Print

[VE7RF]

My theory is: Tube amp + solid state rig = buffer/cheap insurance.

That is  NOT a good rule.

Using an unnecessary keying interface just slows down the amp relays going into transmit, not a desirable situation.

Just stick a volt-meter in the amp keying jack with amp in Operate to measure the voltage, then put the meter in mA current measuring mode to see what the keyed current is when the key jack is shorted via the meter.  Compare to the specs of the radio being used.

A  $1.85 switch transistor from Mouser, rated for 300 vdc @ 500 ma, switches in just 3.6 microseconds. Is that fast enough ?   That isn't gonna delay anything in the amp.  IF the 'keying interface' consist of a mech relay, then yes, that adds  a bunch of delay.... and guarantees hot switching the TR relays in the amp.

The opto isolator's I use are slower.  100 microseconds to turn on, (.1 msec)..... and 750 microseconds (.75 msec) to shut off.  Still plenty fast enough for VOX  SSB / semi break in CW. (Just reduce the VOX delay by 1 msec, non issue).    I like the opto isolators over the switch transistors.   If the switch transistor ever failed shorted, the full OCV would appear in the xcvr.  For 12 vdc, not a big deal, but with 24-170 vdc OCV, it would be a big issue.  The opto's are slower than the switch transistor, but offer 4 kv isolation between their input + output side.  They can be mounted inside, or outside the amp in question.

For stuff like a SB-220, that use 110 vdc, I would toss the oem 3PDT mech relay, and replace with a DPDT  dip relay on the input, and a bigger DPDT dip relay on the output.  To easily obtain the Vdc to run any new relay, the 5 Vac at the cold end of the fil choke can be run through a FWD, and then you end up with a  12.74 vdc to power anything you want....without having to add a small xfmr.   Alternatively, you can obtain 13.8 vdc from the back of most xcvr's.

[VE7RF]

Probably the most dangerous situation I have run across is with the Ameritron AL-84 amplilfier. It has a 12V relay easily driven by most solid state rigs.

The problem is shown below,



The relay has no snubbing diode to prevent a 100 volt spike from occurring when it disconnects. That voltage will exceed the ratings of many transistors.

The solution is a diode across the relay coil.



This is the result with the diode in place.

Is the snubbing diode on the schematic ?   If not, who ever designed it, made a fubar error. If it was left out on the assy line, then no QC was done.

The snubbing diode will work, but will slow down the rls a bit, non issue on a 3pdt mech tr relay.   For faster rls, install a resistor in series with either side of the snubbing diode. Typ value is  2/3 the DC resistance of the relay coil..... but that is typ only used on vac relays, and only when QSK operation is intended. 

[HAMHOCK75]

Is the snubbing diode on the schematic ?   If not, who ever designed it, made a fubar error.

Here is the AL-84 manual and schematic.

http://www.radiomanual.info/schemi/ACC_PA/Ameritron_AL-84_user.pdf

[VE7RF]

Is the snubbing diode on the schematic ?   If not, who ever designed it, made a fubar error.

Here is the AL-84 manual and schematic.

http://www.radiomanual.info/schemi/ACC_PA/Ameritron_AL-84_user.pdf

Tnx.  Snubbing diode not shown on schematic.  Chief engineer major design error flaw.  He/she ..'forgot' to read all about  "relay 101".

Cheap bastards used a  fubar half wave rectifier for the 12 vdc supply.   Drake did the same thing on their 24 vdc supply in their L4/L4B/L7 amps.    Easily changed over to a FWB, or if more vdc required....a FWD.

[W4JFA]

My theory is: Tube amp + solid state rig = buffer/cheap insurance.

That is  NOT a good rule.

Using an unnecessary keying interface just slows down the amp relays going into transmit, not a desirable situation.

Just stick a volt-meter in the amp keying jack with amp in Operate to measure the voltage, then put the meter in mA current measuring mode to see what the keyed current is when the key jack is shorted via the meter.  Compare to the specs of the radio being used.

A  $1.85 switch transistor from Mouser, rated for 300 vdc @ 500 ma, switches in just 3.6 microseconds. Is that fast enough ?   That isn't gonna delay anything in the amp.  IF the 'keying interface' consist of a mech relay, then yes, that adds  a bunch of delay.... and guarantees hot switching the TR relays in the amp.

The opto isolator's I use are slower.  100 microseconds to turn on, (.1 msec)..... and 750 microseconds (.75 msec) to shut off.  Still plenty fast enough for VOX  SSB / semi break in CW. (Just reduce the VOX delay by 1 msec, non issue).    I like the opto isolators over the switch transistors.   If the switch transistor ever failed shorted, the full OCV would appear in the xcvr.  For 12 vdc, not a big deal, but with 24-170 vdc OCV, it would be a big issue.  The opto's are slower than the switch transistor, but offer 4 kv isolation between their input + output side.  They can be mounted inside, or outside the amp in question.

For stuff like a SB-220, that use 110 vdc, I would toss the oem 3PDT mech relay, and replace with a DPDT  dip relay on the input, and a bigger DPDT dip relay on the output.  To easily obtain the Vdc to run any new relay, the 5 Vac at the cold end of the fil choke can be run through a FWD, and then you end up with a  12.74 vdc to power anything you want....without having to add a small xfmr.   Alternatively, you can obtain 13.8 vdc from the back of most xcvr's.

Jim, Do u have a link and/or schematic for that opto circuit? I bought a couple opto's just for this purpose but haven't got around to building it yet. I'll need to switch an old Henry which has a high voltage spike when it disengages. I did build one with a HV switching transistor and with the help of Phil but wanted to experiment with the opto.
Bob. W4JFA

[K4TLJ]

I put this circuit in an SB-200 and it worked very well.

[VE7RF]

I put this circuit in an SB-200 and it worked very well.

Mine is basically wired the same way, except I didn't use a 33 ohm resistor.
Make sure you wire a rvs connected diode across the coil in the TR relay in the amplifier.

[KD6VXI]

IRF640s are cheap. 

I used them for my PWM on my homebrew am tx...  And as final output xisters in my 80 and 40 meter RF deck.

I also used it as a switch to switch the SB220 and my Barker & Williamson PT-2500.

They work great, are capable of switching more than I need and are quick as *&^$.

--Shane
WP2ASS / ex KD6VXI

[VK6HP]

Nowadays it's easy and desirable to provide excellent galvanic isolation in the keying line.  There's no place for ARB-704's in 2022.  Photovoltaic (PV) opto couplers, for example, easily avoid the need for common input and output connections and readily cope with both positive and negative amplifier keying systems.

One easily built or purchased PV interface is the KD9SV "Keyall" circuit, described at https://static.dxengineering.com/global/images/instructions/svp-sv-kr_sh.pdf  A wide range of MOSFETS can be used at the output but it's worth selecting low on-resistance devices so that there's negligible key line resistance to affect the grid bias voltage of 30L-1's etc.

Another approach which works for many amplifiers is to use a completely integrated PV solid state relay. For many applications single-chip solutions such as the Panasonic AVQ25x series (https://www.allaboutcircuits.com/electronic-components/datasheet/AQV254--Panasonic/) are suitable.  I often use an AVQ254 to give a 400V, 150 mA unit which is small enough to be built into a connector.  For example, for a TS-590S I housed the chip and input current limiting resistor in the housing of a 7-pin DIN plug, and powered it from the active-high (+12V) "transmit" line from the transceiver. This gives an isolated, active-low, electronic keying line which is not otherwise available in the radio.

73, Peter