9229 / 9239 input range

[Pumuckl]

Hi,

I would like to use these modules for measuring different voltages up to 400V. Since they have an input impedance of 1Meg I should be able use a suitable resistor in series with the Input to extend my range.

This should work good if the 1Meg input is precise and stable.

Since I cannot find a specification in the datasheet I thought I to ask here if someone has experience using external probes or maybe someone involved in the design is reading this forum and can say something about it.

Best regards,

Christian.

PS:

I did see that the 9238 has >1GOhm input Impedance, meaning I could also choose the parallel resistor for the input but I would rather stick to the 10 and 60V modules. They get connected and disconnected a lot and if someone is a bit careless and uses the wrong probe I'd rather have 100V overvoltage protection.

[santo_13]

I would rather not rely on the input 1Meg impedance to implement a voltage divider due to the accuracy is unknown.

Instead, I would recommend implementing an external attenuator + buffer circuit and feeding into the AI channel. The issue is that there are sample and hold circuits which require a low impedance signal source, if you insert a 9Meg series resistor to get a 1:10 divider, the sample & hold circuits cannot perform properly.

[crossrulz]

---

@santo_13 wrote:

Instead, I would recommend implementing an external attenuator + buffer circuit and feeding into the AI channel.

---

Fully agree here.  In the past, I used a voltage divider and then an op-amp in a voltage follower configuration.  We then used the DAQmx Scale to have DAQmx properly read and log the data.  Granted, that was with an older PXI DAQ board, but the concepts work the same.

[Pumuckl]

Since I finally got my 9229 a couple of days ago I wanted to do a follow up on this post. I did a first test to determine how far off the input impedance is, and to my surprise it is spot on.

I made up a simple probe using 9x1meg 1206 resistors. Then I took voltage readings of a 50V DC source with and without the probe. The calculated voltage value, using 1meg as input resistance and the measured probe resistance landed me within 0.2% of the reading without the probe. Since I do not have any precision resistors at hand I did this on the bench with a fan to have a somewhat temperature controlled environment.

Even without a compensation capacitor the 3db point is at 5kHz, which is probably sufficient for most measurements I plan on doing. Adding ~3pF will extend the frequency range beyond what is reasonable for a 50ks device.

I think I will ditch the active route and draw up a passive probe with stable resistors and a compensation capacitor. If I can find the time in the near future to do that I will post the results here.

Best regards,

Christian.

[Henrik_Volkers]

Some hints:

If you dare to measure line voltages: Keep in mind that spike in the kV range are not uncommon. So it's a good idea to stick with a resistor chain (or use resistors with a higher voltage rating)  , same if you want to add the 3pF,   I would think about 3 x 10pF with at least 400V rating.   

If you make a PCB: keep good clearances .. and a thin trace is a cheap fuse 😉

[Pumuckl]

...and I will be adding a low capacitance TVS at the input, otherwise the fuse will most certainly not blow before the overvoltage protection of the module.

Don't worry about me, I should be fine. Almost everything I do is connected to mains and I am the guy my colleagues go to when they fail their surge or other EMC tests. Whenever I get around to actually do something I probably post it here for you guys to have a look and provide feedback.

...about the "fusetrace"

This is actually a good idea and I have used it in a couple of products. This works very good for surge events or short circuits but events where the current heats up the trace without blowing it can lead to catastrophic failures of the PCB. I have seen PCBs that started to degrade and conduct under the fuse area and as a result the area up to the input connector burnt away. If you implement such a fuse be sure that you have no opposite polarity trace near your fusetrace so if the PCB degrades no current can flow anywhere. This is done best with a slit isolating the fusetrace or at the edge of a PCB.

Also, for practical reasons I will probably use a fuse with a holder. I don't think I will be the one blowing that fuse, but I can never be sure about our young guys. They manage to break everything.

Best regards,

Christian.