Potential Transformers, HV measurement. / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Potential Transformers, HV measurement.

previous 1 2 3 4 next last

Move Thread

LAN_403

radiotech

Tue Sept 14 2010, 08:03PM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

In the tranformer you are testing is there 180 degree displacement between the two halves of the secondary?

ie <---------------*--------------->

Patrick

Tue Sept 14 2010, 08:23PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

yes each one should contribute 6000 volts 180 deg out of phase. so they should add up too about 12,000 V

i wonder if the stray capacity of the fluke73III (<50pF, 10.20Mohm) and the 525.1 Mohm HV resistor stray C arnt dividing right, therefore when i keep the meter near or at ground that capacity doesnt mater,(keeping the meter-side grounded is safer anyway.) this may explain the bogus measurement.

radiotech

Wed Sept 15 2010, 03:21AM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

With you DMM ,can you measure the current throught the 520 meg resistor? It should be 24 uA with 12000 volts across it. If you meter has a 200uA AC range. it should be possible. By doing this you can forget voltage division and just concentrate on current. Any meter capacitance shunting the current terminals can be ignored. If you can, crank the NST slowly up noting current using that current times 520 megs as voltage.. Put a small neon lamp across the current terminals.

Patrick

Wed Sept 15 2010, 03:23AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

radiotech wrote ...

With you DMM ,can you measure the current throught the 520 meg resistor? It should be 24 uA with 12000 volts across it. If you meter has a 200uA AC range. it should be possible. By doing this you can forget voltage division and just concentrate on current. Any meter capacitance shunting the current terminals can be ignored. If you can, crank the NST slowly up noting current using that current times 520 megs as voltage.. Put a small neon lamp across the current terminals.

dont have a 200 uA range though.

i do have a 40mA range though, and variac.
should i make a new HV resistor to produce 10-15mA @ 12kV in? EDIT: nope on the 10mA, too many watts

EDIT: i measured the stray cap of the resistor by placing it on copper sheet zero the meter, then coonect meter to one end of the resistor then the other lead to the copper sheet , i get 13.1pF or there about. i dont know if thats how to measure capacitence of resistors though.

when i assume 13pf for the resistor and 40pf for the meter i get very close to 18kv, in my simulator. i think the capacitence is a divider that isnt even close to the division ratio of the resistor+meter so the meter is trying to Vrms' a current + voltage wave thats not in phase. there by either peak reading or bogus reading.

radiotech

Wed Sept 15 2010, 09:52PM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

I think you are on the right track with the resistor. The measurement of "stray capacitance" at 60 Hz might be done by comparing the resistance's know value with its actual found impedance. and the deciding on which model to use.

You could say it is a resistor in series with a capacitor, or one in parallel. working back thru |Z|= sqr(R^2 + Xc^2) or |Z| = R/(sqr(1+(39.5*R^2*C^2)) respectively. with angle= arctan 1/377*R*C or -377*R*C respectively.

A voltage across two terminals is |E| ie, o-----|E|-----o so if the Fluke reads 10 Volys, it is 10 volts no matter how much capacity is across the terminals of the Fluke. However you
*Cannot* know that was 10 volts before you measure or after you stop.

Patrick

Thu Sept 16 2010, 12:11AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

this is complicated

i may be able to get my professors to let me use a HP 34401A, how should i connect this for an accurate uA measurement? it must be grounded, so i guess ill be making 6000 + 6000 = 12000, kind of measurements as measurements across the full secondary while center tap grounded would suicide the meter.

this is a link to the manual (HP 34401A), the pdf is too large to post directly here on the forum.

high voltage probe design:

Fluke 79 ii manual and specs:
]79_29___umeng0200.pdf[/file]

EDIT @ 7pm:
wooooo hooooo !!!! i measured 12uA with the Fluke 79 ii .

sec A to center-tap ground: (through 525.1M)
11uA @ 108 Vac --> (11uA x 525.1M = 5776V)
13uA @ 123 Vac --> (13uA x 525.1M = 6826V)
14uA @ 136 Vac --> (14uA x 525.1M = 7351V)
though only 2 sig figs are justified (5.8, 6.8, 7.3 kV) it seems that each 1uA = 525V

sec B to center-tap ground: (through 525.1M)
12uA @ 108 Vac --> (12uA x 525.1M = 6301V)
14uA @ 123 Vac --> (14uA x 525.1M = 7351V)
15uA @ 136 Vac --> (15uA x 525.1M = 7876V)
though only 2 sig figs are justified (6.3, 7.4, 7.9 kV) it again seems that each 1uA = 525V (V per T ?)

