NI3000 with five dead serial ports but two work
Hedberg
Posts: 671
in AMX Hardware
Just informed that an installation from 2005 with an NI3000 was having problems. Looks like serial ports 1 - 5 are dead but 6 and 7 are working fine. I suspect that this is the "famous" capacitor issue that has been discussed several times (I've read those threads and thanks to all contributors for lots of good info) but am wondering about the fact that there are still two ports working. Can anybody verify that the bad capacitor issue kills all seven ports? I know that we have a number of these installed with four or five port in use and if 1-5 failed we would assume that all seven had failed -- that's why I'm asking.
In any event, we're going to re-install current master and device firmware which will probably just waste an hour and then we're going to replace it and ship it back to AMX as I understand they may repair it even if out of warranty.
Thanks for any thoughts
edit:
This is the first time we've seen this issue (probably the capacitor). Suppose that operating continuously for 5 1/2 years without issue isn't too bad. This is the earliest installation that we did with a NI3000. Had a WAP200G of the same vintage fail there just about a month ago.
In any event, we're going to re-install current master and device firmware which will probably just waste an hour and then we're going to replace it and ship it back to AMX as I understand they may repair it even if out of warranty.
Thanks for any thoughts
edit:
This is the first time we've seen this issue (probably the capacitor). Suppose that operating continuously for 5 1/2 years without issue isn't too bad. This is the earliest installation that we did with a NI3000. Had a WAP200G of the same vintage fail there just about a month ago.
0
Comments
Just sayin.
So, for the ports that are still working for you, what mode are they running in?
Roger McLean
Swinburne University
Maybe some devices will work with the screwed up ports and others won't. In the case we are looking at, the two working ports are connected to Sony Evid-70 cameras. Maybe other devices wouldn't work.
Edit:
Oh, Oh, Oh! the Evid 70 cameras are both operating RS422, not RS232. I think that makes sense why those are working and the other five (all RS232) are not. Didn't occur to me until I re-read the messages in this thread as there is no difference in the program between RS232 and RS422 -- it's all in the wiring.
Roger McLean
Swinburne University
Well, I'm the guy with the keyboard. The guys with the screwdrivers and soldering irons are going to have to decide what to do with the master. I do know that they have the technical ability to replace a capacitor, but given the way this whole thing sorts out, my guess is a trip back to AMX for repair rather than a local repair. If it were my master, I'd look at it, but they'd go bonkers if I started taking a master apart with a soldering iron in my hands. I'm sure that AMX would rather have their techs be the ones to do the repair, also. Probably about 10 minutes total in Dallas.
As for RS422 versus RS232, it seems like a mighty big coincidence that the only two ports working are the two configured for RS422. It seems very likely that the capacitor in question is the problem and that it only affects RS232.
edit: speaking of NI3000 masters, there have been several sell for scary low prices on Ebay within the last couple weeks. NI3100 masters seem to draw a much higher price. Perhaps the NI3000s are getting a bad reputation. Kind of like the Cre*stron 10" touchpanels that you can no longer get batteries for, the ones that are adapted from a Viewsonic panel of some sort. Beautiful devices, when new and properly configured, but selling for just about nothing these days--the damned things aren't worth crap if you can't get a replacement battery.
A safe and quick way to check would be to swap a few ports around in your program and get the local techs so swap the associated cables. (Dump the modified program to the master without rebooting, then have the tech change cables and power-cycle to load the new program.) If your RS422 camera works on what was previously a "problem port", and your other RS232 device still doesn't work on the known-working ex-Camera port, you know that the issue is the capacitor for RS232.
Your clients, your call. I'm just the kind of person who likes to know exactly why things don't work.
Roger McLean
Swinburne University
We're probably going out there tomorrow -- maybe the next day. First thing we're going to do is to reload firmware for both the master and the 5001 device though I have very little confidence that this will have any effect -- it's what the boss wants to try first. Then, because it won't take much time at all, we'll try putting the cameras on a couple other ports. No re-soldering or anything will be required, just reposition the DB9 connectors and change the device definitions in the program. Though I'm not normally allowed to operate a screwdriver myself, the guys who do operate the screwdrivers have respect for my opinions on such matters and normally don't mind messing things about at my suggestion-- within reason, of course.
