| |
G8MNY > TECH 24.11.11 12:32l 187 Lines 9806 Bytes #999 (0) @ WW
BID : 17993_GB7CIP
Read: GUEST
Subj: QRO 1kW HF Metered Dummy Load
Path: IZ3LSV<I0OJJ<IK6ZDE<IK2XDE<IR2UBX<IW2OHX<HB9ON<IW8PGT<CX2SA<UA6ADV<
GB7CIP
Sent: 111123/2243Z @:GB7CIP.#32.GBR.EU #:17993 [Caterham] $:17993_GB7CIP
From: G8MNY@GB7CIP.#32.GBR.EU
To : TECH@WW
By G8MNY (Updated Aug 08)
(8 Bit ASCII graphics use code page 437 or 850, Terminal Font)
This is a peak reading meter, QRO dummy load, & with a low level RF take off.
SCHEMATIC
RF ________ _________
in o)ÄÄÄÂÄÄÄÄÄÂÄÄÄÄ´Peak ÃÄÄÄÄ>o´Range CalÃÂÄÄÄÄÂÄÄÄÄÄÄÄÄÄÂÄÄÄÄÄÄÄÄÄÄ¿
³ÚÄÄÁÄÄ¿ÚÄÁÄÄ¿ ³Detector³Range ³ÃÄÄÄÄÄÄÄÄÄ´³ ÚÄÄÁÄÄ¿ ÚÄÄÄÄÁÄÄÄÄÄ¿ ÚÄÄÁÄÄ¿
RF o)´Atten³³1kW ³ ÀÄÄÄÄÂÄÄÄÙ SW À´ Speed upÃÙ ³Meter³ ³ Meter ³ ³Meter³
mon ³ÀÄÄÂÄÄÙ³Load³ ³ ÀÄÄÄÄÄÄÄÄÄÙ ³Damp ³ ³Protection³ ÀÄÄÂÄÄÙ
³ ³ ÀÄÄÂÄÙ ³ ÀÄÄÂÄÄÙ ÀÄÄÄÄÂÄÄÄÄÄÙ ³
ÀÄÄÄÁÄÄÄÄÄÄÁÄÄÄÄÄÄÄÄÁÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÁÄÄÄÄÄÄÄÄÄÁÄÄÄÄÄÄÄÄÄÄÙ
LAYOUT
It uses a commercial 52ohm 1kW coaxial 5cm dia x 30cm long load, mounted in a
perforated screwed together panelled steel metal box 43 x 13 x 13cm.
Ú - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ¿ -40dB RF
³ 75³__ Sniff
³-----¿ RF HOT Earth clamp Ã-|__| BNC
_³ | |~|___Clamp ===ÄijijU EARTH
³M³ PEP | |S³°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°° ² ºAtten
³E³ | |U³ 15mm copper tube spacer! pin ± º__
³T³METER| |PÀÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜÜ/_/_³ ³COAX
³E³ | |PÚßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßßß ~\~³___³SO239
³R³ CAL | |O³<tube ! ßÜ º
³_³ | |R³°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°°° ßÜ ºdiodes
_³ CCT.| |T|~~~ 52ohm tube load ===~~³~³U EARTH
³ ³-----Ù ~ C1³
³_³_³: ³
Range À - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Ù
Switch
The input socket is bolted through a 2cm wide U shaped copper strip, that is
clamped on to the outer of the dummy load with a capacitor type clamp. The
socket centre is soldered to the pick off attenuator R, diode probe stack, &
also a 4mm banana plug pin that engages into a banana socket soldered to the
15mm dia copper centre tube. This tube forms the centre of 5cm dia coax to make
the low radiating load. It is supported inside the load @ the pin end with a
card spacer to hold it central for assembly.
The copper tube's far end is split & flatted out to make 2 long tags that are
bolted to the RF hot clamp. This clamp has an insulated paxalin (copper less
PCB) sheet support from the case bottom with a metal right angle strip.
CIRCUIT - - - - - - - - - - -
| ÚÄR5ÄÄRV1Ä¿ CAL CCT |
Peak Detector 1kW³ C2 ³ BOARD |
Range oÄÄÁÄÄÄ´ÃÄÄÄÄ´ |
Ú´>ÃÄ´>ÃÄ´>ÃÄ´>ÃÄ´>ÃÄ´>ÃÂÄoÄÄ> | ³ Meter |
Input³ D1-6 ³ | ÚÄR6ÄÄRV2Ä´ Damper |
o)ÄÁÂÄÄÄ¿ ³ oÄÄ´ C3 ÃÄÄÄÄÂÄÄÄÄÄÄÄÄÄÂÄÄÄÂÄÄÄÄ¿
239³ R1 ³ ³ 100WÀÄÄÄ´ÃÄÄÄÄ´ +³ | _³_ ³ +³
³ ³ R4 ³ | ³ === | \_/ ³C6 _³_
³ R2 ³ DUMMY C1 === oÄÄÂÄR7ÄÄRV3Ä´ C5³ | D6 ³ === / \
o)ÄÄ´ ³ LOAD ³ 10W ³ C4 ³ ³ | D7_³_ ³ ³Meter³
BNC³ R3 ³ ³ | ÀÄÄÄ´ÃÄÄÄÄÙ RV4 | \_/ ³ \___/
³ ³ ³ ³ | ³ | ³ ³ ³
RFÀÄÄÁÄÄÄÁÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÁÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÁÄÄÄÄÄÄÄÄÄÁÄÄÄÁÄÄÄÄÙ
Sniff Attenuator - - - - - - - - - - - Meter
Cal & Speed up Protection
COMPONENTS
No Resistors Capacitors Diodes Preset R Misc
1 * 3k9 3W 0u33 400V BYW29 200V 35nS 1M Meter 100uA 1kê
2 * 3k3 2W @ 23n 1kV " 330k BNC chassis mount
3 * 75R @ 68n 400V " 100k SO239 chassis mount
4 * 52R 1kW @ 150n 400V " 10k Switch 1 pole 3 way
5 2M2 0.25W @ 10u 10V " Perf Boxing panels
6 820k 0.25W $ 1nF 25V " Nuts & bolts
7 220k $ 1N4001 4mm Plug & Socket
8 $ 1N4001
* = non inductive type, @ = for my meter ballistics, $ = mounted on meter
RF TAKE OFF
This give about -40dB output (-46dB terminated) for scopes or analysers, & uses
just 3 Rs that are laid down close to the metal panel to reduce extra RF
pickup. The output impedance is 75ê. It should be noted that the peak reading
diodes will cause some even harmonics to appear on suddenly increasing signal
levels as the peak dummy load RF actually gets slightly clipped.
