Method and apparatus for detecting arcing in electrical connections by monitoring high frequency noise

• US 5,223,795 A
• Filed: 07/30/1992
• Issued: 06/29/1993
• Est. Priority Date: 07/30/1992
• Status: Expired due to Term

• Current Assignee: Siemens Industry Incorporated (Siemens AG)
• Original Assignee: Frederick K. Blades
• Inventors: Blades, Frederick K.
• Primary Examiner(s): Strecker, Gerard R.
• Assistant Examiner(s): DO, DIEP N
• Application Number:
  US07/921,829
• Time in Patent Office: 334 Days
• Field of Search: 324/520, 324/521, 324/522, 324/525, 324/536, 324/613, 324/650, 361/111, 361/113, 340/650, 340/658
• US Class Current: 324/536
• CPC Class Codes: G01R 31/1272 of cable, line or wire insulation, e.g. using partial discharge measurements locating faults in cables G01R31/083
• CPC Class Codes: G01R 31/52 Testing for short-circuits, leakage current or ground faults
• CPC Class Codes: G01R 31/58 Testing of lines, cables or conductors testing of electric windings G01R31/72
• CPC Class Codes: G01R 31/66 Testing of connections, e.g. of plugs or non-disconnectable joints testing for incorrect line connections G01R31/55
• CPC Class Codes: H02H 1/0015 Using arc detectors
• CPC Class Codes: H02H 1/04 Arrangements for preventing response to transient abnormal conditions, e.g. to lightning or to short duration over voltage or oscillations; Damping the influence of dc component by short circuits in ac networks

Abstract

An arc detector for detecting potentially hazardous arcing in electrical connections comprises detection and signal processing circuitry for monitoring high-frequency noise on the power line characteristic of arcing and distinguishable from other sources of high-frequency noise. If high-frequency noise is present and a gap occurs every half-cycle of the power frequency, arcing is determined to be present, and an alarm is given.

Claims

1. A method for detecting the presence of electrical arcing on an AC power line of frequency F and period T=1/F, comprising the following steps:

monitoring the line voltage or the line current as a waveform;

monitoring the waveform for the presence of high-frequency noise, said high-frequency noise being substantially higher in frequency than F;

monitoring the waveform for a gap of duration t_b wherein substantially no high-frequency noise is present, t_b being less than or equal to T/4;

determining that arcing exists if high-frequency noise is substantially continuously present except for one such gap during any time interval equal in duration to T/2; and

providing an output responsive to said determination that arcing exists.

2. The method of claim 1 comprising the further steps of:

determining whether such gaps are detected in multiple successive time intervals, each of duration equal to T/2; and

, if so,

determining whether the gaps occur at substantially the same relative temporal position in each of said intervals; and

,

if not, inhibiting the output responsive to said determination that arcing exists.

3. The method of claim 2 wherein the number of said intervals is in the range of 10-100.

4. The method of claim 1 comprising the further step of adjusting the duration of t_b to control sensitivity to extraneous noise on the line.

5. The method of claim 1 wherein t_b is in the range of 1°

to 60°

of the line frequency period T, one full period T being 360°

.

6. The method of claim 1 wherein the range of high-frequency noise monitored extends from about 10 KHz to about 1 MHz.

7. The method of claim 1 comprising the further step of determining that high-frequency noise is present only if the noise is determined to be substantially random.

8. The method of claim 7 further comprising the following steps, performed to determine whether the noise is substantially random:

filtering the high-frequency noise component from the waveform;

comparing the instantaneous level of the high-frequency noise to a fixed level;

counting the number of times the instantaneous level exceeds the fixed level during each of several intervals of equal length on the waveform; and

determining that the noise is substantially random if said counted number of times differs substantially from interval to interval.

9. The method of claim 1 comprising the further step of determining whether the noise exhibits a wide bandwidth.

10. The method of claim 9 comprising the following further steps, performed in order to determine whether the noise exhibits a wide bandwidth:

filtering the high-frequency noise through two or more non-overlapping bandpass filters;

measuring the amplitude of the noise passing through each of said bandpass filters; and

determining that the noise is wide bandwidth if the amplitude of the noise passing through each of said bandpass filters is substantially the same.

