Another problem I ran into was at a different arena with LEX Products Power Gate switches. The 400 amp breaker trips at 320 amps instantaneous load (not 5 minutes or 10 minutes at this load, it trips at 80% load RIGHT FUCKING NOW).
Breakers don’t trip without a reason, it’s just that we don’t always understand why they have tripped. A possible explanation why the Lex Company Switch tripped at 80% load could have to do with the elevated rms and peak current drawn by non-linear loads. The Lex company switch, like most company switches that use the newer electronic breakers, use the ABB T-Max with the “peak sensing” PR221DS-LS/I trip unit. Since the rms value of the harmonically distorted current drawn by non-linear loads can be 30-40% higher than that of linear loads, peak breakers will “nuisance” trip prematurely. The problem is not the breaker but the lighting technician that is distributing the load based upon paper calculations, ignoring the poor power factor of the loads they are powering. Even if the electricians distributed their loads using true rms current values that they measure, the elevated peak current of non-linear loads, which can be 20-25% higher than normal, will prematurely trip a peak-sensing electronic breaker at a 75-80% of it rated value.
The source of the harmonics could be non-power factor corrected amps, power wedges, LED Lighting, HMI Lighting, or dimmer packs. As more and more of the lights on a rig consist of non-linear LED lighting fixtures rather than linear incandescent fixtures, peak current has increased relative to rms current. For example, over half of the LED fixtures that I have tested at random (from the inventories of Boston area rental and lighting sales companies) were not power factor corrected (pfc.) With power factors ranging from .45 to .63, these fixtures draw roughly twice the current than one would expect using Ohm’s Law (W=VA) and can have Crest Factors as high as 2.7 where a normal sinusoid has a Crest Factor of 1.414.
(http://www.screenlightandgrip.com/images/generators/CF_Color_Blaze.jpg)
Even those LED fixtures that were power factor corrected generated harmonic currents when dimmed or switched to only one color emitter. For instance, the pfc of the Color Blaze above dropped from .99 to .60 when switched to only the red channel (to see which LED lights are power factor corrected or not, use this link - http://www.screenlightandgrip.com/html/emailnewsletter_generators.html#anchorHigh Output AC LEDs (http://www.screenlightandgrip.com/html/emailnewsletter_generators.html#anchorHigh Output AC LEDs) - to see some of the results of my tests.)
Dimmer racks will also generate higher peak currents when dimming incandescent fixtures than the fixtures do when they are not dimmed.
Where it is somewhat counter intuitive that a lamp draw a higher peak current when dimmed than it does when operating at 100%, let’s look at the effect that dimming a 1kw incandescent lamp with an SSR dimmer pack has on the current waveform drawn by the lamp to see why this is true. As this series of power quality meter readings for several dim levels demonstrates, the harmonic distortion increases as the duty cycle of the SSR decreases and the lamp dims (source IATSE Local 481 Power Quality Workshop.)
Dimmer at 100%:
(http://www.screenlightandgrip.com/images/generators/CF_1kw_dim_100.jpg)
Dimmer at 85%:
(http://www.screenlightandgrip.com/images/generators/CF_1kw_dim_85.jpg)
Dimmer at 50%:
(http://www.screenlightandgrip.com/images/generators/CF_1kw_dim_50.jpg)
Dimmer at 30%:
(http://www.screenlightandgrip.com/images/generators/CF_1kw_dim_30.jpg)
Dimmer at 15%:
(http://www.screenlightandgrip.com/images/generators/CF_1kw_dim_15.jpg)
As is evident in this series of power quality readings, as the duty cycle of the SSR gate grows shorter, the current drawn by the 1kw becomes progressively more distorted eventually reaching a THD of 68.6% when the lamp is dimmed to 15%. It is also worth noting that, as the lamp is dimmed, the peak current drawn by the lamp increases. While counter intuitive, this fact becomes apparent if we look below at the Crest Factor of the current drawn by the 1kw lamp as it is dimmed (also from my 481 Power Quality Workshop.)
