pmiller056

That looks like a mains supplied switch mode power supply combined with an inverter for a fluorescent lamp backlight - in this case 2 tubes. From your description, it is the backlight inverter that is at fault. The transformer for this inverter is the lump that has the 'QC OK' sticker on it in the last photo. The inverter produces a very high voltage to strike the lamps on no load, with a much lower voltage to run the lamps at a few mA. If it is easy to do so, have a look at the lamps - if the ends are blackened due to wear, they're probably end of life, in which case I'd consider scrapping the monitor. The lamps used are likely to be custom made for the manufacturer and replacements unobtainable. Another symptom of end of life backlight fluorescent lamps is they have a pinkish orange tint to the light instead of a brilliant bluish white. Peter

Turpentine Substitute (aka Paraffin / Kerosene) and if that fails, White Spirit. Try on an inconspicuous place first to make sure that the solvent doesn't remove the surface finish. Both available from your local DIY shed. Pay attention to hazard and safe use warnings on the packaging.

For some projects I've used a small SMD to DIP adapter PCB, allowing the modern component to be placed on an older design PCB. Unfortunately, in this case it will need to be a custom adapter PCB to go from the current 16 pin SMD device to the older 8 pin DIP pcb. If you choose to do this be aware that there appear to be a number of different versions of the SMD device, so check the actual part number to be used before starting to design anything.

If the camera is moved around a lot, consider using a HDMI to CAT5 converter system at each end (Secure the camera end converter to the camera body). Depending on converter system used you *may* be able to use a longer cable length. CAT5 patch lead is much more flexible than a HDMI lead and is quick, cheap and easy to replace if anything goes wrong with it.

There are a number of retrofit LED replacement lamps available that will fit into an existing SOX lamp head (I think Big Clive has done a review and tear-down of one on his Y**T***e channel). Most of these LED replacements will need the SOX specific ballast and control gear removed and wiring altered to suit. A local council has relamped an entire street from SOX to LED this way - it looked weird at first. The columns and lamp heads were relatively new (<10 yrs) and in good condition which made this change worth it. If you can get replacement SOX lamps, they're now very expensive for what they are so the conversion to LED becomes more compelling.

The IC's are in sockets which is not good for long term reliability because the contacts do tarnish. Tarnished IC socket contacts cause all manner of weird and unexplainable faults. A lot of problems can be temporarily fixed by unplugging and re-plugging each IC several times. The long term fix is to remove the sockets and solder the IC's directly in. IC sockets are considerably less reliable than the ICs plugged into them. The very high control voltage is most likely to be caused by a short-circuited pass transistor in the voltage regulator circuit.

I've seen a number of different Voltage/Current requirements for ceiling panels - some of which make no immediate sense. If you wish to dismantle and rebuild an existing one, it's straightforward to make up what you need. Generally when you buy a new ceiling panel, as well as the standard fixed controller, you can get the option of a matching dimmable controller. Different suppliers support different control protocols, you may be able to get DMX but 0-10V is a fairly universal offering.

Tim - Noted and thank you.

Thank you for your replies. The new dimmer packs will be used for temporary or event type work rather than a fixed installation, so they're always going to be more vulnerable to damage. I'll go with the PTC and Transzorb solution suggested by Tim for simplicity. Previous damage has been noted where a laptop PC used as the DMX controller has been plugged into the earthed dimmer while operating (why's nothing working, oh, oops!). Upon investigation, the DMX line was wired directly to the line receiver IC in the dimmer with no other components around. I suspect that the leakage current (or connection transient) from the double insulated (unearthed) laptop power adaptor in use at the time may have been sufficient to do the damage. Surface mount repair work is now a fact of life - I'd still prefer not to have to do it! Thanks Peter

I've got a new dimmer pack for LED strips. The DMX input circuit uses a surface mounted RS485 line receiver. In the past I have had to occasionally change the DMX receiver IC on equipment. This has been easy because the IC has been socketed or easily solderable by hand with simple tools. Does anyone have any useful protection circuits or devices that can be used between the DMX line and the new dimmer pack to reduce the risk of damage in use? Thanks

Some people detest them, but a Behringer LC2412 does provide 12 channels of 0-10V analogue out on a 15 pin D connector. It's available new and there's usually a few available used on 'the bay' at any given time - remember to check 'completed listings' to check what they are actually selling for, rather than what sellers are wishing for. The documentation is available online to allow you to make up suitable leads to go to your dimmer packs. They also provide DMX output so you can build a system with a mix of dimmer pack communicaton protocols. This desk is OK for conventional lamps but get something else for movers and other clever stuff.

