Voltage, but no current?

[kiilljoy]

In preparing to design a show in a space I'm not familiar with, I was trying to troubleshoot/confirm the functionality of a handful of dimmers. Of the six that I was told didn't work, I discovered that four of them did actually seem to work (have to wait until show time to see they are actually trustworthy, but they did pass initial inspection). One seems to have a short either in the line or on the dimmer end. The other simply doesn't turn on either by the manual switch at the dimmer pack or from the control system.

I noticed something odd though as I was testing. In checking the dimmers, I turned my Extech 411 multimeter's dial to the 200 mark with the "~" above it for AC voltage, and I get a number. I was getting a voltage reading between the hot and neutral even though the dimmer was not on, and when I plugged a load into it (a 6" fresnel), the load did not turn on.

I got this result with working dimmers and with non-working dimmers. If I turned the breaker off at the pack, I would get a voltage reading of 0. The packs are 2 24x10amp CD-80 cabled to boxes of 3 circuits in the grid. Any ideas about what is going on? Am I doing something wrong, or is it something else? It seems to me that if there is no current (as must be the case since the light didn't come on) shouldn't there also not be a voltage reading? I don't specifically recall what the voltage reading was when the circuit was hot, but I was getting readings of between 108v and 114v on various circuits.

[derekleffew]

This warrants being included in the "Question of the Day," and thus is being moved there.

[mnfreelancer]

Hot-neutral potential explained by dimmer pre-heat?

[avare]

I noticed something odd though as I was testing. In checking the dimmers, I turned my Extech 411 multimeter's dial to the 200 mark with the "~" above it for AC voltage, and I get a number. I was getting a voltage reading between the hot and neutral even though the dimmer was not on, and when I plugged a load into it (a 6" fresnel), the load did not turn on.

Easy to explain electrically. The dimmer's voltage control device does not have infinite resistance. When "off" the resistance is around 50 to 100 megohms. When the tester with an impedance of 10 megohms is placed across the output, the tester and voltage control device form a series circuit. Ohm's Law.

Andre

[TupeloTechie]

Well, Since its a Question of the Day style thing I will give it a shot with my limited physics knowledge... please correct me if I'm wrong.

Isn't Voltage a measure of Potential Energy, and Current the flow of Energy?

So isn't it possible to have voltage in the circuit but because of a switch inside the dimmer not have current flowing through the circuit, which would explain why you got a reading but the light would not turn on.

[kiilljoy]

Easy to explain electrically. The dimmer's voltage control device does not have infinite resistance. When "off" the resistance is around 50 to 100 megohms. When the tester with an impedance of 10 megohms is placed across the output, the tester and voltage control device form a series circuit. Ohm's Law.

Andre

Does this mean that any tester that uses impedence to measure circuit properties will return a result like this when measuring the voltage on a dimmer? If that is so, is that effect common knowledge?

[avare]

Does this mean that any tester that uses impedence to measure circuit properties will return a result like this when measuring the voltage on a dimmer? If that is so, is that effect common knowledge?

I am not certain about what you are referring to as "common knowledge." It is Ohm's Law. I think a point overlooked, including in my response, is that this is not a normal situation. Something is wrong somewhere. In modern dimmers the voltage will be zero. What I provided was an obvious plausible explanation.

Aside from the switching device(s) in the dimmer, there may be leaking (defective) filter capacitors as just one example.

Potentially,
Andre

[waynehoskins]

The SCR/SSR in the dimmer rack is in a pulse-width modulation circuit, which achieves the effect of "dimming" by turning the load on and off really quickly; it's the ratio of on-time to off-time in a 60-Hz cycle (that is, the width of the pulse) that determines the "intensity". This isn't a truly linear relationship, but for the purpose of this exercise we can think of it that way.

The incandescent filament has a nonlinear resistance: it's a dead short when it's cold, and it becomes more resistive as it burns hotter. Therefore if you hit a dead-cold filament with a full-on pulse, it draws lots of current and also is then most likely to fail, being under the most stress then .. which is why your bathroom lightbulb burns out most often when you hit the switch rather than while it's been burning.

