X Series Board Driver

[LarryL]

I have been reviewing the XSeries MHDDK documentation and examples as I figure out how I can use the X series PCIe-6351 to replace 

the M series PCIe-6251 in our in-house control application.  As a start, I use the example aiex2 for anaalog input test. Now the problem is:

After armed the AI subsystem

aiHelper.getInTimerHelper(status).armTiming(timingConfig, status);

It can't get the following condition met

if (device.AI.AI_Timer.Status_2_Register.readSC_Q_St(&status) == nInTimer::kSC_Counting)

The device.AI.AI_Timer.Status_2_Register.readSC_Q_St(&status) returns 1 and nInTimer::kSC_Counting is 5

Therefore the measurement can't be started.

Any ideas?

Thanks,

Larry

[JoeFriedchicken]

---

LarryL wrote:

The device.AI.AI_Timer.Status_2_Register.readSC_Q_St(&status) returns 1 and nInTimer::kSC_Counting is 5

---

It looks like the start trigger signal is not being detected.

 tInTimerValues.h

    typedef enum {

       kSC_Idle                           = 0,

       kSC_WaitforStart1                  = 1,

       kSC_PreCounting                    = 2,

       kSC_PreWait                        = 3,

       kSC_WaitForStartOfSample           = 4,

       kSC_Counting                       = 5,

    } tInTimer_SC_Q_St_t

[LarryL]

Hi Joe,

That's where my problem is. I went through the aiex2.cpp code line by line and couldn't find why the start trigger signal is not detected. The following is some of the code, can you see any problem???

aiHelper.reset(status);

device.AI.AI_Timer.Reset_Register.writeConfiguration_Start(kTrue, &status);

aiHelper.programExternalGate(

       nAI::kGate_Disabled,// No external pause signal

       nAI::kActive_High_Or_Rising_Edge,// Don't care

       status);

aiHelper.programStart1(

       startTrigger,// Set the PFI line

       startTrigPolarity,// Make line active on...

       kTrue,// ...edge (not level)

status);

aiHelper.programStart(

       nAI::kStartCnv_InternalTiming,// Drive the clock line from internal oscillator

       nAI::kActive_High_Or_Rising_Edge,// Make the line active on rising...

       kTrue,// ...edge (not high level)

       status);

if(deviceInfo->isSimultaneous)

{

// Program the convert to be the same as START

aiHelper.programConvert(

       nAI::kStartCnv_InternalTiming,

       nAI::kActive_High_Or_Rising_Edge,

       status);

printf ("isSimultaneous\n");

}

else

{

// Program the convert clock to start on the sample clock

aiHelper.programConvert(

       nAI::kStartCnv_InternalTiming,// Drive the clock line from internal sample clock

       nAI::kActive_Low_Or_Falling_Edge,// Convert on falling edge

       status);

       printf ("convert clock to start on sample clock\n");

}

// Program the sample and convert clock timing specifications

timingConfig.setAcqLevelTimingMode(nNISTC3::kInTimerPostTrigger, status);// This is a finite measurement with samples after the trigger (eg post-trigger samples)

timingConfig.setUseSICounter(kTrue, status);// Use SI for internal sample clocking

timingConfig.setSamplePeriod(samplePeriod, status);

timingConfig.setSampleDelay(sampleDelay, status);

timingConfig.setNumberOfSamples(sampsPerChan, status);

timingConfig.setRetriggerRecord(kTrue, status);// Set the measurement to be retriggerable

if(!deviceInfo->isSimultaneous)

{

timingConfig.setUseSI2Counter(kTrue, status);// Use SI2 for internal convert clocking

timingConfig.setConvertPeriod(convertPeriod, status);

timingConfig.setConvertDelay(convertDelay, status);

printf("not simultaneou timing configuration\n");

}

aiHelper.getInTimerHelper(status).programTiming(timingConfig, status);

aiHelper.programFIFOWidth(nAI::kTwoByteFifo, status);

aiHelper.getInTimerHelper(status).clearConfigurationMemory(status);

for(u16 i=0; i<numberOfChannels; ++i)

