High Speed I/O

jcr

Lifetime Supporting Member
Join Date
Oct 2006
Location
Central Florida
Posts
91
I'm looking for a solution. I need to be able to control I/O in the 20-30 microsecond range. I have a customer who needs to control a laser (Shutter Open/Close) at very high speeds. Does anyone have experience controlling I/O at this rate?

Beckhoff claims to have very fast I/O that is repeatable down to 10 microseconds. Has anyone used this I/O? Are there any other companies out there offering solutions in that speed range?

Thanks,

jcr
 
I'm looking for a solution. I need to be able to control I/O in the 20-30 microsecond range. I have a customer who needs to control a laser (Shutter Open/Close) at very high speeds. Does anyone have experience controlling I/O at this rate?

Beckhoff claims to have very fast I/O that is repeatable down to 10 microseconds. Has anyone used this I/O? Are there any other companies out there offering solutions in that speed range?

Thanks,

jcr

+++++++++++++++++



Siemens makes a high speed boolean processor.
Its called a FM352-5. It will do your application. You program an
FPGA and your scans are 1 microsecond. Yes micro

I've used it on an cigarette filter machine.
 
i know a manufacturer that makes lines for lumber mills that uses beckhoff just for the fast processing speed.
I havent tried but i heard they should be really easy to program as well.
 
VIPA also make some high speed I/O modules for their S7 Compatible processors. I don't know if they're as fast as you need but worth a look.

Nick
 
I think that 20-30 microsecond (not millisecond) is outside any current PLC's update range, even Beckhoff.
You need to consider special fast processing modules, like the one suggested by PLC.
 
Out of interest, I've just looked up some statistics on the VIPA Speed7 CPUs and fast I/O cards. The Speed7 processors implement a a high speed bus exiting the left hand side of the CPU called the Speed-Bus for comms with the high speed peripheral cards.

The 317-4NE12 CPU processes bit and word instructions in 0.01 micro Seconds/instruction (approximately 3 times faster than a Siemens 317-PN/DP). The fast DI card (312-1BH70) has an input delay time of 6.12 Micro Seconds but the fast DO card doesn't quote a propagation delay.

I'm not sure what the Speed-Bus Update rate would be but it has a bandwidth of 58 MBit/S so it should be quite snappy if you're only updating 1 or 2 cards. So if you gave the CPU nothing else to do, it might just about achieve the required speeds - but that would be a dedicated controller.

Nick
 
20μsec pulse length corresponds to 25kHz pulsetrain of 50% duty ratio.
Try any PLC with high-speed positioning output, and perform 1 step positioning with 25kHz frequency setting.
Most probably, it will generate a single 20μsec pulse you need.
Or, perform jog pulse in Micrologix.
 
Last edited:
Scratch that. Yes 0.025 µs for bit instructions on a 317 CPU.
But it wont work.

Even if you have so low execution times, how do you make the program update in micro-second area ?
One problem is that there are no timers that will work in microseconds. That is definitely a limitation.
Another problem is that the smallest possible timed interrupt is 500 µs. (edit: 500 µs is for a 319. 317 can go as low as 1 ms).
Running the program freely in OB1 without interrupts will depend of how fast the CPU updates its other "house-keeping", including communication tasks. Dont know how fast a "near-empty" 317 will run, but doubt it will go as low as the required 20-30 µs range.
 
Last edited:
Manglemender, I am pretty certain that the instuction timings for Vipa and Siemens CPUs you mention, must be in milliseconds not microseconds.

See for yourself:
CPU:
http://www.vipa.de/en/products/system-300s/cpus/?no_cache=1&tx_sbinclude_pi1[page]=317-4NE12.html
DI:
http://www.vipa.de/en/products/system-300s/signal-modules-digital/?no_cache=1&tx_sbinclude_pi1[page]=321-1BH70.html
DO:
http://www.vipa.de/en/products/system-300s/signal-modules-digital/?no_cache=1&tx_sbinclude_pi1[page]=322-1BH70.html

I should say that I haven't used any of these products simply because most people who want a S7 PLC want a Siemens product and not a third party one.

Nick
 
Even if you have so low execution times, how do you make the program update in micro-second area ?
One problem is that there are no timers that will work in microseconds. That is definitely a limitation.
Another problem is that the smallest possible timed interrupt is 500 µs. (edit: 500 µs is for a 319. 317 can go as low as 1 ms).
Running the program freely in OB1 without interrupts will depend of how fast the CPU updates its other "house-keeping", including communication tasks. Dont know how fast a "near-empty" 317 will run, but doubt it will go as low as the required 20-30 µs range.

Jesper,

Please don't missunderstand me. I agree with you that this is a job for a specialist piece of kit and not a general workhorse PLC.

Having looked at the specs, it was very impressive to see what the VIPA systems can do; apparently they have analog I/O that is fast enough to modulate MP3 files without loss of quality.

Nick
 
I can see that maybe there is a possibility with a SPEED7 in stead of a Siemens S7.
But I still doubt it. With propagation delay of 6 µs for the input, plus unknow propagation delay for the output, plus CPU housekeeping time, plus program scan time, it will soon be more than 20-30 µs.
 

Similar Topics

Do i have to use interrupt subroutine, or immediate read high speed input, for Unitronics Samba plc or reading only the correponding register in...
Replies
2
Views
121
Hello Folks, Has anyone configured a Momentum high speed counter on Unity 13.1. We need the wiring diagram for Momentum High speed counter and...
Replies
0
Views
73
Does anyone have any experience with working on piezoelectric motor control? I want to use off the shelf components to hopefully change the...
Replies
5
Views
731
i am bench testing a 1734 -VHSC24 Point I/O High Speed counter module, i cannot find any examples of wiring the outputs from the module. does...
Replies
4
Views
1,368
Hi, I have a setup of one master s7-1200 - three slaves s7-1200 On each of the slaves there is a HSC on a motor for positioning. It generates...
Replies
8
Views
1,069
Back
Top Bottom