or another.

I think you might be onto something. I am including a screen shot of MAX's device tree. I will check the settings of the VXI-MXI-2 card. It seems strange to me that NI would ship it set to something other than LA=1.

Many Thanx. GReg...

[Julianna]

Thanks for the screen shot. From the VISA instrument descriptor that the VXI-MXI-2/B is being seen as LA 225. This is the reason the other cards are not displayed in MAX (because their LAs are lower than this). Definitely check the LA the VXI-MXI-2/B is set for. You are correct that the default setting is LA 1. If the switches are not set to 225, you could reload the factory settings for the EEPROM. In MAX, right click on the VXI-MXI-2/B and select Hardware Configuration. Under the Common Configuration section, use the pull-down menu to select Default Configuration, and press OK. This will reload the factory defaults to the User side of the EEPROM. If this doesn't fix the problem, I would recommend having the card repaired (a bit mi
ght be stuck). You could also change the LA of the other cards to above 225 to make sure they are properly detected.

[gsiemens]

I changed the LA of the VXI-MXI-2/B to 1 (one) using the DIP switches. This did not help. I have included another screen shot showing MAX. Boy I am confused.

In addition, I can still get the RM to lock up as described in my first post by inserting either of the two troublesome cards into the backplane. Lockups go away when card(s) is(are) not present.

I have an extremely short timeframe to solve this problem. The existing VXlink computers will have to be replaced by March 6 and there is programming to be done before that can happen. This does appear to be poor planning on our part, but in our own defense, we were overcome by circumstance. We did not forsee problems with such a fundamental setup issue. AAARGH!

Could the ch
assis cause these issues? The HP cards have two other "hard coded" configuration settings:

IRQ - Choices are 4 DIP switches - 0,1,2,X
Card ID - Choices are 4 DIP switches


I have another VXI-MXI-2/B card to try. Thanx. GReg...

[Julianna]

Greg,

You mention that 2 cards cause Resman to hang. Which cards cause this? I looked on the Agilent site, and the modules you are using are VXI modules. However, the chassis you are using follows the VME rather than VXI backplane specification. Agilent has several chassis they call VXI/VME chassis. These chassis meet the power and cooling specifications of VXI, but use the VME backplane. Therefore, VXI features such as the reference voltage lines, CLK10, and MODID are not available. If any of your VXI modules rely on one of these VXI features, then problems would arise because the feature is not implemented. So, yes, the chassis could be causing the problem. Unfortunately, I cannot determine whether the HP card
s rely on any particular VXI signal. I would suggest talking to HP about that aspect of the issue.

[gsiemens]

The cards that are causing the lockups are the E1351/53A (16 channel FET MUX) channel multiplexer cards. They connect the sensors to the voltmeter (E1326) in our systems. The card-chassis layout is in the attachment.

The cards initially causing the lockups were LA 25 and 33. I pulled LA 25 and 33. I changed addresses of the other 1351 cards (LA 26 and 34) to 25 & 33 and no lockups. I changed the LA of the pulled cards to 26 and 34 and reinserted them - these cards still caused a lockup singly with the other cards, together with the other cards or alone with the VXI-MXI-2/B. Occasionally this configuration would work with all of the cards in the chassis, but I cannot repeat this consistently.

I tried using a
second VXI-MXI-2/B. The RM will only finish and not lock up if the new VXI-MXI-2/B is alone in the chassis. Oh boy! Put the first VXI-MXI-2/B in and it was as before - will not tolerate having the two "problem" 1351A cards in the chassis.

I am going to try another chassis. Also I will try another cable and PCI-MXI-2 card as well.

I do see VXI/VME logo on the 1302B chassis. The lack of features of the chassis doesn't seem to be a problem. Two of these DAQ systems (VXlink + 3 x E1326 + 9 x E1351 + 2 E1330B) with VXlink interfaces have been running 24/7 for about 4 years under Win95a! We've had some issues with stability and crashes, but I don't think that they were due to card-chassis problems. As I understand it, all of the devices are operating in a simple message based mode. Ask for data - read the data (SICL commands iwrite and iread).

Should I be changing the setup of the MXI-2 interface because we are running a simple system? On the other hand, should I make chan
ges because we are on a VME backplane? I remember seeing some jumper on the VXI-MXI-2/B about CLK10. Is it time to involve NI tech support? Thanx again. GReg...

[gsiemens]

UPDATE SUCCESS

I think I figured it out thanx to Julianna's last post. The VME/VXI chassis I have does not send a CLK10 for the VXI bus. So I changed the VXI-MXI-2/B to:

1. Become the slot 0 device (forced W3 to Slot 0 not auto).
2. Use the on board oscillator to generate the VXI bus CLK10 (W1 default).
3. Force it to drive the MXI bus CLK10 from the VXI bus CLK10 (S1 IN position).

Now upon power up of the chassis, I have a yellow VXI indicator lite on the VXI-MXI-2/B. I then ran RM with one card in the chassis with the VXI-MXI-2/B. The card's LA was detected along with its' description (LA = 24 type E1326B). All of this now appears in MAX's tree. WaaHoo!!

Now I will insert all of the other cards
for the test system and see if they work. Thank You Julianna. GReg...