Axcent3 controller time lags

trx250r87

Every AXT-CA10 and Axcent3 controller setup that I have loses time. After about a month, each clock is behind a few minutes. Does anyone know how to stop this?
ERIC

John Nagy

Some things respond best to brute force. Years ago I wrote a routine that read the time periodically and added a calculated amount based on the hours of use (it was disk I/O that interrupted the clock) and reset it. It wasn't for an AMX, but it reduced the annual error to a couple minutes total.

Obviously, newer systems (NetLinx) has built in time services to get the atomic clock time. Maybe not available on an older system.

true

Easiest thing to do would be to do this.

Another thing you can try is replacing crystals until you get something closer to ideal.

DHawthorne

After visiting clients whose old G3 panels were showing times that were hours off, and the client never complained, I came to the conclusion that although I thought the time display on the panel was cool, no one else paid any attention to it. I have seen similar things with the built-in clock on telephone systems, microwaves, you name it. So I would first ask myself, "is this an important clock that the customer actually uses?" If the answer is no, just get rid of it. It's not like there is a shortage of digital clocks in the world. I can see four from where I am sitting at the moment ...

trx250r87

These are used in classrooms where there is only 1 analog clock on the wall. Our systems are programmed to turn off at specific times in the evening to eliminate wasted projector lamp hours when professors don't shut the system down.
ERIC

annuello

trx250r87 wrote: »

These are used in classrooms where there is only 1 analog clock on the wall. Our systems are programmed to turn off at specific times in the evening to eliminate wasted projector lamp hours when professors don't shut the system down.
ERIC

I know the feeling. If projectors were cars some people would leave it in the driveway with the engine running. One way around the issue would be to install PIR motion sensors in the room. If the system is running an there is no movement for 90 minutes (or whatever you feel is appropriate), shut down the system. If there is movement during the countdown, reset the countdown to 90 minutes. We do this in all our rooms and it works pretty well. Just don't use the cheapest PIRs available, where the sensitivity calibration may drift over time. (We use dual spectrum IR/Microwave types.)

If you really want to pursue time-based shutdown you will probably have to look at some form of network connection and NTP. The easiest way to do this is to migrate to NetLinx controllers which can be time-synced to an NTP server. We do midnight shutdowns in addition to our PIR shutdown. Also, if a projector is left "blanked/VideoMute" for 30 minutes it is also turned off.

Probably the best supporting argument for migrating to NetLinx masters would be remote lamp/error reporting. Rather than getting end-user reports of "It had some error" (or no report from them at all), you can log what is actually happening. If your projector has a rich set of errors you can pin-point the actual problem. (I like the range of Epson errors.) The lamp reporting also allows you to perform preventative maintenance since you can order lamps before the current one dies.

A last-resort option for your clock-drift issue would be (assuming that you have a spare serial port on the Axcent controller): send a clock string to the Axcent from a 3rd party device and parse the string in your Axcent. Once parsed, adjust the clock:

send_command 0,"'CLOCK ',<your time info, correctly formatted>"

If the batteries haven't been changed in the last 5 years this would be contributing to clock-drift.

Roger McLean
Swinburne University