Interfere betwwen 6503 output lines in output mode

[stnspring]

Hi,

       We have developed a control box using PCI-6503. The basic layout is as this: all three ports are set to putput mode, and each line control 28V external devices through relays. For each of the 24 lines, it is connected to LC110 through a 2.2k ohm resister. What we found is that, for single line testing, it works fine. But after several (usuallyly >8) lines are set TTL high, when we set the next line to TTL high, all previous TTL high-lines are all droped to low. That is, only the current line is set to TTL high.

       What is the reason for this? Is it caused by current overdrawn? But we already connect 2.2k  ohm resister to limit current under 2.5mA.

       Looking forward to expert's reply and suggestions.

       Thanks.

[Nicholas_B]

Greetings spring,

Thank you for contacting National Instruments.  It does sound like your setup is drawing too much current from the PCI-6503 device.

As you noted, the spec for the PCI-6503 output current states that it must not exceed the typical 2.5mA, indicated by the information in this manual:  http://www.ni.com/pdf/manuals/374938a.pdf  The note at the bottom of page A-2 in the manual makes reference to a resistance of at least 2k ohm to limit the current below 2.5mA (5V/2kohm = 2.5mA).  This resistance refers to the equivalent load resistance of the circuit formed by the digital output line and the load.  If you are using a 2.2kohm resistor in parallel with your LC110 or Relay device, you must make sure that the equivalent resistance of this entire load circuit is at least 2kohm.  The formula to determine the equivalent resistance, Req, of two resistors in parallel, R1 and R2, is 1/Req = 1/R1 + 1/R2.  In your case, R1 would be 2.2kohm and R2 would the the input impedance of the LC110 or Relay device.  To ensure a Req of at least 2kohm, R2 must be at least 22kohm.

Please take a look at the specs for your LC110 or Relay device to make sure its input impedance is high enough.  If it is not, you can use a different R1 value to make Req at least 2kohm, given the R2 of your device. 

Regards,

Nicholas B, Applications Engineer, National Instruments

[stnspring]

Hi, Nicholas B,

      Thank you for your reply. In fact, in our device, the 2.2kohm resistor in not parallel with LC110, it serves as a current-limit resistor and is chained at the LC110 input. That is why I am get confused. In this case, the resitor on pin output will definitely be greater than 2.2kohm. and it should wrok fine.

      In fact, for each of signle line output, there is no problem. The problem occurs when multiple lines are set. So we think that this may be caused by the overall current limit of PCI6503 (what is this limit?).  When total current is over a threshold, different lines will get interfered. After some tests. we think this may be the reason. And we also found in the mannual saying that PCI6503 is a logic device and had better not used with current-driving curcuit such as LED and mechinical relays. So currently we have put a driving layer (CD4050) between PVI 6503 and LC110. This seems working fine in our first-step test. We will further test it.

    Thank you very much.

    Best regards.   

[Nicholas_B]

Hi spring,

The Intersil 82C55 datasheet specifies that each I/O line can source 2.5mA each.  The link to that document is here:  http://www.intersil.com/data/fn/fn2969.pdf#search=%22intersil%2082C55%22

You are correct when you say that the 6503 should not be used with current-driving circuits such as LED and mechinical relays... this is mentioned in the note referenced on the bottom of page A-2, as per my previous post.  Using the intermediate driving circuitry is a good idea, just be sure that this circuitry does not sink more than 2.5mA from each line.  I would suggest using an ammeter to monitor the current coming out of each I/O line.   This will give us a better idea of exactly how much current we are sinking from the device at any given time. 

Hope that helps, please let me know if you run into any more issues!

Regards,

Nicholas B, Applications Engineer, National Instruments