notice how Sec B seems to read just a bit higher, it did this in both the series current, and voltage division measurements. this sort of gives me confidence. [ NST secondary = <A--------*--------B> ]

Also i could not reverse the resistor+meter [ from R+Vm, to Vm+R ] posistion and get the same mesurement, the meter always had to be ground referenced. other wise i would get 12uA one way and a jumpy 40-80uA the otherway. if the capacitence were not relevent then the two series posistions and measurements would yield the same result, but did'nt.

(making a ground referenced "half-measurement" is safer for me and the instrument anyway.)

________________________________________________ _____________________________________

from this experiment i have learned four profound lessons:

first, measuring current in the uA range through a precision high Mohm resistor is more accurate, stable, and precise (easier too) then resistive voltage division.

second, the meter should always be ground referenced (lo noise, better human saftey anyway.) with the resistor sourcing current from the high side.

third, the stray hand capacitence doesnt effect the uA meter at all, while the volt meter flickers counts wildly up and down from my hand cpacitence and from a grounded sheet copper i moved closer and farther.

fourth, the impdance derivations radiotech provided, indicated that the meter had an impedance of 8.4M but fluke guarantees its greater then 10M, and ifound it to be near 10.20M, also the real voltage division ratio was a precise 51.4, while a derived/imperical 68.7 was found to work for secondary B and the exact same setup yielded 63.5 division ratio for secondary A, obviously a wild variance, probably due to high impedance holding up stray capacitience, to a false high voltage.the divider should have a constant ratio to be useful,but it doesnt,it varies with the voltage its measuring, so if we draw a gain plot im sure we would get wild variations up and down all over.
____________________________________________ _________________________________________

i am not done with this thread, not by a long shot, i need others to look at what ive done and i need feedback, but ive learned so much, i would like to compare my series current measures with a PT, just too see how accurate and precise my conclusions are but i dont have a PT.

next up i will build a 60M ohm 9 kv resistor! should give me 122uA @ 7,000V my meter will then be accurate to
(+- (1.5% of reading + 4 LS counts)) thats about 6090V or 5910V for a 6000V input thats pretty accurate i think.....i mean an uncertainty of 180V per 6kV, thats dam good, good enough for a calibration target...

Patrick

Fri Sept 17 2010, 09:27PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

61.21 Mega Ohm @ 9kV made. will post table data soon.

radiotech

Fri Sept 17 2010, 10:54PM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

Here's some HV resistor methods.

Patrick

Sat Sept 18 2010, 01:44AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

sec A to center-tap ground: (through 61.21M)
095uA @ 108 Vac --> (095uA x 61.21M = 5815V)
107uA @ 123 Vac --> (107uA x 61.21M = 6549V)
118uA @ 136 Vac --> (118uA x 61.21M = 7223V)
125uA @ 146 Vac --> (125uA x 61.21M = 7651V)
Now 3 sig figs are justified (5.8, 6.55, 7.22, 7.65 kV) it seems that each 1uA = 61.2V

secA:---61M----525M
108V---5.8------5.8 kV
123V---6.55----6.8 kV
136V---7.22----7.3 kV
146V---7.65----x.x kV

sec B to center-tap ground: (through 61.21M)
097uA @ 108 Vac --> (097uA x 61.21M = 5937V)
109uA @ 123 Vac --> (109uA x 61.21M = 6672V)
120uA @ 136 Vac --> (120uA x 61.21M = 7345V)
127uA @ 146 Vac --> (127uA x 61.21M = 7774V)
Now 3 sig figs are justified (5.9, 6.67, 7.35, 7.77 kV) it seems that each 1uA = 61.2V

secB:---61M----525M
108V---5.9------6.3 kV
123V---6.67----7.4 kV
136V---7.35----7.9 kV
146V---7.77----x.x kV

radiotech

Sat Sept 18 2010, 02:30AM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

Another thing to look at is the apparent turns ratio deviation from the lowest excitation to the highest. This might indicate the shifting of the core operating point. Seems to be 1.03 with the last set of numbers a tad lower. Your PT should be compared the nst for this.

previous 1 2 3 4 next last

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob