DB0FHN

G8MNY  > TECHNI   24.08.20 16:54l 167 Lines 7275 Bytes #-1527 (0) @ WW
BID : 39358_N1URO
Read: DJ6UX DF7EAV GUEST DL9YCC OE7FMI
Subj: 2M 65W PA with 2x M6774H
Path: DB0FHN<OE2XZR<OE5XBL<F1OYP<IZ3LSV<IR1UAW<I0OJJ<GB7CIP<N3HYM<OK2PEN<
      VE3CGR<N1URO<N2NOV<K5DAT<K3CHB<N1URO<VE3CGG<GB7CIP
Sent: 200824/0728Z @:GB7CIP.#32.GBR.EURO #:2745 [Caterham Surrey GBR] $:2745_GB

>From g8mny%gb7cip.#32.gbr.euro@n2nov.ampr.org Mon Aug 24 03:30:37 2020
Received: from n2nov.ampr.org by n2nov.ampr.org ([JNOS]) with SMTP
	id AA146831 ; Mon, 24 Aug 2020 03:30:37 EDT
Message-Id: <2745_GB7CIP@k5dat.bbs>
>From: g8mny@gb7cip.#32.gbr.euro
X-JNOS-User-Port: Telnet   (k5dat @ 50.50.110.146)  -> Sending message

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

By G8MNY                            (Updated Oct 14)
(8 Bit ASCII graphics use code page 437 or 850, Terminal Font)

My local club had a source of these Class C VHF PA blocks. The specification
for them is a max of 500mW drive for 33W out on 12.5V, but the H type is
150-175MHz! So there usefulness on 2m was unknown, although some 2m kit was
reported to use them.

         ÚÄÂÄÄÄÄÄÄÄÄÄÄÄÄÄÂÄ¿
Heatsink ³ ³             ³ ³  I tested one on a heatsink with just coax
is ground³o³ M 6 7 7 4 H ³o³  connected & it produced 33W, so all looked good.
         ³ ³             ³ ³
         ÀÄÁÂÄÂÄÄÄÄÄÂÄÄÄÂÁÄÙ  DC1 powers the 1st low current stage,
            ³ ³     ³   ³     DC2 @ 8A Max for full output.
           in DC1  DC2  out

A 1.5W handheld can do more drive than that, so I used 2 of them to make a 2m
PA with a good power level out of junk bits.

                                    ÚÄÂÄÄÄÄÄÄÄÄÄÄÄÄÄÂÄ¿
A PCB (double sided) was cut        ÞÛÝ             ÞÛÝ
to fit over the heat plate &        ÞßÝ M 6 7 7 4 H ÞßÝ
under the screws & connection       ÞÛÝ             ÞÛÝ
wires, to provide connection        ÞÛÝÂÄÂÄÄÄÄÄÂÄÄÄÂÞÛÝ
pads, through links were put        ÞÛÝÛ ÛÞÛÛÛÝÛÞÛÝÛÞÛÝ
in to. An old 900MHz phone PA       ÞÛÝß ßÞÛÛÛÝßÞÛÝßÞÛÝ
casting was used as heatsink        ÞÛÛÛÛÛÛÛÛÛÛÛÛÛÛÛÛÛÝ
& case.

2 PA MATCHING PRINCIPLE
Assuming the PA blocks are identical in power level, gains, & signal phases
(time delay). Then there are a couple of ways to connect them, I have used the
simplest using just coax.

To match the 2 PA blocks to the 50ê input & output, 4x 1/4 wave 75ê coaxes are
needed to make the connections. Each 75ê 1/4 wave transforms the 50ê to approx
100ê & 2 of these paralleled up give 50ê again.

                 50êÚÄÄÄÄÄ¿50ê
       ÉÍÍÍÍÍÍÍÍÍÍÍ͵BLOCKÆÍÍÍÍÍÍÍÍÍÍÍÍ»
       º ¬ Wave 75ê ÀÄÄÄÄÄÙ ¬ Wave 75ê º
50ê InĶ                               ÇÄ50ê Out
       º ¬ Wave 75ê ÚÄÄÄÄÄ¿ ¬ Wave 75ê º
       ÈÍÍÍÍÍÍÍÍÍÍÍ͵BLOCKÆÍÍÍÍÍÍÍÍÍÍÍͼ
                 50êÀÄÄÄÄÄÙ50ê

I measured the velocity factor of a good length of the 75ê coax (0.6 - 0.90 the
speed of light depending on the type of coax) with an open circuit teed on a
signal generator & scope/(Detector) & adjusted the frequency to give a null
which was 22MHz for my length. This meant I needed 22/145 times this length to
get the same @ 2m (46cm for my white foam filled coax).

