Login: GUEST @ JH4XSY.14.JNET1.JPN.AS [Tsuchiura]
home | newest check | boards | help index | log | ps | userlogin | send sysop | slog | status forward | bcm news | users | version | remove cookie
G8MNY > TECH 08.07.26 17:31l 92 Lines 4690 Bytes #7 (0) @ WW
BID : 64430_GB7CIP
Read: GUEST
Subj: Kelvin Varley Voltage Divider
Path: JH4XSY<N3HYM<VE3CGR<LU9DCE<LU1DBQ<VK6ZRT<GB7CIP
Sent: 260708/0824Z @:GB7CIP.#32.GBR.EURO #:64430 [Caterham Surrey GBR]
From: G8MNY@GB7CIP.#32.GBR.EURO
To : TECH@WW
By G8MNY (Updated Nov 07)
(8 Bit ASCII graphics use code page 437 or 850, Terminal Font)
I came across this intriguing circuit in an old Fluke Differential Valve
Voltmeter. It divides the reference 500V into 50,000 x 10mV steps, with just 5
2 pole 10 way switches & 49 Rs.
10's 1s 0.1s
ﾚﾄﾄﾄﾄﾂﾄo ﾚﾄﾄﾄﾄﾂﾄo ﾚﾄﾄﾄﾄﾂﾄo
ｳ 8K ｳ 1K6 ｳ 320 0.01s
ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ﾚﾄﾄﾄﾄﾂﾄo 10
ｳ 8K ｳ 1K6 ｳ 320 ｳ 64
ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo 9
ｳ 8K ｳ 1K6 ｳ 320 ｳ 64
ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo 8
100's ｳ 8K ｳ 1K6 ｳ 320 ｳ 64
ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo 7
+500V ｳ 8K ｳ 1K6 ｳ 320 ｳ 64
ﾄﾄﾄﾄﾄﾂﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo 6
40K ｳ 8K ｳ 1K6 ｳ 320 ｳ 64
ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo 5
40K ｳ 8K ｳ 1K6 ｳ 320 ｳ 64
ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo 4
40K ｳ 8K ｳ 1K6 ｳ 320 ｳ 64
ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo 3
40K SW5aｳ 8K SW4aｳ 1K6 SW3aｳ 320 SW2aｳ 64
ﾃﾄo<ﾄﾄﾄﾙ ﾃﾄo<ﾄﾄﾄﾙ ﾃﾄo<ﾄﾄﾄﾙ ﾃﾄo<ﾄﾄﾄﾙ ﾃﾄo<ﾄﾄﾄﾄﾄﾄﾄﾄ
40K | 8K | 1K6 | 320 | 64 |
ﾃﾄo 100V ﾃﾄo 10V ﾃﾄo 1V ﾃﾄo 100mV ﾃﾄo 1 Selected
40K | 8K | 1K6 | 320 | 64 10mV Output
0Vﾄﾄﾄﾂﾁﾄo<ﾄﾄﾄﾄﾄﾄﾄﾄﾁﾄo<ﾄﾄﾄﾄﾄﾄﾄﾄﾁﾄo<ﾄﾄﾄﾄﾄﾄﾄﾄﾁﾄo<ﾄﾄﾄﾄﾄﾄﾄﾄﾁﾄo 0 Voltage
ｳ SW5b SW4b SW3b SW2b SW1 |
ｳ___________________________________________________________0V
HOW IT WORKS
The last 0.01V decade divider is conventional, with 10x 64ohm (11 steps). This
is then fed from 2 taps apart of the 0.1s (11 320ohm Rs) switch giving 10
steps. Each step has a 100mV, as the 0.01 chain load is 10x 64 = 640, across 2x
320 = 640 making the 2 taps apart look like 320 to the other 9 320 in that
chain keep the 10:1 divide ratio perfect. This only works for 11 resistor
chains with 5x the previous value. The other divide banks do the same thing.
All resistors were special 0.1% stable types. In practice the 6 high value 40K
Rs are all made with a calibration 1k preset + 39K5 for good accuracy.
Note the output Z is quite variable from 64 ohms for 10mV to 50k for 255V. So
the dialled up voltage is only with no load like a valve volt meter.
MAKING YOUR OWN
Of course standard E12 values could be used & a correction R put across each
stack, to calibrate the load to be correct to the next stack...
33Kx6 0.5W, 2M2+6K8x11, 150K+1K5x11, 33K+330x11, 22K+68x10 gives 0.2% chains.
10's 1s 0.1s
ﾚﾄﾄﾄﾄﾂﾄo ﾚﾄﾄﾄﾄﾂﾄo ﾚﾄﾄﾄﾄﾂﾄo
ｳ 6K8 ｳ 1K5 ｳ 330 0.01s
ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ﾚﾄﾄﾄﾄﾂﾄo 10
ｳ 6K8 ｳ 1K5 ｳ 330 ｳ 68
ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo 9
ｳ 6K8 ｳ 1K5 ｳ 330 ｳ 68
ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo 8
100's ｳ 6K8 ｳ 1K5 ｳ 330 ｳ 68
ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo 7
+500V ｳ 6K8 ｳ 1K5 ｳ 330 ｳ 68
ﾄﾄﾄﾄﾄﾂﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo 6
33K ｳ 6K8 ｳ 1K5 ｳ 330 ｳ 68
ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo 5
33K ｳ 6K8 ｳ 1K5 ｳ 330 ｳ 68
ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo 4
33K ｳ 6K8 ｳ 1K5 ｳ 330 ｳ 68
ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo ｳ ﾃﾄo 3
33K SW5aｳ 6K8 SW4aｳ 1K5 SW3aｳ 330 SW2aｳ 68
ﾃﾄo<ﾄﾄﾄｴ ﾃﾄo<ﾄﾄﾄｴ ﾃﾄo<ﾄﾄﾄｴ ﾃﾄo<ﾄﾄﾄｴ ﾃﾄo<ﾄﾄﾄﾄﾄﾄﾄﾄ
33K ｳ 6K8 ｳ 1K5 ｳ 330 ｳ 68 |
ﾃﾄo 2M2 ﾃﾄo 150K ﾃﾄo 33K ﾃﾄo 22K ﾃﾄo 1 Selected
33K ｳ 6K8 ｳ 1K5 ｳ 330 ｳ 68 10mV Output
0Vﾄﾄﾄﾂﾁﾄo<ﾄﾄﾄﾁﾄﾄﾄﾄﾁﾄo<ﾄﾄﾄﾁﾄﾄﾄﾄﾁﾄo<ﾄﾄﾄﾁﾄﾄﾄﾄﾁﾄo<ﾄﾄﾄﾁﾄﾄﾄﾄﾁﾄo 0 |
ｳ SW5b SW4b SW3b SW2b SW1 |
ｳ___________________________________________________________0V
The input voltage can be what you like e.g. 200V to 1KV, just use the
appropriate number of 33K Rs (V/100 +1) to keep the system calibrated.
Why don't U send an interesting bul?
73 De John, G8MNY @ GB7CIP
前のメール | 次のメール