11. The method of claim 1 comprising the further step of monitoring the temporal position of the gap relative to the waveform and inhibiting the output if said position is outside of a predefined portion of the waveform.

12. The method of claim 11 wherein said predefined portion is the vicinity of the zero-crossing of the waveform.

13. The method of claim 1 wherein the measured waveform is that of the line voltage, and said method comprises the further steps of:

monitoring the temporal position of the gap relative to the waveform;

determining by said position the reactance of the load in series with the arc; and

providing an additional output indicating the reactance of the load.

14. The method of claim 13 wherein if said position of said gap is centered in the interval between 45°

preceding each zero-crossing of the waveform and 45°

following each zero-crossing of the waveform, an output is provided indicating that the load is substantially resistive.

15. The method of claim 13 wherein if said position of said gap is centered in the interval between 45°

and 90°

following each zero-crossing of the waveform, an output is provided indicating that the load is inductive.

16. The method of claim 13 wherein if said position of said gap is centered in the interval between 45°

and 90°

preceding each zero-crossing of the waveform, an output is provided indicating that the load is capacitive.

17. A method for detecting the presence of electrical arcing on an AC power line of frequency F and period T=1/F, comprising the following steps:

monitoring the line voltage or the load current as a waveform;

monitoring the waveform for the presence of high-frequency noise, said high-frequency noise being substantially higher in frequency than F;

monitoring the waveform for the occurrence of a pattern, the pattern comprising;

a first interval of duration t_a, t_a being less than T/2, wherein high-frequency noise is substantially present, and followed immediately by

a second interval of duration t_b, t_b being less than T/2, wherein high-frequency noise is substantially absent, and followed immediately by

a third interval of duration t_c wherein high-frequency noise is substantially present, t_a +t_b +t_c being no greater than T/2;

determining that arcing exists if said pattern occurs only once in any interval equal in duration to T/2; and

providing an output responsive to said determination that arcing exists.

18. The method of claim 17 comprising the further step of monitoring the temporal position of the start of said second interval relative to the waveform and inhibiting the output if said position is outside of a predetermined range.

19. The method of claim 18 wherein said temporal position is in the vicinity of the zero-crossing of the waveform.

20. The method of claim 17 wherein t_c is equal to zero and the sum t_a +t_b is less than or equal to T/2.

21. The method of claim 17 wherein t_a is zero and the sum t_b +t_c is less than or equal to T/2.

22. The method of claim 17 wherein t_a +t_b +t_c is T/2.

23. The method of claim 17 comprising the further steps of:

determining whether arcing exists in multiple successive time intervals, each of duration equal to T/2, and if so,

determining whether said second interval of the pattern begins at substantially the same relative temporal position in each successive interval, and if not, inhibiting the output responsive to said determination that arcing exists.

24. The method of claim 23 wherein the number of said successive intervals is in the range of 10-100.

25. The method of claim 17 comprising the further step of adjusting the duration of t_b to control sensitivity to extraneous noise on the line.

26. The method of claim 17 wherein t_b is in the range of 1°

to 60°

of the line frequency period, one full period being 360°

.

27. The method of claim 17 wherein the range of high-frequency noise monitored extends from about 10 KHz to about 1 MHz.

28. The method of claim 17 comprising the further step of determining whether said high-frequency noise is substantially random.

29. The method of claim 28 wherein said step of determining if the noise is substantially random is performed by performing the following steps:

filtering the high-frequency noise component from the waveform;

comparing the instantaneous level of the high-frequency noise to a fixed level;

counting the number of times the instantaneous level exceeds the fixed level during each of several identical intervals of the waveform; and

determining that the noise is substantially random if said number of times is substantially different during each interval.

30. The method of claim 17 comprising the further step of determining whether said high-frequency noise exhibits a wide bandwidth.

31. The method of claim 30 wherein the following steps are performed to determine whether the noise exhibits a wide bandwidth:

filtering the high-frequency noise through two or more non-overlapping bandpass filters;

measuring the amplitude of the noise passing through each of said bandpass filters; and

determining that the noise exhibits a wide bandwidth if the amplitude of the noise passing through each of said bandpass filters is substantially equal.

