Erratic C-Bus Indicator LED's

Hi all,

I have a site that's only about 1 year old. Recently, the buttons have begun being non responsive when pressed in that the lighting load states do not change, even though their respective status LED's track as normal. This issue comes and goes randomly. The C-Bus Indicator on all DB equipment blinks erratically at times and remains ON solidly at others. When ON solid, the switches work and I can scan the network with Toolkit. When blinking intermittently, Toolkit scans usually fail. This behavior is exhibited on all C-Bus switches, whose button indicators may also flash On-Off at times, after being pressed.

I've enabled the software burden and clock in the 5500PACA and 2 more clocks in adjacent, respective 5502DAL's. There are a total of 3 x 5500PS's powering 6 x 5502DAL's, 1 x 5500PACA, 30 x Reflection switches and 6 x 30-mech master switches. There's a power supply at either end of the network and one in the middle. Voltage levels at most places along the network sits at around the 28V level.

I tried replacing one of the Reflection switches the client reported difficulties with and observed all issues disappeared...for about half hour. The C-Bus Indicator blinking seems to come and go with the changing of the wind. Incidentally, I also noticed correct network log messages from the switches which did not affect the relevant 5502DAL back at the DB, to effect the necessary loads.

What makes this issue complex is its intermittent nature. The client had also confirmed that other works had also recently taken place on site, in particular air conditioning system mods that required ceiling tiles to be lifted... Mmmmm!!
I just want to make sure that its not a C-Bus issue before I start blaming others. Can a short or high resistance fault across the C-Bus network wiring cause behavior such as this? I tried a bunch of other tests on site with little to no remedy. Next, I'll begin taking the network apart and work backwards from there.

Any suggestions would be most appreciate and really sorry for the long-winded explanation...

 

The behaviour you describe is resembelent of a powersupply issue..

A quick calculation of what you have is:

3x350ma power supplied = 1050ma

6x32ma for dali gateways = 192ma
1x32ma for pac
30x22ma for reflection switches = 660ma
6x18ma for 30m masters = 108ma
Total consumption = 992ma

Pretty close to the supplied power.. If one of your powersupplies is faulty intermittently that could cause headaches.. Could also be a dry joint or crimp on one of your powersupplies..

If it was me i would add a fourth psu to the system and monitor it to see if that resolves the issues..

 

Thanks Roosta.

Point taken.
However, upon looking at my project again, actual component qty's are as such:
18 x Reflection (@22mA) = 396mA
11 x 30-Mech (@18mA) = 198mA
6 x 5502DAL (@32mA) = 192mA
1 x 5500PACA (@32mA) = 32mA
Total = 818mA

3 x 5500PS (@350mA) = 1,050mA

That's around 232mA headroom, which I would've thought should be ok, provided the power supplies are placed strategically. Sorry about the initial bum steer on qty's, as I was going on memory (ailing at that, as well!!).

Also take you point about dry joints but why would this present after 1 yr of perfectly normal operation?

It seems more like crap being flooded onto the network at times.
I should also mention that when I tried connecting a 5500PC to this network (with clock and burden also enabled) it would also trigger this same type of phenomenon??

 

Hey Roosta, Ashleigh and Grasshopper,

You guys are already on the right track but I thought I would add an interesting situation I had that may hold some bearing.

I had a site I was called to done by another integrator about 2 years before. The site had only about 50ma of headroom when I calculated the required power supplies and this was contributing to a Black & White Touch screen dropping off the network, mostly in the evenings. This was the problem I was there to fix with my 'service call'

I put in a brand new power supply and got called back in the next day, the site had started intermittently not working at all, with many switches flashing at times and generally much slower response to sometimes, none at all. The C-Bus LED indicators in the C-Bus board were all flashing in a crazy fashion (looking like when a re-clocking is happening). It did it every few minutes and the whole network would crash.

The C-Bus diags would go from zero errors per seconds for several minutes and then for 30-40 seconds, you would have 50pps or more of errors.

Replaced new power supply with replacement, all the problems went away immediately. Obviously power supply had an intermittent component.

You might have the same, so try a few new supplies as a quick test if you have them.

All the best,

Brad


PS. This was about a year ago, so your power supply may have been from the same batch as the one I got. Next 3 were fine though.

 

Thanks guys...

All really useful suggestions..

I'll be there next week armed with power supplies and patience.
I'm almost inclined to stick my head in the ceiling space and start looking for any signs of obvious disruption, before doing anything else.

Will keep you all posted on the outcome, thanks once again..

 

Okay, so that's now Day 2 been and gone, since going back to site and applying the useful suggestions provided earlier. I'm cautiously optimistic that the problem's been solved as I've not heard boo from the client. I'll follow up in a few days all the same.

I ended up changing one of the 5500PS's and that seemed to stop the erratic drop-out of C-Bus comms. I measured loaded and unloaded voltage levels at each of the three units on this network. They each sat at around 28V (loaded) and 35V (unloaded). However, the unit in question would fluctuate from a loaded voltage of 27.5 to 30V constantly, almost periodically?? All other units had stable levels. I though this alone was probably a good place to start, so I replaced it with a new unit. Thankfully, all problems ceased immediately. For good measure, I also ran the C-Bus diag. application and ran a reliability test for about an hour. All units had a PERFECT result.

Thanks once again oh wise (and helpful) forum dwellers...