• Variable RF Attenuator

    From G8MNY@21:5/101 to TECHNI on Mon Jul 8 10:30:13 2019
    R:190708/0923Z @:LW2DQC.1653.BA.ARG.SOAM #:13695 [13697] FBB7.01.35 alpha R:190708/0923Z 25058@LU3DVN.USH.TDF.ARG.SOAM LinBPQ6.0.17
    R:190708/0914Z 32626@N3HYM.MD.USA.NOAM BPQ6.0.18
    R:190708/0920Z @:CX2SA.SAL.URY.SOAM #:9734 [Salto] FBB7.00e $:11695_GB7CIP R:190708/0920z @:F4DUR.FRHA.FRA.EU $:11695_GB7CIP
    R:190708/0920z @:F1OYP.FAQI.FRA.EU $:11695_GB7CIP
    R:190708/0820z @:ON0AR.#AN.BEL.EU $:11695_GB7CIP
    R:190708/0919Z @:VE2PKT.#QBC.QC.CAN.NOAM #:13756 $:11695_GB7CIP
    R:190708/0918Z 20015@OK2PEN.SP.BRA.SOAM [Sao Jose dos Campos] $:11695_GB7CIP R:190708/0916Z @:GB7CIP.#32.GBR.EURO #:11695 [Caterham Surrey GBR]

    From: G8MNY@GB7CIP.#32.GBR.EURO
    To : TECH@WW

    By G8MNY (Updated Mar 10)
    (8 Bit ASCII graphics use code page 437 or 850, Terminal Font)

    While working on an old pulse generator that goes up to 50MHz, I was interested to know how the variable attenuator works. It used a ganged log-antilog pot to make quite usable approximation to this table.

    Attenuation Log Var Anti Var
    2k5 40dB 2500.0Ω 0.0Ω
    50Ω o)──┬───Log<───┬──(o 50Ω 35dB 1406.0Ω 0.4Ω
    Output│ │ Var │ │ Input 30dB 790.0Ω 1.05Ω
    │ 51Ω 51Ω │ 25dB 443.0Ω 2.5Ω
    │ └────┬─────┘ │ 20dB 248.0Ω 5.0Ω
    │ 3k │ 15dB 136.0Ω 10.65Ω
    │ Antilog<─┐ │ 10dB 70.7Ω 22.5Ω
    │ Var │ │ 7dB 44.8Ω 40.0Ω
    └───────────┴────┘ 6dB 37.7Ω 151.0Ω
    5dB 30.4Ω 64.0Ω
    3dB 17.6Ω 120.5Ω
    2dB 11.6Ω 215.0Ω
    0.1dB 0.5Ω 3000.0Ω
    HOW IT WORKS
    At maximum loss the 2k5 in series with the signal & the 3k is shorted out, the input & output Z is mainly due to the two 51Ω.

    o)──┬──2k5─┬──(o o)──┬─44.4─┬──(o o)──┬─44.4─┬──(o o)──┬──0.5Ω─┬──(o
    │ │ │ │ │ │ │ │ │ │ │ │ │ │ 0.1dB │ │
    │ 51Ω 51Ω │ │ 131Ω 131Ω│ = │ 51Ω 51Ω │ │ 51Ω 51Ω │
    │ │ 40dB │ │ │ │ 7dB │ │ │ │ 7dB │ │ │ └───┬───┘ │
    │ │ │ │ │ │ │ │ │ 80Ω 80Ω │ │ 3k │
    └──┴──────┴──┘ └──┴──────┴──┘ └──┴──────┴──┘ └──────┴──────┘
    approx

    At minimum loss the 2k5 is a short & the antilog to earth is nearly open circuit @ 3k. The cleaver bit is mechanically arranging that log & antilog tracks produce the correct resistance to maintain the 50Ω match at all other settings. Note 7dB is half way around the pot with about equal tracks!

    To work well to 50MHz all the leads are quite short & the Rs are not wire wound types!


    Why Don't U send an interesting bul?

    73 de John G8MNY @ GB7CIP

    --- Mystic BBS v1.12 A43 (Linux/32)
    * Origin: HAMRADIO telnet lu9dce.dynu.com (21:5/101)