PXIe-7962R on-board RAM timing

Hello everyone,
I am working with PXIe-7962R on a project for which I have written custom VHDL code and inserted it into the design using HDL Node.
I need to access the on-board (512MB) RAM and I am doing this by connecting my custom unit with the memory pins within LabVIEW.
However, I need to at least know some basic information about the timing of the RAM, because the timing is tight and I can't just poll the read strobe pin. I need to know beforehand how long the first read and each subsequent read takes so that I can schedule the operations I need to perform.
Does anyone have the information I request or is there a datasheet with the timing diagram of the on-board RAM?
Thank you for your help!

Hello,
Are you developing a custom adapter module? Can you provide me a list of your hardware and software? We'll likely need to get the PSE involved if you are doing custom module development.
Based on your description (accessing the pins "directly", just as you do with the DAQ module inputs/outputs by dragging and dropping them on the schematic), it sounds like are using I/O nodes from the CLIP. If you are interacting with the memory through the FPGA diagram, the latency will be controlled there. If you are accessing it directly through VHDL in your CLIP, then I honestly don't know.
Can you give me complete hardware/software configuration in regard to the FlexRIO and adapter module?
Could you also provide a screen shot of the code you are using to access the DRAM? Is it directly from your CLIP/VHDL, or on the block diagram of the FPGA VI?
Regards,

Similar Messages

Simultaneously triggered acquisition: Multiple PXIe-7962R + NI-5734 cards, across multiple PXIe-1085 chasses

Hi all,
I need to do high-frequency, high-channel-count acquisition.
Specs
Signals: 10 MHz, 60 channels
Controller: PXIe-8135 (Windows 7)
Chasses: PXIe-1085 (2x)
Input cards: PXIe-7962R FlexRIO base with NI-5734 digitizer (15x)
Description
Each FlexRIO card will store data in circular buffers, and continuously monitor the voltage levels. When any one channel exceeds the user-defined threshold, all 60 channels need to be triggered simultaneously to pass waveform data (200 μs pre-trigger, 300 μs post-trigger) to the host OS for logging.
In other words, every FlexRIO card needs the ability to fire the trigger, which is then received by all 15 FlexRIO cards (the card that fires the trigger also needs to receive the trigger).
Question
Can I achieve this with only a single PFI line? Or do I need some other technique?
I've found some examples, but they only show single-source and single-card triggering, e.g. https://decibel.ni.com/content/docs/DOC-30182 uses one analogue input to trigger the acquisition of 2 channels on the same card.
Any advice would be much appreciated. Thanks!

Assuming you are using LV 2013 or 2014 you should download version 1.3.4 of the FlexRIO Instrument Development Library (FIDL). This will install 5734 Acquisition Engine examples which use the MultiRecord Library to stream data to the DRAM in a circular buffer, allowing you to recover pre-trigger samples. The example also demonstrate how to implement the Syncrhonization Library that synchronizes triggered acquisition to within a single sample period across multiple devices.
Getting the example to work across two chassis' may be difficult, but I believe it is possible. Though to do so you will need a timing and sync module in each chassis to distribute the triggers and reference clock that the sycnrhonization library requires.
National Instruments
FlexRIO Product Support Engineer

1 beep at startup, DRAM refresh failure? Board/RAM problem? Bios needs changing?

I am new to PC building, and I have just put together my first system. So I have everything assembled, system turns on, and I hear 1 short beep. I can go into my motherboard BIOS screen, and even begin installing Windows 7. However, after about 4-5 minutes, the computer will just turn itself off. According to my motherboards beep codes, 1 short beep means a DRAMS refresh failure. Is this RAM related? Now I tried to unhook the ram, switch the spots on the board, and plug it back in, but I got the same problem. Is this perhaps faulty RAM? Or is a motherboard/power supply issue? Here are the components I am using:
Patriot Gaming Series 4GB (2 x 2GB) 240-Pin DDR3 SDRAM DDR3 1333 (PC3 10666) Desktop Memory Model PGS34G1333ELK
MSI H55-GD65 Motherboard
Intel Corei5 750 Processor
eVGA GTS 450 Video Card (01G-P3-1450-TR)
Cooler Master 650W Power Supply ATX 12v
I am new when it comes to all this, so is it possible I need to set ram timing in my BIOS? As far as I know, the 2 ram sticks should be compatible with my motherboard.
Any help would be great, and I can supply more information if need be.

When you the system continues to booting something (e.g. Windows Setup CD), then that one beep is probably a power-good signal or indicates an installed USB device, however, it cannot really be a sign of a critical error, otherwise the system would simply halt when it beeps and you would probably see nothing on the screen at all.
Test with one memory module only and make sure the memory command rate is set to 2T. If you still experience sudden shut-downs or reboots, please check the CPU Temperature in BIOS Setup (H/W Monitor section) to make sure this is not a simple overheating issue. Often, such reboots indicate a problem with the PSU. So, you might want to check that as well.

Error Ram timing to Tight?

I have a msi neo - fis2r working with geil 3200 ddr... i put in 2-2-2-5-8burst and 2.65 voltage.. It booted windows but said ram timing to tight? please help

give it some more vdimm, try 2.8 that should solve your problem.
GeiL ram loves voltage, if that doesnt fix it, put your timings at 2-4-4-7 or something like that

STILL no luck reaching 1n RAM timing -Command Rate- on X48C Platinum

I think my brain has finally fused.... I have tried EVERYTHING and I mean everything to get my RAM to operate at 1N timing -Command Rate 1- and it wont take it....it just wont. At 1st I was trying DDR2 800s on the Qualified Ram list in the manual and they always ran at 2N -Command Rate 2- After much frustration and various other types of DDR2 that also were stuck at 2N timing I moved to DDR3 1333 -Patriot-. They work just fine but guess what!?!? 2N timing and nothing but! I have tried every setting....every BIOS file right up to the the most current BETA I have seen 7.1B5 -715- and this board simply wont enter 1N timing! I've sent MSI support a ticket on the matter and they recomended 7.1B3 -713-. Despite the fact that I already tried that one I thought I'd give due diligence and give it a go again. Same result...2N timing at any setting, big surprise.
I'm at my witts end here. There cant possibly be a combination I'm missing... the condition persists thru multiple RAM chip manufacturers, Speeds, Types and Flashes.... theu my frustration I chuckled to myself in the thought that this Command rate setting is a spoof...a joke setting meant to blow the minds of ppl like me with WAY too much time in front of a computer
Is there anyone out there, anywhere that has ACTUALLY, FOR REAL had 1N timing work for them? Is there a new beta out? I'll take an ALPHA at this point.... This board is wicked fast and stable and I know its askin for my cake and eating it too but if I could just get this ONE issue ironed out this would make this board the ultimate and a justification to why I defected from NFORCE back to Intel based boards.
Thanks alot for any input or ideas! I appreciate any suggestion!
Zyro
Specs
E6700 2.66@ 3.37
MSI X48C BIOS 7.0 --> 7.1B5 -and all inbetween-
2G Patriot 1333 @ 6-6-6-18
Sapphire HD3870 512
Enermax FMAII 535W

Thanks for the feedback,
I know these modules CAN operate at 1N timing -in various other boards i verified this, none of them AMD...I have feared AMD for over 10 years lol- as did all the other chips I tried...again I say I tried BOTH DDR2 and DDR3 the limitation is something in the board not the chip(s). I tried using the XMS profiles advanced and normal got 2N timing. Using the settings the RAM wants to use still get 2N timing. If I dont overclock my cpu the DDR3 runs at 1066 -normal I know but still only 2N timing-
@ Remus. Thanks for your comment
I would like to respond to your comments directly if I may:
1) If 1N timing is unstable on intel boards -I dont buy that personally, I've set many intel boards spanning many chipsets to Command rate 1 with no troubles- maybe the NEXT bios iteration should have the feature removed if it truely is borked -ASUS removed Nvidia GPU-Ex support from their NForce 6 series for that reason although I never had an issue with it-
2) If there is "only" a 3% gain then this board is a lil strange. My previous board, ASUS P5N32-E -stripped striker extreme basically- saw marked improvement, not a two fold effect like I hoped but still much higher than 3% when the command rates were reduced to 1T. I know the other boards I had in the past did better at 1N timings as well but I didnt examine my results or go as in depth with my tuning as I did with the p5N32-E I dare say if U only saw 3% increase then there was some other issue -Inferior BIOS rev? OS problems? System tuning incorrect \ overdone?- causing a bottle neck.... cuz 3% increase when you consider the CPU cycle rate to the RAM bein cut in half seems VERY wrong. I'm not expecting two fold performance....I'm just expecting a feature that is available to be changed, to work. Else why is it there? And IF.....big IF..... there is only a 3% increase....or even a .05% increase.... I want it! Paid for the feature after all
Thanks alot again for all those that have taken the time to visit this thread! Keep it coming!
Zyro

Kernal Panics/Logic Board/RAM

My Powerbook recently died whereby I kept getting the "You must restart your computer message...."
I ran the Apple Hardware Test and it appears that the issue is Defective or incompatible RAM, as many forum messages have expressed.
Now this is fine, but I took it into my local Apple Shop who said that, given the particular specifications of the G4 12inch, RAM is split between Logic Board and somewhere else (sorry, I forget), and that I could only fix the problem by replacement of Logic Board AND RAM. Price quoted was £660. Quote was given without them turning Laptop on, although to be fair I had brought the Error Message in full into the Shop.
So my question is, could it be possible that the Logic Board is not defective, and how do I go about testing this?
I want to exhaust all avenues before I put this to bed, because given the Price quoted it would be nonsensical to get it fixed.
Secondly, is there anyway I can extract all my photo's/music from my Laptop if I have to concede defeat on this?
Cheers in anticipation,
mcmikeyc

Hi MC,
Weidly, ran Hardware Test again last night before reposting and this time came up with following Error Message:
2NVD/5/4: 2122
Do anyone know what this message means?
First thing I'd say was take it apart and examine the built in with a microscope. You'd be looking for bad/broken solder joints or solder "whiskers".
I would be possible to replace it if it was a problem if you were in the states. We have people that do that here. I'm surprised it can't be done in UK or at least we haven't heard of people doing it.
I've also heard of people running without it.
It's right there on top of the board but the top case and the top shield has to come off.
Therefore I need to replace logic board. Any ideas of cost (£660 seems high given that I could replace/upgrade Laptop for less)?.
£660 seems outrageous!
Even DT&T services would be cheaper it's just the shipping. How do these guys in Hong Kong do it? I just bought a iBook charger for $27 shipped. Shipping was like $4 from Hong Kong?
Richard

IBook 1.33Ghz on board RAM

How much RAM should come attached to the mother board? I read 512MB but when I look at my memory it shows two different entries. On for on board and one for PC2700. Shouldn't 512 be built in? Does that mean there is a 256 chip in my expansion slot?

Stephen,
The early and late 2004s apparently had 256 MB RAM and the mid 2005 had 512 MB built in. The way to tell the difference in the later model is (according to the service manual):
"The only noticeable physical difference between this model and the earlier G4
model is the appearance of the foot sockets. Check the foot sockets on the bottom case to see this difference. For this model, the foot sockets are uniformly flat with no "dimples."
However, the foot sockets are compatible with earlier models, so this physical difference is not an absolute indicator of the model type. Be sure to confirm the model type using serial number verification."
You can run the serial number here:
http://www.chipmunk.nl/klantenservice/applemodel.html
That should clear up the matter.
Also the earlier models used PC2100 and the later PC2700.
If you look deeper in the "about this Mac" under "more info" under "memory" it will tell you what's in each socket.
Richard
Message was edited by: spudnuty

P965 Platinum Limited - RAM Timing issue (Corsair 6400C4)

