Differential Keyer

Gotta have QSK. Gotta key modern and ancient rigs from one keyer. What's a guy to do?!

CW just isn't much fun without break-in operation. Now, it doesn't have to be "I can hear even while the key is down!" QSK but it has to be pretty dang good. And, it has to happen even when running a full gallon.

My amp is an ancient Amp Supply LK500-ZC with the QSK option. Which means it has a differential keying circuit built in and vacuum T/R relays. Cool stuff!

But, I also run an old Drake T4X-C transmitter and it uses grid block (negative) keying which the keying circuit in the Amp can't tolerate. All my other rigs use positive (typically 12 VDC) keying. And, I like to be able to switch among rigs at the flip of a switch without worrying about what keyer is connected and how it's configured. What a dilema!

The solution seemed to be some sort of external differential keying circuit that would be able to key any rig in my shack. A quick check of the marketplace revealed a solution or two but for kinda large sums of money for what surely must be a simple enough device.

I had to have a good think over this one...

First, it seemed to me that I needed relays to drive both the amplifier and the rig's key lines. With a relay of suitable specifications I'd be able to key any sort of key-line, positive or negative. But, I also wanted something that could keep up with me. Back when I was young I could send and receive at upwards of 50 WPM but now that I'm one of The Ancients, I putter along at about 30 or so. Still, I knew I would need a fast relay.

Second, I wanted something dirt simple. Low parts count usually translates into high reliability.

Third, I wanted whatever I built to be powered by 12 VDC. Duh..

So, armed with a good search engine I set off in search of the key (pun intended) component. The mystical, magical relay.

And, I stumbled upon what are known as reed relays. Reed relays are pretty quick -- I found many that have make times of 1 millisecond! There was nothing suitable in my local junk box so I turned to the big junk box (ebay) where I found the Hamlin HE3321A0400 SPST 5VDC Reed Relay for a buck apiece. What a deal!

While waiting for the relays (I ordered 10 of 'em based on the rule that if you need two then you really need four. Or more..) I pondered the idea of a differential keying circuit. What I needed was a circuit that would key the amplifier a few milliseconds before keying the rig. And, at the end of a dot or dash, would unkey the rig a millisecond or two before unkeying the amplifier. And, at least nominally, preserving the original duration of the dot or dash. Hmm..

What I came up with seemed pretty simple. Almost too good to be plausible. It would require only two relays, two resistors and two capacitors. And, maybe some ancilliary stuff like an LED to indicate the power was on. That kind of thing..

Here's what the circuit looks like:

Differential Keyer schematic

The idea behind this circuit is really simple. Two identical relays are used, one to key the amplifier and one to key the transmitter. The magic is in the R/C circuits connected to each relay.

When the key is closed, current flows through R1, charging C1. When the voltage across C1 reaches a critical value, K1 closes and keys the amplifier. Further, R1 serves to limit the current through the 5 VDC relay coil to a safe value. When the key is opened, C1 discharges through the resistance of the coil and when the voltage reaches a critical value, the relay drops out, unkeying the amplifier.

The same sequence of events happens in the other circuit with C2, R2 and K2. The beauty is that the value of R1 is lower than the value of R2. This causes C1 to charge more quickly than C2. Thus, the amp is keyed before the transmitter.

My initial thinking was that C2 would have to be a larger value than C1. This would cause K2 to remain keyed longer than K1 since it would take longer to discharge through the relay coil than C1. But, while fooling around with the values I found that I could get my desired timing with C1 and C2 being equal value. (I suspect that the capacitors may well be different but are within their design tolerance of plus/minus 20 percent.)

Anyway, fiddling around a bit with the values of R1, C1 and R2, C2 allows one to achieve a variety of reasonably useful timing values.

Here's what I ended up with (I don't have a camera for my scope so please forgive the poor quality of these images): Leading edge delay

The upper trace is the keyline to the amplifier and the bottom trace is the keyline to the transmitter. In the picture above you can see that the top trace goes to ground 3.12 milliseconds before the bottom trace goes to ground.

Trailing edge delay

Here you can see the trailing edge delay. Notice that the bottom trace (the transmitter line) goes high about 1.43 milliseconds before the top (amp) trace. And, over all you can see that the top trace is at ground potential (keyed) longer than the bottom trace.

My goal was to get 3 ms delay on key down and 1.5 ms delay on key up. Using the values shown I ended up with 3.12 ms on key down and 1.43 ms on key up. Close enough for me..

And, hey, it works! Now why are these things so expensive commercially?