PEAK READING
The 6 fast recovery tab type diodes, give a 1200V PIV 8A RMS 35nS detector.
They have there tabs sawn off (cutting with tin snips I found damaged them).
Other slower diode types easily blow up in this peak reading application! The
diodes are soldered in series (1/2 about face & upside down) & have individual
heat shrink over the tab remains so there in no chance of shorting out.
ÚÄÄÄÄ¿ all sleeved
³ () ³ x6 ÚÄÄÄÄ¿ÚÄÄÄÄ¿ÚÄÄÄÄ¿ LOAD
ÃÄÄÄÄ´ gives ³____³³____³³____³ /
³____³ |-+|-+|-+|-+|-+|-+|
| | C1/ ³ÜÜÜܳ³ÜÜÜܳ³ÜÜÜܳ
ßßßß ßßßß ßßßß
ÄÄ´<ÃÄ´<ÃÄ´<ÃÄ´<ÃÄ´<ÃÄ´<ÃÄÄ
C1 is mounted on a tag around the input socket so there is next to no lead
inductance to give false readings. There is very high pulse current into C1 on
applying sudden RF, as the source Z is actually 25ê & there could be >300V
peak, that is 12A peak current on a 1kW sudden RF pulse! The meter load is only
100uA max, so there is no real loss due to that loading. C1 has to store the
peak value until the next speech peak not just the RF peak, hence it is very
large value for an RF detector.
The long DC lead away from C1 to the meter switch is glued to the metal base
plate to kept it away from the dummy load R4, so there is little RF pick up &
no chance of it getting burnt too.
The resultant DC + diode losses is then exactly the AC peak value.
2 But sine wave 2
Vrms crest power Vpk
POWER = ----- is 2x RMS so... POWER = ----
R 100 (for 50ohm)
Leads to & from the calibration circuit matrix board are also kept close to the
front metal case as possible to reduce RF pick up from the load.
CALIBRATION
From above, the 3 meter scales can be accurately calibrated from 1st
principles. The meter scale 0-100% needs to be drawn in Watts. e.g. a square
law scale...
%SCALE VOLTS WATTS VOLTS WATTS VOLTS WATTS CORRECTED
0 0 0 0 0 0 0 0
5 17 3 5 0.3 1.7 0.03 0.1
7 22 5 7 0.5 2.2 0.05 0.2
10 32 10 10 1 3.2 0.1 0.3
14 45 20 14 2 4.5 0.2 0.5
17 55 30 17 3 5.5 0.3 0.5
23 71 50 23 5 7.1 0.5 1
32 100 100 32 10 10 1 2
39 123 150 39 15 12.2 1.5 2.5
45 141 200 45 20 14.1 2 3
50 158 250 50 25 15.8 2.5 3.5
54 173 300 54 30 17.3 3 4
63 200 400 63 40 20 4 5
71 224 500 71 50 22.4 5 6
78 245 600 78 60 24.5 6 7
84 265 700 84 70 26.6 7 8
89 281 800 98 80 28 8 8.7
94 300 900 94 90 30 9 9.5
100 316 1000 100 100 32 10 10
CAL RV1 CAL RV2 CAL RV3
The diode loss of about 3V is almost insignificant except on the 10W scale &
the correction can be added in if you wanted to make that scale a special one.
A 1kW variac on mains with a DVM meter will confirm the calibration to 1000W,
as the RF power meter is quite flat from about 20Hz to 30MHz.
METER BALLISTICS for PEP measurement.
An attempt has been made in this circuit to use both meter damping to stop
overshoots with C5 & RV4 (terminate the meter movement) & also proved meter
speed up capacitors C2,C3 & C4 across the series calibration ratio Rs. My bul
on "Meter Damping & Speed Up" explains the principles in detail.
The capacitor values used are ONLY correct for my meter movement. Using a slow
RF step pulse of known square envelope (scope it), you can determine best
values for both of these, so there is no overshoot, but still a very fast rise
time.
CONCLUSION
The result here is a very accurate peak reading meter up to 30MHz. Above that
the coax line load & diode detectors all start to go wrong & can't be trusted!
If the RF has a high harmonic content (e.g. worse than -20dB) then there will
be quite an error in the power reading as the peak voltage will be wrong.
I have used the load for testing Ham HF amps up to 1kW & soak testing 531kHz AM
Tx at up to 800W PEP when it did get very hot!
See my tech bul "AF 2 Tone Test Osc Design" & "PA instability in ICOM IC735".
Why Don't U send an interesting bul?
73 De John, G8MNY @ GB7CIP
Read previous mail | Read next mail
| |
