Print

[ZS5WC]

I built this little circuit two years ago to prevent very high keying voltages and currents from TL-922 on modern transcievers.
Works a charm!.
Also all the mods Harbach, earthing , soft start , etc have been applied and Vacuum relays were installed.
Amp keys silently, you have to watch the "on Air" lamp to confirm PTT.







Keying current is negligible, and circuit is inherently stable.
Could even be applied to other positive keying amps.

73 de William, ZS4L / ZS5WC

[ZS5WC]

 :)I notice the picture resolution is not wonderful.
If anyone would like a copy via email, my details are correct, and on QRZ.

I see various designs on the net using PNP transistors, but you still have about 120v on the base through the BE junction.
This circuit keys at 12v, and very small current.
73 de William, ZS4L / ZS5WC

[AA7IS]

I just use a simple 12 volt relay with a 1n4xxx diode across the coil.
99% of modern transceivers will pull one side of a simple 12 volt relay to ground for tx with no problem.
I'm wondering what all of the other circuitry is for?

[ZS5WC]

 ;)A relay will work just fine, however since vacuum relay keying is silent, I wanted a noise free solid state option installed within amp.
Circuit works by unbiasing NPN transistor with PTT ground, switching Mosfet hard on, and supply is regulated to the circuit.
The other parts are biasing and decoupling components.
Many ways of skinning a cat-just wanted to share my take on it.
73 de William
ZS5WC / ZS4L

[VE7RF]

;)A relay will work just fine, however since vacuum relay keying is silent, I wanted a noise free solid state option installed within amp.
Circuit works by unbiasing NPN transistor with PTT ground, switching Mosfet hard on, and supply is regulated to the circuit.
The other parts are biasing and decoupling components.
Many ways of skinning a cat-just wanted to share my take on it.
73 de William
ZS5WC / ZS4L

The TL-922 came with that god awfull clunker of a mech relay.   With vac relays operating in typ 1-3 msec..and < 1 msec rls, the last thing you want is to use a mech relay to key em. When u put a back emf diode across the coil, it slows it way down on RLS.

Myself, I use an opto-isolator.  4 kv isolation between input side of opto and output side.  100 us operate.... and 750 us rls.

I tossed the oem 3pdt mech clunker relay in my L4B's..and replaced with 3 vac relays.  Input, output, and cutoff bias. Yes, there is several ways to pull this off.  With the opto, I can sink..or souce the input side...and ditto with the output side.
On that 6M amp, the input side is sunk..and the output side is sourced.

On my yaesu MK-V, it has 1/2 power output coming out of the coax connector after just 11 msecs.  I don't want ANY PO coming out of the xcvr, until the vac relays have keyed on.   On the MK-V I can dial in a 0-30 msec delay for RF ( relays in the amp get a 0-30 msec  'head start'    but that only works on  CW mode, not SSB.  To get around that issue, I use the oem yaesu spst tiny mech relay to grnd the TX inhibit lead.  That spst relay operates in 5 msecs.    That gives the vac relays in the amp a 5 msc head start.   Once the TX inhibit lead is grounded, only then can RF emit from the xcvr.

On my yaesu 1000-D,  the same spst mech relay has a 15 msec operate time..and no provisions for any 0-30 msec delay on CW.

Those oem spst mech relays were originally used to key an amp.  I just put em to better use.  I use the 'electronic grnd' on the DIN jack of the yaesu's to key( sink) the input of the opto isolator.... that's inside each amp.  To obtain  the 9-18 vdc (15 vdc nominal)  to operate the opto's input side, I simply rectified the 5 vac filament voltage (cold side of fil choke) and ran it through a doubler.   5 x 2.828 =  14.14 vdc. 

The drake 5V @ 30 amp fil xfmr has an extra 22 vac @ 1 amp CCS winding on it, to power the oem 24 vdc relay.   Originally they used a 1/2 wave rectifier.  Most folks convert to a FWB.    I changed mine to a doubler.  (+62 vdc).

I bought the 12 vdc coil version of the gigavac GH-1 vac relays..and wired all 3 x 12 vdc coils in series.  Then apply +62 vdc ( via a 175 ohm, 25 watt metal finned resistor...with a 50 uf @ 150 vdc cap in parallel with the drop resistor). Drop resistor only dissipates 3.9 w ..and only on TX.  If any of the 3 x coils ever went open, all 3 x relays won't operate.

Then it's all silent, no hotswitching...and stupid fast.