I have 2Gigs of the Corsair CM2X1024-6400C4 RAM running on my P965 Platinum Limited. The RAM is rated to run at 4-4-4-12 timings at 2.1v. SPD defaults to something like 5-5-5-18. I've run several passes of Memtest on these sticks, both pass with no issues.
Now, here is what happens: when I manually set the timings in the BIOS to 4-4-4-12 (up to 2.1v), I have issues coming out of standby. Infact, it only half wakes up, Windows never pops and my monitor remains in a sleep state (yes, I've tried changing the setting for this in the BIOS). The thing that really stands out to me is that regardless of the utility I use to check this, when I set the timings to 4-4-4-12, they all report 3-4-4-12 in Windows. I even checked the 1.51 beta BIOS that popped up last week, its still the same. Setting it to 4-4-4-12 in the BIOS gives reports of 3-4-4-12 in Windows, and I still have the standby issues.
I've done some poking around, and it sounds like this has popped up on the P965 Neo boards as well. I figured I'd throw this up here, see what kind of responses and feedback I can get on this issue. Is it my board, or is it the board, ya know?

Quote from: Maesus on 02-April-07, 14:54:24
Duke Togo, can you publish your full spec? (please read the moan guide, link at my signature)
try set 5-4-4-12, will you get 4-4-4-12?
No, it comes up as 5-4-4-12. The problem only seems to be with 4 coming up as 3.
Intel Core 2 Duo E6600, 2400 MHz
MSI P965 Platinum Limited Edition (MS-7238)
Corsair CM2X1024-6400C4, DDR2-800, 1GB (x2)
EVGA GeForce 8800 GTS, 640MB DDR3, Superclocked
Seagate ST3160812AS, 160 GB, 7200 RPM, SATA-II (x2)
Creative Audigy 2 ZS (SB0350) Sound Card
Microsoft Windows XP Professional SP2
Antec NeoPower NeoHE 550, ATX12V, 550W, +3.3V@24A, +5V@20A, +12V1@18A, +12V2@18A, +12V3@18A

Question about RAM timing etc. on MSI P965 Platinum ? 4th question

Hi.
Just board new Cellshock mem. Now i want to chance the default DRAM voltage on the board from 1.90v to the
spcific ram (mem) setting. They are 4-4-4-12 2.1 to 2.2v
But the voltage settings is grayed out. How to chance it to 2.2v
thx

Quote from: MuggeGHQ on 03-March-07, 07:00:13
Found out now. Thx buddy
But maybe you can help me out with another thing
My new ram is Cellshock DDR2 800 2GB kit standard for the ram is 4-4-4-12 2.1 - 2.2v
What i did in the bios for now was.
Disable the thing you just said (cant remember what you called it) and set it manualy to 4-4-4-12
Then back to cell menu and set the FSB to standard 266 first. And the ram to 800 and 2.0v
Now in the CPU-Z it shows 400Mhz / 2:3 / 3-4-4-12 / 2T
Is that a good setting or ? i tought it would say 4-4-4-12 when i adjusted it to it manually.
I heard these ram could go op to around 1200Mhz .. Damn. its true.
So could you point me in the right direction.
thx man
There is a flaw in the BIOS that when you manually set the latencies in the BIOS it does not set properly. You can change the DRAM Voltage (Vdimm) with it set to SPD by using page up / down on that field. I think DDR2 800 will show up as 400 mhz in CPU-Z. I will check later.

HT2434 Can a logic board, RAM, or processor core be compromised if an attacker gains root access?

If an malicious attack on my computer gained root access from a poisoned program (java applet, flash, unix executable) Could that attacker permanently poison my logic board/PRAM-VRAM or have access from the processor if everything else was (scrubbed)??
Someone gained keychain access on day one in the first 10-15 min. while I was setting things up thinking that the apple firewall was on by default. (Former PC user-oops) I was nailed by a vicious email attack-lucky that it only worked on the yahoo side and it was not opened by (mail) on my MBP. I was also nailed by a brutal PDF, Safari, Quicktime attacks while using Snow Leopard (Mid 2010) In OSX LION, someone was constanly changing my password after waiting for me to restore everything. Then it was a directory services hack. This is why I am concerned. I still see strange activity in console.
Thank you for your time! Mac is still enjoyable to use but the popularity has made it a huge target.

