Blinking Lights

Status
Not open for further replies.

markstg

Senior Member
Location
Big Easy
Test results with 350 mfd capacitor.

Capacitor current: 32A
Line Current on Compresssor start amps: 92A (reduced from 120A)


Unfortunately it was a start capacitor, and i couldn't get into the house to see if lights blinked, before it blew.

Need to find a run capacitor.
 

gadfly56

Senior Member
Location
New Jersey
Occupation
Professional Engineer, Fire & Life Safety
I think kwired has already put it best, but I'll try an analogy.

After start up, your AC motor is steadily downing beer by the pint. During start up, it's trying to chug a 55 gallon drum. You are trying to help things out by throwing another tablespoon out there with your capacitor. It's unlikely to make a noticeable difference.
 

kwired

Electron manager
Location
NE Nebraska
Test results with 350 mfd capacitor.

Capacitor current: 32A
Line Current on Compresssor start amps: 92A (reduced from 120A)


Unfortunately it was a start capacitor, and i couldn't get into the house to see if lights blinked, before it blew.

Need to find a run capacitor.
Capacitors have inrush current when first put across an A/C voltage as well. Your mentioned capacitor is drawing more current then the motor load it is supplementing, my guess is you have also overcorrected and will have a leading power factor, on this particular circuit anyway, when running.

I still don't think it will help with the voltage dip when starting this load, and if too large of a capacitor is used could even be worse at startup.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
The measurements themselves showed that the line current was noticeably lower at starting then without the cap. Whether that makes a noticeable difference in flicker and what problems the overcorrection while running cause are still open questions.
Do not confuse the one time DC discharge energy from the cap with the reactive current it supplies every cycle.
 

kwired

Electron manager
Location
NE Nebraska
Do not confuse the one time DC discharge energy from the cap with the reactive current it supplies every cycle.
I tried to say something like that earlier. A capacitor is storage device for DC circuits. For AC circuits it is an impedance, and a place for exchange of apparent power when used to correct power factor. Power factor at locked rotor and power factor at normal running load will not be same, and it will only reduce apparent power. It will have some resistive component to it which is real power, and again the inrush at motor startup is not all reactive power, the real power portion of that current is still going to load down the supply.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
140717-1130 EDT

The light flicker is a result of a transient condition (motor turning on). Power factor is a term that that relates to steady state conditions. It is hard to justify the concept of power factor during the occurance of a time varying non-linear transient load change (startup of a motor). Some form of transient analysis is needed.

.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
140717-1130 EDT

The light flicker is a result of a transient condition (motor turning on). Power factor is a term that that relates to steady state conditions. It is hard to justify the concept of power factor during the occurance of a time varying non-linear transient load change (startup of a motor). Some form of transient analysis is needed.

.
Transients occur on two time scales, sub-cycle and multi-cycle.
The capacitor will have little or no effect on sub-cycle transients. But for multi-cycle transients (starting current lasting from a large fraction of a second to several seconds) a good part of the transient analysis can be approximated by steady state formulas in which the coefficients are changing with time.

If the motor draws a high current for 20 cycles, for example, before the rotor movement starts to affect the motor current through induced back EMF, it can be useful to look at the power factor for that 20 cycle period. Whether you use transient analysis or approximate as steady state, the current for one cycle will still consist of resistive, leading and lagging current contributions from the motor and the capacitor.

It is also important to differentiate between lights "flckering" (very short term) and lights "dimming and returning to their original brightness" (blinking?) (brightness reduction for a noticeable fraction of a second or more.) And the return to full brightness may be perceived as an increase in brightness. Or the brightness may actually increase long term if there is a high neutral resistance and the motor load is 120V.
Let's see it the OP can clarify what is happening to the light intensity.
 
Last edited:

markstg

Senior Member
Location
Big Easy
I would characterize the light decrease as dimming, not a flicker, flicker being where the lighting intensity varies over the interval of light instability. The time duration of the light anomoly is less than 1 sec.
 

markstg

Senior Member
Location
Big Easy
If the motor draws a high current for 20 cycles, for example, before the rotor movement starts to affect the motor current through induced back EMF, it can be useful to look at the power factor for that 20 cycle period. Whether you use transient analysis or approximate as steady state, the current for one cycle will still consist of resistive, leading and lagging current contributions from the motor and the capacitor.

