"Memory is full" No it isn't!

[Charcinders]

I'm writing a signal-generating app in LV6. I sometimes get "Memory is full" error from part of program where I'm making a 1500 point sinewave, multiplying it by a 1500-point envelope and sending it to AO Write. Windows resource display shows 50Mb free! Suggestions?
System: Dell Inspiron 8000 / Win Me

[Labviewguru]

,

You should probably post your code. It sounds like a DMA buffering issue to me, without seeing the code. If you post the code, someone should be able to help point you in the right direction.

[Charcinders]

OK, here is the bit where the error happens. This isn't the full program - that's about 1Mb total. As you can see, I'm not doing anything difficult. Perhaps what I should have asked is: is there any way to get Windows to allocate more memory to Labview?

[Labviewguru]

Okay, I have a new policy.

I will not follow up to questions where I have been given a rating like this. If someone doesn't like my answer, then why should I bother following up?

Now, if the original poster wants to follow up with me, they can email me.

This motivates the person who originally posts the question to become involved. I have already requested that the rating system be modified to allow tracking. Until such time; however; I will assume that the person rating the answer was the person asking the question, and not follow up when an improper rating is given. If the person asking the question didn't do the rating, post that in a follow up, or rate it yourself.

This forum is based on integrity. If there is someone out there vio
lating this, then please leave the forum, or correct your actions. This is not a place for personal issues to be raised, especially anonymously. If someone out there has a problem with me or my postings, let's discuss it like adults. Don't play childish games. This forum will not tolerate it.

[Dennis_Knutson]

I agree that someone is abusing the forum. I think NI should either eliminate the rating system or change it so that the rater is identified (and with a legitimate name and email).

[Labviewguru]

Dennis,

You, and anyone else that feels this way should email webmaster@ni.com with your concerns.

I have contacted the person in charge directly (so as not to flood him with emails, I will not release his name - but I believe he receives the emails to webmaster@ni.com regarding DZ) and he has indicated that this issue is being addressed. I am sure he would like to hear from others about this issue. He indicated that anonymity is going to be eliminated from the rating system. Other suggestions should be submitted.

I personally think this is a great resource, and am always troubled when people violate the integrity of this forum. I am making a stand to keep this forum as it is, and also to try to expand it. I always te
ll my clients about this forum.

[Labviewguru]

Charcinders,

I have looked at your code. As I do not have hardware, I cannot run it. Could you please send the text, or screenshot, of the exact error you receive? Also, if you could run it in Highlight Execution, and let me know (or send a screen shot of) exactly which node is causing the problem, I can be of further assistance.

So far, what I can tell is that the AO Config may be called too often. Also, you are calling the AO Write in a loop with no data input. I am not sure what the effect of this is.

DAQ and AO are rather tricky things. You need to be very careful about setting them up. As I have not worked on any DAQ in quite some time, I cannot immediately advise you in this situation, especially since I have no hardware. Again, if I knew
exactly where the error was occurring, I could be of further assistance.

[Charcinders]

Thanks for that. The machine I was working on is in Atlanta, and I'm now back in the UK. Don't know if they are still getting the error. As I recall the message reported the error happened at node 0xD4, which probably doesn't help...
Calling AO write in a loop with no input allows the program to wait until the waveform output has finished.
I should add that I've used these procedures for waveform output in several other programs on this machine and others without problems, and this program works most of the time, so I suspect that if you could run it you wouldn't see an error.

Any thoughts on tweaking memory allocation?

PS. I am not the phantom rater!