Not completely true, I had a NI-4100 fail on some comm ports and it was only a cap. I believe it may be the device that is attached to the comm port and the load it places on it that is a factor. Oddly enough when I removed the load and shorted the comm port pins 2-3 the RMS issue would disappear.
Take note that the newest device firmware requires master firmware 3.5+ (which is duet firmware).
If you want to stay with a non-duet firmware on the job, DO NOT update the devices to 1.20.xx. I think 1.13.7 is the newest non-duet device firmware.
I made that mistake yesterday and have not been able to get the devices to downgrade to 1.13, so I am stuck with duet firmware on the 3000 when I get it back from AMX with a new capacitor.
Yes, I noticed that the available device firmware requires the Duet master firmware so we're going to install the Duet firmware (I understand that it has to be installed twice to be successful) and then the device firmware. That's the plan. Given a choice, I'd rather not go with the Duet firmware, but we've been installing masters with Duet firmware all over the place and haven't had any difficulty so I'm not anticipating any from this update.
1. It's the well-known capacitor problem
and
2. The capacitor problem prevents RS232 but not RS422.
The master is headed back to AMX for repair.
Thanks to all who commented on this -- I feel like I've learned something. New experiences, as they say.
Harold
Hi,
I've just had word of 4 controllers all with the RS232 port failure.
Could be the capacitor issue but could a firmware upgrade have caused this? If so is rolling back the firmware an option?
Currently the master is on v3.50.439
A colleague has recently updated firmware in this area. He is not contactable at the moment so I'm looking for some assistance.
Meanwhile, in one of the rooms, I'm going to try moving some devices to other available ports.
Cheers
John D
The firmware version that you supplied is for the master (CPU), not the NI device (serial ports/relays/etc). I've seen no issues with master firmware versions.
I've found one "fatal" issue when updating Device firmware from versions prior to v1.20.7, but it is dependant on the type of hardware that you have plugged into the serial ports. If the external hardware is chatty (i.e. It generates it's own RS232/422/485 traffic) this can cause the internal I2C bus to lock up and fail during a firmware update. The master is trying to process the data coming into the system while also trying to update 4 or so ICs over the internal bus. End result is a saturated internal bus and failed firmware update, requiring the hardware to be sent back to AMX for repair. In my case it was a lighting system that pumped out 13 bytes every 0.7 seconds. I expect some video-conference systems would cause the same issue.
AMX have reworked the device firmware as of v1.20.7 to prevent external traffic from saturating the bus during firmware update. However, you still need to unplug the chatty devices if you are upgrading from an earlier version to that one. Once you've got v1.20.7 you don't need to unplug for future upgrades. I've tested v1.20.7 with the same problem lighting system (by sending fw v1.20.7 again) and it seems to work fine.
If none of that seems to apply, it could be the capacitor issue. Or even better - both! :-0
Roger McLean
Swinburne University
Hi Roger,
Appreciate your feedback. It is very helpful.
The issue is now resolved independently of the factors I initially thought were involved.
The systems that had the problem had been programmed with the old 1:1:0, 1:2:0 etc. device addressing protocol for the RS232 ports. Somehow the firmware update had made this unusable.
Once I changed this to 5001:1:0, 5001:2:0, the problem was resolved.
Thanks again
John D
From that information, it would seem that if the same device does work on one port and not another (and it's not really just random edge-condition intermittent everywhere), it's not likely the capacitor, but a port issue of some other kind.
These old units are really debatable as to being worth repair. Consider your hourly rate before tinkering endlessly...
The tx device is creating the carrier voltage, and the failed cap drops the Amx tx voltage. The connected device still produces the correct carrier voltage and so the Amx rx side still works
What's wierd in my case is the port (#1) does work on tx but not on rx. Have an Autopatch Precis 18x18 connected to it. Commands to switch and volumes work, but no feedback at all...
Is it good to be sent back for repair?
And by the way, in our case it appears the serial programming port isn't working when this happens. I'm looking at 2 NI-4100s from different jobs with the same symptoms.