32. Apparatus for detecting the presence of electrical arcing on an AC power line of frequency F and period T=1/F, comprising:

means for monitoring one of the line voltage and the load current and for extracting high-frequency noise therefrom, said high-frequency noise being substantially higher in frequency than F;

means for measuring the amplitude of said noise and comparing the amplitude to a predetermined value to determine if high-frequency noise is present;

means for determining whether a pattern is exhibited by said noise in any interval equal in duration to T/2, said pattern comprising at least one interval wherein high-frequency noise is present and another interval wherein high-frequency noise is absent, and, if so, determining that arcing exists; and

means activated to produce an output signal responsive to the determination that arcing exists.

33. The apparatus of claim 32 comprising further means to monitor the temporal position of the start of said interval wherein high-frequency noise is absent relative to the waveform, and to inhibit the output signal if said temporal position is outside of a predetermined portion of said waveform.

34. The apparatus of claim 32 wherein said pattern comprises:

a first interval of duration t_a, t_a being less than T/2, wherein high-frequency noise is present, followed immediately by a second interval of duration t_b wherein high-frequency noise is absent, the sum t_a +t_b being less than or equal to T/2.

35. The apparatus of claim 32 wherein said pattern comprises:

a first interval of duration t_b, t_b being less than T/2, wherein high-frequency noise is absent, followed immediately by a second interval of duration t_c, the sum t_b +t_c being less than or equal to T/2, wherein high-frequency noise is present.

36. The apparatus of claim 32 wherein said pattern comprises:

a first interval of duration t_a, t_a being less than T/2, wherein high-frequency noise is present, followed immediately by a second interval of duration t_b, t_b being less than T/2, wherein high-frequency noise is absent, followed immediately by a third interval of duration t_c wherein high-frequency noise is present, t_a +t_b +t_c being less or equal to than T/2.

37. The apparatus of claim 32, further comprising means for:

determining whether said pattern occurs in multiple successive time intervals, each of duration equal to T/2, and if so,

determining whether the interval of the pattern during which no high-frequency noise is present begins at substantially the same relative temporal position in each successive interval, and if not,

inhibiting the output responsive to said determination that arcing exists.

38. The apparatus of claim 32 wherein said means for extracting the high-frequency noise is a high-pass filter.

39. The apparatus of claim 32 wherein said means for extracting is a band-pass filter.

40. The apparatus of claim 32 further comprising means for amplifying said extracted noise.

41. The apparatus of claim 40 wherein said means for amplifying further comprises limiter or automatic gain control means to extend the dynamic range of said means for amplifying.

42. The apparatus of claim 32 wherein said means for determining that arcing exists is a microprocessor.

43. The apparatus of claim 32 further comprising an annunciator responsive to said output signal for providing one or both of audible or visible alarm signals.

44. The apparatus of claim 32 further comprising means responsive to said means for determining that arcing exists for interrupting the current if arcing is determined to exist.

45. The apparatus of claim 32 wherein said means to determine if high-frequency noise is present further comprises means for determining whether the noise, if any, is substantially random, and if not, to conclude that high-frequency noise is not present.

46. The apparatus of claim 45 wherein said means to determine if the noise is substantially random comprises:

means for comparing the instantaneous amplitude of the high-frequency noise to a fixed level;

means for counting the number of times the instantaneous amplitude exceeds the fixed level during each of several identical intervals on successive cycles of the waveform; and

means for determining that the noise is substantially random if said number of times is substantially different each interval.

47. The apparatus of claim 32 wherein said means to determine if high-frequency noise is present further comprises means for determining whether the noise exhibits a wide bandwidth, and if not, to conclude that high-frequency noise is not present.

48. The apparatus of claim 47 wherein said means for determining whether the noise exhibits a wide bandwidth comprises:

means for filtering the high-frequency noise through two or more non-overlapping bandpass filters; and

means for comparing the relative amplitude of the noise passing through each of said bandpass filters, and for determining that the noise exhibits a wide bandwidth if the amplitude of the noise passing through each of said bandpass filters is substantially identical.