Topband: Upon further examination of powering 26.5V relays with 12V

[Thomas]

In my work I often use a SPICE circuit simulation program to analyze 
customer application circuits. So I decided to take a look at the K6XX 
and PE1MWB voltage doubling circuits proposed as solutions for powering 
26.5V relays from 12V supplies. Since I do a lot simulation work I know 
first hand that the simulation are only as good as the models used for 
the simulation. Therefore, I did a lot of tweaking of the relay and 
circuit component variables to arrive at my conclusions.

As mentioned the K6XX and PE1MWB circuits are very similar with the 
exception that the PE1MWB circuit has a convenient transistor switch at 
the input. However, simulation of the two circuits showed that the 
latter resulted in an extremely short 24V pulse - on the order of 5us in 
duration. Once the pulse ceased the voltage immediately fell below 12V. 
When we consider the fast switching speed of a vacuum relay being 
roughly between 6 and 12ms, 5us hardly seems significant  in duration to 
be useful. The duration of the 2X voltage pulse is very much affected by 
the switching speed of the transistors used in the circuit.

The K6XX circuit with its smaller 1k base resistor and modeled with a 
medium power, 2N4951 transistor switching the relay produced a 
significantly longer near 24V pulse. A 2N4401 NPN transistor was added 
at the input like the PE1MWB circuit. The relay pulse very quickly 
jumped to nearly 24V when the 2N4951 momentarily turned ON. But instead 
of the voltage falling off in 5us it ramped downward falling to about 
15V after 1ms. The duration still isn't long compared to the relay 
switching time, however, when integrated the power supplied to the relay 
is much greater for the K6XX circuit.

I think it is important to step back and remember what K6XX intended his 
circuit to be used for - increasing amplifier relay speed. The relay 
used in the application already has the required coil voltage available 
and the short burst of over-voltage is being used to make it switch 
faster. That is different than trying to switch a relay with a voltage 
source very briefly supplying the required coil voltage and then 
dropping to half or less.

The next step would be to evaluate these circuits with pulse generators 
and oscilloscopes to determine how they truly behave and how effective 
they are in accomplishing the intended task. I suspect the simulation 
results are close and in which case I question how effective the 
circuits are in this application. Rick, N6RK had suggested using a boost 
regulator and I had considered this myself. Currently, I am considering 
the MC33063A boost/buck regulator. It will easily supply the required 
coil current and voltage in the boost mode, with a wide range of lower 
input voltages. This may prove to be a better solution.

'73, Thomas - AC7A