about pin diode linearity

[omegahijack]

hello everyone

I use some pin diode to do my experiment.

now I have some problem.

The linearity of pin diode always is good.

but now I need a very poor linearity pin diode to do my experiment .

so how can i make a poor linearity pin diode by using external circuit to do

this? or some other kinds diode that wideband and have nolinearity

characterist?

thx for help

 

[flatulent]

Try using an ordinary diode.

 

[biff44]

By linearity, I assume you mean high intercept point, or low ability to generate harmonics.

If you WANT to generate harmonics or intermod products, a PIN diode is a poor candidate because of the thick "I" region. For microwave signals, by the time electrons travel from one side of the I region to the other, the sine wave has already switched polarity, and the electron is not travelling back to the starting point. In other words, since the fast period of the microwave signal does not give enough time for an electron to travel across the "I" region and get collected, there is not RF current flow--hence a pretty linear on/off switch.

You can make a PIN diode be non linear by making the I region thin. A standard junciton diode, especially one with a little forward DC bias current (say 10 uA), will be pretty non-linear. Schottky diodes can be pretty non-linear. Reverse biased diodes have non-linear capacitance vs. voltage curves, and therefore generte intermods or harmonics readily. Many varactor diodes are doped to exacerbate this effect.

A lot of fiberoptic comonents are highly non-linear, and can cause excellent harmonic generation, such as Mach Zender modulators.

Is this what you are asking for?

Rich

 

[omegahijack]

By linearity, I assume you mean high intercept point, or low ability to generate harmonics.

If you WANT to generate harmonics or intermod products, a PIN diode is a poor candidate because of the thick "I" region. For microwave signals, by the time electrons travel from one side of the I region to the other, the sine wave has already switched polarity, and the electron is not travelling back to the starting point. In other words, since the fast period of the microwave signal does not give enough time for an electron to travel across the "I" region and get collected, there is not RF current flow--hence a pretty linear on/off switch.

You can make a PIN diode be non linear by making the I region thin. A standard junciton diode, especially one with a little forward DC bias current (say 10 uA), will be pretty non-linear. Schottky diodes can be pretty non-linear. Reverse biased diodes have non-linear capacitance vs. voltage curves, and therefore generte intermods or harmonics readily. Many varactor diodes are doped to exacerbate this effect.

A lot of fiberoptic comonents are highly non-linear, and can cause excellent harmonic generation, such as Mach Zender modulators.

Is this what you are asking for?

Rich

thanks for help biff44

I need some nonlinearity component which can generate nonlinearty characterist

at each frequency (I mean broadband). That is why I choice pin diode but I didn't

notice, it is high linearity.

So what kind nonlinearity component can achieve my condition (broadband and

nonlinearity at each frequency)

I can't use "diode" becasue it's not broadband.

or have some kind external circuit can achieve my condition??

thx for help

 

[karesz]

Hi omegahijack,
What about some SRD (Step Recovery Diodes) application?
www.aeroflex.com/ams/metelics/micro-metelics-prods-silicon-diodes.cfm

But I think you are not right with the sentence:
"I can't use "diode" because it's not broadband."
There are lot of highest frequency (MW) diodes existent...
K.

 

[omegahijack]

Hi omegahijack,
What about some SRD (Step Recovery Diodes) application?
www.aeroflex.com/ams/metelics/micro-metelics-prods-silicon-diodes.cfm

But I think you are not right with the sentence:
"I can't use "diode" because it's not broadband."
There are lot of highest frequency (MW) diodes existent...
K.

thx karesz sorry about the sentence

rewrite:

I need a diode which have "same" non linearity characterist at each frequency

(ex. insertion loss....)

 

[biff44]

You have replied twice, but have not provided any additional information about what you are trying to do. One might need to know power level, frequency range, desired non-linear effect,etc, before answering further.

 

[omegahijack]

You have replied twice, but have not provided any additional information about what you are trying to do. One might need to know power level, frequency range, desired non-linear effect,etc, before answering further.

I am really sorry , I am not good in english.

I need a component. This component is wideband. For example, its s21 is flat in 500mhz ~ 3000mhz.

and It's nonlinearity(intercept point) is same in 500mhz~3000mhz too



[/img]

 

[Borber]

Try BAT15-03W or BAT62-03W for higher voltages.

 

[karesz]

omegahijak,
My english is-I belief- more bad, but my be we can communicate with us...
What about over thes way/solution?
**broken link removed**
K.

 

[biff44]

500-3000 mhz is a reasonable bandwidth. You could do all of the follwing.

Put antiparallel shottky diodes across the transmission line (from transmission line to ground), and drive the RF input with enough power so that they clip.

If you want different order of harmonics, you can put antiparallel diodes in series.

You can take a broadband amplifier (say one to 10 Ghz) and overdrive it so that it clips.

How about these ideas?

 

[omegahijack]

500-3000 mhz is a reasonable bandwidth. You could do all of the follwing.

Put antiparallel shottky diodes across the transmission line (from transmission line to ground), and drive the RF input with enough power so that they clip.

If you want different order of harmonics, you can put antiparallel diodes in series.

You can take a broadband amplifier (say one to 10 Ghz) and overdrive it so that it clips.

How about these ideas?

biff44 thx a lot you give a good idea

I have some question about the antiparallel shottky diodes.
As we know , this circuit will be some inertion loss maybe S21 = -3dB

1.does this circuit 's insertion loss will be the same magnitude
at each frequency in 500MHz~3GHz ??

2.does this circuit 's nonlinearity will be the same magnitude
at each frequency in 500MHz~3GHz ??

I know this circuit will generate nonlinearity in each frequency in 500M~3G

but does it the same magnitude

at each frequency in 500m~3G??

 

[biff44]

Will it be the same at all frequencies? No. Will it be within maybe +/- 4 dB? Probably.

I would assume that the circuit would get lossier at the high end of the frequency range, so I would build a microwave loss equalizer...something with maybe 5 dB of loss at 500 MHz, and 0 dB of loss at 3 Ghz.

If you know ahead of time what the frequency is, you can simply use a variable attenuator and a PROM control circuit to linearize the intermods vs input frequency.

To get really flat intermods vs input frequency with a broad instantaneous bandwidth, you probably have to use some sort of device that has ultrawide bandwidth. Like take a 12 Ghz bandwidth Mach-Zender amplitude modulator, drive it with a CW laser diode, on the modulation port put your RF signal PLUS a DC bias to drive it into a very non-linear region. Then convert the modulated laser light back to RF with a photodetector. Stuff like that.