[Amps] Vac relay sequencing.. [WAS contact bounce]

[Jim Thomson]

The  RJ-6  is just the 'longer life'  version of the  RJ-2's  that I use all the time.   Both relays  have  identical specs.   BOTH  have a 18 msec op  time.. and a 9 msec rls  time.    The gigavac eq of em is called a 'G-2'   or a  'G-2 ham'  [also has identical specs, and don't be fooled by the g-2 ham specs, the g2-ham is a carbon copy of a G-2 ..except price. 
Don't mess with sequencers.. esp the arrl  ones, [a] they don't work, [b]  u don't  need em.. and esp with a RJ-6.. RJ-2... G-2... G-2 ham   relay.   
You want the  INPUT relay to be FAST  RLS.     Just install a 200-400 ohm  2-3 watt  mof  resistor in series with a 1N5408/6A10 diode.... [either end of diode] .. and put that combo  across the relay coil.  [ diode is reversed] 
You want the OUTPUT relay to be SLOW  RLS.      Just install a  1N5408 /6A10  diode, RVS polarity , across the output relay coil. 

These relays, stock,  are  SLOW to operate as is [ 18 msecs !]     You don't want to slow down the input relay  at all, otherwise you end up with an input relay that operates in say 24 msecs.. and an output relay that operates in 18 msecs. 

You can't  use these  relays in qsk mode.. even sped up, they are too slow.  The best you can do is   super fast VOX  ssb/cw. 

They  BOTH  have to be sped up.. with an overvoltage.   I  use  +120 vdc on mine..... but anything from  +50 vdc  to  +120 vdc  will work nicely.    This  will increase the Operate  speed  from 18 msecs.. down to 7-9 msecs. 

IF  your  stock relay  supply is a FWB... OR  a simple  1/2  wave rectifier,  the  simple fix is  to  change it  to a DOUBLER, which requires  just 2 x diodes and 2 x  big lytics.   Then at least you have  +48  to +50 vdc  to play with.    You can then use just one  drop resistor  for BOTH parallel relay coils.   To speed em up even faster, install a 50-100 uf  lytic directly across the drop resistor. 

Another option is  to replace both relays   with new  gigavac  G-2  ham versions.    You can get ANY of the gigavac relays  in either a  12 vdc coil....  OR  a 26.5 vdc coil.    Your  stock  +24 vdc  relay supply  will easily speed up a 12 vdc  coil.  If a doubler is used.... the  +48  to  +50 vdc   will really speed up a  12 vdc  coil.   And again, put a 50-100 uf  lytic  across the dropping resistor. 

We used a small xfmr  from the mouser catalog, ran a doubler on it.. to get  + 120 vdc.   The xfmr had enough taps  to get a lot of  different voltages.    Also,  we used TWO  RJ2-B's.. with all contacts in parallel... to handle more RF on the output side...... yet  we used  a RJ1-A    [ or gigavac GH-1]  on the input side.      We also used a RJ1-A  for the cut off bias.    ALL  4 x relay's sped up with +120 vdc.      ONE drop resistor used for the paralleled coils  for the RJ2-B's.... and  a  2nd  [different value]    drop resistor  for the paralleled  coils  of the  RJ1-A's .    On some configs, I use a   single  RJ2-B  on the  input  Rf side,  and a pair of em on the output side.... and a  RJ1-A  [ or a similar fast spst vac relay]    for the t/r  cut off bias... [ just shorts a paralleled pair of  100 k  3 watt mofs. ]

Some  XCVR's like  my yaesu- MK-V.... have a menu option  for CW  mode only.. that delay's  RF  from the  xcvr  from 0-30 msecs.  [ I set it to 30 msecs].   No dit's, dashes  gets truncated  cuz it's a digital delay.  What's  going over the air is delayed by 30 msecs.     That gives the relay's in the  amplifier, a  30 msec  head start.    Ok this is  fine on CW.. but what abt ssb?   On  SSB, there is RF  coming out of a MK-V.. and also my 1000-D  after 5 msecs.    To solve the  SSB  problem,   I use the  TX inhibit feature.    I also use the... 'electronic grnd'   from the band data jack of the mk-v /1000-D  to key an opto isolator  [ P+B  ODC-15A.]   Output of the ODC-15A   keys the vac relays in the amp.  [ODC-15A is rated for  8-20 vdc  @  10-15 ma on the input side... and  200 vdc @ 1A  CCS  on the output side.   The  ODC-15  is rated  for  60 vdc  @ 3A  on it's output side....both versions  have a 100  micro second operate time.. and a 750 microsecond RLS time]    I also use the stock, 15 msec mech relay, spst, that's  inside BOTH the  mk-v  and and the  1000-D.... to  supply a grnd for the TX inhibit lead !    TX RF  gets  inhibited  by 15 msecs !... [ regardless of mode]    On any  1000-D... the  grnd  from the  RCA  TX  jack, is  simply routed back into the   'TX inhibit lead'  pin on the  band data jack.    On any MK-V.... same deal.. EXCEPT a 1000 ohm, 2-3 watt mof  has to be installed between the  TX inhibit pin and the +13.8 vdc pin... both on the band data jack.  The +13.8 vdc  can also be used  to sit on one side of the opto isolator's inputs.... and ...'TX electronic grnd' on band data jack is used to sink the opto on TX.    You cough into the vox... vac relays trip in the amp... THEN the 15 msec  spst relay in the yaesu operates.. DEAD LAST.. and  TX is no longer inhibited...and then RF comes out of the xcvr.....  NO  hotswitching. 

End result of all this is.... smooth  VOX   SSB/CW... when using RJ2-B/C's.. [ or RJ6's../ G-2... G-2 ham].   Smooth as  a baby's  bum .. and NO hot switching.    The trick here is... the  RJ2/6, STOCK  has a slow  18 msec  OP  time.... so even with the  15 msec  TX inhibit delay on SSB.... it's  still  NOT fast  enough.. stock.    It HAS  to be sped up.    The  1000-D  has NO  provision  to delay RF  on CW... like the 0-30 msec menu item the MK-V has.  The best you can do on a 1000-D is.. is to use the  15 msec  mech delay.. and use it in conjunction with the TX inhibit lead..[ALL modes].    You won't hot switch any  YC-156's  [ or Yu-148's,  3x3's]  with this setup.   It's  100%  bullet proof.. and field tested.

As a  side note, if you order a new Gigavac  G-2 ham... get the version that has threads on all 3 x cylinders.   Then no soldering!  They come with silver plated machine screws,  which makes  installing either strap.. or  round  'sta-kons'.. a  snap.        




IF you want high power  [ YC-156 etc]  QSK  cw.... then you only have ONE option...  to use a  PAIR  of  RJ1-A's ,  with all contacts in parallel, on the output side.. [ and coils in parallel].. and also a single  RJ1-A   on the input RF side.   You may as well use a 4th  RJ1-A  for the  t/r  bias... and  simply parallel all 4 x  coils up.... and speed all 4 x of em up.     Again, if  gigavac GH-1 ham  relay's  used,  get the  12 vdc  coil version for all 4 x relays, and they will just scream on operate.   Then all 4 x relays will operate in 1.8 msecs.   

Sri  for the  diatribe, but I have  extensive experience  using the RJ-2  and  RJ-6  vac relays.   Your  RJ-6, as configured above,  will  last 25 x million operations.... per jennings. 

later.... Jim   VE7RF