Control/Dimming ETC Idea48 desk, Colortran D192 dimmers

[derekleffew]

... or possibly bad grounding in the adapters. ...

There's often lots of confusion and some controversy/debate over the terms: ground, common, shield, electrical ground, chassis ground, potential, etc. in reference to signal or data wiring.

To illustrate this point, some may want to check out this current thread DMX Grounding on LightNetwork.

[USER]adickerson[/USER], as [USER]dramatech[/USER] said above, as long as all of your cables and adapters check out with pin to pin continuity on each end, no continuity between pins on the same connector, and nothing connected to the shell (EVER!), you can rule out wiring as the source of your problem.

 

[adickerson]

It's fixed.

I'm not sure whether I should laugh or cry, and quite frankly I'm a bit embarassed that solution was so simple...

There is a set of dipswitches on the back of this board, which the previous owner for some reason went to great lengths to hide (it was camouflaged with a small piece of plastic glued over it, of course the EXACT same color as the casing). Dipswitch 3 chooses whether you are operating up to 256 dimmers or 256 to 512 dimmers. A simple flip of this switch fixed my problem - we only have 48 dimmers and it comes from the factory with the default set to 256 to 512.

While this doesn't explain the problems that were occurring between the Elation board and the dimmers, but at this point I'm just happy to have the ETC working.

Thanks, all, for your great insights and ideas. I learned much from this experience, and certainly got to brush up on my soldering skills.

 

[techno89]

It's fixed.

I'm not sure whether I should laugh or cry, and quite frankly I'm a bit embarassed that solution was so simple...

There is a set of dipswitches on the back of this board, which the previous owner for some reason went to great lengths to hide (it was camouflaged with a small piece of plastic glued over it, of course the EXACT same color as the casing). Dipswitch 3 chooses whether you are operating up to 256 dimmers or 256 to 512 dimmers. A simple flip of this switch fixed my problem - we only have 48 dimmers and it comes from the factory with the default set to 256 to 512.

While this doesn't explain the problems that were occurring between the Elation board and the dimmers, but at this point I'm just happy to have the ETC working.

Thanks, all, for your great insights and ideas. I learned much from this experience, and certainly got to brush up on my soldering skills.

Thats not silly and if it was glued over it really isn't obvious. It's weird that the other person did that and I guess I could see why but it was odd to say the least.

 

[derekleffew]

...Dipswitch 3 chooses whether you are operating up to 256 dimmers or 256 to 512 dimmers. A simple flip of this switch fixed my problem - we only have 48 dimmers and it comes from the factory with the default set to 256 to 512. ...

Glad your problem is solved.

ETC folk (yes, looking at you, DavidNorth and STEVETERRY, and others):
Now I'm curious as to what that switch actually does to the DMX output, and why the dimmers care what position it's in. DMX is DMX, right? Transmission of 512 values at 44 times per second?

Along the same lines, I've never received a good answer to another question:
Say on an Express, if one changes the # of dimmers from the maximum of 1024, does anything happen to the DMX? Faster/slower? Does the console's processor run better with fewer dimmers and thus fewer calculations to make? I've always been pretty anal about setting to the exact number of dimmers I had, but does it really matter?

 

[dramatech]

DMX is DMX, right? Transmission of 512 values at 44 times per second?

Derek, I am not sure this helps answer your question about the DMX output, but a Lightronics TL5024 on the optional output only puts out 256 channels of DMX. I have a DMX tester with a digital readout, that verifies the number of channels and the transmission speed. I am pretty sure that when any of my other boards have an adjustable dimmer output, that are set for less than 512, that the DMX output is just what has been set, and not 512.
I have confirmed this on a Lightronics TL3256 and a Colortran innovator. I just got the tester, and haven't thought to test the ouputs of the other boards that I own, or have access to. My curiosity will damand that I do that ASAP.

Tom Johnson

 

[mstaylor]

So are you saying the number of channels drop but the speed stays the same? I have never tested it but it seems that whether you are running a small amount of dimmers with a reduced packet size or a small amount of dimmers with a full 512 packet, the speed stays the same, although it affect the signal strength.

 

[dramatech]

So are you saying the number of channels drop but the speed stays the same? I have never tested it but it seems that whether you are running a small amount of dimmers with a reduced packet size or a small amount of dimmers with a full 512 packet, the speed stays the same, although it affect the signal strength.

Excellent questions, that I am not prepared to answer at this time. I have had my tester for such a short time, and didn't really use it with the goal of finding out the processing speed. I do know that I have seen different speeds indicated, but didn't relate to channel count. I am in the middle of a very heavy hang at the moment, but as soon as an opportunity arrives, I will run it on the 6 boards that I own and on a couple that I have access to. Will get back to you.

Tom Johnson

 

[mstaylor]

Sounds great, it's one of those things that would be interesting to know. It might make a good dianogtics tool in the future.