(http://www.screenlightandgrip.com/images/generators/CF_1kw_Crest_Factors.jpg)
(Right to left: Crest Factor of 1kw lamp at 100%=1.4, 85%=1.6, 50%=2.0, 30%=2.3, 15%=2.7)
At full the lamp draws a nice sinusoidal current waveform with a, to be expected, Crest Factor of 1.414. As the lamp is dimmed the Crest Factor increases until at 15% it reaches 2.7. If we were to calculate the Peak Current drawn by the 1kw lamp when dimmed to 30% we see that it is 24% higher than it is when not dimmed (see calculations below.) As the Crest Factor increases so does the current peak and amplitude of the current drawn by the 1kw.
Peak Current at 30% = 6.01A x 2.389 CF = 14.36A
Peak Current at 100% = 8.2A x 1.414 CF = 11.59A
14.36/11.59 Peak = 1.238 or 24% increase
With this level of harmonic distortion it is not surprising that breaker sensing errors occur, especially if the breaker is equipped with a peak detecting electronic trip unit like that in the Lex Company Switch mentioned above. Errors occur because peak detecting trip units calculate rms current by dividing the Peak Current by 1.414 to arrive at an rms value. This method accurately measures the heating effect of current when the current sine waves are perfectly sinusoidal. However, as current becomes increasingly non-sinusoidal, peak current measurement does not adequately reflect the true heating effect of the current and peak detecting breakers have a tendency to overprotect. Since, as we just saw above, the peak of the distorted current is usually higher than normal, this type of circuit breaker may trip prematurely at a lower RMS current.
Put another way, over protection, or “nuisance tripping”, occurs because the peak value of distorted current (fundamental plus harmonics) results in a calculated rms greater than the actual rms of undistorted current. If we take as an example the waveform of the 1kw dimmed to 30% above, the true rms value is 6.01A. But, based on the peak value of 14.36A (6.01A x 2.389 CF = 14.36A), above, the rms pure sinusoidal calculation performed by the breaker would yield 10.15A rms (14.36A /1.414 = 10.15A.) A peak sensing breaker rated at 10A rms would then not trip when the 1kw is not dimmed (8.2A), but would “nuisance trip” when the same light is dimmed to 30% because it erroneously senses the rms load to be more than 10A because of the high peak of the distorted waveform. From this we can extrapolate that between the distorted current drawn by non-pfc amps, power wedges, LEDs, and dimmed incandescent fixtures, Peak Current relative to rms current could increase to the point where it would trip the peak sensing unit in the Lex Company Switch prematurely.
Guy Holt, Gaffer
ScreenLight & Grip
www.screenlightandgrip.com
Breakers don’t trip without a reason, it’s just that we don’t always understand why they have tripped. A possible explanation why the Lex Company Switch tripped at 80% load could have to do with the elevated rms and peak current drawn by non-linear loads. The Lex company switch, like most company switches that use the newer electronic breakers, use the ABB T-Max with the “peak sensing” PR221DS-LS/I trip unit. Since the rms value of the harmonically distorted current drawn by non-linear loads can be 30-40% higher than that of linear loads, peak breakers will “nuisance” trip prematurely. The problem is not the breaker but the lighting technician that is distributing the load based upon paper calculations, ignoring the poor power factor of the loads they are powering. Even if the electricians distributed their loads using true rms current values that they measure, the elevated peak current of non-linear loads, which can be 20-25% higher than normal, will prematurely trip a peak-sensing electronic breaker at a 75-80% of it rated value.
The source of the harmonics could be non-power factor corrected amps, power wedges, LED Lighting, HMI Lighting, or dimmer packs. As more and more of the lights on a rig consist of non-linear LED lighting fixtures rather than linear incandescent fixtures, peak current has increased relative to rms current. For example, over half of the LED fixtures that I have tested at random (from the inventories of Boston area rental and lighting sales companies) were not power factor corrected (pfc.) With power factors ranging from .45 to .63, these fixtures draw roughly twice the current than one would expect using Ohm’s Law (W=VA) and can have Crest Factors as high as 2.7 where a normal sinusoid has a Crest Factor of 1.414.
