adccommunitymod (AutomationDirect) asked a question.
Created Date: January 12,2001
Created By: sid k
**** This post has been imported from our legacy forum. Information in this post may be outdated and links contained in the post may no longer work.****
I am about to bid on a control system re-design that currently uses a bunch of analog circuits. I would like to do as much as possible in PLC with Analog I/O modules. I ran a test to see how fast I could read an A/D channel, mask it (12 bits), and send it to a D/A module. Analog input was from a Function generator and both the Fn gen input and the D/A output were viewed on a dual-channel scope. The only code in the ladder was A/D and D/A initialization (pointer method) on Scan ONE (SP0) and one line of SP1 to read the A/D, mask it, and sent it to the D/A (then of course an END line). I was dismayed to see a 50 mS delay from in-to-out!! This seems MUCH slower than I expected. DirectSOFT reports the PLC scan time to be min 7, max 10 mS. I am confident in my measurement techniques and I am familiar with pointer addressing the Analog modules. Any words of wisdom out there. Thanks, Sid DL250, F2-08AD-2, F2-02DA-2
Created Date: January 12,2001
Created by: sid k
I am about to bid on a control system re-design that currently uses a bunch of analog circuits. I would like to do as much as possible in PLC with Analog I/O modules. I ran a test to see how fast I could read an A/D channel, mask it (12 bits), and send it to a D/A module. Analog input was from a Function generator and both the Fn gen input and the D/A output were viewed on a dual-channel scope. The only code in the ladder was A/D and D/A initialization (pointer method) on Scan ONE (SP0) and one line of SP1 to read the A/D, mask it, and sent it to the D/A (then of course an END line). I was dismayed to see a 50 mS delay from in-to-out!! This seems MUCH slower than I expected. DirectSOFT reports the PLC scan time to be min 7, max 10 mS. I am confident in my measurement techniques and I am familiar with pointer addressing the Analog modules. Any words of wisdom out there.
Thanks,
Sid
DL250, F2-08AD-2, F2-02DA-2
Created Date: January 12,2001
Created by: Russ
Hi Sid,
I haven't used the F2-08AD-2 module myself, so I might be completely wrong about this, but have you read the General Specifications section on the module? It says it requires 3/ms per channel ( asynchronous ) Data Acquisition Time. Also, on page 5-7 you can see the diagram where an analog mux is servicing the channels through a single A/D converter. On page 5-9 the discussion talks about the timing of the module and related filter settling time in relation to how fast the data can be converted. Again, I might be completely wrong, but it looks like to get the fastest performance you need to disable as many unused channels as possible ( via the jumpers on page 5-5 ).
Regards,
Russ O 'Rourke
Created Date: January 13,2001
Created by: sid k
Good evening Russ,
I removed the jumpers to enable one channel only and the response time dropped to about 8 mS. Two channels showed about 15 mS...
Bummer. Looks like I'll be using more op amps than I had wanted. At least I can replace the relays and do some of the slower analog with the PLC. Thanks for the suggestion.
Sid
Created Date: January 13,2001
Created by: Tom Jenkins
Not to sound like a smart aleck, but can the final control device (valve, VFD, whatever) really respond faster than 50 ms?
Created Date: January 13,2001
Created by: sid k
--------------------------------------------------------------------------------
Tom,
Well, no, not the "final control device ". However, there are analog linear position sensors (position-to-volts (0 to +10))that are monitoring various machine positions and 1) decisions being made (currently by analog comparators)that cause control devices to actuate... these would be ok with 50 mS response times. and 2) two of the dc position signals are inputs to a "fast " (about 15 Hz bandwidth) hydraulic servo loop and I am concerned that the 50 mS delay would began to adversly affect loop stability.
Some of the analog inputs are completely unconcerned about delay...these are from operator inputs from pots. Upon thinking about the test I ran after Russ's suggestion, one possibility I see is rather than doing it all with one 8-channel A/D, use two A/Ds... one for the fast stuff and one for the slow stuff. Also, maybe using one of the available $400 touch-panels to do operator inputs. Nicer than pots but with a learning curve.
Created Date: January 22,2001
Created by: RHanrahan
Sid,
I am using the TouchPanel to replace all the pots I had on my door. Works great. I am also sending an Analog signal out to a vibrator control card. I haven't noticed that it went to slow. I have been using it to handle 2 speed control for the vibrators instead of sending a 24vdc to each of the 2 inputs on the vibrator card. The pots used to be hard wired to the cards. Made my wiring much simpler too.
Created Date: January 22,2001
Created by: sid k
"R "
Which TouchPanel are you using? Which PLC?
How difficult was the TouchPanel programming?
Thanks,
Sid
Created Date: January 22,2001
Created by: RHanrahan
Sid,
I have been using the DirectTouch touch panel but am thinking of switching to the EZTouch. It looks like a better product. With the DirectTouch, I had to write a bunch of custom buttons. It doesn't handle screen transitions well. The EZTouch appears to do more also. I have been using the DL-240 and DL-250 with the TouchPanel. The DirectTouch Panel forced me to write the screen data thru the PLC program to be able to do the things I wanted(using the same help screen for multiple screens, for example.) If you haven't any specific questions, I would be glad to help.
Bob Hanrahan