quogueelectric
Senior Member
- Location
- new york
Some of sq d heater sections are designed to wrap the feed through the ct multiple times to get the right ratio to trip.wptski said:
GFCI & heater problems
- Thread starter Greg Swartz
- Start date
- Status
mdshunk
Senior Member
- Location
- Right here.
I don't know what you're talking about. I've got meters that are older than me that I can wrap through the loop a dozen times to get the required sensitivity and still pass a baseball through.wptski said:
Think outside the box...
mdshunk
Senior Member
- Location
- Right here.
Maybe you're blackballed by the internet Gods? I just tried it again, and it loads in about 4 seconds. Better luck next time.wptski said:How come I'm getting a "page not found" on that link?
wptski
Senior Member
- Location
- Warren, MI
I got beat up pretty badly in this forum about the same subject of leakage current and GFCIs because the probe I was using at the time wouldn't read zero, so it was declared inaccurate by the majority here.mdshunk said:
You might be looking for something like 5-7ma. If you think that you can unearth some old relic of a current probe and increase its specs by looping some wire your sadly mistaken. You'd need a probe with good specs to begin with.
wptski
Senior Member
- Location
- Warren, MI
I was in the mist of replying, opened IE a second time and the link wouldn't work. Just tried it with IE opened twice and it worked. Something to do with replying at the same time?:-?mdshunk said:
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
080519-1230 EST
Just ran an experiment on an old Fluke Y8101 AC Current Transformer. The meter is a Fluke 27 with a resolution of 0.1 uA. The ratio is 1 to 1000.
Calibration is about correct at 10 A (1 turn --- meaning clip probe around 1 wire). At 10 MA calibration is way off. Reading is 1.7 MA. At 2 turns it is 3.8/2 = 1.9 MA (actual meter reading 3.8 uA). At 5 turns it is 11.7/5 = 2.3 MA.
.
Just ran an experiment on an old Fluke Y8101 AC Current Transformer. The meter is a Fluke 27 with a resolution of 0.1 uA. The ratio is 1 to 1000.
Calibration is about correct at 10 A (1 turn --- meaning clip probe around 1 wire). At 10 MA calibration is way off. Reading is 1.7 MA. At 2 turns it is 3.8/2 = 1.9 MA (actual meter reading 3.8 uA). At 5 turns it is 11.7/5 = 2.3 MA.
.
Greg Swartz
Senior Member
- Location
- Colorado Springs, Colorado
Not to throw a curve ball...
Not to throw a curve ball...
Not to throw a curveball in Marc's age old meters...
But, I went to the customer's house today and talked at length to him, only to find out (insert foot here) that the BREAKER was tripping...
So, we talked some more. As it turns out, he is using a GFCI that I installed for an IR heater for him, for his pond pump also. The IR heater works fine. Never had a problem. The pool pump trips the breaker every so often. 3 times Saturday night, and 2 times in the week before that.
The other GFCI previously installed (by another contractor) is working fine with his other pool pump.
So, I went and installed a new breaker. I told him that the IR heater (1500VA on 20A circuit) works fine, and so does the pump. If there are any more problems, they are with the pump. Since I did the install last August, and no other problems, I am quite confident of that.
So, I replaced the breaker, and as I was leaving, he stopped me and asked... "So, how much do I owe you..."
I told him it was on the house, but not next time...
(*tears in my eye*) I just love customers like that...
Greg
Not to throw a curve ball...
Not to throw a curveball in Marc's age old meters...
But, I went to the customer's house today and talked at length to him, only to find out (insert foot here) that the BREAKER was tripping...
So, we talked some more. As it turns out, he is using a GFCI that I installed for an IR heater for him, for his pond pump also. The IR heater works fine. Never had a problem. The pool pump trips the breaker every so often. 3 times Saturday night, and 2 times in the week before that.
The other GFCI previously installed (by another contractor) is working fine with his other pool pump.
So, I went and installed a new breaker. I told him that the IR heater (1500VA on 20A circuit) works fine, and so does the pump. If there are any more problems, they are with the pump. Since I did the install last August, and no other problems, I am quite confident of that.
So, I replaced the breaker, and as I was leaving, he stopped me and asked... "So, how much do I owe you..."
I told him it was on the house, but not next time...
(*tears in my eye*) I just love customers like that...
Greg
mdshunk
Senior Member
- Location
- Right here.
The Simpson 150-2, the meter I was thinking of as a "for instance", actually includes that process in the instructions. Thanks for your opinion anyhow. Looping though a CT is a pretty dog-gone common thing to do.wptski said:
mdshunk
Senior Member
- Location
- Right here.
What's the date of the last NIST tracable calibration?gar said:
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
080519-2131 EST
Marc:
None.
