Harmonics Losses through Voltage Dividers

Status
Not open for further replies.
aelectricalman

aelectricalman

Senior Member
Location
KY
Has anyone heard of voltage dividers causing harmonic skewing during power quality testing? I understand the use of CCVT's will skew harmonics because of the capacitive coupling but what about when there is no capacitive coupling in the XFMR? What if you just have resistive elements? If I remember correctly (from school) the resistive elements will not cause any harmonic concerns during testing. With this being said, my intention is to pq monitor a 7900V shielded line with a Dranetz PX5 or a Fluke 1750. I will have to include PT's or V dividers for the potentials. The intent of my monitoring is to find the harmonic spectrum and Vthd/Ithd provided by the utility and on the load side of the customers 24 pulse drive.
 
A resistive divider will work fine for harmonic rich waveforms up to the point where stray capacitance becomes significant. This will happen sooner at very high resistance values like the 10 megohms in a scope probe. Normally not a problem at power line harmonic frequencies.
A transformer will have a constant ratio (= turns ratio) up to the frequencies where the capacitance becomes an issue, but may mess up at low frequencies where the core saturation is an issue.
An auto transformer is more likely to deliver a constant ratio than an isolation transformer would.
If you do not need isolation, I would recommend a resistive divider over a PT.
 
140716-2146 EDT

Attached is a photo of a homemade high voltage probe good for in excess of 30,000 V and compensated to about 1 MHz. This was made in 1955 for research on automotive ignition systems.

Resistive dividers for wideband measurements use a combination of a capacitive divider in parallel with the resistive divider. The ratio of both the resistive and capacitive dividers are made equal. This is done experimentally by adjusting the input capacitance of your instrument. Usually the adjustment capacitor is in the scope probe connector, or in the shank of the probe.

You might possibly build just a capacitive divider for AC measurements. This would be a function of how high an input resistance you could create at the input of your instrument.

.
 

Attachments

  • PICT3681.jpg
    PICT3681.jpg
    56.4 KB · Views: 0
GoldDigger said:
A resistive divider will work fine for harmonic rich waveforms up to the point where stray capacitance becomes significant. This will happen sooner at very high resistance values like the 10 megohms in a scope probe. Normally not a problem at power line harmonic frequencies.
A transformer will have a constant ratio (= turns ratio) up to the frequencies where the capacitance becomes an issue, but may mess up at low frequencies where the core saturation is an issue.
An auto transformer is more likely to deliver a constant ratio than an isolation transformer would.
If you do not need isolation, I would recommend a resistive divider over a PT.
Click to expand...

Could you please take a second to explain the phenomenon of stray capacitance? Its been a while. Is this from harmonics caused by improper grounding? or something different? At any rate, thanks for the great info!!!
 
If I take a simple straight insulated wire and run it next to a grounded surface there will be some capacitance from that wire to ground. That capacitance will be in parallel with any load resistance attached to the wire and can usually be ignored.
But if the load is a high resistance that is part of a voltage divider, the capacitive reactance may be significant at high frequencies.
Also, if I have a very long run of wire the capacitance to ground may pull enough current at high harmonic frequencies to trip a GFCI. Or the capacitive coupling to an adjacent unused wire may develop a phantom voltage reading seen by a non-contact or high impedance voltage detector.
All of these would be the result of "stray" (unintended) capacitance.
 
Status
Not open for further replies.
Top