RAM is not the issue in a Mac Pro with Xeon processor, unless you get a distinctive kernel panic, machine check. Otherwise, every Read from memory is checked by Hardware and errors are corrected on the fly, with essentially no slowdown. In the Mac Pro silver tower or dark cylinder, Memory problems do not fester undetected.
panic(cpu 8 caller 0xffffff800d8dcc1d): Kernel trap at 0xffffff7f8df3d00c, type 14=page fault
      Kernel Extensions in backtrace:
         com.apple.iokit.IOSCSIArchitectureModelFamily(3.6.7)
         com.apple.iokit.IOSCSIParallelFamily(3.0)
            dependency: com.apple.iokit.IOSCSIArchitectureModelFamily(3.6.7)
         com.highpoint-tech.kext.ru1022(1.0.3)
            dependency: com.apple.iokit.IOSCSIParallelFamily(3.0.0)
            dependency: com.apple.iokit.IOPCIFamily(2.9)
         com.apple.iokit.IOSCSIBlockCommandsDevice(3.6.7)
            dependency: com.apple.iokit.IOSCSIArchitectureModelFamily(3.6.7)
            dependency: com.apple.iokit.IOStorageFamily(1.9)
Page fault is a reference to memory out of bounds (you do not own). It is most often caused by software.
If you remove the hexadecimal addresses from the stack trace as shown above, the list of what extensions were at the "scene of the crime" becomes much more clear. That is why I could see that your external (SCSI architecture) drives were implicated, and thomas_r. could call out the highpoint driver as being involved.

Replaced Logic Board, RAM, and hard drive connecting wire, but Macbook Pro 15' (Mid 2010) now freezes and shuts down randomly

I have a mid 2010, Macbook Pro '15. Summer of 2012, it kept shutting down and making the three beeping noises, so I took it in for repair. I was told that my Logicboard needed to be replaced, so I replaced it, and took my Macbook abroad with me. Halfway through the summer, the three beeping noises resumed, and since I was abroad, I couldn't use my 90 day warranty and take it back for repair. By the time I returned to the US, my warranty ended, and I was told that the RAM actually needed to be replaced. So I replaced that.
In May of this year, my laptop wouldn't start and instead showed a folder with a question mark. I took it to repair and was told it was the hard drive wire that needed to be replaced, so I replaced that. Less than two weeks after, my Mac began freezing whenever I had a YouTube video or song playing- it made short, staticky sounds and then shut down. Now, it shuts down randomly even when I'm not playing videos or songs, and regardless of if it is running on battery or plugged in. Since I got my hard drive wire replaced, I've also noticed that the laptop takes an incredibly slow time to start up.
The last the Genius Bar technicians looked at my computer when replacing the harddrive wire, they hadn't said anything else was wrong with my laptop.
I've spent almost $1000 total fixing my macbook and am quite sick of going through this over and over again. Can anyone help me know what went wrong this time? (I have no apple care) Thank you in advance.

Check to see if you've any panic logs -> http://support.apple.com/kb/HT2546.
If so, post the text of the latest.
Clinton

MOVED: 1 beep at startup, DRAM refresh failure? Board/RAM problem? Bios needs changing?

This topic has been moved to Intel Core-iX boards.
https://forum-en.msi.com/index.php?topic=142656.0

When you the system continues to booting something (e.g. Windows Setup CD), then that one beep is probably a power-good signal or indicates an installed USB device, however, it cannot really be a sign of a critical error, otherwise the system would simply halt when it beeps and you would probably see nothing on the screen at all.
Test with one memory module only and make sure the memory command rate is set to 2T. If you still experience sudden shut-downs or reboots, please check the CPU Temperature in BIOS Setup (H/W Monitor section) to make sure this is not a simple overheating issue. Often, such reboots indicate a problem with the PSU. So, you might want to check that as well.

