[Prev][Next][Index][Thread]
Coil Efficiency, and Solid State Wattmeter
From: David E. Sharpe[SMTP:sccr4us-at-erols-dot-com]
Sent: Saturday, June 28, 1997 12:37 AM
To: Chip Atkinson
Subject: Coil Efficiency, and Solid State Wattmeter
John, Malcolm, ALL
I did test run on spark length versus input power
on the 1kVA magnifier that a college student and
I built as a project. These results are extremely
preliminary and incomplete, but I think exciting
if for no other reason, giving hard evidence when
transformer saturation is occuring.
Circuit Specs:
Input:
AC 120,1ph, 60hz, 20A
Ballast X (none)
Ballast R: 1500W bathroom heater, tapped to limit Ip
Wattmeter: Optoisolated, solid state w/ 0-15v meter
1.4kW FSD
Variac: 0-140V, 20A
Rotary Gap: 0-120VAC, 3A Variac, running series wound
vacuum cleaner motor through FW rectifier
HV
Transformer: 120Vpri, 11kV sec, 460VA (nameplate),
non-shunted, 7.5% Z(measured) [out of
RADAR xmtr we think]
HV EMI filter: 1mh + 1mh (each HV side)
1200pf each side at + junction to equip ground
safety gap across doorknob caps
(XL = 10 XC; low pass "T" filter)
Rotary Gap: 4 active, 6 total "BI-PHASE" super series
rotary, 0.015" gap clearance,
Rotary electrodes, 10/32 round head 304 SS
Fixed electrodes, 1/4-20 steel carridge bolts
Interrupt rate (actual) 3-4 per half cycle
Cap: Virgin LDPE, 0.060" conpression clamped by 1/4"
switchgear grade fiberglass plates, flat plate
arranged as 200,400,800,1600,3200,6400pf in 1
2.2 gal container. 2 containers bussed in parallel.
2 capacitors in parallel with two banks in series with
tank primary and spark gap across HV EMI/RFI output
Dielectric, Exxon Voltrex N61 Xfmr oil
Effective Tank C = 10800 pF
Circuit Topology: Tesla Equidrive Magnifier
Primary: RG11 Coax 11 turns, tapped from 4.5 to 10.5Turns (1 T inc.)
System tap point 8.5T, (50uH), Fo ~ 216kHz
System useful tune range: 166 --> 614kHz
Driver: 100T #16 THHN, 24"OD, 4.5 TPI (5mH L)
Secondary: 4.5" OD x 10" wire length, #28 close wound (~25mH)
Toroid: 4" x 30" on top of 2" x 12" machine handwheel (approx. toroid)
Combined Fo ~ 216kHz
SPARK LENGTH vs INPUT POWER (W meas)-- 2 independent runs
Notes:
1. Unfortunately no info on input VAC, IAC or interrupt rate was made
2. Gap speed was adjusted for what was visually observed to be "optimum"
performance <GROAN>. This is NOT scientific. The test should have
two pieces (a) leave gap speed fixed (360BPS), input VAC fixed and
measure streamer length and input V,I [kVA], P; and (b) adjust gap
BPS for max streamer length at a given VAC input (record V,I,P
streamer length).
3. Will be rerun since wattmeter was calibrated (replicate experiment)
4. This is a very preliminary experiment, that is why I am sharing info
with the TESLA LIST. FLAMES and ADVICE ARE WELCOMED!!!
Input kW(1) Spark Length(1- in.) Spark Length (2-in.)
0.40 18 22
0.50 19 23
0.60 22 26
0.70 26 31
0.85 30 36
1.00 36 43
1.10* 37 44
*Transformer non-linearity / saturation (IMO) clearly observed.
Discussion:
1. If an exponential correlation is run in Excel 7.0 (default settings)
Spark length per unit input to 1.0kW takes the general form:
L = A*e ^ (1.2034*(b))
Where: L = Length of streamer (in.)
A = scaling coefficient
e = 2.718...
b = Input Power (kW)
For both of these above curves to have an exponential
correlation to
the 4th decimal place (and an R^2 value of 0.9901) suggests a
"NAILED" equation :^)
Even more remarkable, if I recollect correctly, this data was
taken on
different days at about 10pm., with two gap speeds being of
unknown
correlation to one another (independent experiments).
2. Transformer saturation shows up as subtlely as a giant bar fight
(IMO) !
As BEN FRANKLIN would say, DO THE EXPERIMENT... OVER AND OVER AND...!
Regards
DAVE SHARPE, TCBOR