Jump to content
Light-O-Rama Forums

Multiple Network Expansion


Larry Drum

Recommended Posts

At what point in a displays life does it become beneficial to split out to multiple networks?

Our situation this past season (2010) began to show it might be time to split. We ran 24 controllers and 3 ELL's. The topography of the network was setup like this:
485 booster using both out-feeds. One side simply powered 6 wired controllers, the other side fed 8 wired controllers and the main ELL. The ELL side had 2 receivers with 7 and 3 controllers.

During some of the really intense sequences I was noticing some latency at the far end of the network.

This year (2011) we will be adding 3 more controllers to the 7 on the ELL and 2 more to the 3. We may also be adding up to 3 more houses at both ends of the network. This would mean 3 or possibly 4 more ELL's (Unless Dan gets long range ones)

Is it time to split? It seems like it.

Thanks
Larry

Link to comment
Share on other sites

Larry Drum wrote:

During some of the really intense sequences I was noticing some latency at the far end of the network.

This year (2011) we will be adding 3 more controllers to the 7 on the ELL and 2 more to the 3. We may also be adding up to 3 more houses at both ends of the network. This would mean 3 or possibly 4 more ELL's (Unless Dan gets long range ones)

Is it time to split? It seems like it.

Thanks
Larry



Your ELLs were possibly the "weak link" resulting in the latency, because they were limited to a data speed of only 57,600. Adding another network would not necessarily solve that.

In another post Dan did mention they were possibly looking at a new wireless system.

If LOR will offer support for E1.31 that should solve your communications latency. Last year a user in Australia controlled a display on the east coast of the USA using E1.31
Link to comment
Share on other sites

I had 40 controllers on one network this year. I don' think we had any perceptable issues with the network in this case.

I do plan to split it next year, but only because with 40 controllers, it just seems like too big a chance of loosing on average, half of the park to a cable issue. If I run two networks, then a single cable fault will take out less than half, and on average, only 1/4 of things.

Then, depending on funding, we may add two more networks towards the park for some new CCB elements.

So, in theory, you should not need to be splitting things up unless you just have crazy busy sequences, but you could easily put the ELL one one, and the hardwired ones on another if you just wanted to see if it had an impact...

Link to comment
Share on other sites

  • 2 weeks later...

JBullard wrote:

Larry Drum wrote:
If LOR will offer support for E1.31 that should solve your communications latency. Last year a user in Australia controlled a display on the east coast of the USA using E1.31


Do you know the name of the user? I would like to attempt this.

Brian
Link to comment
Share on other sites

Hi

During the past season, I was seeing some channels sticking and suspect a comms issue. I am wondering whether I am pushing the limits of running a single network. I was running some 350+ channels on 29 controllers + a CCR. Mixed controller types CTB16D, 16PC, but mostly DC boards. I was also running a small interactive animation - push a button and get some extra flashing lights when you donate. I am running a single network from a com port using the LOR RS232 to 485 adaptor. All coms are cat 5 cable and I try to separate this from the power and lights cables where possible. Total network length would be about 130 metres.

The first problem I noted was that at a few parts of the sequence, the interactive animation would self trigger. Usually within a particular section of the sequence, but not at identical times, or with identical numbers of triggers, or which of the two interactive channels (or both), would trigger. At some other parts of the sequence, some channels would occasionally stay on (or shimmer) until the next time that channel was addressed - changed. This behaviour happened on different boards both AC and DC, but always clears the next time that channel is activated. The channels are often number 1 on their controller. I tried the 3 different network speed settings and did not really see a difference in the responses. While my problems are annoying, most viewers don't notice, so while the show is running, I decided to leave well alone and investigate after Christmas. My first thought was going to be to split the network for the next season.

Does this behaviour sound like a network limit problem? or could it be just noise? Has anybody else come across similar behaviour? and if so, how did you solve it?

Is there a way of checking the level of network activity? or whether you might be getting close to a network limit? Any other suggestions would be welcome.

Regards Geoff

Link to comment
Share on other sites

Geoff Harvey wrote:

Hi

During the past season, I was seeing some channels sticking and suspect a comms issue. I am wondering whether I am pushing the limits of running a single network. I was running some 350+ channels on 29 controllers + a CCR. Mixed controller types CTB16D, 16PC, but mostly DC boards. I was also running a small interactive animation - push a button and get some extra flashing lights when you donate. I am running a single network from a com port using the LOR RS232 to 485 adaptor. All coms are cat 5 cable and I try to separate this from the power and lights cables where possible. Total network length would be about 130 metres.

The first problem I noted was that at a few parts of the sequence, the interactive animation would self trigger. Usually within a particular section of the sequence, but not at identical times, or with identical numbers of triggers, or which of the two interactive channels (or both), would trigger. At some other parts of the sequence, some channels would occasionally stay on (or shimmer) until the next time that channel was addressed - changed. This behaviour happened on different boards both AC and DC, but always clears the next time that channel is activated. The channels are often number 1 on their controller. I tried the 3 different network speed settings and did not really see a difference in the responses. While my problems are annoying, most viewers don't notice, so while the show is running, I decided to leave well alone and investigate after Christmas. My first thought was going to be to split the network for the next season.

Does this behaviour sound like a network limit problem? or could it be just noise? Has anybody else come across similar behaviour? and if so, how did you solve it?

Is there a way of checking the level of network activity? or whether you might be getting close to a network limit? Any other suggestions would be welcome.

Regards Geoff



Geoff

in 2008 I was having major latency issues. I was running four networks and over 2000 channels. I sequenced Reflections and used some extremely tight timings. It wasn't until 2010 when I had time to get back into LOR that I re-tested my system to figure out where the issues were. Until I rebuild my system, I will not know for sure if the changes I make are successful. Here is what I am planning:

I am putting my neighbors house onto a showtime player. (due to a 150' data cable) - this will also free up a network and will cut out 5 controllers from the show computer

I will no longer be using rs-232 adapters and all networks will be usb adapters.

I will split my idmx units onto 2 networks (I use 7) I originally used one network

I will also be installing terminators at the end of the two yard networks - although LOR says they are not needed, I did see some odd waveforms when placing one of the original networks on a scope

Hopefully these suggestions will help you. Keep us posted as to what you decide to do and your results

Chris
Link to comment
Share on other sites

Geoff Harvey wrote:

The first problem I noted was that at a few parts of the sequence, the interactive animation would self trigger. Usually within a particular section of the sequence, but not at identical times, or with identical numbers of triggers, or which of the two interactive channels (or both), would trigger. At some other parts of the sequence, some channels would occasionally stay on (or shimmer) until the next time that channel was addressed - changed. This behaviour happened on different boards both AC and DC, but always clears the next time that channel is activated. The channels are often number 1 on their controller. I tried the 3 different network speed settings and did not really see a difference in the responses. While my problems are annoying, most viewers don't notice, so while the show is running, I decided to leave well alone and investigate after Christmas. My first thought was going to be to split the network for the next season.

The fact that the interactive triggers are tripping sounds like a network noise issue to me. Ensure you don't have any deviation from the single daisy chain of controllers.

Splitting into two networks should either divide your total noise across the two networks that might allow everything to run error free, or isolate the issues to one of the two networks. You are up towards the region to where it makes sense to separate things, so it is not like you would be doing it for no reason but diagnostics.

Also remember that having clean contacts on all the controllers and data cables is important. There have been more than a few reports of a corroded cable end being the source of all network issues in a display.
Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...