How can I synchronize KineticSystems V200 boards in a PXI/VXI hybrid system?

My current setup is a VXI/PXI hybrid system. I use a PC controller over a MXI-4 connection to connect to my PXI system, and use a MXI-2 connection to connect to my VXI system. I have a PXI-6653 timing card in my PXI chassis and I am routing my 10 MHz clock to my clock out.
I have attempted to use the ClkOut to connect to my VXI-MXI-2 ClkIn front panel and I set the jumpers to broadcast that signal to the CLK10 on the VXI backplane. However, I see no setup option to pull the CLK10 signal off of the backplane with the V200 cards.
So after that, I attempted to run my 10 MHz clock from my PXI-6653 straigt to the CLKIN of my V200 board. I have attempted to use mode 7 for my clock selection, however it throws on OpCode error. I increased my sample rate to 200KS/s, and that got rid of my OpCode (however, I am still getting thrown an error when it attempts to calibrate). I connected a scope to the ClkOut of my V200 board, and it's sending out ~350 KHz (I believe) signal. This signal is sent out regardless of what I'm putting in the ClkIn or clock selection mode. Does anyone here have experience with the V200 boards and timing that could help me out here?
Thanks,
Michael Hill

Hi,
As this is a duplicate post, please continue here.
Thanks,
Todd V.
National Instruments
Applications Engineer
NI Prototyping Community

MXI3 (NI PXI-8330) board not detected. Not powering up, no link light.

MXI3 (NI PXI-8330) board not detected. Not powering up, no link light. Cable looks fine. Any ideas?

PeteMan,
I will need to know in more detail how your system is behaving to give you better tips, so please feel free to send me more information.
For now, let me send you a couple of troubleshooting tips:
Cable connection
Make sure both ends of the copper cable are well connected to the PCI-MXI-3 and the PXI-MXI-3 boards. Be careful when disconnecting the copper connectors. you need to push down the plastic connector so it does not get damaged. Damaged connectors could result in bad MXI-3 links.
If the connector has not been used properly, it could be broken (this rarely happens). Make sure you are not connecting the cable backwards.
Power - Up order.
It seems you are controlling your PXI chassis using a PCI-to-PXI extender (copper cable). Make sure you are powering up the PXI chassis BEFORE powering up the PC.
PXI is built on top of the CompactPCI spec, so it uses the PCI high speed bus. The MXI-3 extender is really a PCI-to-PCI bridge toyour computer, so for your PC, all the slots in the PXI chassis look like more PCI slots. In other words, your computer suddenly has more PCI slots.
It is very important to power up the PXI chassis first before powering up the computer, because at power up the PC enumerates all PCI slots. If the PXI chassis is OFF during boot up, the slots in the PXI chassis will never get enumerated and your PC will never see the board installed in the PXI chassis.
Normal Power up LED sequence
Each MXI-3 board has four LED light on its front panel: two green lights (POWER and LINK) and two orange lights (Tx and Rx). When you power up the PXI chassis you should see one green light and one orange light ON steadily (POWER and ). Then, when you power up the PC, both green lights shoudl be ON steadily and the orange lights shoudl flicker.
If you do not see this behaviour at power up, most probably one of your boards its damage.
If you have another MXI-3 extender kit, try to set up the system mixing the boards from the two kits to try isolate which board is the one that is damaged.
Run the MXI-3 Optimization Software
The MXI-3 Optimizaton Software makes sure that the link between the PC and the PXI chassis has been set up properly. It will show you an error otherwise.
MXI-3 Software Version 1.15 for Windows 95/98/NT/2000/ME
The PC powers up fine if the PXI chassis is OFF, but if I turn on the PXI chassis first and then try to power up my PC, my PC does not boot
There are several things that could be causing this behaviour:
It could be that your PC has an old BIOS that has issues detecting PCI-to-PCI bridges. You could try upgrading the BIOS.
A part of the hardware could be damaged. If you have a main-stream computer (like a DELL or HP PC, for example), test your MXI-3 extender in it to make sure that the hardware is in good conditions. Main-stream computers tend to use new Motherboards with good BIOS.
There are more tips in the following KnowledgeBase:
KB 267FAS5P : MXI-3 Troubleshooting Guide
Hope this helps,
Claudia Lorente
Applications Engineer
National Instruments
http://www.ni.com/ask