To compensate for this, theatrical dimmers incorporate a preheat circuit, so that Off isn't completely off, it's just a little bit above off to keep the filament warm.

So your dimmers that are off aren't actually quite off. Their output is 120 volts (give or take) for probably 1/50 of a cycle. The sample-time-peak-response of your digital voltmeter grabs this peak output voltage, and that's what you're reading.

In short, your dimmers are probably working properly (SCRs and SSRs fail in either always-on or always-off states, never "in-between"), and the problem is on the control side.

[mnfreelancer]

The sample-time-peak-response of your digital voltmeter grabs this peak output voltage, and that's what you're reading.

How would the measurement of a dimmer in this scenario differ if the volt meter were a true RMS meter?

[porkchop]

I might open up the dimmer and give it a good once over for anything that looks out of the ordinary or burned. Ohm's Law (V=IR) is the place to start for mathematical analysis but if theres a low resistance short that is in effect parallel to the hot to neutral connection of your instrument then you might get into a situation where the total current passed is inversely relative to the two legs resistance:

i1 (current through leg one) = Vtotal * (r2/(r1 + r2))
and
i2 (current through leg two) = Vtotal * (r1/(r1 + r2))

Basically this means that although you have 120ish V potential from point A to point be the energy is flowing through the path of least resistance and that's the short (or something like that) not your instrument in such an unbalanced proportion that is seems like the current through the instrument is 0.

To be totally honest this is how I would analyze it mathematically and troubleshoot in a controlled lab situation where the circuit is only powered for a short time. If you have turned this dimmer on for any notable length of time I would expect the low resistance short to burn up and become a high resistance short rather quickly, but for an experiment in electricity math and a last resort if you are SURE that your right on the control side it's not a bad idea.

[waynehoskins]

How would the measurement of a dimmer in this scenario differ if the volt meter were a true RMS meter?

It would be different. Time factors into it, and I'm rusty on the calculations, though I believe it would be a good order of magnitude smaller. If it were a mechanical movement, that too would behave in a time-averaging sort of manner.

It could also be a microprocessor problem: (newer, at least) CD80s, and packs too I think, introduce a rack-level softpatch to let you address physical dimmers differently. That could be the culprit.

It could also be the opposite problem to a low-resistance short: a high-impedance connection. Tighten your terminations while you have it open.

[avare]

So your dimmers that are off aren't actually quite off. Their output is 120 volts (give or take) for probably 1/50 of a cycle. The sample-time-peak-response of your digital voltmeter grabs this peak output voltage, and that's what you're reading.

Where did you get this information? One of the first things I checked was if the multimeter is peak or RMS. When I checked the Extech website, the 411 was shown as being "true RMS." The 411 User's Guide is here for reference. When did Extech change the 411 from peak?

One item missing in the discussion is if preheat was on or off during the initial testing. It has assumed that it was on, the reason for the voltage. Hopefully the OP will double check on that.

Andre

[00AVD]

The dimmer output devices (Triacs or SCRs) will leak a very small amount of current even when turned off. The snubber network across the output device will also contribute to this leakage.

A digital multimeter has a very high impedance, so doesn't damp this leakage much at all. The result is that you can get quite a high reading even when the channel is down at 0%.

As soon as you connect a real load, the voltage will fall to almost zero as the load resistance is many orders of magnitude lower than the leakage impedance.

[kiilljoy]

When did Extech change the 411 from peak?

My tester claims to have RMS. My impression when I bought it was that this particular model has always had RMS. The 410, I think does not.

One item missing in the discussion is if preheat was on or off during the initial testing. It has assumed that it was on, the reason for the voltage.

I am unaware if the dimmers are set for a preheat without any signal from the control board.

It could also be a microprocessor problem: (newer, at least) CD80s, and packs too I think, introduce a rack-level softpatch to let you address physical dimmers differently. That could be the culprit.

The racks look to be in the neighborhood of 5-10 years old. I'm not familiar enough with the different generations to be able to more than guess as to the age. I don't understand how a rack-level soft patch could affect the output from the dimmers. At the time, none of the dimmers were on.

