LED par cans powered from dimmers?

[DrV]

But the waveform on your average dimmer is still chopped. .. not a pure sinewave surely?

 

Therefore unsuitable for SMPS like in an led...

 

In a basic SMPS the input bridge rectifier will only be conducting near the peak of the mains cycle. Provided the dimmer triac switches on before this point in the cycle there will be no difference in the current waveform being fed into the SMPS. However, if you reduce the level of the dimmer then, as the switching point of the triac moves later in the cycle, there will suddenly become a point where the triac switches on after the voltage has reached the level of the reservoir caps. From this point onwards, the triac is switching on into a very low impedance. This will cause current spikes in the triac and in the bridge rectifier. The effects of this will, to some extent, be taken out by the input filter of the PSU and the output filter of the dimmer but neither of them is designed for this so there is a possibility of increased EMI and breakdown of the affected components (filter caps, bridge, triac). So in summary, provided the dimmer is on full you are unlikely to cause a problem but you will, no doubt, have run the gear outside its specification and manufacturer's instructions so would be on dodgy ground in the event of a warranty claim for a related or unrelated failure.

 

Of course, if you have an LED with an electronic PFC input then it gets a whole lot more complicated!

 

Dave

[alistermorton]

In a basic SMPS the input bridge rectifier will only be conducting near the peak of the mains cycle.

 

Really? I'd have said it would be conducting as soon as the input voltage exceeds the forward voltage of the rectifier diodes, which, if they're silicon rectifiers is going to be about 0.7V - now I'm confused.

 

I thought the problem comes if the waveform being fed to the power supply goes into an inductor that doesn't like the choppiness of the triac output, and it's this that we don't know. There are issues about driving certain types of inductive load with certain types of triac dimmer.

[Brian]

Really? I'd have said it would be conducting as soon as the input voltage exceeds the forward voltage of the rectifier diodes, which, if they're silicon rectifiers is going to be about 0.7V - now I'm confused.

 

 

Think about it. If there is no charge on the cap then when power is applied the diode will conduct as soon as the voltage rises about about 1v. HOWEVER, once the cap is charged up then the diode will only conduct once the difference across it is 1v. If the cap stays charged to the peak voltage of the mains then the diode will never conduct.

 

I thought the problem comes if the waveform being fed to the power supply goes into an inductor that doesn't like the choppiness of the triac output, and it's this that we don't know. There are issues about driving certain types of inductive load with certain types of triac dimmer.

 

The input filtering inductor on your typical SMPSU is way too small to worry a Triac dimmer.

[ianknight]

erm....

 

Shouldn't you (or the venue management) be getting in touch with the contractor/supplier for an explanation - at best the configuration will dramatically reduce the capabilities of the rig, much more likely is that it will cause a significant reduction in the lifespan of kit and compromise virtually every show that uses the theatre. The contractors must have planned for how this kit was going to be used and thus explain their procedure so that the system can be used OR if they haven't then they must be brought to task for fitting a system that isn't fit for purpose.

 

Agreed, this will be my next course of action once I can back-up the cause with writing particularly the LED manufacturer's statement of suitability of them being powered from dimmers, and probably a similar statement from Anytronics.

 

It sounds like our 'favourite' kind of install that we get called to when it goes horribly wrong.

 

The contractor has supplied a specification (I imagine) that went something along the lines of "We want a dimmer, a Zero88 desk because our drama teacher had one at her last school and some of those LED things because everyone's going on about them...". Alongside the spec was a low budget which sourced funds from a couple of smaller budgets so the contractor got a price for a Chilli (they're made by the same company after all) and fell over when he saw the cost - his wholesaler then suggested Anytronics.

 

Having said that - I have more faith in the Anytronics kit after all the dealings I've had with them over the last 16 years and I would suspect that if you used the option switches on the rack that it would behave almost faultlessly in switch mode. The issue is going to be the PSU's in the LED's - whilst a lot of LED's use transformer PSUs some of the fancier units use a switch mode supply and they can be very fussy about input waveforms.

 

Ultimately, you may have to 'sacrifice' some socket outlets to hard power and have the contractor supply them from an additional breaker board so that the LED's are happy but I suspect you're going to witness some stomping about before you get there.

[alistermorton]

Once you start drawing power from the capacitor it has to be recharged, though. So depending on load the capacitor could be being recharged a lot or a little, no?

[grantr22]

Sorry, I don't think I was clear enough in my post (was a little late, admittedly).

 

Referring to this bit:

Yes, I was thinking I need to find out what the LEDs are and check with the manufacturer or distributor, though with the dimmers tripping out at about 90% fader suggests that the answer will be that they're not suitable to be powered from a dimmer.

 

I really doubt any manufacturer is going to say that it's viable to power LED kit from anything other than proper hard-power for all the reasons given above - chopped waveforms, the risk of actually dimming the channel on some kit, etc.

 

But I get the impression that you're actually trying to dim the fixture based on this comment - again, I could be wrong, but when you say 'dimmers tripping out at about 90% fader' that rings alarm bells with me.

 

 

 

 

Just to clarify as we're 3 pages in now - in no way am I wanting to dim LEDs by operating the dimmer - very much the opposite. The contractor has installed the rig like this so the risk and definite likelihood of this happening, accidentally or otherwise, has been manifested. Rest assured that the same alarm bells which ring for you ring for me too.