Choosing a PXIe controller for streaming 200 MBps

Warning: This is a long post with several questions. My appologies in advance.
I am a physics professor at a small liberal-arts college, and will be replacing a very old multi-channel analyzer for doing basic gamma-ray spectroscopy. I would like to get a complete PXI system for maximum flexability. Hopefully this configuration could be used for a lot of other experiments such as pulsed NMR. But the most demanding role of the equipment would be gamma-ray spectroscopy, so I'll focus on that.
For this, I will need to be measuring either the maximum height of an electrical pulse, or (more often) the integrated voltage of the pulse. Pulses are typically 500 ns wide (at half maximum), and between roughly 2-200 mV without a preamp and up to 10V after the preamp. With the PXI-5122 I don't think I'll need a preamp (better timing information and simpler pedagogy). A 100 MHz sampling rate would give me at least 50 samples over the main portion of the peak, and about 300 samples over the entire range of integration. This should be plenty if not a bit of overkill.
My main questions are related to finding a long-term solution, and keeping up with the high data rate. I'm mostly convinced that I want the NI PXIe-5122 digitizer board, and the cheapest (8-slot) PXIe chassis. But I don't know what controller to use, or software environment (LabView / LabWindows / homebrew C++). This system will likely run about $15,000, which is more than my department's yearly budget. I have special funds to accomplish this now, but I want to minimize any future expenses in maintenance and updates.
The pulses to be measured arrive at random intervals, so performance will be best when I can still measure the heights or areas of pulses arriving in short succession. Obviously if two pulses overlap, I have to get clever and probably ignore them both. But I want to minimize dead time - the time after one pulse arrives that I become receptive to the next one. Dead times of less than 2 or 3 microseconds would be nice.
I can imagine two general approaches. One is to trigger on a pulse and have about a 3 us (or longer) readout window. There could be a little bit of pileup inspection to tell if I happen to be seeing the beginning of a second pulse after the one responsible for the trigger. Then I probably have to wait for some kind of re-arming time of the digitizer before it's ready to trigger on another pulse. Hopefully this time is short, 1 or 2 us. Is it? I don't see this in the spec sheet unless it's equivalent to minimum holdoff (2 us). For experiments with low rates of pulses, this seems like the easiest approach.
The other possibility is to stream data to the host computer, and somehow process the data as it rolls in. For high rate experiments, this would be a better mode of operation if the computer can keep up. For several minutes of continuous data collection, I cannot rely on buffering the entire sample in memory. I could stream to a RAID, but it's too expensive and I want to get feedback in real time as pulses are collected.
With this in mind, what would you recommend for a controller? The three choices that seem most reasonable to me are getting an embedded controller running Windows (or Linux?), an embedded controller running Labview real-time OS, or a fast interface card like the PCIe8371 and a powerful desktop PC. If all options are workable, which one would give me the lowest cost of upgrades over the next decade or so? I like the idea of a real-time embedded controller because I believe any run-of-the-mill desktop PC (whatever IT gives us) could connect and run the user interface including data display and higher-level analysis. Is that correct? But I am unsure of the life-span of an embedded controller, and am a little wary of the increased cost and need for periodic updates. How are real-time OS upgrades handled? Are they necessary? Real-time sounds nice and all that, but in reality I do not need to process the data stream in a real-time environment. It's just the computer and the digitizer board (not a control system), and both should buffer data very nicely. Is there a raw performance difference between the two OSes available for embedded controllers?
As for live processing of the streaming data, is this even possible? I'm not thinking very precisely about this (would really have to just try and find out), but it seems like it could possibly work on a a 2 GHz dual-core system. It would have to handle 200 MBps, but the data processing is extremely simple. For example one thread could mark the beginnings and ends of pulses, and do simple pile-up inspection. Another thread could integrate the pulses (no curve fitting or interpolation necessary, just simple addition) and store results in a table or list. Naievely, I'd have not quite 20 clock cycles per sample. It would be tight. Maybe just getting the data into the CPU cache is prohibitively slow. I'm not really even knowledgeable enough to make a reasonable guess. If it were possible, I would imagine that I would need to code it in LabWindows CVI and not LabView. That's not a big problem, but does anyone else have a good read on this? I have experience with C/C++, and some with LabView, but not LabWindows (yet).
What are my options if this system doesn't work out? The return policy is somewhat unfriendly, as 30 days may pass quickly as I struggle with the system while teaching full time. I'll have some student help and eventually a few long days over the summer. An alternative system could be built around XIA's Pixie-4 digitizer, which should mostly just work out of the box. I prefer somewhat the NI PXI-5122 solution because it's cheaper, better performance, has much more flexability, and suffers less from vendor lock-in. XIA's software is proprietary and very costly. If support ends or XIA gets bought out, I could be left with yet another legacy system. Bad.
The Pixie-4 does the peak detection and integration in hardware (FPGAs I think) so computing requirements are minimal. But again I prefer the flexibility of the NI digitizers. I would, however, be very interested if data from something as fast as the 5122 could be streamed into an FPGA-based DSP module. I haven't been able to find such a module yet. Any suggestions?
Otherwise, am I on the right track in general on this kind of system, or badly mistaken about some issue? Just want some reassurance before taking the plunge.