 

[starksk]

... ETC folk (yes, looking at you, DavidNorth and STEVETERRY, and others):
Now I'm curious as to what that switch actually does to the DMX output, and why the dimmers care what position it's in. DMX is DMX, right? Transmission of 512 values at 44 times per second?

DMX is not always the same DMX, especially when it comes to timings.

To answer your first question, the switch (dipswitch 3) controls whether the console transmits 512 or 256 bytes in the DMX packet. It defaults in the down position which sets transmits all 512 DMX values. It can be changed to transmit only the first 256 levels by setting the dipswitch in the up position. (For future reference, dipswitch settings and explainations can be found on page 125 of the user manual.)

When that dipswitch is changed, besides the packet length, the other timing effect that is changed is the Break to Break timing (time between the start of one packet and the start of the next packet, not to be confused with the Interpacket time which is the time from the end of one packet to the start of the next packet). Because the packet is smaller, it is understandable that the Break to Break timing is just under 10 milliseconds faster in the setting that only trasmits 256 levels. Neither the Mark After Break (MAB) time nor Break Length time is affected.

...
Along the same lines, I've never received a good answer to another question:
Say on an Express, if one changes the # of dimmers from the maximum of 1024, does anything happen to the DMX? Faster/slower? Does the console's processor run better with fewer dimmers and thus fewer calculations to make? I've always been pretty anal about setting to the exact number of dimmers I had, but does it really matter?

The reason you haven't received a great (possibly even good) answer to this question is that it depends on a number of factors. The major factor being which software version you are running. The next major factor is what speed you have selected in the output configuration.

For version 3.1, there is the possibility for a slight difference in the Break to Break timing when running in Max or Fast speed settings. When I say slight, I am talking 2-4 milliseconds maximum. The reason for that possibility lies in the fact that it is a software driven speed so interbyte and interpacket timings can vary slightly depending on what the software is doing.

In the Medium and Slow speed settings, there is no difference whatsoever in the output speed based on the number of dimmers defined in the setup page. This is related to the fact that the software has more time to process its internal functions.

In all speeds, the packet length stays at 512 bytes regardless of the number of dimmers defined on the setup screen.

As for whether the console will run better/faster with a smaller number of dimmers defined, while technically yes, the difference is so minute that the performance gain is insignificant and will not be noticeable by humans.

For more detailed information about DMX Speed and ETC consoles, please check out our wiki article.

 

[mstaylor]

Good info, thanks.

 

[derekleffew]

Thanks, Kirk.

So if I brought three 96 racks into adickerson's space, and flipped the Idea's dipswitch#3 switch DOWN to the 257-512 position, would the 48 house dimmers work?

 

[SteveB]

Thanks, Kirk.

So if I brought three 96 racks into adickerson's space, and flipped the Idea's dipswitch#3 switch DOWN to the 257-512 position, would the 48 house dimmers work?

My read on the manual and Kirks post is that dipswitch 3 UP limits output to 256 addresses, while switch DOWN sets it to full 512 output, not turning off data for addresses 1-256.

 

[starksk]

SteveB is correct. Up, it should output 1-256 only; Down, it should output 1-512.

 

[derekleffew]

Steve and Kirk, I fear you have misunderstood me. I was merely quoting from the manual:

3 Down 257-512 dimmers
...Up ....Up to 256 dimmers

I never meant to imply that the console would send out only values for dimmers 257-512.

But I still don't understand how adding 209 (or more up to 464) dimmers to what's existing would magically make the house dimmers function properly.

 

[SteveB]

Steve and Kirk, I fear you have misunderstood me. I was merely quoting from the manual:
I never meant to imply that the console would send out only values for dimmers 257-512.

But I still don't understand how adding 209 (or more up to 464) dimmers to what's existing would magically make the house dimmers function properly.

I'm guessing the answer is in this statement from Kirk "When that dipswitch is changed, besides the packet length, the other timing effect that is changed is the Break to Break timing"

Thus I'm assuming that the conversion that is done to the Colortran D192 control modules, from what was originally CMX and AMX, to allow for DMX in place of CMX, narrows the window that the D192 pack listens to, with possibly some data getting lost, but not all data ?, and for whatever reason, the D192's sees the Break and MAB, thus know to start to listen to valid DMX, but with a truncated set of following bytes, things get scrambled. Remember as well that the on-board processing for the D192 Control Module was designed for original D192, which was 192 dimmers/addresses. The conversion chip and resistor configures the pack for 512 addresses of DMX, not 256.

In truth, I don't know enough of how the D192 packs works to know what's going on, and cannot contact Steve Short to ask. Dave North might know.

 

[derekleffew]

...Remember as well that the on-board processing for the D192 Control Module was designed for original D192, which was 192 dimmers/addresses. ...

I'm pretty certain D192/Colortran protocol/CMX was always 512 dimmers, according to a 1983 Patchman brochure. The "192" in D192 refers to up to 192 dimmers in a single rack, unlike the measly 96 in the competition's (and later bested by the Teatronics MD 288).