For most practical purposes, T26 and T27 are interchangeable. Same filament height from base and envelope dimensions. One of them has the filament in a flat plane, the other doesn't - so one works (optically) better in a profile spot projecting gobos. Not sure which is which.....

KiCad. Free to download and use with a reasonable component library. It is Electronic/Electrical CAD software - I doubt you will need the PCB design parts. Adding parts to component library is straightforward. You can drag parts around whilst retaining connections. Good online community for help. Available for Windows/Mac/Linux.

Use a Rotovator (or similar) to dig it in reasonably deeply with minimal effort.

Car HT ignition cable. Don't use the interference suppressed type becuase the resistance along it is significant - use the copper cored type.

There are widely available, pre-made buck converters based on the LM2596 which are good for a 3A load. These are switch-mode converters, so they don't get hot unless you are using them incorrectly. 5A converters are readily available as well. Most of these converters will work happily with a 24V input which makes power distribution even less bulky.

Thank you for the helpful market overview comments from various contributors. I'll have to radically re-think how I'm going to control pixel mapping along with the rest of the show lighting. The changes will have to be made at some point anyway..... If you have a stock of existing (Pat tested) 12V power supplies for LED tape, but having to use 5V pixel tape, why don't you use small 12V to 5V buck/step-down converters periodically along the length of 5V tape? Using 12V for power distribution keeps wire sizes a bit more manageable for longer runs.

Thank you for your suggestions. I'm more uncertain now.... The Enttec, Madrix, Qolorpix units are way beyond our budget and in addition there's the infrastructure needed to make them work. The SP108 is interesting and I'll have a play with one. Many user reviews of the SP108 complain of poor reliability controlling the device. The K1000-C controllers (and similar derivatives) look promising, but again poor documentation and software let them down. There appear to be many suppliers of these devices, but I get the impression that no-one has any real clue about what they are selling or the software and firmware involved. The K1000-C devices do have an RS485/DMX input port but no obvious way of doing anything useful with it. Shame. At the moment, it looks like I'll have to polish my Arduino skills

A related question. Does a device or project exist where I can set a DMX address and depending on the DMX value received, output different preset scenes (static or animated) to a pixel tape installation? Ideally the device/project should take a pre-programmed SD card with all the pre-set scenes. Minimal cost for communtiy theatre work is preferred because our lighting desk is simple and only sends out one universe of DMX with no additional clever networked stuff.

My day job is with a higher education institution. For certain product categories (particularly CPC only items), there is no discount. I've found that Onecall tends to bias search results towards Farnell - however you may get better pricing through Onecall by feeding it the equivalent CPC part numbers. In my experience Onecall are Ok with non-account orders for personal use, provided that the order is paid for at the time with a card and delivered to a place that does have an account.

The motor will be a fixed speed synchronous motor. It is not economically practical to control the speed of this type of motor by electronic means. If you only need fixed speed operation, change the motor pulley for a smaller diameter (1/2 diameter = 1/2 speed). If you do this you will need to either change the drive belt for a shorter one or use some of the unused holes by the motor shown in your photo to move the motor further away from the driven shaft. If variable speed operation is required, the simplest and cheapest way will be to change the motor for a DC motor+gearbox, but you will then need to add a speed controller and appropriate power supply. Plenty of online sources (UK and overseas) for the relevant parts. With this solution, significant metalwork modification will be required to make it all fit together. The empty holes and taped up wiring shown in the photos strongly suggest that this system has had previous significant modifications.