drnikitin,
The reason you did not find the spec for the rearm time for
the 5133 is because the USB-5133 is not capable of multi-record acquisition. The rearm time is a spec for the reference
trigger, and that trigger is used when fetching the next record. So every time you want to do another fetch
you will have to stop and restart your task.
To grab a lot of data increase your minimum record size. Keep in mind that you have 4MB of on board
memory per channel.
Since you will only be able to fetch 1 record at a time,
there really isn’t a way to use streaming.
When you call fetch, it will transfer the amount of data you specify to
PC memory through the USB port (up to 12 MB/s for USB 2.0 – Idealy).
Topher C,
We do have a Digitizer that has onboard signal processing
(OSP), which would be quicker than performing post processing. It is
the NI 5142
and can perform the following signal
processing functions. It is
essentially a 5122 but with built in OSP.
It may be a little out of your price range, but it may be worth a
look.
For more
information on streaming take a look at these two links (if you havn’t
already).
High-Speed
Data Streaming: Programming and Benchmarks
Streaming Options for PXI
Express
When dealing with different LabVIEW versions
it is important to note that previous versions will be compatible with new
versions; such as going from 8.0 to 8.5.
Keep in mind that if you go too far back then LabVIEW may complain, but
you still may be able to run your VI. If
you have a newer version going to an older version then we do have options in
LabVIEW to save your VI for older versions.
It’s usually just 1 version back, but in LabVIEW 8.5 you can save for
LabVIEW 8.2 and 8.0.
ESD,
Here is the link
I was referring to earlier about DMA transfers. DMA is actually done every time you call a
fetch or read function in LabVIEW or CVI (through NI-SCOPE).
Topher C and ESD,
LabVIEW is a combination of a compiled
language and an interpreted language.
Whenever you make a change to the block diagram LabVIEW compiles
itself. This way when you hit run, it is
ready to execute. During execution LabVIEW
uses the run-time engine to reference shared libraries (such as dll’s). Take a look at this DevZone article about
how LabVIEW compiles it’s block diagram (user code).
I hope all of this information helps!
Ryan N
National Instruments
Application Engineer
ni.com/support

PXIe-7962R on-board RAM timing

Similar Messages

Maybe you are looking for