(http://www.screenlightandgrip.com/images/generators/CF_Color_Blaze.jpg)
Even those LED fixtures that were power factor corrected generated harmonic currents when dimmed or switched to only one color emitter. For instance, the pfc of the Color Blaze above dropped from .99 to .60 when switched to only the red channel (to see which LED lights are power factor corrected or not, use this link - http://www.screenlightandgrip.com/html/emailnewsletter_generators.html#anchorHigh Output AC LEDs (http://www.screenlightandgrip.com/html/emailnewsletter_generators.html#anchorHigh Output AC LEDs) - to see some of the results of my tests.)
Dimmer racks will also generate higher peak currents when dimming incandescent fixtures than the fixtures do when they are not dimmed.
Where it is somewhat counter intuitive that a lamp draw a higher peak current when dimmed than it does when operating at 100%, let’s look at the effect that dimming a 1kw incandescent lamp with an SSR dimmer pack has on the current waveform drawn by the lamp to see why this is true. As this series of power quality meter readings for several dim levels demonstrates, the harmonic distortion increases as the duty cycle of the SSR decreases and the lamp dims (source IATSE Local 481 Power Quality Workshop.)
Dimmer at 100%:
(http://www.screenlightandgrip.com/images/generators/CF_1kw_dim_100.jpg)
Dimmer at 85%:
(http://www.screenlightandgrip.com/images/generators/CF_1kw_dim_85.jpg)
Dimmer at 50%:
(http://www.screenlightandgrip.com/images/generators/CF_1kw_dim_50.jpg)
Dimmer at 30%:
(http://www.screenlightandgrip.com/images/generators/CF_1kw_dim_30.jpg)
Dimmer at 15%:
(http://www.screenlightandgrip.com/images/generators/CF_1kw_dim_15.jpg)
As is evident in this series of power quality readings, as the duty cycle of the SSR gate grows shorter, the current drawn by the 1kw becomes progressively more distorted eventually reaching a THD of 68.6% when the lamp is dimmed to 15%. It is also worth noting that, as the lamp is dimmed, the peak current drawn by the lamp increases. While counter intuitive, this fact becomes apparent if we look below at the Crest Factor of the current drawn by the 1kw lamp as it is dimmed (also from my 481 Power Quality Workshop.)
(http://www.screenlightandgrip.com/images/generators/CF_1kw_Crest_Factors.jpg)
(Right to left: Crest Factor of 1kw lamp at 100%=1.4, 85%=1.6, 50%=2.0, 30%=2.3, 15%=2.7)
At full the lamp draws a nice sinusoidal current waveform with a, to be expected, Crest Factor of 1.414. As the lamp is dimmed the Crest Factor increases until at 15% it reaches 2.7. If we were to calculate the Peak Current drawn by the 1kw lamp when dimmed to 30% we see that it is 24% higher than it is when not dimmed (see calculations below.) As the Crest Factor increases so does the current peak and amplitude of the current drawn by the 1kw.
Peak Current at 30% = 6.01A x 2.389 CF = 14.36A
Peak Current at 100% = 8.2A x 1.414 CF = 11.59A
14.36/11.59 Peak = 1.238 or 24% increase
With this level of harmonic distortion it is not surprising that breaker sensing errors occur, especially if the breaker is equipped with a peak detecting electronic trip unit like that in the Lex Company Switch mentioned above. Errors occur because peak detecting trip units calculate rms current by dividing the Peak Current by 1.414 to arrive at an rms value. This method accurately measures the heating effect of current when the current sine waves are perfectly sinusoidal. However, as current becomes increasingly non-sinusoidal, peak current measurement does not adequately reflect the true heating effect of the current and peak detecting breakers have a tendency to overprotect. Since, as we just saw above, the peak of the distorted current is usually higher than normal, this type of circuit breaker may trip prematurely at a lower RMS current.