Also, the picture I posted above (from an eBay listing) shows a "Universal Control Module" from the factory with AMX and DMX, not a retrofitted/upgraded version of the original that you and I both love. As I recall, the original Control Module for packs had only a single thumbwheel and addresses 0-9, thus allowing a maximum of 120 dimmers in a stream. (I recall it also had no Analog In, but had an "Expansion" DB25 port that would control the pack set of 12 dimmers if that pack had the Expansion (Analog) module. Please check yours and post a picture, K?) We don't know which version adickerson has.

I'm not trying to be argumentative, just curious as to whether it's the ETC Idea console or Colortran dimmers that are "at fault" here (not that it matters much, as we as talking about ~25 year-old equipment). And the original problem has been solved.

Perhaps the following is pertinent?

...Colortran dimmer control modules were pure hardware implementations with no processor--just a TTL state machine connected to a UART and a ramp-and-comparator firing scheme.

Also, there is some mythology about the actual bit rate of CMX. It has always been 153.6 kbps.

Finally, people's confusion over the name CMX is understandable. The was no "CMX" until after 1986 when DMX512 was introduced. Then the Colortran marketing department created "CMX" instead of the prior "Colortran Protocol".

ST

 

[SteveB]

Ahh Fuzzy Memory, gotta love it.

You might be correct Derek, about how many addresses D192 was configured for. Not sure why my memory was 192, possibly because the C-Tran DAC's (Digital to Analog converters) had a limit of analog output cards - 6 is what I recall and that would make for 192 dimmers in 6 cards of 32 ea., but I'd have to go look at the DAC we just took out of service.

But you have to love Google as well. Typing in Colortran D192 yielded a bunch of stuff. Here's one from Pathway Connectivity:

" Colortran's D192 dimmer racks for example, the rack's control electronics can be modified to receive DMX512. This works fine most of the time, but the modified receiver tends to be pretty fussy about the DMX flavor it will accept."

Hmmm....

 

[derekleffew]

Neato!

But I'm guessing that adickerson has packs with the Universal Control Module as I pictured. Which should read correctly any flavor of DMX512 thrown at them. Maybe PhilHaney can step in here, as I believe he worked for Colortran during this time.

 

[DavidNorth]

Wow! Lot's of questions and answers in this thread. No doubt I will not get all of them so let me know if there is more you'd like to know.

D192: Yes, it was named for the rack's ability to have 192 dimmers in one chassis. The D192, or Colortran protocol, was packetized very similar to DMX with the primary difference being that the baud rate was 153.6k as opposed to 250k for DMX. There was support for 512 dimmers in the protocol. The change to control modules [which I highly recommend be done at Lite-Trol] is an oscillator and a resistor responsible for the required new packet timeout for the faster baud rate. Remember that the controller in the installation rack is a state machine and is highly reliant on oscillators and analog components to properly operate.

DMX: Yes, five places exist where timing can be modified: Interpacket, Interbyte, Post Start, Break and Mark After Break. This would be one of two reasons why DMX is not always DMX. Add in to this differences in hardware implementations and grounding schemes over the years and it's a little surprising most stuff in fact works together at all. Eventhough the new standards tighten up acceptable hardware implementations, and all versions of the standard place the burden of accepting odd timings on the recieving device, we all know these devices have gotten cheaper and hence sometimes porrly designed. The common solution is to now burden the transmitter [console] with solving timings mismatches.

Express consoles: Indeed DMX refresh rate got a little slower as we went from v2.0 code to the v3.x series. Changing speeds between Max/Fast/Med/Slow really only added some timing in between packets [Interpacket] and added time to Break and MAB, but Interbyte time was always close to zero [or 8usec depending on your terminology]. Unfortunately most cheap DMX receivers need extra Interbyte time, although not likely the case in a Colortran pack/rack. To be honest, I've never verified what DMX timings change when an Express has its number of outputs changed. Larry [one of the sw engineers] had told me a long time ago that it only sets some limitations in patch and displays without affecting anything in timing. I do know that no mater the settings, both output ports always send 513 bytes [counting the Start Code].

Idea: Not sure what the timing difference is between 256/512 outputs. Based on the very slow/small processor in the console, it is highly likely that changing the number of outputs affects not only output speed but overall console operating speed. This is indeed why the switch was installed - that is, to get the console to run faster set at 256, not to speed match to receivers. I do not remember the data rates. No doubt this change is just enough to get the Colortran dimmers to receive the DMX. My guess would be an increase in both Interpacket and Interbyte time got them to sync up.

Does that cover enough? Let me know,

David

 

[philhaney]

Neato!

But I'm guessing that adickerson has packs with the Universal Control Module as I pictured. Which should read correctly any flavor of DMX512 thrown at them. Maybe PhilHaney can step in here, as I believe he worked for Colortran during this time.

Sorry lads. It's over my head. Milton over at Doug Fleenor Design would know, along with some of the other folks who have contributed posts to this thread.