Is there any solution that can meet my requirements?

[kevinn]

Hello, sir,
I am a graduate student in Xi’an Jiaotong University, China and now I am doing some research on the UHF PD(partial discharge) detecting. And we have got the support of a company to develop a set of on-online measurement on turbogenerators. As all the other measurement systems, we obtain the signals from our sensor devices, then send the signals to the amplifier, the switches, at last to the industrial computers. The main problem we met now is that we can’t do well in the aspect of data collection which is your strongpoint. So I hope that you can give me some solutions or suggestions to me.
The following are the characters of the input signals:
The PD signal is a transient pulse and
the rise-time is very short(about several ns or several tens of ns). This means the frequency is very high(about several hundred MHz). The interval between the generation of one pulse and the oscillating attenuation to zero is about 1μs and the interval between one pulse and next pulse is about several tens ofμs(see figure 1). What we concern are the voltage magnitude information and the phase information. So we want to get the max value(about several mv~several v )of the voltage magnitude and we wonder to know when the max value happens.
I don’t know whether the NI’s products can meet my requirements and I will be very grateful if you can give me some advice. If there is any problem or something which can’t be understood about this, please let me know.

[Ben]

Hi Kevinn,

The only thing that NI offers that can come close to what you are asking would be their high-speed scope cards. After reading your spec and reviewing your diagram, I would suspect the scope is probably going to be about 10-100 times to slow.

With this in mind, I would recomend finding an o'scope that can capture multiple discharges with sufficient resolution to capture the transients. Make sure the scope has a GPIB interface and can capture the w'form and transfer it to a PC. When you set up the hardware, make sure one of the channels has some type of reference signal that can be used to get the phase info you are after.

The draw back in what I have described lies in the speed of the GPIB etc. With the resolution you need and the size of the buffe
r required to capture multiple discharges, the data buffer is going to be rather large. This buffer has to be pumped throught the GPIB pipe. This will establish a limit on how often the acquisition can occur.

Let us know if there is anything else we can do to help.

Ben

[Ben]

Please see my answer to your other posting.

Ben