PIN diode selection

[anvoice]

I'm trying to build the attached schematic to be used as a quarter wavelength switch. Unfortunately I'm having trouble finding a PIN diode with the appropriate characteristics for D1 and D2. Specifically, it seems very difficult to find a diode to operate at my required frequency, which is around 0.5-1MHz. I tried several places including Mouser, which has a tiny selection of diodes which operate from 100kHz and up, but the 100kHz in particular is an smd item that costs about $6 a shot, which is a bit steep. I was hoping to source from ebay, but even when I find datasheets for the ebay diodes they often don't specify frequency. Is there a good source of PIN diodes I can look through? Preferably they'd operate from 500kHz, but even 1MHz would be ok.

[w2aew]

How about something like the BAR-64 series from Infineon - looks to have a long enough carrier lifetime for your application.  Available from Mouser:
https://www.mouser.com/ProductDetail/Infineon-Technologies/BAR-64-03W-E6327?qs=mzcOS1kGbgfTLMy%2FY7i0og%3D%3D
https://www.mouser.com/datasheet/2/196/Infineon-BAR64series-DS-v01_02-EN-1107806.pdf

[David Hess]

If I could not find a suitable PIN diode, then I would test various inexpensive varactor diodes.

[anvoice]

How about something like the BAR-64 series from Infineon - looks to have a long enough carrier lifetime for your application.  Available from Mouser:
https://www.mouser.com/ProductDetail/Infineon-Technologies/BAR-64-03W-E6327?qs=mzcOS1kGbgfTLMy%2FY7i0og%3D%3D
https://www.mouser.com/datasheet/2/196/Infineon-BAR64series-DS-v01_02-EN-1107806.pdf

You're right, this one looks pretty good. Many thanks.

If I could not find a suitable PIN diode, then I would test various inexpensive varactor diodes.

I think w2aew found about the only properly specced PIN diode for my application. Truth be told I've never heard of varactor diodes before, goes to show how much I still need to learn.

Thank you!

[w2aew]

How about something like the BAR-64 series from Infineon - looks to have a long enough carrier lifetime for your application.  Available from Mouser:
https://www.mouser.com/ProductDetail/Infineon-Technologies/BAR-64-03W-E6327?qs=mzcOS1kGbgfTLMy%2FY7i0og%3D%3D
https://www.mouser.com/datasheet/2/196/Infineon-BAR64series-DS-v01_02-EN-1107806.pdf

You're right, this one looks pretty good. Many thanks.

If I could not find a suitable PIN diode, then I would test various inexpensive varactor diodes.

I think w2aew found about the only properly specced PIN diode for my application. Truth be told I've never heard of varactor diodes before, goes to show how much I still need to learn.

Thank you!

Another option for LF/HF applications would be high voltage rectifiers such as 1N4007 - sometimes they're built as PIN structures, or at least have a long enough carrier lifetime to serve reasonably well as LF/HF switches.

I'm assuming that you already saw my video on the PIN-diode based T/R switch based on that same topology...

[anvoice]

I'm assuming that you already saw my video on the PIN-diode based T/R switch based on that same topology...

I did indeed. Super helpful video: I just figured out it was you though.

I actually found some diodes on ebay that appear to operate at 1MHz and above: they're Infineon BAR65-03W and BAR63-02V. Those might work if I can get away with the higher frequency. Unfortunately I don't know which characteristics are preferable besides the low minimum frequency, i.e. should I minimize resistance?

If I may be so immodest as to ask you another question since you made that T/R switch video: do you know how close the capacitance and inductance need to be to the theoretical values for the equivalent circuit to the 1/4 waveguide? I tried asking on this forum earlier but it looks like nobody that looked knew the answer.

[w2aew]

I'm assuming that you already saw my video on the PIN-diode based T/R switch based on that same topology...

I did indeed. Super helpful video: I just figured out it was you though.

I actually found some diodes on ebay that appear to operate at 1MHz and above: they're Infineon BAR65-03W and BAR63-02V. Those might work if I can get away with the higher frequency. Unfortunately I don't know which characteristics are preferable besides the low minimum frequency, i.e. should I minimize resistance?

If I may be so immodest as to ask you another question since you made that T/R switch video: do you know how close the capacitance and inductance need to be to the theoretical values for the equivalent circuit to the 1/4 waveguide? I tried asking on this forum earlier but it looks like nobody that looked knew the answer.

The overall Q is quite low, so they don't have to be super critical.  Often tweaking the windings of the toroid is enough to get sufficient rejection. 

On the PIN diodes, they can be used at lower frequencies when the signal level is low enough compared to the bias level.

[anvoice]

Got it. I was considering a variable inductor but I can't find one on ebay with a frequency specification. Instead I will opt for a micrometals mix 2 toroid ferrite core and some 20awg wire, and wind one myself.

[Rainwater]

How is the absolute maxium power through the diode determined from the specifications in the data sheet?
(Values taken from page 2 of the above linked pdf)

Edit: after a long search this is the best explanation i've found
https://www.quanticpmi.com/application-notes-how-to-specify-pin-diode-switches

Page 3
Thermal Model
The maximum allowable power dissipation, P_D, is determined
by the following equation:
P_D = (T_J - T_A) ÷ θ
where T_J is the maximum allowable junction temperature
(usually 175°C) and T_A is the ambient or heat sink temperature.
Power dissipation may be computed as the
product ot the RF current squared multiplied by the diode
resistance, R_S.
For CW applications the value of thermal resistance, θ,
used is the average thermal resistance, θ_AV.

[David Hess]

Another option for LF/HF applications would be high voltage rectifiers such as 1N4007 - sometimes they're built as PIN structures, or at least have a long enough carrier lifetime to serve reasonably well as LF/HF switches.

For a 2 meter transmitter hunt where I was hiding, I once made a series-shunt power attenuator using 1N4007s so that I could vary the transmitter power randomly.  I used a linear feedback shift register to drive a resistor DAC which fed a diode anti-log generator to drive the PIN diode attenuator and produce a reasonable dynamic range of output power.  Input power to the attenuator was about 1/2 watt.

If I did it today, I would replace the LFSR and DAC with a flicker noise generator.

[geggi1]

J-FET can also be used as RF switches.
https://www.rfwireless-world.com/Terminology/FET-RF-Switch-vs-PIN-diode-RF-Switch.html
This might be better suited for your application.

[David Hess]

J-FET can also be used as RF switches.
https://www.rfwireless-world.com/Terminology/FET-RF-Switch-vs-PIN-diode-RF-Switch.html
This might be better suited for your application.

The PIN diode method worked fine and was trivial to implement, and it was a power attenuator so even if JFETs could be used, modern ones are not powerful enough.  That I could use cheap 1N4007s just made it even better.

The better alternative would be to use a multiplier or operational transconductance amplifier for continuous gain control, but the PIN diode attenuator had the advantage of operating directly with a 0.5 watt signal so no power amplifier after the attenuation stage was required.

I also know how it could be done with matched power MOSFETs, but a PIN diode implement is both less expensive and simpler.