PA SCHEMATIC
     ÚÄÄÄÄÄÄ¿
     ³RF PTTÃÄÄÄÂÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄ¿
     ³ Sense³   ³          ÚÄÄÄÄÄ¿          ³
     ÀÄÄÄÄÄÂÙÚÄÄÁÄÄ¿ ÉÍÍÍÍ͵BLOCKÆÍÍÍÍÍ» ÚÄÄÁÄÄÄ¿
RIGÍÍÍÍÍÍÍÍÏ͵DRIVE³ º 75ê ÀÄÄÄÄÄÙ 75ê º ³OUTPUTÃ(o AERIAL
   50ê       ³RELAYÃĶ                 ÇÄ´RELAY ³   SO239
  Coax       ÀÄÄÒÄÄÙ º 75ê ÚÄÄÄÄÄ¿ 75ê º ÀÄÄÒÄÄÄÙ
                º    ÈÍÍÍÍ͵BLOCKÆÍÍÍÍͼ    º
                º          ÀÄÄÄÄÄÙ          º
                ÈÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍÍͼ
                         50ê Rx coax
CIRCUITS                                                  ÚÄÄÄÄ>To Handheld Rig
  2x 1N4148                        ÚÄ100ÄÄ¿               ³
     ÚÄ´>ÃÄÄÂÄÄÄÄÄÂÄÄÄÂÄÄÄÄÄÄÄÄÄÄÄÄÁÄÄ´<ÃÄÁÄÄÄÄÂÄÄÄÄo\oÄÄÄÁÂÄÄo-oÄÄ< +12V
     ³     ===  ³/e  ===30n           //TX     1k   ON     ³  13A   Car lighter
   * ÃÄ´<ÃÄÄÁÄÄÄ´PNP _³_               LED    _³_          ³     ÚÄ< 0V
  4pF³      1n  ³\                            \_/=ON      12V   _³_
    ===           ÃÄÄÄÄÄÄÂÄÄÂÄ470ÄÂÄÄÄÄÄ¿      ³  LED    to PA
1W>ÄÄÁ>Drive  ÚÄÄÄÁÄÄ¿   ³u1³     ³  ÚÄÄÁÄÄ¿  _³_        Blocks
Rig    Relay  ³Output³  === ³    +³  ³Drive³
      Common  ³Relay ³   ³ _³_   === ³Relay³5V
              ÀÄÄÄÂÄÄÙ12V³ /_\ 10u³  ÀÄÄÂÄÄÙ
                  ÃÄÄÄÄÄÄÁÄÄÁÄÄÄÄÄÁÄÄÄÄÄÙ
                 _³_     1N4148

A tiny amount of drive RF current is fed through the 4pF to voltage doubler to
turn on the PNP (* I tried different values, & this worked with my PNP & relay
load @ 100mW low drive power). The 2 changeover relays are both plastic
encapsulated ones, the drive one is a smaller 5V DIL type, they are both held
in place with a local earth tag soldered to the coil connection & the coaxes
ground there too.

The RF detector & relay drive PNP transistor are mounted UGLY fashion around
the drive relay.

To stop the output relay operating when RF power is present, I delayed the
driver relay by about 1/100 second, with the 470 & 10uF. I checked the drive
relay operated on low RF power 100mW & low voltage of 11V OK, even with the
series Tx LED!

The diode across the relays stop back EMF from damaging the PNP, the u1 cap
across the output relay coil reduced VHF RF pickup on the wiring when 65W is on
the contacts!

    For each PA block
 ÚÄÂÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÂÄ¿
 ³ ³                 ³ ³           Power filtering capacitors are
 ³o³   M 6 7 7 4 H   ³o³           applied to each module keeping the
 ³ ³                 ³ ³           leads short for good VHF practice.
 ÀÄÁÂÄÂÄÄÄÄÄÄÄÄÄÂÄÄÄÂÁÄÙ75ê Coax   A ferrite bead (FB) is used to feed DC1.
ððððÙ ³ _FB     ³   Àðððððððððððð
  _³_ ÃÄÄÄÂÄÄÄÂÄÁÄ¿ _³_            The 1N4001 is only there in case
      ³ ~ ³  _³_ +ÃÄÄÄÄÄÄÄ+12V     of power rail reversal & 2 of them
   u1=== === /_\ ===               should be able to blow the 13A fuse.
      ³6n8³   ³   ³1000uF
      ÁÄÄÄÁÄÄÄÁÄÄÄÁ¿
           1N4001 _³_

LAYOUT
Fitting in all the coax          ÚÄÄÄÄÄÄÄÄÄÄÄ¿
is the main problem with         ³  [±±±±±]  ÿSO
this design, as my finned        ³  ³_____³[]ÃÙ239
blackened dye cast box was       ³ PNP    O/PÃ Tx LED
is only 18 x 13 x 3cm.           ³[]       =ð³]Fuse
                                 ³Drive      Ã On LED
Heat conductive paste is         ³  [±±±±±]  ØÍÍÍÍÍÍÍÍÍÍÍ12V Lead
put under the modules &          ³  ³_____³  ] On Sw
the temperature remains          ³           ØÍÍÍÍÍÍÍÍÍto Rig
quite cool for short overs.      ÀÄÄÄÄÄÄÄÄÄÄÄÙ
(Fan might be needed for heavy use)

RESULT
With 13.8V & just over 1W drive output was a good 68W, 13A drawn & with 200mW
about 50W needing only 10A.

Harmonics from these modules are quite low, I measured them at >-30dBc (near
the limit of my testgear @ VHF), but I did add a RF pi low pass filter @ the
aerial socket. 

Tx relayÄÄÄÄÄÂÄ()))ÄÂÄÄÄÄo)SO239  The 23pF caps are 2 Tx grade ceramic
       ³23pF===  L ===23pF³       the L is 2 turns, dia & wire gauge
       ÁÄÄÄÄÄÁÄÄÄÄÄÄÁÄÄÄÄÄÙ       select for 1:1 SWR match on test.

If the modules are misterminated of course anything can happen & as with many
PAs loads of sprogies might result!

Through/Rx loss was 0.5dB with 1m of driver coax & BNC in circuit. A switched
Mutek well filtered preamp was fitted in the Rx path, so the PA not powered up
the drive power was available, this overcomes any Rx insensivity as well as
proving good out of band Rx protection.


See my tech buls. "RF Modules List", "Scope RF Trick", "Coax Traps", "Non relay
instant RF Switching", "DC/RF Sensing PA Switching", & "2m Cavity Filter".


Why don't U send an interesting bul?

73 De John, G8MNY @ GB7CIP






Read previous mail | Read next mail