Single vs Multi Channel Discrepancy in VI for GPIB Controlled Instrument

[Akello]

Hello all,

I am hoping to get some help with a VI that I am modifying. The VI controls an Agilent 34970A via an NI GPIB-ENET 100/1000. I downloaded the driver for this particular instrument and got it working such that I can select a channel (or multiple channels) and monitor the voltage(s) when I run it. 

The problem I am having is that the current set up requires the channels to be initialized each time a reading is taken. This happens in the first two subVIs of the case structure in the attached screen capture. This is unacceptable as the VI is meant to control over 200 channels at pretty high frequencies and I don't want a mechanical switch to have to flip on and off every time a reading is taken.

This VI initially did not have the case structure around the first two subVIs and the switching was constant every time a reading was taken. The modification was added with the consideration that the channels would be initialized in the 0th iteration while all subsequent iterations (1 or default in the case structure) would have a direct connection in the case thereby bypassing the initialization.

The unfortunate problem is that this fix only works for a single channel. When a single channel is selected, the switch is heard once at the beginning and is silent for all subsequent readings. When multiple channels are selected, however, the instrument goes back to the pre-case structure state and the switch is heard every time a reading is taken. I have tried everything I could think of but I cannot figure out why selecting multiple channels has this effect much less fix the problem. 

Any help is greatly apreciated.

Thank you,

Yusif Nurizade

[Dennis_Knutson]

I don't understand what you are saying. With multiple channels, you should hear relays switching. That is what multiple channels and a scan list mean. Every time you initiate a scan with the trigger, each and every relay in the scan list will open and close. That is how a multiplexer works. One point at a time is connected to the DMM. If you don't want mechanical switches, you are using the wrong instrument. Read the manual to see what the fastest scan rate is and whether it meets your 'high frequency' requirement.

[Akello]

To my understanding the relays switch when a scan is initiated I should hear a switch, however, for subsequent readings there is no reason for the switch the go off. The channel has already been selected and the MUX is taking readings. This is what happens when I select only one channel: I hear a switch the first reading and all subsequent readings occur without the switch going off. I want to extend this to multiple channels; the switches go off during the first reading and stay off for multiple readngs.

In fact the single channel had the switch going off each time a reading was taken before I put in the case structure. Wth the case structure in place, it only goes off when initiated. Does that clarify the situation or am I way off base?

Thanks,

Yusif Nurizade

[smercurio_fc]

You're pretty much way off base. A mux (multiplexer) does not take readings. It selects one of many signals to route to a single point. Thus, it has to switch. Otherwise you'd be connecting all of the channels simultaneously, and have a nice little bonfire.

[Dennis_Knutson · Accepted solution]

You are way off base. As I said, each time you perform a scan, each and every relay is going to open and close. You seem to be under a serious misunderstanding of the hardware. Look at the schematics included in the manual. There is a mux with multiple relays that connect multiple points to a single DMM. Only one point at a time can be connected to the DMM. Under your belief, keeping all of the relays closed after the initial sweep would short all of your connections together.  With a single point, of course there is no need to open/close multiple relays. 

[Akello]

Fair enough. In that case why did the single channel stop switching when I added the case structure?

I saw it work without switching, besides the initial, on the single channel and wanted to extend that to multiple channels. This MUX can support over 200 channels, if I have it switching constantly on all of them I get the feeling I will be replacing some parts very soon!

Please advise,

Yusif

[Akello]

Sorry your last post answered my last post's question before I had a chance to see it.

This leaves me in a bit of a quandary since I am trying to have it run many channels and the constant switching will have me replacing parts pretty often. Guess I should take a look at the switch's life cycle.

Thanks for pointing me in the right direction, please excuse the novice, 

Yusif Nurizade

[Dennis_Knutson]

I believe the ROUT:SCAN command will first clear all relay closures.

The life of the relays is specified in the manual I believe. Otherwise, check with Agilent tech support.

[Akello]

I'll be checking the manual/calling Agilent shortly.

I think we've gotten beyond my level of comprehension; could you elaborate on the ROUT CAN command? What would be the benefit of clearing all relay closures?

Yusif

[smercurio_fc]

The 34970 isn't the actual switch module. You have to put modules into it to provide the multiplexing. For example, the 34901A is a 20-channel multiplexer. Thus, you'd need to look at the specs for the modules, not the 34970. For example, the 34901A has a rated life of 100,000 operations for the switches under rated, resistive load. Rated load is 300V, 1A, 50W. It is 100,000,000 under no load. It sounds to me like you need to be looking at solid state switching. You did not indicate what you mean by "high frequency".