Put another way, over protection, or “nuisance tripping”, occurs because the peak value of distorted current (fundamental plus harmonics) results in a calculated rms greater than the actual rms of undistorted current. If we take as an example the waveform of the 1kw dimmed to 30% above, the true rms value is 6.01A. But, based on the peak value of 14.36A (6.01A x 2.389 CF = 14.36A), above, the rms pure sinusoidal calculation performed by the breaker would yield 10.15A rms (14.36A /1.414 = 10.15A.) A peak sensing breaker rated at 10A rms would then not trip when the 1kw is not dimmed (8.2A), but would “nuisance trip” when the same light is dimmed to 30% because it erroneously senses the rms load to be more than 10A because of the high peak of the distorted waveform. From this we can extrapolate that between the distorted current drawn by non-pfc amps, power wedges, LEDs, and dimmed incandescent fixtures, Peak Current relative to rms current could increase to the point where it would trip the peak sensing unit in the Lex Company Switch prematurely.
Guy Holt, Gaffer
ScreenLight & Grip
www.screenlightandgrip.com
Lovely.
It was a 100% incandescent load on professional solid state dimming. Full "bump on" test at 100% was 318 amperes/phase leg A, B; 321 amperes phase leg C.
I am suggesting to the venue that they directly engage LEX Products.
The issue my friend ran into was left unknown. They ended up switching to generator power both times, and it was left as a mystery.
Yes, sometimes breakers go bad, but don’t assume that because a rig will trip the 400A breaker of an arena company switch, but not the 400A breaker of a generator, that it means that the breaker on the company switch is defective. Another possible explanation is the fact that a rig consisting of predominantly non-linear loads with Switch Mode Power supplies, (non-pfc amps, power wedges, LED Displays, LED Lights, HMI discharge lights, etc.), will draw appreciably higher Peak Current, that is more likely to trip a peak sensing breaker, on grid power than on generated power. The RMS Current drawn by a SMPS can also be appreciably higher (20% more) on grid power than on generated power. The reason is the way in which the diode-capacitor front-end of SMPSs react to system impedance.
(http://www.screenlightandgrip.com/images/generators/ProSound_Hard_vs_Soft-2.5kw_SMPTs.jpg)
As illustrated above, the front end of SMPSs first convert AC into DC by feeding the AC input through a bridge rectifier, which inverts the negative half of the AC sine wave and makes it positive. The rectified current then passes into a conditioning capacitor that removes the 60 Hz rise and fall and flattens out the voltage - making it essentially DC. The DC is then fed to the Switch-mode converter, which shapes it into the desired waveform (square wave in the case of HMIs, high frequency sinusoid in the case of Fluorescents, or DC of a specific voltage.)
(http://www.screenlightandgrip.com/images/generators/ProSound_Hard_vs_Soft-2.5kw_Peak_Current.jpg)
(Top: AC Waveform, Middle: Rectified AC (dashed line) and Capacitor Charge (solid line.) Bottom: Current waveform drawn by linear loads (dashed line) and current waveform drawn by SMPTs (solid line.) Note high peak of current drawn.)
What accounts for the greater Peak Current? To obtain a low ripple on the DC output the smoothing capacitor of SMPSs must be relatively large. Since the smoothing capacitor can only charge when input voltage is greater than its stored voltage, the instantaneous line voltage is below the voltage on the capacitor most of the time. Consequently the time the capacitor has to charge is only a very brief period of the overall cycle time while voltage ascends to its’ peak. That is because, after peaking, the half cycle from the bridge drops below the capacitor voltage, which back biases the bridge, inhibiting further current flow into the capacitor. Thus the rectifiers conduct current only for a small portion of each line half-cycle and only during the ascending portion of the supply voltage waveform - which pulls the current out of phase with the voltage. And since, during this very brief charging period, the capacitor must also charge fully, pulses of current are drawn for short durations. The end result is that the current drawn from the mains is a series of narrow pulses that peak before the voltage peaks and whose Peak Current typically 5-10 times higher than the resulting DC value.