As soon as you connect a real load, the voltage will fall to almost zero as the load resistance is many orders of magnitude lower than the leakage impedance.

I'm back in the space on Thursday. I will check this and report back.

[David Ashton]

Wow, what a wealth of mis-information in one thread, firstly a lamp is not a dead short when cold, it is around 1/10 of its hot resistance, second a dimmer is normally set to around 5v when the desk reads 0, so that as you bring up the fader you are straight into the working part of the dimmer curve, this is not what is usually referred to as pre-heat [which is a totally useless ritual previously argued on this and other fora] but is an unavoidable side effect of the way dimmers are set up.The accurate input to the thread was the explanation that the scr's leak a small current into the high resistance of your meter but will drop to a very low voltage as soon as you add a load.
In summary the only valid voltage output reading you can take must be with a load on.The other important fact is that even with the desk at 0 the dimmer can put out enough current to kill you, always unplug the light to change lamps.

[waynehoskins]

Wow, what a wealth of mis-information in one thread, firstly a lamp is not a dead short when cold, it is around 1/10 of its hot resistance, second a dimmer is normally set to around 5v when the desk reads 0, so that as you bring up the fader you are straight into the working part of the dimmer curve, this is not what is usually referred to as pre-heat [which is a totally useless ritual previously argued on this and other fora] but is an unavoidable side effect of the way dimmers are set up.The accurate input to the thread was the explanation that the scr's leak a small current into the high resistance of your meter but will drop to a very low voltage as soon as you add a load.
In summary the only valid voltage output reading you can take must be with a load on.The other important fact is that even with the desk at 0 the dimmer can put out enough current to kill you, always unplug the light to change lamps.

Well, yes, not dead short, but very low compared with its hot resistance. That is what I meant by "dead short". Happy to provide misinformation.

But yes, I agree definitely on the meter impedance being the culprit.

[STEVETERRY]

In preparing to design a show in a space I'm not familiar with, I was trying to troubleshoot/confirm the functionality of a handful of dimmers. Of the six that I was told didn't work, I discovered that four of them did actually seem to work (have to wait until show time to see they are actually trustworthy, but they did pass initial inspection). One seems to have a short either in the line or on the dimmer end. The other simply doesn't turn on either by the manual switch at the dimmer pack or from the control system.

I noticed something odd though as I was testing. In checking the dimmers, I turned my Extech 411 multimeter's dial to the 200 mark with the "~" above it for AC voltage, and I get a number. I was getting a voltage reading between the hot and neutral even though the dimmer was not on, and when I plugged a load into it (a 6" fresnel), the load did not turn on.

I got this result with working dimmers and with non-working dimmers. If I turned the breaker off at the pack, I would get a voltage reading of 0. The packs are 2 24x10amp CD-80 cabled to boxes of 3 circuits in the grid. Any ideas about what is going on? Am I doing something wrong, or is it something else? It seems to me that if there is no current (as must be the case since the light didn't come on) shouldn't there also not be a voltage reading? I don't specifically recall what the voltage reading was when the circuit was hot, but I was getting readings of between 108v and 114v on various circuits.

What you are seeing has nothing to do with preheat (even if the dimmer has it). It is caused by the leakage current of the thyristor (triac or SCR), as well as the capacitive snubber network typically used with this type of device to stop false turn-ons due to dv/dt spikes.

The bottom line: you cannot measure the output of an SCR dimmer accurately without a significant minimum load (I like 25 watts) in parallel with your meter. And, it goes without saying, but you need to use a true-RMS-responding meter for measuring the output of any phase-control dimmer due to its non-sinusoidal voltage waveform.

ST

[derekleffew]

More than once (twice IIRC; never in the past 18 years!) I've gotten a "tingle" while repairing/replacing a female pin connector, and being too lazy to walk to the dimmer room to LOTO the rack. This was before I even knew what LOTO was.