But clearly as it stands the device is not well suited to making low current measurements. The 10 A test shows expected results, and obviously there are flux coupling problems at low currents. The kind of flux error that is occuring at 1 MA is not obvious at 10 A without a careful measurement, and we may not be working with a linear device at these low current levels. In other words flux coupling is probably not linear with flux density. Also there are core losses of concern. 1 MA provides very little energy to support core losses.
The core materials used and the use of a split core reduces the coupling from primary to secondary.
A device with a 150 A full scale rating and looking at 1 MA is 1 part in 150,000. I am not referring to resolving 1 MA at 150 A just the relative scale.
It was clear that as I increased the current level the calibration constant was improving, but only slowly.
.
Marc:
None.
But clearly as it stands the device is not well suited to making low current measurements. The 10 A test shows expected results, and obviously there are flux coupling problems at low currents. The kind of flux error that is occuring at 1 MA is not obvious at 10 A without a careful measurement, and we may not be working with a linear device at these low current levels. In other words flux coupling is probably not linear with flux density. Also there are core losses of concern. 1 MA provides very little energy to support core losses.
The core materials used and the use of a split core reduces the coupling from primary to secondary.
A device with a 150 A full scale rating and looking at 1 MA is 1 part in 150,000. I am not referring to resolving 1 MA at 150 A just the relative scale.
It was clear that as I increased the current level the calibration constant was improving, but only slowly.
.
Greg Swartz
Senior Member
- Location
- Colorado Springs, Colorado
If I remember my electromagnetics class right... it is not linear. There is a constant involved, but I believe that constant is squared if not cubed (area & volume)gar said:
http://www.iit.edu/~smile/guests/gsmxsec1.htm <--all about Maxwell and his equations. Good suff!
parts or ratio?gar said:
If I am understanding you correctly, 1MA:150A == 1:150,000 ?
Greg
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
080520-0758 EST
Greg:
Generally one considers a transformer as a linear device from 0 input up to the area where saturation becomes significant. This may not be true for a ferromagnetic device at very low flux densities. Looking at the magnetization curves for some materials it may indicate that permeability drops substantially at low flux densities. This would cause an increase in leakage flux and thus an increase in the primary to secondary impedance.
When one has a very high permeability material, iron, the flux coupling between primary and secondary is very good and small variations in permeability are not important. Pure iron is 275,000, silicon steel is 66,000, and Supermalloy is 1,050,000 from Attwood "Electric and Magnetic Fields". But these are or may not be the values at very low flux densities.
Core loss is a complicated area and Attwood discusses this on p 350. He also discusses "The Closed Ring --- Differential Permeability" on page 355. This is concerned with leakage flux with a non-uniformly wound coil on a core.
Yes I was referring to the ratio of 1 MA to 150 A.
Never really gave any consideration to the following question:
Is an ordinary resistor, a piece of wire, linear in respect to its e-i curve for very low current levels?
Marc:
If you take your old current transformer and pass one wire thru it with 1 MA of current what is the output reading, the scale factor, and what is the full scale rating of the transformer?
.
Greg:
Generally one considers a transformer as a linear device from 0 input up to the area where saturation becomes significant. This may not be true for a ferromagnetic device at very low flux densities. Looking at the magnetization curves for some materials it may indicate that permeability drops substantially at low flux densities. This would cause an increase in leakage flux and thus an increase in the primary to secondary impedance.
When one has a very high permeability material, iron, the flux coupling between primary and secondary is very good and small variations in permeability are not important. Pure iron is 275,000, silicon steel is 66,000, and Supermalloy is 1,050,000 from Attwood "Electric and Magnetic Fields". But these are or may not be the values at very low flux densities.
Core loss is a complicated area and Attwood discusses this on p 350. He also discusses "The Closed Ring --- Differential Permeability" on page 355. This is concerned with leakage flux with a non-uniformly wound coil on a core.
Yes I was referring to the ratio of 1 MA to 150 A.
Never really gave any consideration to the following question:
Is an ordinary resistor, a piece of wire, linear in respect to its e-i curve for very low current levels?
Marc:
If you take your old current transformer and pass one wire thru it with 1 MA of current what is the output reading, the scale factor, and what is the full scale rating of the transformer?
.
weressl
Esteemed Member
Greg Swartz said:
I am not certain what pond heaters are and what NEC article would they fall under but here are some thoughts.
Some heaters are recognized by the NEC as having higher leakeage current than normal circuits would do, so there is allowance for those circuits to be protected by special high mA trip protectors.
You may try to segreagte the heaters into individual circuits and feed each one of those with separate GFCI's.
You may try to disconnect some of the branches, starting at the end, and see if the heaters hold up.
Measuring the leakeage current or reducing the number of heaters on a single circuit are both crapshoots, since the leakeage current changes with the weather, soaked ground, humidity, etc. so you may make your customer aware of this. This way you show diligence and he would not be reluctant to call you or give up on your efforts of trying to do your best to remedy the problem.
- Status