Given the way SMPS convert AC to DC, source impedance has the effect of attenuating the Peak Current and thereby reducing the current drawn. With low source impedance, the duration of the charging cycle is short and so the current drawn by the capacitor is high. Higher impedance does not allow as much current to be drawn in the same interval, and so the time it takes to charge the capacitor is extended. As a result, the Crest Factor of the current pulse drawn is reduced as well as the harmonics. With the reduction of Peak Current and current harmonics, comes an improvement in Power Factor and a more efficient use of power.
(http://www.screenlightandgrip.com/images/generators/ProSound_Hard_vs_Soft-2.5kw_Ballast_Sch.jpg)
(Conversion of AC to DC and then to an alternating square wave in an HMI ballast)
In a Power Quality workshop I developed for IATSE Local 481, I demonstrate this using a 2500W HMI lamp with a non-power factor corrected electronic ballast (see ballast schematic above) As is evident by the PQM reading on the right below, a 2.5kw HMI will draw 35.5A on grid power. Where as, on generated power the 2.5kw draws only 28.5A (the left PQM reading above.) What accounts for the 7A or 20% difference in current drawn? Generators typically have 15-20% internal reactive impedance. This is much higher than the local utility connection where the source impedance is established by the closest distribution transformer, with only 2-4% impedance being typical.
(http://www.screenlightandgrip.com/images/generators/ProSound_Hard_vs_Soft-2.5kw_Load.jpg)
(Left: Generated Power. Right: Grid Power)
Upon closer examination, we saw similar improvements in other measures of power quality and efficiency. For example the side-by-side comparison of the Crest Factors of the current drawn by the same ballast on the two power sources below shows an improvement in the Crest Factor as well. When powered by the grid the current drawn by the ballast has a Crest Factor of 2.5.
(http://www.screenlightandgrip.com/images/generators/ProSound_Hard_vs_Soft-2.5kw_CF.jpg)
(Left: Generated Power. Right: Grid Power)
When powered by the generator the current drawn by the ballast has a Crest Factor of 2.0. Given the relatively short charging period, the smoothing capacitor draws current in abrupt high amplitude pulses, which accounts for the high Crest Factor on the low impedance of the utility grid. Given a longer charging period, the same smoothing capacitor draws current over a longer period and so the current waveform is less peaked and has a lower Crest Factor.
(http://www.screenlightandgrip.com/images/generators/ProSound_Hard_vs_Soft-2.5kw_THD.jpg)
(Left: Generated Power. Right: Grid Power)
With the change in current waveform comes an improvement in the harmonics generated by the ballast as well. When powered by the grid, the current drawn by the ballast has a Total Harmonic Distortion of nearly 75%. When powered by the generator, the current drawn by the ballast has a THD of 55%. With the reduction of Peak Current and current harmonics, comes an improvement in Power Factor. When powered by the grid, the ballast has a Power Factor (PF) of .64. When powered by the generator, the ballast has a PF of .79.
(http://www.screenlightandgrip.com/images/generators/ProSound_Hard_vs_Soft-2.5kw_PF.jpg)
(Left: Generated Power. Right: Grid Power)
As we have seen here the introduction of impedance into the system reduces the crest factor as well as the harmonics. With the reduction of peak current and current harmonics, comes an improvement in efficiency and in Power Factor. So that the power consumer can know the efficiency of an electronic device under the worse case scenario, the PF given for an electronic devise in its technical specifications is for operation of the device on stiff utility power.
Does this mean that you should always use generated power if possible? No. The increased efficiency of electronic devices operating on generated power comes at the expense of voltage waveform distortion. Without the system inertia of the utility grid (low impedance) to nullify harmonic-induced electrical distortions, generated power is more susceptible to voltage waveform distortion from non-linear loads. When designing power distribution systems for non-linear loads, you must weigh the advantage of higher efficiency against the consequent voltage waveform distortion (use this link - http://www.screenlightandgrip.com/html/emailnewsletter_generators.html#anchorVoltage%20Waveform%20Distortion (http://www.screenlightandgrip.com/html/emailnewsletter_generators.html#anchorVoltage%20Waveform%20Distortion) - for more details about voltage distortion caused by non-linear loads.)
Guy Holt, Gaffer
ScreenLight & Grip
www.screenlightandgrip.com