{

// Set channel parameters

chanConfig[i].isLastChannel = (i == numberOfChannels-1) ? kTrue : kFalse;// Last channel?

chanConfig[i].enableDither = nAI::kEnabled;// Dithering helps increase ADC accuracy

chanConfig[i].gain = gain;// Set the gain (used by the hardware)

chanConfig[i].channelType = termConfig;// Set the terminal configuration

chanConfig[i].bank = nAI::kBank0;// AI channels 0..15 are on bank0

chanConfig[i].channel = i%maxChannel;// aiN where N = i%maxChannel

chanConfig[i].range = range;// Set the range (used by the scaler)// Push the channel configuration to the device

aiHelper.programChannel(chanConfig[i], status);

}

aiHelper.getInTimerHelper(status).primeConfigFifo(status);

device.AI.AI_Timer.Reset_Register.writeConfiguration_End(kTrue, &status);

aiHelper.getInTimerHelper(status).armTiming(timingConfig, status);

device.AI.AI_Timer.Status_2_Register.readSC_Q_St(&status);

printf("device.AI.AI_Timer.Status_2_Register.readSC_Q_St is: %x\n", device.AI.AI_Timer.Status_2_Register.readSC_Q_St(&status));

printf("nInTimer::kSC_Counting %x\n", nInTimer::kSC_Counting);

Thanks,

Larry

[JoeFriedchicken]

---

LarryL wrote:

That's where my problem is. I went through the aiex2.cpp code line by line and couldn't find why the start trigger signal is not detected. The following is some of the code, can you see any problem?

---

I think what you're looking for is in the comment at the top 😉

[LarryL]

Hi Joe,

Can you be more sepcific where it is? If I can find, I wouldn't ask for help.

Thanks,

Larry

[JoeFriedchicken]

Sure, aiex2 uses an external start trigger as described in the synopsis at the top and "External Connections" subsection. The default behavior is to trigger on PFI0. If you want to use a different the terminal, you can change line 85:

const nAI::tAI_START1_Select_t startTrigger = nAI::kStart1_PFI0;

If you are already providing an external signal, verify that it is securely connected and has the proper electrical characteristics according to the specifications.

[LarryL]

Spoiler

Thanks, Joe.

 It helps a lot.

[LarryL]

Hi All:

I have made the X series board driver working. However, I got some strange results. The driver I made is working very well on desktop PC and it is not working well with industrial PCs. We have tested it on 3 different PC hardware and got 3 different results (see attached screenshots). Does anyone has some ideas/suggestions?

The first two following figures are from two indutrial PCs and the third one from desktop PC. The X series cards and driver used were all same for three PCs  

Thanks,

Larry 

[JoeFriedchicken]

---

LarryL wrote:
I have made the X series board driver working. However, I got some strange results. The driver I made is working very well on desktop PC and it is not working well with industrial PCs. We have tested it on 3 different PC hardware and got 3 different results (see attached screenshots). Does anyone has some ideas/suggestions?

---

Hi Larry,

I have a few questions:

1. In your graphs, what are the units for the X axis and Y axis?
2. Which data are you plotting? It looks like analog input, but you didn't say.
3. What are the settings for the measurement? If your data are from the analog input subsystem, what settings did you program into the configuraiton FIFO?
4. What is the environment like for each computer -- are they all in the same physical place?
5. What are the electrical connections for each computer -- what kind of sensor are you using and to which pins have you connected it?

[LarryL]

Hi Joe,

Thanks for the response.

Here are the answers for your questions:

1. the unit on graph is micron for X axis and sample point (time) for Y axis

2. the data plotted is actually the following error. we did test on one of our control system and got all data. we also had reference and actual moving data. But the following data shown better.

3. all the data is from encoders. the problem wasn't coming from actuator encoder. it was actually from reference. our control reference signal is generated from encoder inputs. the glitch you saw on the graph is where the reference had glitch.

4. the industrial PC was in same place and we switched the processor boards. the desktop PC was next to industrial pc.

5. all the connections were the same to 3 set up pcs.

Thanks,

Larry