To follow my earlier comments, I have designed and built small industrial control systems where people will get hurt if the system fails to work correctly. This may provide context to my comments and approach to the problem...... Microcontroller pins are electrically very fragile. Provided the microcontroller pins don't go too far and never directly out of the system enclosure there's usually no problem. Successful microcontroller systems (as used in household appliances and vehicles) always isolate and protect the microcontroller from the outside world. These designs also feature a lot of software and hardware self checking as they operate which is well beyond the scope of this topic. A Programmable Logic Controller for industrial control systems is an extreme example, but at the heart of it is a microcontroller working away. A PLC is designed to operate and survive in an electrically hostile environment - the techniques used for interfacing in these systems are good examples to follow. Directly wiring microcontroller pins around a building (as originally suggested) is asking for reliability problems. The two biggest issues will be Electrostatic Discharge (ESD) and unwanted electrical signals picked up from the environment. ESD may not cause immediate failure of a device but will often injure the device leading to premature hardware failure. ESD events frequently cause program crashes. A common cause of ESD is someone unintentionally 'zapping' a connected pushbutton. Unwanted signals will be picked up from the environment because the wiring acts as an antenna/aerial for noise, similar to hum and buzz picked up in a microphone cable running close to power wiring. Mobile phones and 2 way radios used nearby will create havoc if poorly designed microcontroller interface circuits are used. A good reference for this topic is the relevant pages from Horowitz and Hill's 'The Art of Electronics'. Any edition will do. If you have not previously encountered this book, it's a non-mathematical, comprehensive practical guide to digital and analogue electronics which is easily readable. There are 3 different editions available and each is very different reflecting current custom and practice at the time of writing. +1 for not using a book with Arduino. Read switches, flash lights and then see where your ideas take you. If you get it wrong, read the online help, rewrite the broken bit of code and try again. If you are stuck, ask your favourite search engine. If you are really stuck, ask in one of the many relevant online forums. The white smoke only escapes when the outside world is incorrectly connected!

Having a system that you consider to be disposable does not help when it suddenly fails in a strange way 1 hour before curtain-up in 2 years time. The cost of a lost show is many, many times the cost of the failed system. Your coding IS entry level coding, however coming back to and understanding your own code many months after completion is hard. It is even harder and more time consuming when you have to understand someone else's code on an unfamiliar platform and development system. Been there and done both. It becomes nearly impossible when you are trying to do it with a very expensive clock ticking loudly (the cost of a lost show). I'd much prefer to fault find and maintain any system where the ONLY required documentation is one A4 sheet of a circuit diagram tucked inside the box for the master station. A USB stick helpfully left inside is not instantly useful and needs other equipment to make it work. Please also remember that microcontrollers along with the required software infrastructure to make them work evolves VERY rapidly. A system that is readily available today probably will be obsolete and only maintainable with difficulty in as little as two years time.

For clarification... the design brief for the system I mentioned above included a requirement for a very long installed service life (10yrs+) with minimal maintenance. It was also assumed that the original designer and installer would not be available if anything did go wrong. These are the main reasons why a non-microcontroler system was chosen, the loss of some clever functions was an acceptable compromise to meet this and other design goals. A 'master go' function is useful - the installation has only 4 outstations, so it is assumed that the person at the master station has sufficient fingers to push 4 big chunky buttons nearly simultaneously. It required, it is easy enough to add a 'master go' function with an extra pushbutton and some diodes. To keep the system simple a latching standby function was omitted. The user workaround (bear in mind this is a community theatre which will happily evolve its own way of doing things!) is that the master station flashes the red standby at the outstation, which will respond with a green flash back to the main station, acknowledging the standby signal. We generally don't use a headphone system (history, paracticalities and personalities!) so being able to signal in both directions with this cue light system is very useful. For example, if a scene change is going to take longer than rehearsed, the crew can signal back to the master station with a red light alerting the SM, then signal with a green light when ready. Needless to say there are other possible combinations (similar to inter-signal box communications using bells in old railway systems) which are generally not wanted or needed in our application. Oddball temporary relocations and extra stations can be quickly added with one 3 pin xlr microphone cable per pushbutton and indication (no RJ45 infrastructure in our building).

For live performance work, equipment used must be totally reliable and if it fails, fail gracefully to a safe state. For this reason, simpler systems are usually preferred. Are you making this too complicated with a microcontroller in the middle of it? From experience, making something like this work installed in a real building is a least an order of magnitude harder than making it work on a nice cosy benchtop. For a local community theatre, I made a very simple cue light system. Each indicator circuit consists of parallel connected (local, remote) pushbuttons operating series (local, remote) connected LED lamps. Either pushbutton when pushed lights the connected LEDs. The LEDs are series connected to provide confimation that the other one is also lit when a button is pressed. This circuit is duplicated for Red and Green channels and repeated as many times as required for all the outstations. CAT5 wiring can be used for the system hardwired, or via RJ45 connectors. Becaue the power demand is so low (power is only drawn when a button is pressed), two PP3 batteries happily power the system for a long time until they fail due to old age. Faults are instantly self-evident and can be qiuckly fixed with little more than a screwdriver and a meter. If budget can be persuaded, use 22mm industrial pushbuttons and lamps (or illuminated pushbuttons) - they are very reliable and good size to see and use in dimly lit backstage areas.