PID in base version. $1000 for PID?!!!!!

[acolbourn]

This is such a fundamental and basic function, it shouldn't be a $1000 add-on.   Us programmers making careers writing Labview need it to spread more and I think not including basic control functionality like this could be seriously limiting its growth!  Many programmers like me can't actually buy it because there is no way I can justify the cost to my boss.  Instead I'm wasting programming time making my own and ending up with an inferior result which just makes LabView look bad.  

[altenbach]

Well, it IS included with the LabVIEW Real Time module and that's where it is supposed to be used. I don't think it is reasonable to do a control application in a general purpose OS like Windows.

It is also much more than just PID, there is also some fuzzy stuff in there and much more... 😉

All that said, I agree that it should be included as part of plain LabVIEW. People will then try to used it, notice that it is not deterministic enough, and buy the RT hardware and software at an even higher cost. 😄

I remember when the g-math toolkit was an add-on back in LabVIEW 4.0 (hey, I still have a license for it!). Now it's bundled in at no additional cost.

[acolbourn]

Good point, it could lead to bigger hardware sales.  Haha, ya I figure eventually they will include it like the g-math and the event structure! Hopefully enough people agree that they give in!

[vitoi]

If your controlling something like temperature in an oven, Windows is fine. If the $1000 add on has fuzzy logic and stuff, why not just include plain PID in standard LabVIEW.

[SteenSchmidt]

I don't get it; You can't justify the cost but you end up using precious time making an inferior piece of code that doesn't do the job? That's exactly why you should fork out the $1000 and use your time making stuff that actually makes sense for you to be doing.

If you need it and can't make it yourself in less time than it takes you to earn its cost, just buy the toolset (whichever it is). I don't own cows for milk, I buy my milk (I know this isn't true for everybody, but for me that is the case ).

I'm not saying the PID toolset should cost $1000, or more, or less, or be free, I'm just speaking in general terms here. That toolset is obviously worth something for you...

Cheers,

Steen

[Norbert_B]

To add and contradict altenbach:

It is true that "trying then buying" makes more sense (fun) for most of us.

But to be honest, when adding LV PID to all LV packages, i foresee that people tend to create questions:

Hey, i have that PID thing, but it does not run with 10kHz on my machine. There must be a bug!

Telling those people that they don't have the correct hardware setup and they ought to buy some other stuff (LV RT+dedicated hardware) is sometimes quite difficult.

On the other hand, spending 1k in addition to LV Core for doing PID might also be difficult to understand, as this suggestion shows.....

just some thoughts,

Norbert

[Manzolli]

Here at our company, I know some people doing long term components testing (1000 hours the longest, 41 days and 16 hours) using LabVIEW's PID in many Windows machines (2000, XP Vista and 7). They are doing this for about 10 years. Incredibly people also uses the machines controlling these tests to read emails, access Internet, etc. Since the processes to be controlled is very slow (temperature and fluency),  the PID under Windows is just fine for their case.

I agree with altebach that Windows should not be trusted for this kind of task. I said it to the guys there. But no theory is better that real world. They are fine for more than 10 years...

But PID is a tool that many people use nowadays, like FFT.  It should come in the base package. As any other product, has limitations. NI should clarify these limitations in the documentation.

[PhillipBrooks]

The PID / Fuzzly Logic Toolkit is included with the purchase of a LabVIEW Real Time Module. In order to program the module, you nead LabVIEW RT. You will probably also need signal conditioning hardware for your inputs.

If you aren't using LabVIEW RT hardware, you can certainly purchase stand-alone process controllers that have RS232/485, Ethernet and/or GPIB interfaces for far less than the cost of an NI hardware/software solution. These have been around for a long time, have direct interfaces for your measurement inputs (current loop, RTD, etc...) and can failsafe to a specific state.

[lvABC]

acolbourn,
Have you checked this document?

http://zone.ni.com/devzone/cda/epd/p/id/4973 

Hope it helps!

[acolbourn]

Hey lvABC,

Thanks, I haven't seen that before but it will be a perfect reference!  If labview even just included that in the base version as a subvi I think it would make a lot of people really happy and save a lot of time!