5V input drops to 1.5V

[PizzaHawk]

I am trying to connect a number of push button sensors to a USB-6225. I have all my sensors connected per the manual (5 volt supply with a 50k ohm resistor connected to the input port with one side of the sensor and the other side of the sensor connected to the DGND). My first tests showed that when I engaged the switch i would have 5V at the terminals of the 6225 and the NIMAX program reported an active input. Maybe a minute later the 5v drops off to about 1.5V and the NIMAX reads no input on the channel. When I disconnect the wire from the 6225 board I see the 5V, it immediately drops back to 1.5VDC on re-connection. The 5V power supply is part of large manufacturing tool and is not being strained at all by these extra sensors. .5A out of the 5VDC supply at highest and the Voltage out reading doesnt even flicker when I activate these switches. I have tried a second USB-6225 and it never makes it to 5V when a switch is activated, its immediately 1.5VDC. I've hit a roadblock on what to try next. Hoping maybe one of the fine folk here has seen something similar and can at least point me in the right direction or give me a list of things to try come monday. 

[santo_13]

Instead of several sentences to describe an electrical circuit, a proper circuit diagram will make it simple to communicate and easy to identify mistakes.

In general, whenever voltage drops from ideal, it means higher load is applied than it can supply.

[PizzaHawk]

Back at work so I can supply some images.

I have 8 push button sensors that I am connecting per the USB-6225 Manual

Hand drawn schematic of my setup that shows some branches. Parentheses are the terminal point im connecting to and a current reading math says I should have on one branch and all together. So even if all were connected I shouldnt be straining the power supply. 

Reading deeper into the spec sheet and the manual itself, Im seeing a discrepancy that might have to do with my problem. Plus another article reads similar to what im seeing, so im thinking I may be looking at a floating input (somehow)

While the manual schematic (above) shows a pull up resistor, the Spec sheet calls for a pull down.

My concern is that switching to a pull down would leave the input pin without any current limiting. The following section points out to not exceed 250uA on the inputs.

Any Thoughts and input would be appreciated. Should i try the pull down configuration or is that a sure fire way to smoke the card? Feel im missing something stupid simple.

[PizzaHawk]

Further clarification: Still seeing only 1.5VDC at the terminal point when a switch opens. 6225 Board power supply voltage (Terminal 125) reads a steady 5VDC with switches open or closed.

[santo_13 · Accepted solution]

As you discovered, 6225's inputs contain a 50k pulldown resistor, so your signal must be active high (connect to 5V).

Typically, you don't need the current limiting resistor to limit to 250uA, those are the datasheet spec for input leakage current.

The risk that you need to mitigate is when the DIO was configured (intentionally or accidently) as output and a switch is pressed, now, you need to limit the current to 16mA to not to damage it.

Next step, to protect against this fault, choose a current limiting resistor that results in above 2.2V at the terminal with the inbuilt (20-50k pulldown) and a 5V across this current limiting resistor doesn't go beyond 16mA.

[PizzaHawk]

(Insert colorful and vulgar cussing here). Spend a week fighting this because they cant add "Internal" before "Pull down resistor" Thank you for the help. Now I will spend a few more hours cussing the NI tech writers and engineers while I make the new connections.

[santo_13]

A choice of 10k series resistor from 5V to the switch will suffice.

Now, this 10k with a 20-50k resistor divider will generate 3.33 - 4.16V at the terminal and since the VIH is 2.2V, it will register as logic High.

If the DIO was configured as output and at logic Low, if the switch is pressed, 10k with 5V across will lead to only 0.5mA and will not damage the DIO channel.

[PizzaHawk]

Ill see your paint shop masterpiece and raise it dry erase scribbles:

This is how setup is currently wired. Due to the voltage I was reading earlier I had the equivalent resistance within the Nidaq somewhere in the 22k range. I was contemplating a 2k resistor as it would give me close to 5V and 2.5mA. I can see the risk here that if someone were to switch to output it would have an immediate path to ground. May have to go with the risk based on the fact that when any of our monitoring systems get set up, the nidaq cards gets tucked away and people are too scared to touch it more than power cycling a card.

[santo_13]

---

@PizzaHawk wrote:

Ill see your paint shop masterpiece and raise it dry erase scribbles:

This is how setup is currently wired. Due to the voltage I was reading earlier I had the equivalent resistance within the Nidaq somewhere in the 22k range. I was contemplating a 2k resistor as it would give me close to 5V and 2.5mA. I can see the risk here that if someone were to switch to output it would have an immediate path to ground. May have to go with the risk based on the fact that when any of our monitoring systems get set up, the nidaq cards gets tucked away and people are too scared to touch it more than power cycling a card.

---

are you going with the active low approach due to constraints or choice? you can put that NC switch series with the 2k pullup to make it active high and you're safe.

[PizzaHawk]

I did this primarily because that is what the 6225 manual showed as a digital input connection. How it acts works well for me though. When the valve actuates, sensor opens, reading goes high. Makes it more intuitive for my techs (and me), I see a high signal it means the component is active.