ESR Communication Failure

[engineerdave]

Nice find.
Think I'd still be looking for loose connections.
I'd still be banging my head.

[Phase Loss]

Thanks for the explanation! Definately something I need to get into more and read about, I appreciate your response.

http://www.embeddedsys.com/subpages/..._art_RS485.pdf

Here is an explanation of RS-485 networks that's real technical. Page 2-3 talk about the importance of termination. I read it and kinda get it, but lots of computer geek stuff going on in the article that I don't get.

To keep it simple all we need to know in the Refer field is that the network should be terminated...not so much why it should

[neumonic]

Thanks for the links guys.

Phaseloss I think I like your explanation better then that article haha I tried to follow it but lots of computer lingo in there , im gonna try and re-re-read it.

[markettech]

Thanks for the links guys.

Phaseloss I think I like your explanation better then that article haha I tried to follow it but lots of computer lingo in there , im gonna try and re-re-read it.

If you're anything like me, you'll read it 4 or 5 times and not get it.

Then, at 0200 some morning you'll jolt up out of bed and be amazed because somehow it suddenly makes sense.

I think its called an epiphany....

[neumonic]

If you're anything like me, you'll read it 4 or 5 times and not get it.

Then, at 0200 some morning you'll jolt up out of bed and be amazed because somehow it suddenly makes sense.

I think its called an epiphany....

Haha I know what you mean....I do believe I had me one of them epiphanies last night , but on a totally unrelated subject

[jeremyjp]

I'm pulling my hair out on this one - looking for advice/input.

Here's the story (its a little long winded) -

Got called out to a store today on an E2 alarm - 'ESR Board #4 Device Offline'. The board went offline at approximately 1035 and returned to normal several minutes later. It went offline again twice more - all 3 failures occurring within 20 minutes of one another - and all returning to normal.

Of course, when I arrive later in the day - there are no active alarms, and the device hadn't missed a beat since the hiccup earlier in the day.

In poking around the controller, I see this particular board has been a problem child for quite some time with multiple call outs for this issue dating back to January 2011 - and as indicated by the maintenance log, has been changed 3 times since April of this year by the previous servicing contractor.....the most recent being September 18th.

For point of reference, the controller is mounted in a back room, with a couple Hill/Phoenix pump houses sitting on the roof (low and medium temp). The LT system has a few MultiFlex boards, including 3 ESR's while the MT system also has a few MultiFlex, including 2 ESR's.

I checked the daisy chain wires/connections obviously - also found and repaired a loose spade connector on the secondary side of #4's transformer.

I tightened EMS electrical connections on rack, as well as electrical terminal blocks which feed power to fuse blocks/transformers....several connections were less than snug.

The 460V supply wiring to the rack had been wire tied to a bundle of shielded cable and I noticed a few areas where high voltage wiring was running parallel with network cabling in the wire chases. Both of these issues were taken care of.

I pulled, poked, prodded, and otherwise molested the wiring on the entire EMS side of the rack - but the entire time I'm on site, the friggen thing doesn't miss a beat - voltage looks good the entire time and she's communicating like a newlywed couple for the duration....so I pack out, do invoicing - and I'm outta there.

I no sooner walk in the door at home, and monitoring calls me saying board #4 went offline again! (How they got a hold of my work number is also puzzling, but I'm focusing on one problem at a time right now....)

#4 lost communication twice about half hour after I left - each time coming back online within a matter of a few minutes. It went offline a third time about 5 minutes before I got back to the store........and reestablished communication 2 minutes before I got to the controller.

It's now dark on the roof and raining like a sumb!tch....and I'm back for a second time on a friggen board that refuses to cooperate with me and fail while I'm present.

Scraping the bottom of the barrel of tricks, I double check baud rate and board address on all MultiFlex boards in both racks.

I was concerned that maybe the factory installed daisy chain terminal points at the network cable plug could present a problem if there was a hairline crack at the top of the male connector or a loose/broken wire........so I cut them out and landed the daisy chain wiring directly into the network cable plug.

An hour on site and no comm loss - no nuthin. Once again I type up a narrative and get signed out. Manager leaves (as by now the store has been closed for a couple hours) and I decide to look through my E2 manual to see what I may have overlooked.....which burned up another half hour.

By the time I call monitoring to check out, the time is a touch after 0100. As I give the store # to the monitor-er......he tells me #4 went off line again at 0052.

WTF!!!

I need to be back there by 0600 and I'm not fuggen leaving until this baby is put to bed.

Like I said at the beginning, if you've got any thoughts on this ('this' being the communication failure.....not the technician's lack of ability ) I'd be glad to hear them.

Nice descriptive writing skills!!! was able to follow the whole way

[Blackmira]

Is it the bottom board on a weathepack? Water damage will produce similar results as I have found.....

[markettech]

Is it the bottom board on a weathepack? Water damage will produce similar results as I have found.....

Nope, it was the last board in the daisy chain - in the condenser electrical panel.....not terminated.

[Blackmira]

jumpers up?

[markettech]

jumpers up?

The terminating resistance jumpers were actually set to the 'in' position instead of the 'out' position on the MultiFlex board - or, as mentioned previously.....not terminated.

[mseng]

CPC boards do not need to be terminated when they are first installed; however, they should be. The low voltage communication will work until it gets some age on it. I have seen them work for a couple of years and then start failing.