PXI Platform - DC-DC Converter Output Ripple vs Load

[Cisco25]

Hi,

This interesting NI tutorial relates about measuring different DC-DC converter parameters:

http://www.ni.com/white-paper/9622/en 

Unfortunately there is nothing related to the output ripple.

I am building an automated testing setup to characterize the output ripple vs the output current load of a Linear Technology DC-DC converter.

The goal would be to try different types of output capacitors on a wide temperature range and measure the line ripple for each of them.

I was investigating the NI PXI platform as we already use it for other applications.

So far, the use of two PXI-4130 sounds a promising starts.

The missing component would be an accurate ADC module with sampling capacity above 2Ms/s, the DC-DC switching at 400kHz.

Accuracy for the ripple voltage measurementwould be also very important, in the order of 0.1mV.

What modules and/or setup would you recommend?

Thank you.

Regards,

Francois

[WillPwr]

Hello!

I'd be happy to recommend something to you, although I do have a few questions first. Since you're looking to characterize the ripple vs. current load, what level of current will you need to measure?

I'm also assuming you're using the 4130 SMU as both a source and measurement device. Please let me know if this is not the case.

[Cisco25]

Hi

Thanks for the response.

Yes you're right, the 4130 would be used for both source and sink. I have seen that the 4130 can go up to 10W in sinking (I am not sure about the source though), as my DC-DC converter output is 5V, the load current will be limited at 2A (it could go up to 5A). Do you have a better device to recommend with more input power?

Any recommendation for the ripple measurement?

Thanks.

Francois

[WillPwr]

Hello!

Thank you for the information. Unfortunately, we don't offer a device that can measure current at the speeds you're looking for. You can use the 4130 (as you mentioned) to souce and sink the current, but you'll need an external shunt resistor and a digitizer or other high sample DAQ card, such as an X-series DAQ like the PXIe-6366.

I also wanted to verify - is 2MS/s going to be enough for your application? If you have 400k to measure, you'll only be geting 5 samples/cycle, which is relatively low. It may be fine, but I just wanted to verify. Note that the 6366 could handle up to 3.33MS/s.

Also, if you're measuring multiple current levels (say, 1mA, 10mA, 100mA, etc), then you'd need a different shunt resistor for each range. You might also consider a switch, too, if you'd like to automate switching between each level.

[Steve_B]

The PXI-5922 has a flexible resolution vs sample rate and would be worth looking at. It has a +/-5V input range, which may not be enough for the 5V output, but it could be AC coupled. 

[Cisco25]

Will,

I got the chance to get hands on the NI USB-6366, I did try at 1.25MS/s and the result is far from accurate. 2MS/s still wouldn't be enough I think.

Could you clarify why should I need a shunt resistor/switch? Isn't the current switch integrated in the PXI 4130 module used to sink output current?

I'd like to measure output voltage ripple and I think that hooking up a DAQ to my dc-dc output should be fine as long as it can provide AC coupling.

Steve,

The PXI-5922 could definitely be a solution for the ripple measurement. Thanks very much.

[WillPwr]

Hello, Francois!

The 4130 can't sample as fast as you'd otherwise need it to, hence the use of the shunt resistor. The internal shunt on the 4130 won't suffice at those speeds. Does this make sense?