G8MNY  > TECH     03.03.21 11:00z 111 Lines 5480 Bytes #156 (0) @ WW
BID : 25623_GB7CIP
Subj: PEP Meter modification
Sent: 210303/0959Z @:GB7CIP.#32.GBR.EURO #:25623 [Caterham Surrey GBR]

By G8MNY                                    (Rewrite Aug 07)
(8 Bit ASCII graphics use code page 437 or 850, Terminal Font)

I have implemented a PEP Power meter in a large HF Valve PA, so that full
carrier tuning is not required (used /P on generator). The circuit is loosely
based on an old Radcom article (Jan 1986 page 46) that describes the principle
of a conversion without much modification, & best of all maintaining the
original calibration!


          _³_ D1    \³/      ³   +³  47u   D4
                  ÚÄRV3Ä¿    ³   === C4
                  ³ 50k ³    ³   _³_
     >ÄÄÄÄÄ¿     1M    1M 3³\³8               IC LM358
ORIGINAL   ÃÄ100kÄÁÂÄÄÄÄ)ÄÄ´+ \ 1  D2          5³\
RF SWR    RV1      ³    ³  ³ A  >Ä´>ÃÄ100kÄÂÄÄÄÄ´+ \ 7  D3
DETECTORS 200    C1³    ÃÄÄ´- /            ³    ³ B  >´>ÃÂÄÄÂÄÄÂÄ> +VE
SWITCHING  ³      ===   ³ 2³/³4  C2        ³  ÚÄ´- /  ³   ³  ³  ³    SWR
          RV2    u1³    ÃÄÄÄÄ)ÄÄÄ´ÃÄÄÄÂÄÄÄÄ´+ ³6³/    ³   ³D5³ ===   METER
          20K      ³    ³    ³   u1  10M  === ÀÄÄÄÄÄÄÄÙ   ³ _³_ ³u1
           ³       ³   100k  ³        ³  C3³1u            ³ \_/ ³C5

As I did not have this specific 0V input IC, but plenty of normal dual FET ones
(more RF immune than bipolar types), & I had an AC supply on the PA, I decided
to use a simple straight forward ñ dual power rail design...

             _³_   PCB           ³      _³_   ³
             \_/                 ³  12V /_\' ===100u
  PTT         ³                  ³       ³    ³
 RELAY +30V<ÄÄÅÄÄÄÄRÄÄ¿          ³       ÃÄÄÄÄÅÄÄÄÄ>0V
CIRCUIT       ³   ____³_____     ³      _³_   ³
        100u === ³Fail relay³    ³  12V /_\' ===100u
              ³   ~~~~³~~~~~     ³       ³    ³

           Mains Fail Relay
METER  >ÄÄÄ' Ä>>ÄÄÄ¿PEP        ³\³+12V                PEPÚÄÄÄ>>Ä 'ÄÄ> +VE
FROM METER        100  ³   === ³ A  >Ä´>ÃÄ1kÄÂÄÄÄÄ´+ \   ³         SWR METER
SWITCH             ³   ³ n47ÃÄÄ´- /          ³    ³ B  >Ä´         (211R in
                  RV1 ===   ³  ³/³-12V       ³  ÚÄ´- /   ³         my case)
                multi- ³22n ÃÄÄÄÄÄÄÄÄÄ4M7ÄÄÄÄ´  ³ ³/     ³
                 turn  ³    ³   X101     C1 === ÀÄÄÄÄÄÄÄÄÙ
                preset ³   47K           u1  ³

The meter was used for 3 RF power functions as well as DC metering in this PA
circuit, so the modification had to be faithful to all meter uses.

As there was HT metering, I found the modification was a SAFETY hazard when the
PA is switch off, falsely indicating no HT! So I added a small double pole
relay (fed from the +30V via a suitable R to get the right relay voltage) to
revert the metering back to normal on no power. (useful for calibration by
shorting the relay coil)

In a valve PA there is a lot of RF around, & finding a low RF field location
for this circuit is important for correct operation. With the simple power
supply on the same strip board as the op-amps, the number of RF carrying wires
is kept to a minimum of just 4, reducing RFI problems. So only the input op-amp
needed RF protection, with a 22n across the dummy meter resistor & a n47 across
the op-amp input pins. I tested that this was enough by putting a series 10n in
the meter input lead to bring in all the RF, & see if the meter moved at all,
it did not. I also tested the circuit with an AF osc via a high value resistor
to the input, to mimic peak SSB signals & checked it gave a flat peak reading
response to > 2kHz. (the 22n & n47 affect this).

A major difference to the original published design, was to add gain of 101x
on the 1st op-amp & then attenuate by 101 with a series resistor (Rm=100x the
meter value) to the meter. Without this, the op-amps are comparing tiny meter
signal levels. So this system masks any small offset voltages, op-amp leakage,
& RF pickup in the system. With ñ12V supply there is plenty of amp headroom
even with 101x gain. An accurate DMM is needed to test the 4M7, 47k, Meter,
RV1+100 & Rm values. (see my "Odd Resistance Finder" bull for making up Rm)

The input load of the Multiturn Preset RV1 + 100R is set to be exactly the same
resistance as the meter, in practice this can be finely tweeked later on for
the accurate calibration.

Silicon signal diode D1 allows op-amp A only to charge up C1 @ 5mA max peak
current, hence C1 is a small value to charge nearly instantly via the 1k, to
the peak +ve voltage the original meter x 101, would see from the detector etc.

Op-amp B then accurately follows the C1 voltage & applies it to the meter via
the 100x Rm resistor (1/101) maintaining the calibration.

Fine Calibration of RV1 is just to switch between normal & PEP (short relay
coil) on a steady "hum less" carrier & adjust for the identical power
reading. Check this at different power levels & bands.

See also my bul on "Meter Damping & Speed Up" & "QRO 1kW HF Metered Dummy Load"

Why Don't U send an interesting bul?

73 de John G8MNY @ GB7CIP

Read previous mail | Read next mail