Control/Dimming Dirty DMX Problems

[pat811]

Hi everyone!

I am having a semi-large problem with some LED fixtures in a theatre I work at. They are DaisyLite RGB 575's, which is code for super China fixture, but they are bright and do the job. I have experienced problems with their DMX a few times, randomly when I plug the data in they will start strobing colors and just generally freaking out. The solution has been to replace one of the XLR cables being used to daisy chain them, and it fixes the problem and I don't experience it again for the rest of the time they are being used. However the other day everything in the theatre started strobing and freaking out, all the conventionals and even the house lights. We had no idea what was happening and we decided to pull one the LEDs down at plug it directly into the board so we could try and see if the dimming system was bad or the board, and as soon as we unplugged the data to the LED I all of a sudden had control of everything and the issue was solved. So we tried plugging the LEDs back in and everything was fine for about 30-45 seconds and then everything went dark. Eventually I just by-passed the first LED in my chain and everything was fine, until the next day and the freaking out started happening again. And now I don't have control of them at all. So we decided to strike them from the show we have this weekend but I would love to get some opinions.

I called a local rental shop to get their opinion and they recommended getting some kind of DMX conditioner(?) to put in between my board and my DMX lines? Any other suggestions to help solve my dirty dmx problems? One thought that I had was buying 3-pin DMX cable, but I'm not sure the difference between 3-pin XLR and 3-pin DMX?

P.S. I did try terminating my signal and no dice.

 

[sk8rsdad]

My guess is the fixture is chock full of cold solder joints since it was likely assembled by a labourer who was paid piecemeal. I had an early batch of original Pucks with a similar "feature". They worked fine for a few months but started misbehaving. A quick trip through the reflow oven and mine are all working fine. If you don't have access to a reflow oven then the job can be done with a soldering iron but it can be pretty tedious.

 

[hobbsies]

What does a cold solder do differently than a correct solder? Is it a reference to the soldering iron not being hot enough and thus not making a good connection?

 

[derekleffew]

... getting some kind of DMX conditioner(?) ...

An Opto-splitter serves two purposes:
1. Splits the signal for more convenient cabling or to allow more fixtures on a chain, and
2. Isolates each output to prevent a fault on one from propagating to others.

... I'm not sure the difference between 3-pin XLR and 3-pin DMX? ...

XLR is a connector type. DMX is a transmission protocol. Microphone cable should not be used for DMX512 transmission.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[sk8rsdad]

What does a cold solder do differently than a correct solder? Is it a reference to the soldering iron not being hot enough and thus not making a good connection?

A cold solder joint does not have permanent electrical contact. It can be caused by a cold solder iron, or not allowing the joints to reach the correct temperature, or movement of the pieces being soldered before the solder has solidified. The joint will often work fine until oxidation sets in.

 

[ejsandstrom]

I did not see a quantity of fixtures. For me that has been a huge problem solver. Especially with the cheap lights. 10 fixtures on a universe, every thing works fine, 11 fixtures, one or 2 problems, 13 lights and I get Disco Fever. Back down to 10 and everything works great. And even the fixture type has had a big impact as well. I added a Chauvet Legend to a universe that was working fine. I pulled 3 lights out and added the one and every thing freaked out. I put it on a different universe with less lights and everything was fine.

I think there maximum number of fixtures is like 20 or 25. You could just have to many.

 

[derekleffew]

... I think there maximum number of fixtures is like 20 or 25. You could just have too many.

From DMX512 - AusChristmasLighting Wiki :

Node Limit

The RS485 standard specifies the maximum number of nodes (dimmers, controller, etc) you can have in a single DMX chain. This is specified as 32 "load units". Traditionally, one RS485 receiver (sometimes referred to as a DMX chip) was equivalent to one "load unit". This meant that you could have 31 receivers in the chain, as the transmitter also counts as a unit.

Newer variants of the venerable MAX485 IC have a load rating of 1/4 or 1/8. This means that you can use 4 or 8 times as many of them on a single line. In practice it is best to assume that each device still represents one load unit, as there is no easy way of knowing without checking the IC type and looking up its data sheet.

Please note that this node limit applies to each single DMX chain. On an active splitter, each output is independent and capable of driving another 31 load units. ( the splitter counts as one )

But yes, I've also seen systems work better with fewer fixtures per chain, especially when using long runs of non-DMX512 complaint cabling. An opto-splitter can help to isolate problematic fixtures, or units with questionable morals that don't like to play well with others.
.