This was the basis for my calculation in an earlier post.
 

ATSman

ATSman
Location
San Francisco Bay Area
Occupation
Electrical Engineer/ Electrical Testing & Controls
Thyristor Timing Issues

Thyristor Timing Issues

The problem is the failure to apply dynamic power factor correction (SCR controlled capacitor) to prevent flicker: the capacitor switching by thyristor should be applied in this case. This paper sheds light on the topic:http://www.epcos.com/blob/530844/download/2/pdf-dynamicpfcan105.pdf

Sahib,
Thanks for this interesting article. Years ago I was involved with an investigation of power fuses blowing on robot welding equipment at an auto assembly plant. Using a PQ recording instrument made by Reliable Power Meters (since then I think they were bought out by Fluke) and after weeks of measurements we determined that the random firing of the thyristors in the welding equipment was causing the intermittent blowing of the input fuses to these units. We ended up recommending electrical interlocking be installed between each unit preventing multiple units from firing at the same time. This was implemented and the fuse problems disappeared.
I am wondering if installing these devices by Epcos could have accomplished the same thing.
 

markstg

Senior Member
Location
Big Easy
I also think it is saying that Static capacitors although can reduce to some extent the flicker, they cannot stop it. Which is also what the others have been saying.

So unless someone thinks it still worth trying, I'm going to save my $50 on run capacitors, and spend my money on a hard start kit which I hope will lessen the flicker (A proven method and won't mess with my warranties). Also, even if the capacitor mitigated the flicker to be unnoticable, I cannot leave the capacitor in the circuit all the time, since this presents another can of worms (overvoltage, capacitor life, unkown effects on electronics). On the test I noticed my Wi Fi stopped and needed to reset the router, and my digital thermostat dispay was showing heating and cooling.

thanks for all the input.
 

Sahib

Senior Member
Location
India
markstg:
An incandescent bulb light output is very sensitive to voltage fluctuation, even though the voltage fluctuation is small. So if you replace it with CFL, the flicker may become tolerable to you.
 
Last edited:

Sahib

Senior Member
Location
India
Sahib,
Thanks for this interesting article. Years ago I was involved with an investigation of power fuses blowing on robot welding equipment at an auto assembly plant. Using a PQ recording instrument made by Reliable Power Meters (since then I think they were bought out by Fluke) and after weeks of measurements we determined that the random firing of the thyristors in the welding equipment was causing the intermittent blowing of the input fuses to these units. We ended up recommending electrical interlocking be installed between each unit preventing multiple units from firing at the same time. This was implemented and the fuse problems disappeared.
I am wondering if installing these devices by Epcos could have accomplished the same thing.

Perhaps the fuses were undersized or the high inrush current was caused by wrong switching instant of thyristors (the EPCO thyristor modules switch the capacitors at the zero crossing of the current, thus avoiding inrush currents).

The multiple units firing at the same time might have occured to reduce the overall peak current draw from the supply.
 

kwired

Electron manager
Location
NE Nebraska
I would characterize the light decrease as dimming, not a flicker, flicker being where the lighting intensity varies over the interval of light instability. The time duration of the light anomoly is less than 1 sec.
Do they dim briefly then return to/near same level they previously was? That would be because of voltage drop during the increased demand occuring while your AC unit is starting. The motor is drawing more power during this short time then it is drawing while running, true power factor will have some contribution but the bulk of the draw is still true power. You are only going to see significant lesser effects by reduced voltage starting or other method that will actually reduce power draw while starting.

I also think it is saying that Static capacitors although can reduce to some extent the flicker, they cannot stop it. Which is also what the others have been saying.

