Peak extraction

[johnnydas]

Hello everybody...

I would like to inquire you regarding a certain device that I am looking for.

In the company I work for we are using a load cell to measure the force applied in one of our machines and an indicator to view the resulting data.
The force readings follow, more or less, a sinusoidal pattern with a frequency of 15-20 Hz.
We are taking multiple readings between two consecutive "peaks" so the number of readings per minute is approx. 10000.

What we are now looking for is a signal processor that could be programmed so that it can extract the peaks from the data and furthermore provide us with the average value of x consecutive peaks (x is variable). The above peak extraction and averaging should not be done in a PC (due to stability reasons) but in the specified equipment

This indicator should be able to connect to both a PC and a PLC so that the resulting signal can be processed furthemore.

I would appreciate it if you could propose a device like that.

Thanx everyone....

 

[nbucska]

You are useng some Analog-dig. converter. How many bits?
Type ?

It would be -- unless you need 100-s or more -- simplest
to use a PC as signal processor using the parallel port to interface to the ADC.

The programming is trivial.
<nbucska@pcperipherals.com>

 

[johnnydas]

We are using an indicator made from the dutch company Penko and the whole problem is that due to company policy we cannot use a pc to do the job as no pc is stable to provide 100% reliability and this thing is going to be used as part of an automatic control procedure.

I already have the software for the data extraction ready but I can only use it for trials for the above reason. There is another way to create a small circuit that rapidly set to zero the maximum recorded value but this, of course, is not CE certified and cannot be used either....

I have found programmable indicators that can do it though and this is what we are propably going to do after all.

Thanx anyway....

 

[nbucska]

There is no way to assure 100% reliability -- there is always a possibility of failure -- only the probability can be reduced...

You could use 2 or more parallel processors and let the main controller to decide which is reliable.

You can buy for about $160 a 2.6&quot;x2 PC compatible processor.

<nbucska@pcperipherals.com>

 

[johnnydas]

The problem is not the processor but the whole system. Both processors will be using the same software (Windows NT) and windows are not that famous for stability.

Now if there was another way to have two systems running the same software and a controller to decide which of the two is stable then I guess that would be better.
But I don't think that something like that exists.

 

[nbucska]

I suggest using 2 PC stamps ($160 ea, 2&quot;x2.6&quot and write
your own software. You can add self-calibration, etc.

These would use DOS and definitelly not WINDOWS. No disk
either but flash memory.

Need more ? send more info, spec's, etc.
<nbucska@pcperipherals.com>