[timsabre]

According to the Anytronics manual there are dip switches which allow you to put each dimmer channel into switch mode (ie full on or full off).

 

If you do this for the channels powering the LEDs this will prevent damage to the LED kit. However if you turn the channel off you may find the LEDs flash on every few seconds, see my post #9 way back up there.

[grantr22]

It sounds like our 'favourite' kind of install that we get called to when it goes horribly wrong.

 

The contractor has supplied a specification (I imagine) that went something along the lines of "We want a dimmer, a Zero88 desk because our drama teacher had one at her last school and some of those LED things because everyone's going on about them...". Alongside the spec was a low budget which sourced funds from a couple of smaller budgets so the contractor got a price for a Chilli (they're made by the same company after all) and fell over when he saw the cost - his wholesaler then suggested Anytronics.

 

Having said that - I have more faith in the Anytronics kit after all the dealings I've had with them over the last 16 years and I would suspect that if you used the option switches on the rack that it would behave almost faultlessly in switch mode. The issue is going to be the PSU's in the LED's - whilst a lot of LED's use transformer PSUs some of the fancier units use a switch mode supply and they can be very fussy about input waveforms.

 

Ultimately, you may have to 'sacrifice' some socket outlets to hard power and have the contractor supply them from an additional breaker board so that the LED's are happy but I suspect you're going to witness some stomping about before you get there.

 

Not quite on the spec scenario though the brief was for it to be user friendly for all, experienced or otherwise.

 

A separate breaker board supplying hard power to a pair of sockets on each bar (e.g. one either end) is what I was expecting to see, the LEDs have powercon in/out so they can easily be daisy-chained.

 

According to the Anytronics manual there are dip switches which allow you to put each dimmer channel into switch mode (ie full on or full off).

 

If you do this for the channels powering the LEDs this will prevent damage to the LED kit. However if you turn the channel off you may find the LEDs flash on every few seconds, see my post #9 way back up there.

 

Yes I wondered if the setting of the dip switches had perhaps been overlooked by mistake, the contractor was pushed for time towards the end of their available hours from what I understand. The LEDs do flash a little when I switch off the breaker but that won't be a problem.

[sleah]

Drifting slightly....

 

Just looking at a dimmer at 100% for a moment. This sounds a lot like the issue of some electrical equipment not working properly from an inverter (12vdc > 240vac).

I have some experience of connecting IT/AV type kit to inverters (both pure sine & modified) and the results vary.

There is absoltely no guarentee of knowing up front if it will be happy or not, let alone if it will remain happy.

 

In my limited experience, based on both inverters and dimmers, I would say if something is designed for hot, 100% power, then that's what it should get rather than 'guessing' if it will work on a 100% locked dimmer (of any flavour). With the exception of relay packs of course, but they're not dimmers anyway ;)

It could well work one minute, but not the next.... :blink:

[timsabre]

No, 100% on a triac dimmer (which those Anytronics Contractors are) is just the mains switched straight through. It's not like an inverter where the mains waveform has been regenerated. There will be no problem as long as the dimmer is set to switch mode and kept in the "on" state.

[adam2]

Once you start drawing power from the capacitor it has to be recharged, though. So depending on load the capacitor could be being recharged a lot or a little, no?

 

Yes, but even if the capacitor has to be recharged "a lot" this still takes place over a very small part of the mains waveform.

On no load the capacitor will charge up to the peak voltage of the AC supply and stay at vitually that voltage.

At full load the voltage accros the capacitor will drop, but not by that much, and recharging will take place very quickly.

 

Actual measured voltages accross the capacitors of switched mode power supplies are typicly about 300 to 350 volts on UK mains.

[Shez]

Well the install was signed off yesterday...

 

Excellent. That means all the information you need will be contained within the documentation that the contractor must have left as required by the wiring regulations.

It's a shame there isn't a timescale stipulated in those regs - after one building project on this site, the main contractor took nearly 12 months to deliver the O&M pack! And I've been waiting a couple of months now on another one that supposedly contains the serial numbers necessary to use some software that was part of the install...

[ImagineerTom]

English law / regulations are brilliant - in the absence of a stated timetable in any document or agreement (and if there's not an umbrella definition in national law or regulations) then it is automatically assumed that whatever must be done, must be done in a "Reasonable" time. It's deliberately a subjective word as (obviously) the reasonable amount of time for rebuilding a house after a flood is very different to the reasonable amount of time for sending copies of an invoice. Since the paperwork you mention must have been submitted at the point of handover there's not going to be a court in the land that would accept a delay of anything beyond a couple of weeks was "reasonable" so it's probably time to go in all-guns-blazing Shez; if they haven't compiled the information now then the project will have gone from the memories of the people who did it so your documentation will be getting less and less accurate at the weeks go by.

[Brian]

It's a shame there isn't a timescale stipulated in those regs...

But I think there is even if it isn't explicitly stated. My arguement would be that they can't hand over an installation until they've completed all the NICEIC paperwork which should, on anything other than a standard installation, be accompanied by supporting documentation. I'd go so far as to say that unless they issue that documentation at hand-over then they are probably in breach of EAWR. Think about it, all that paperwork is in effect the instruction manual for the installation.