The doorbell just rang, and it was my postal person, with my personal copy of Practical Health and Safety Guidelines for School Theatre Operations. Dr. Randall W. A. Davidson [aka "Dr. Doom"]. Risk International Publishing, 2005. ISBN 0-9774525-0-6. Library of Congress Catalog# 2005938919. Available for purchase, USD 75 +10 S/H, here. ALL EDUCATORS: If you don't have a copy of this book, get it! Buy it yourself, or have your library order it for you. Institutional librarians often have funds available, but need guidance how to spend them.

[mbenonis]

The doorbell just rang, and it was my postal person, with my personal copy of "Practical Health and Safety Guidelines for School Theatre Operations." Dr. Randall W. A. Davidson [aka "Dr. Doom"]. Risk International Publishing, 2005. ISBN 0-9774525-0-6. Library of Congress Catalog# 2005938919. Available for purchase, USD 75 +10 S/H, here. ALL EDUCATORS: If you don't have a copy of this book, get it! Buy it yourself, or have your library order it for you. Institutional librarians often have funds available, but need guidance how to spend them.

Yes, I agree completely - get a copy for your school any way you can. I have a copy on my shelf courtesy of my school system, and it is a great guide to how to work safely in the theatre. Despite being targeted at secondary schools, its information is definitely not limited to these schools.

[kiilljoy]

Alright folks, as promised, further data. And yes, I realize I have ignored the significant digit...

Hypothesis: For modern theatrical dimmers, a load is needed to accurately measure the voltage output of the dimmer.

Method: Each data point was gathered by using the control system (Horizon 98 in this case) to set the selected dimmer (No. 15) using a linear dimmer curve to one of the levels listed below. First, the dimmer was checked without a load measuring the voltage difference between the hot and neutral from inside the female socket. Then the voltage was checked with across the male hot and neutral leads of a 500W load (The tester and load are parallel in the circuit). All tests were done with an Extech 411 that is supposed to be able to measure true RMS.

Data:
No Load
0% -- 111.4v
5% -- 124.0v
50% -- 124.4v
100% -- 124.4v

With Load
0% -- 0.0v
5% -- 21.5v
50% -- 94.8v
100% --121.5v

Conclusion: From the data, it appears that the hypothesis is correct. Though the voltage reading with a load does not go up in a linear fashion, it does appear to increase in accordance with a pattern. The reading taken without an electrical load in parallel does not form a regular pattern, and even produces an anomaly in the voltage reading between tests at 50% and at 100%. With the control system dimmer curve set to linear, those two reading should differ, yet they do not.

[derekleffew]

... Though the voltage reading with a load does not go up in a linear fashion, it does appear to increase in accordance with a pattern. The reading taken without an electrical load in parallel does not form a regular pattern, and even produces an anomaly in the voltage reading between tests at 50% and at 100%. With the control system dimmer curve set to linear, those two reading should differ, yet they do not.

Anyone care to enlighten killjoy?

[kiilljoy]

I hate being wrong. It seemed straight-forward with the information I have. What's the problem?

I was trying to make the point that if the readings were reliable, I would expect a certain relationship between a reading at 50% vs. a reading at 100%. Namely, that they would differ from each other, with 100% being larger (significantly) than 50%. Since that relationship did not exist, the readings must not be reliable. This does assume that the end intensity is controlled by the dimmer adjusting the voltage.

Perhaps the dimmer curve is irrelevant. Perhaps the pattern of the reading without a load does have a pattern, yet is too subtle for my limited data set and tester to detect. Perhaps 124v is beyond the capability of the Strand CD80s. Perhaps the voltage readings with a pattern DO go up in a linear fashion. What? Don't let me wallow in ignorance! I have but the merest bits of information to enlighten my path.

[dramatech]

Part of your problem in the linearity that you are expecting to see is that most dimmer rack microcontrollers insert a dimmer curve, and not all companies use the same curve. When you set the control board to linear, the rack still uses the curve that it has been programmed for. I'm sure that there are other contributing factors when measuring switched sinewaves. Someone else would need to answer that, but I also suspect that the detailed answer is a bit over the top for most readers on this forum and really not all that important. If the dimmer works use it, if the curve seems wrong change it in the board.

[derekleffew]

Correct, dramatech. For more elaboration, see Dimmer Curves.