[AA7IS]

Yep I use a simple 12 volt relay like hams have been doing for many many many years with zero problems, but I fully understand wanting to flex ones perceived knowledge by attempting to make the very simple somewhat complicated, fully understood.

[VE7RF]

Yep I use a simple 12 volt relay like hams have been doing for many many many years with zero problems, but I fully understand wanting to flex ones perceived knowledge by attempting to make the very simple somewhat complicated, fully understood.

Typ 3pdt mech tr relay is 15-17 msecs..then another 2-15 msecs with the 12 Vdc relay u are using to key the amp.  Unless a delay is used, u end up hotswitching the TR relay in the amp when using VOX SSB / semi break in on CW.   Usually a non issue with PTT / footswitch.

[ZS5WC]

 :)100% on board there with VE7RF.
Yes, the old clunker C/O relay was so loud , the sound would reverberate around the shack!.
Hot Keying, that is a valid point I failed to mention.. and like it or not most rigs have a slow attack on the first syllable so you sometimes get this rf spike whilst AMP is still keying. Completely destructive in the long term.
Luckily most modern rigs have a TX onset delay..
Interesting mods mentioned and some I might want to try with future designs.

Man, its so pleasant to key and still hear yourself think..HiHi.
73 de William, ZS4L / ZS5WC

[W4JFA]

;)A relay will work just fine, however since vacuum relay keying is silent, I wanted a noise free solid state option installed within amp.
Circuit works by unbiasing NPN transistor with PTT ground, switching Mosfet hard on, and supply is regulated to the circuit.
The other parts are biasing and decoupling components.
Many ways of skinning a cat-just wanted to share my take on it.
73 de William
ZS5WC / ZS4L

The TL-922 came with that god awfull clunker of a mech relay.   With vac relays operating in typ 1-3 msec..and < 1 msec rls, the last thing you want is to use a mech relay to key em. When u put a back emf diode across the coil, it slows it way down on RLS.

Myself, I use an opto-isolator.  4 kv isolation between input side of opto and output side.  100 us operate.... and 750 us rls.

I tossed the oem 3pdt mech clunker relay in my L4B's..and replaced with 3 vac relays.  Input, output, and cutoff bias. Yes, there is several ways to pull this off.  With the opto, I can sink..or souce the input side...and ditto with the output side.
On that 6M amp, the input side is sunk..and the output side is sourced.

On my yaesu MK-V, it has 1/2 power output coming out of the coax connector after just 11 msecs.  I don't want ANY PO coming out of the xcvr, until the vac relays have keyed on.   On the MK-V I can dial in a 0-30 msec delay for RF ( relays in the amp get a 0-30 msec  'head start'    but that only works on  CW mode, not SSB.  To get around that issue, I use the oem yaesu spst tiny mech relay to grnd the TX inhibit lead.  That spst relay operates in 5 msecs.    That gives the vac relays in the amp a 5 msc head start.   Once the TX inhibit lead is grounded, only then can RF emit from the xcvr.

On my yaesu 1000-D,  the same spst mech relay has a 15 msec operate time..and no provisions for any 0-30 msec delay on CW.

Those oem spst mech relays were originally used to key an amp.  I just put em to better use.  I use the 'electronic grnd' on the DIN jack of the yaesu's to key( sink) the input of the opto isolator.... that's inside each amp.  To obtain  the 9-18 vdc (15 vdc nominal)  to operate the opto's input side, I simply rectified the 5 vac filament voltage (cold side of fil choke) and ran it through a doubler.   5 x 2.828 =  14.14 vdc. 

The drake 5V @ 30 amp fil xfmr has an extra 22 vac @ 1 amp CCS winding on it, to power the oem 24 vdc relay.   Originally they used a 1/2 wave rectifier.  Most folks convert to a FWB.    I changed mine to a doubler.  (+62 vdc).

I bought the 12 vdc coil version of the gigavac GH-1 vac relays..and wired all 3 x 12 vdc coils in series.  Then apply +62 vdc ( via a 175 ohm, 25 watt metal finned resistor...with a 50 uf @ 150 vdc cap in parallel with the drop resistor). Drop resistor only dissipates 3.9 w ..and only on TX.  If any of the 3 x coils ever went open, all 3 x relays won't operate.

Then it's all silent, no hotswitching...and stupid fast.

Jim, I bought some opto's a while back to build a key interface for my Henry then life got in the way. Do u have a schematic handy using the opto? My email is good on QRZ.
Thanks. Bob