The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl Design >> RF Design >> about mixer design https://designers-guide.org/forum/YaBB.pl?num=1251981806 Message started by manfred on Sep 3rd, 2009, 5:43am |
Title: about mixer design Post by manfred on Sep 3rd, 2009, 5:43am Hi everyone: I want to design a mixer, whose speccifications is as follows: Frf=2500MHz, FLO=2700MHz, so Fif is 200MHz conversion Gain=10dB, NF<15dB(SSB), P1dB>-15dBm, the input port is not needed to match to 50ohm, but the output port need drive offchip saw filter whose impedance is 100ohm(differential). the power should be < 10mA. but I have two problems : 1. are these specifications reasonable? (0.35um SiGe process) 2. the output port need to drive 100ohm impedance, what type of output stage is better? emitter follower or common emitter or any others? any reply is appreciated. Best regards manfred |
Title: Re: about mixer design Post by aaron_do on Sep 4th, 2009, 8:34am Hi, I haven't worked with SiGe, but since you have NPN transistors, I guess your gm/IDS is 1/Vt =40. So if you have 10 mA, you should get gm of 0.4 and if you use this to directly drive your load you should be able to achieve a conversion gain somewhere close to 40 = 26 dB. Since this is more than enough, you can even use some resistive degeneration to improve the linearity. The NF seems to be a given for that power consumption but again, i haven't worked in SiGe. You could add a buffer stage but I guess it would need most of the 10 mA budget in order to drive the 100 ohm load and therefore your NF would degrade. The advantage would be better reverse isolation and you could make your conversion gain independent of the load impedance. My last comment is that if you do put most of the power into the buffer stage, you might have to worry about the IF bandwidth achievable by the actual mixer stage. Anyway although i haven't worked with SiGe, most of the specs seem achievable. Hope it helps, Aaron |
Title: Re: about mixer design Post by manfred on Sep 5th, 2009, 9:50am thanks Aaron! yes, I also think most of 10mA maybe used by the output stage. but I don't know what type of output stage should be adopted? and the IF is about 210MHz, so the Bandwidth is also a problem. now I adopt traditonal Gilbert mixer core + emitter follwer output stage. but the gain is 4dB at the output of core, it become smaller at the follwer output which drive 100 ohm saw filter. cheers manfred |
Title: Re: about mixer design Post by loose-electron on Sep 5th, 2009, 11:33am 0.35 SiGe? Probably IBM (6HP) or Jazz.... Suggest looking into lowering the impedance of the mixer and using that (unbuffered) to drive the SAW directly. Why? Lower impedance equals less noise, better linearity from higher current and fewer active amplifiers in the system. If that doesnt fly, you will need to pay the penalty for a buffer amplifier. Emitter followers are lots better at HF than Source followers but still some questions on zero db gain (actually a little loss) and additive noise, linearity issues, and all that. |
The Designer's Guide Community Forum » Powered by YaBB 2.2.2! YaBB © 2000-2008. All Rights Reserved. |