1756-DHRIO slow analog MSG/block transfer

davej

Member
Join Date
Dec 2004
Location
Milwaukee, WI
Posts
9
Hi All,

First time caller, long time listener...

Have a new system in the field with a new controllogix 1756-DHRIO talking to 6 existing chassis of SLC I/O (1 ASB and 3 chassis on each DHRIO channel). Have about 12 slots of analog I/O (NI4 and NO4I) on the remote chassis, so I'm using MSGs to read/write.

Mostly things work ok, but occasionally we are getting LONG delays on the analog reads and writes. Eg. we clear the analog output write value and it takes 10-15sec for the mA at the card to clear. By "occasionally" I mean we actually notice it several times an hour, but it may be more often than that.

Is this 'normal'? I knew RIO was slower than local but 15sec seems excessive.

I don't think there's any errors in the MSG programming but I'm not ruling that out cuz I don't use them much, I've attached a screenshot.

No obvious faults or errors on the card during normal operation, HOWEVER, the DHRIO has twice in the last week stopped responding with an "850C" error code. Had to cycle power to clear. Our AB rep said this could be either excessive noise (which I think we've ruled out), or a card failure. Maybe we just need to try a different DHRIO?

Any feedback greatly appreciated, this project is making me seriously consider a change in career :-(

--- Dave

Capture.jpg
 
The obvious thing that comes to mind is that you are firing these MSG blocks off continuously - i.e. as soon as they are DN re-start them. Looks like you are swamping the message buffer.

I would put them on a timer, or better still, sequence them so you only have one or two started at any one time. You could put the MSG instructions inside a periodic task that has an incrementing scan counter, and fire each message on successive scans (count = x). There is very little point in trying to read the analog cards any faster than the card's RTS (Real-Time sample) rate, you will just be reading the same data repeatedly.
 
Large numbers of block transfers are a challenge in the 1756-DHRIO. You're under the traditional limit of sixteen analog modules, after which I usually recommend the 1756-RIO scanner instead.

The logic that you show isn't going to work. Running a bunch of block transfers using auto-retriggering logic in parallel like that is like running the 440 meter hurdles. It's elegance in motion while it's working but when a runner trips a hurdle the whole thing gets messy in a hurry.

I don't know how many block transfers the 1747-ASB can process simultaneously; I'm going to guess it's just one.

I would separate out the MSG logic per 1747-ASB adapter and program them so that they go in sequence, rather than in parallel. Add some logic to measure the time between the /EN and /DN bits for each Message.

I prefer to manually force my timeouts on a system like this; the default 30 seconds is far longer than any sort of error/retry could account for. I just use a timer; if the /EN has been true for 2 seconds then I set the /TO bit.
 
I lied about the number of analog cards, I actually have 18. I'm hoping the "traditional" limit of 16 analog modules that Ken mentioned is not hard and fast.

Found a Rockwell reference doc that says on p.19 that a 1747-ASB "handles one block transfer request per logical rack at a time." That helps a bit, I can do 3 at a time with my HW setup.
http://samplecode.rockwellautomation.com/idc/groups/literature/documents/rm/1785-rm010_-en-p.pdf

Here's a snippet of what I came up with. I'll try it on real hardware on Monday, unless someone pokes some holes in it before then.

Wish RA would hurry it up with the SLC Ethernet/IP card that's supposedly in the works!

--- Dave

Capture.jpg
 
That "traditional limit" was based on the Cached Connections limit of the controller via DHRIO modules on a single channel, since that was the most common approach for retrofitting PLC-5 systems.

ControlLogix CPUs can handle 32 cached Message connections of any sort. The 1756-DHRIO can handle 16 cached Message connections on a RIO channel, or all 32 on a DH+ channel.

RA Knowledgebase Article ID # 41853 (TechConnect) has a concise summary of these limits.

You can do as many non-Cached connections as you want, as long as you're willing to take the penalty on backplane buffers and on speed.

Just shooting from the hip, I wouldn't try to do this with an AOI. Yes, it's elegant, but Message instructions are really the sort of thing that you want to be able to troubleshoot individually.

If you have other Messaging that the controller needs to do, I'd go ahead and switch to un-cached RIO block transfers only. If the RIO block transfers are the entirety of your messaging, then choose two or more that you're going to make un-cached.

I personally think the 1747-AENT is a little like Bigfoot. I'm curious about whether it will be sighted in Chicago next month.
 
Once upon a time in dark laboratory on a rainy night. There was a sound at the window a engineer turned and caught a glimpse of a module the mythical 1747-AENT.
He told others about it no one would believe him. He has set out on a mission to find this mystical, mythical creature. Rumor has that this engineer will have an unavailing in November at the McCormick Place west.
 
Someone sent me a Rockwell Powerpoint which shows this module. It requires RSLogix 5000 version 20 though.
 

Similar Topics

We have a remove PLC rack that is being used to collect data from older equipment via a 1756-DHRIO module. This module occasionally faults out...
Replies
1
Views
391
I am trying to send some datas from PLC-5 to control logix plc through 1756-DHRIO. When I go online to PLC-5, Message instruction gives error &...
Replies
15
Views
3,297
:banghead: Remote I/O The occasional times i have had to deal with blue hose/PLC5 I/O addressing it always wrecks my head. Now i am working on...
Replies
9
Views
3,245
Called to a site and am neck deep currently ;) running, but issues. Phase 2 completion has been running for almost 2 years no issues. Controller...
Replies
3
Views
1,615
Hello all. I have a job to add a 1794-IE8 Card in slot 5 on an existing Flex IO Rack which uses a 1794-ASB communications module, which...
Replies
8
Views
2,956
Back
Top Bottom