I'm having trouble with PFI<24..31> signals and can't program to higher drive

[spti123]

Hi all,

I'm having trouble setting up my PFI<24..31> signals on my PXIe-6556.  According to the specifications, I can drive these signals to +/-85mA.  I have not been able to achieve that using the static PFI method using NI-HSDIO.  I see the voltage dropout rather quickly.  Using a 500 ohm resistor to ground on this signal, the voltage collapses from 3.3V to 3.0V with only approximately 6mA of current across resistor. 

What am I doing wrong? 

[Kyle_A. · Accepted solution]

Hi spti123,

According to V = iR, i = 3.3 V / 500 ohm = 0.0066 A, or 6.6 mA of current. You'll need a lower value resistor with an appropriate wattage rating to see 85 mA sourced from PFI 24-31. 

To get 85 mA at 3.3V, you'll need a 3.3 V / 0.085 A = 38.824 ohm resistor (or less) that can handle 3.3 V * 0.085 A = 0.2805 W of power (at least). Power is V * I = P.

[spti123]

Thank you for the fundamentals.

I see approximately 30mA at 1.8V for a 56ohm resistor.  I also see only 60mA shorting to ground, so I'm not able to achieve the 85mA.  Is there a drive strength setting for these particular PFI signals like they do with PPMU pins?

[Kyle_A.]

There isn't any current limiting being done on these PFI lines. What application are you trying to perform that requires 85 mA? Are you seeing the same behavior on all lines? Also, are you attempting to do other operations on the board besides driving current from the PFI lines? I'd just make sure you have a power budget on your chassis that can afford plenty of overhead, to be safe.

How are you measuring the current when you are driving GND? Is your measurement device potentially limiting current based on a setting it is being used at? 30 mA at 1.8V is close enough to be acceptable, given a wide tolerance 56 ohm resistor (5% is 2.8 ohm, so on the high side we are seeing potentially 59 ohm, and the measurement seems to indicate a 60 ohm resistance). 

[spti123]

Hey there, thanks for the quick feedback!  I'm solely driving PFI27 but  have tried PFI25, and PFI26 individually.  We were hoping to drive a relay with each PFI signal (specifically 24-31). As a direct source to these relays that have a 500ohm coil resistance, we thought using the 3.3V with drive strength of 85mA would be ideal instead of sacrificing a DIO pin.  Seeing the output voltage collapse tells me there is a power limitation either due to the VHDCI cable or the CB-2162 breakout board or this is the spec'ed max power achievable, so I wasn't sure reading the specs, I couldn't find the max power output on a single PFI channel. 

To answer your other questions:

I'm using an HP 34401A DMM to measure current in series to a selected resistor. No, it's set to the correct range. 

I believe the intention of these PFI signals was for the use on relays. 

I am also using 1% tolerance high power resistors.

[Kyle_A.]

Yes, the PFI lines were designed to drive relays. The coil load may be too much for the lines to achieve 85 mA, they would need to source more voltage. We do not have a power spec on the lines. Is there a minimum current that the coils need to activate? I'm assuming 6 mA is not enough. Usually, the relay manufacturer would provide an ideal source for driving the relay, whether it specify a minimum voltage or current required to activate. 3.3V may not be a high enough voltage for that particular coil.

As another note, be sure to employ any flyback diodes as required, since coils could potentially flyback current and damage the device. The lines are somewhat protected, but I wanted to mention that as a safety precaution as needed.

[spti123]

Sorry I may have confused you.  I was planning on using relays with 500 coil resistance with flyback diodes,  my concern was the fact that the PFI lines are collapsing from 3.3V (default) to 3V while sourcing only about 6mA from the PFI line.  Since the spec declares that they can source 85mA, I assumed it can source 85mA at 3.3V very much like and LDO without going out of regulation.  Perhaps using the DIO signals with PMU function would be a better choice in this application since we can configure the drive strength up to 32mA and the voltage.

[Kyle_A. · Accepted solution]

To explain the circuit, it is simplified a 3.3V power rail connected to the 50 ohm output impedance, which acts like a series resistance. Given this, we do not behave like a LDO or a current supply, and any load connected to the line is going to behave much like a voltage divider rather than a regulator.

The case of tying the line directly to GND, you will see 3.3V / 50 ohm to get 66 mA of output current. The only way to get the full 85 mA is to tie the line to a -0.9 V reference, which is not very practical I understand. Talking to a HW engineer here, the spec is mainly based off the buffer we use and what it is capable of. I think it makes more sense when it is understood that the line is bidirectional, and it can source/sink 85 mA when being acted upon by an external circuit.

Given all of this, I think it won't be suited well enough to drive a 500 ohm load to the specifications you need. You'll need to have some external circuit that can source 85 mA, or another device such as an SMU. The PPMU on the 6556 can source 32 mA (and beyond, we are not current limited, but anything above 32 mA is not within specification and sustained operation outside of spec could damage the board over time). You can use the PPMU as a source, since it acts more like an SMU and has a sense feedback loop to regulate a voltage/current on the line.

Sorry for the inconvenience, hopefully you have enough information to move forward. Thanks.

[Kyle_A. · Accepted solution]

<removed, duplicate>