So unless someone thinks it still worth trying, I'm going to save my $50 on run capacitors, and spend my money on a hard start kit which I hope will lessen the flicker (A proven method and won't mess with my warranties). Also, even if the capacitor mitigated the flicker to be unnoticable, I cannot leave the capacitor in the circuit all the time, since this presents another can of worms (overvoltage, capacitor life, unkown effects on electronics). On the test I noticed my Wi Fi stopped and needed to reset the router, and my digital thermostat dispay was showing heating and cooling.

thanks for all the input.
The main component of a hard start kit is a capacitor - it's intent is to provide more capacitance in parallel with the existing motor run capacitor - this will result in more starting torque. The motor will start faster, but will still have a high current demand when first energized - which is your problem with the lights dimming. The starting time will decrease - it is possible the starting current will actually increase with the hard start kit. After all power is work/time, speed up the starting time but with same work accomplished means power will go up for that duration as compared to the longer duration for a slower start.

You are still accelerating from zero to full speed, just in a shorter time period. Reduced voltage starting is also accelerating from zero to full speed, but during a longer time period. Same work, different time period for either case means different power demand.

At my house I have a 5 ton, 2 stage scroll compressor condensing unit, and my house lights blink on the unit starting.
Specs are 118 LRA, 230V, 23 FLA.

Lamps are incandescent 130 V.

Measurements at the unit disconnect switch are:
Normal Voltage 246 V

Starting Voltage: 231 V
Starting Amps: 121 A

Running Amps: 10 A (Low Stage)

Starting time: 1 second.
Start is on Low stage.

This is on a 1 year old house and has done this since installed.
There is no issue when the compressor goes from low stage to high stage.

Neutrals have all been checked, Meter, Panelboard, AC unit/disconnect, and are all tight.

If I place a hard start (start assist) kit (Capacitor and Start Relay) on the compressor, will it reduce the line current to the unit, so that my service doesn't see the full compressor start current? If so, any guess at how much less.

This was the OP. What has not been mentioned yet is transformer size, how much other load is connected to it, size and length of supply conductors. My experiences have been you need to address those issues when you are having voltage drop issues. Had a customer with a commercial garage one time that complained whenever the air compressor started the lights dimmed. POCO transformer was maybe a couple hundred feet of conductor away before even hitting the customer's individual lateral that was maybe about 100 feet long. Talked POCO into installing a closer transformer and it made a big difference.
 

markstg

Senior Member
Location
Big Easy
This was the OP. What has not been mentioned yet is transformer size, how much other load is connected to it, size and length of supply conductors. My experiences have been you need to address those issues when you are having voltage drop issues. Had a customer with a commercial garage one time that complained whenever the air compressor started the lights dimmed. POCO transformer was maybe a couple hundred feet of conductor away before even hitting the customer's individual lateral that was maybe about 100 feet long. Talked POCO into installing a closer transformer and it made a big difference.



50KVA, 7 Houses on overhead lateral, my service drop from lateral tap to service is 80ft., lateral distance tap to transformer is 120ft. POCO conductors are aluminum don't know size. POCO not interested in my blinking lights with such small voltage drop.
 

kwired

Electron manager
Location
NE Nebraska
50KVA, 7 Houses on overhead lateral, my service drop from lateral tap to service is 80ft., lateral distance tap to transformer is 120ft. POCO conductors are aluminum don't know size. POCO not interested in my blinking lights with such small voltage drop.
So you have a fairly decent sized source, about 200 feet of conductor, the overhead portion is likely 1/0 at the largest, the underground probably 4/0. I can almost assure you if you cut that distance in half you will see pretty significant result in the amount of light dimming you notice. It is supported just with some of the data from your OP alone, you are throwing an instantaneous current of 121 amps on the system, that is causing voltage to drop about 15 volts, but most of that current diminishes within a second or so. You did not mention running voltage that I am aware of but it is probably within a volt or two of what it was initially. That kind of drop that fast will be noticeable from an incandescent lamp, then it goes 95+% back to where it was just about as fast, which will also be pretty noticeable. If you were to soft start that motor, you would still have a drop of maybe a couple volts from no load to full load, but will be gradual enough it is not noticeable via lamp performance.

The other option is to somehow introduce additional power near the load during starting, but a capacitor is not a source or a storage device for AC current.
 
Status
Not open for further replies.
Top