 

[JD]

What does a cold solder do differently than a correct solder? Is it a reference to the soldering iron not being hot enough and thus not making a good connection?

Keeps a lot of service techs in gainful employment!

Actually, the term can refer to a number of conditions. The most literal is that this soldering iron (or dip bath, or whatever) was too cool or of too short of a duration to properly "wet" both the component lead or the land pad on the circuit board. If properly fluxed, a component lead will, upon reaching the correct temperature, become amalgamated with the liquid solder and a metallic path will be formed between the lead, solder, and pad. If this does not occur, the lead will actually remain isolated from the solder blob around it. Some contact may occur, but after a little vibration or a few warming and cooling cycles, the connection will be lost. This is the "classic" cold solder joint. In other cases, a lead may have too much of an oxide coating on it for the flux to dissolve and even though the temperature and duration were correct, a solid connection will not be made. If components and boards are not handled correctly, the surfaces may become contaminated and again, the solder may not take. Lastly, there are things called "ring cracks." often refereed to incorrectly as cold joints. These often occur where heavy leads or terminals are mounted on the boards. The joint may have started off perfectly, but more likely was a little deprived of the proper quantity of solder. Copper, tin, fiberglass, and lead all have different expansion / contraction coefficients. Over time, the fact that things are not expanding and contracting at the same rate may cause the solder, which is softer, to crack and create a ring around the connection, and the joint will no longer conduct. High power connections tend to fail even more as the solder, which has a higher resistance then copper, may actually add to the heating effect.

In any case, the repair is the same. Under a high power magnifier, the defects are found and re-soldered. This can be time intensive.

 

[church]

I have experienced the situation described by the OP with LED lights. I traced the problem to defective capacitors in the internal switch mode power supply. I had 12 LED units in a chain and they would all strobe together and affect other dimmers.

In my situation the problem was caused by one of the LED fixtures, sometimes as soon as it was powered up and sometimes after it had been operating for a period of time. The LED fixtures use switch mode power supplies - the performance of these supplies can be significantly degraded by the electrolytic capacitors selected. You want the lowest Equivalent Series Resistance (ESR) and highest temperature rated capacitors you can buy - also the most expensive. Typically there are two capacitors that are critical to the regulation of the output voltage. As these capacitors degrade the regulation and ripple performance degrades. When I looked at the output of the PSU on an oscilloscope I had 2.5vpk-pk sawtooth variation in the output voltage. The lowest value of the PSU output resulted in the LEDs turning off in effect the LEDs were being strobed at 120Hz. Replacing the failing capacitors with better ones solved this problem.

So why did one faulty fixture impact everything else? In this case the supply voltage to the RS485 transceiver chip was also reduced which produced a changing bus impedance for the DMX signal effectively corrupting the DMX. Once the PSU was fixed the DMX problems went away. I changed the capacitors in all the PSUs after this. The problem with dielectric capacitors is they are temperature sensitive and LED fixtures do produce heat. I had noticed that the problem occured faster if the LED was operating with all four colours at 100% brightness.

Anyway my experience - yours may be different

 

[JD]

I traced the problem to defective capacitors in the internal switch mode power supply.

Long running problem with all switch-mode supplies. The secondary electrolytic capacitor. Almost more of a series inductance in the capacitor. If you cut them open, you will find there are two ways to connect to the roll of foil inside. One, is to tie onto the sheets at either the inner most point or the outer most point. This is the cheapest way to make a capacitor. The result is the roll of foil itself serves as an inductor and the equivalent resistance raises with frequency. Bad, due to the high frequency the supplies operate at. The second method is to stagger the sheets and tie onto the top and the bottom of the roll. This way eliminates most of the inductance effect as all areas of the foil are equal distance from the connection point. This way makes it expensive to produce as the edge of the foil must be able to be bonded to. The primary capacitor simply filters the AC coming in, and although it can be constructed using the cheap method, it still must be rated with a high enough temperature so that it does not dry out or seep in general use. Unfortunately, most of the products from China go for the cheapest way to produce it and get it out the door. They are shaving pennies off of something that costs us hundreds of dollars at our (later) expense.

If you are into electronics and are qualified, it is pretty easy to see which one is bad. If your DC ripple is 120hz (twice the supply due to full rectification), then your primary cap is bad. If your DC ripple is at a very high frequency, your secondary cap is bad. (may be more than one.) Compared to the cost of the unit, the cost of the caps is pretty low. If I found the problem on one unit, I would recommend changing the cap on all of the units.