Damaged diodes in power meter

[neazoi]

This HF power meter works on 5w ok but on 10W it burns out one of the diodes. I am testing it on an actual modern transceiver, I have no idea if there are spikes of higher power at the beginning of the transmission. When the diode is damaged I get infinite vswr.
What can I do about it?

 

[betwixt]

I presume the optocoupler is to expand the scale at low power levels.
The 1N5711 has an absolute maximum PIV of 70V and a maximum current of 15mA so it runs pretty close to maximum rated voltage and if R2 is set to low value it will also exceed the maximum current rating. It isn't a very clever design.

A far better approach would be to use a switch to select high or low power ranges so the voltage is scaled before the rectifier. 1mW to 10W is a 10,000:1 range which is best split into at least two ranges.

Brian.

 

[G4BCH]

Two things.
The absolute maximum reverse voltage for the 1N5711 is 70V, very close the 63V it would see from 10W into 50 ohms. A little over 10W or higher than 50 ohms termination and you will exceed this. There will be perhaps some margin here.
The instantaneous current seen when charging the 100nF capacitors at 10W will be over 1A, its maximum is 15mA, since it survives 5W, (that gives just under 1A peak) it is likely perhaps a combination of the two.

It looks like the design is marginal; find a better diode or add an attenuator.

Looks like Brian beat me too it while composing this

 

[biff44]

are you SURE it is burned out?
that type of power meter is designed to detect pulse AM modulated waves, i.e. is AC coupled to the meter.
are you amplitude modulating the input with a square wave?

 

[vfone]

Try to place in series with each diode a 10 ohm resistor. Can tune the values of the resitors for desired performance..

 

[neazoi]

Thank you all for your replies.
The author stated schotkky diodes there but no type. I tried bat84 and burned one of them (the low one) at 10W. I then tried the 1n5711 and again burned it out. Each time the diode burnt I sensed high vswr on the transceiver.
I checked the diodes with a multimeter (the diode function).
Instead of placing series resistors with the diodes, could you please suggest higher voltage/current diodes? I am not sure I can find that fast diodes (<30MHz) at that current/voltage.
I like the single scale feature of this simple meter as well as the passive circuit.

 

[biff44]

Thank you all for your replies.
The author stated schotkky diodes there but no type. I tried bat84 and burned one of them (the low one) at 10W. I then tried the 1n5711 and again burned it out. Each time the diode burnt I sensed high vswr on the transceiver.
I checked the diodes with a multimeter (the diode function).
Instead of placing series resistors with the diodes, could you please suggest higher voltage/current diodes? I am not sure I can find that fast diodes (<30MHz) at that current/voltage.
I like the single scale feature of this simple meter as well as the passive circuit.

10 watts is a lot.
they are probably PIN diodes or some other type of junction device, and probably have to have a low thermal resistance mounting scheme