05-03-2010 02:45 AM
Good morning,
Since I have to develop a Quench detection system for a superconducting magnet I have to deal with differential voltages in the ranges of (abs value) 8 ~ 9 volts. Depending on the segment where I take the differential voltage a common mode voltage is added and can be instantaneously up to (+ or -) 900 Volts to ground on the most external segments.
The question is the following: is NI 9206 enough ground insulated for this purpose? In principle, if I use it with 16 differential inputs, can I apply
VCM1 + VD = +900 Volt +-9 Volt;
VCM2 + VD = +800 Volt +- 9Volt;
VCM3 + VD = +700 Volt +- 9 Volt
.... and so on untill
VCM10 + VD = 0 Volt +- 9 Volt ?
Take into account that all the voltages are exponentially damped with a time constant of 1.1 seconds [VCMi (time)= VCM(0)*exp (time / 1.1 sec) ]so that within 3 or 4 seconds the discharge is accomplished.
According to 'http://digital.ni.com/public.nsf/allkb/05D3F9C0F740BC2986256E9A00505421' NI 9206 is rated of 600 Volt and CAT I, hence must withstand 2.5kV for about 5 seconds as explained in page 29 at 'http://www.ni.com/pdf/manuals/374231c.pdf', right?
But at page 13 there is a drawing in which all the differential voltages are stacked above each other with one onlz common voltage at the lower side. Is it a general to deal with one only VCM or am I allowed to apply common voltages for each differential input?
Thank you very much for your support.
Best Regards.
Giancarlo Croari
05-03-2010 10:41 AM
The 9206 is not channel-to-channel isolated, which is what you'd need for an application like this. As a result, the maximum voltage difference you can have between any channel is 30V. The 600V rating is between earth ground and all the channels. Thus, you could have all your channels at, say 591V above earth ground with a maximum of 9V above that, but you can't have a difference of more than 30V between any two channels or more than 600V between any one channel and earth ground.
Regards,
05-03-2010 10:54 AM
Many thanks, I am switching to the NI 9205 with a change of the interface between the magnet and Ni9205 itself.
05-04-2010 11:59 AM
The 9205 is also not channel to channel isolated. Every differential pair must be referenced to the same COM.