Electronics > RF, Microwave, Ham Radio
Transistor choices and Active filters for RF?
rwgast_lowlevellogicdesin:
These are two totally unrelated questions. The first is about transistor choises in two different schematics.
So the first one is this pretty simple it called the "Amazing All Band Receiver" Its basically an RF detector with audio out.
http://www.techlib.com/electronics/allband.htm
So they chose to use a 2n4401, DataSheet, now ill admit ive only been doing RF a little while most of my experince is lower speed digital, i.e micros, hombrew z80 etc. In most circuits I design I stay away from transistors when I can and try to use FETS or OpAmps if at all possible. Basically I understand you want to pick a transistor based on gain(hfe) and speed(ft) so the hfe for the 2N4401 is 50 max with an Ft of 250mhz. Based on this im guessing this all band receiver doesn't make it past VHF. My real confusion is why this transistor is there something special about it that I am missing? The Ft for a 2n3904 and 2n222A are higher.. although there hfe is only 300, is the extra gain og the 2N4401 really needed? What if you just threw a much fast power transistor in the thing and used a pot to set bias/gain? is the 2n4401 something I should have on hand as a jelly bean?
Second is the ever popular PARD0T Active Mini Whip, here is the schematic. I really really want to listen to VLF and ELF and especially WWVB. This antenna seems like a pretty decent solution to running a mile long wire. Its also good to 20mhz so I can use it with my SDR up converter. This thing is only good to 20mhz though.... it is using two hi speed amplifiers. I get the JM310 is there to bring the noise floor down most likely because s/n is everything. What I dont get is the 2N5109 (DataSheet) it doesn't have a very high hfe which is understandable since this is all about s/n not overall db gain, but the ft is 1.2Ghz.... That seems like overkill for a VLF/HF amplifier, way beyond nyquist. These guys arent cheap either at $2.75 a pop. I acually plan on ordering a few for other things but why use it here and not just a 40mhz low hfe transistor? Is there something besides Hfe, Ft, that makes this a good choice for the 20mhz mini whip? Also you can get 10 of these for 8 bucks on E-Bay from China, should I trust that? I order from china all the time but I have had bad luck with higher end components being falsely advertised?
Lastly and hopefully an easy answer.. when building filters in RF we all build multi pole passives. My understanding is that there is no place for active filter simply because opamps aren't fast enough. Isnt that kind of outdated? I have 2 TI opamps in front of me rated for 1.7ghz the TI LMH6702 I also happen to know these were used in the front end of a an agilent signal generator so they must work well. Is there an issue with building active filters using chips like this? There only 4 dollars ill pay that for no insertion loss, and even gain! I also have some fairly pricey jfet based opamps that run up to 30mhz, im actually thinking of using these to make a low frequency mini whip...
edavid:
--- Quote from: rwgast_lowlevellogicdesin on July 08, 2016, 07:53:20 pm ---Based on this im guessing this all band receiver doesn't make it past VHF. My real confusion is why this transistor is there something special about it that I am missing?
--- End quote ---
The 2N4401 is only amplifying audio, so the Ft is irrelevant. Another kind of transistor would be fine though.
--- Quote ---Second is the ever popular PARD0T Active Mini Whip, here is the schematic. ... What I dont get is the 2N5109 (DataSheet)
--- End quote ---
Your link is bad.
Anyway, the 2N5109 is used as a high current class A amp, so it won't saturate at high signal levels. If you don't care about strong signals and/or you add a preselector, you could use a cheaper/weaker transistor.
--- Quote ---Lastly and hopefully an easy answer.. when building filters in RF we all build multi pole passives. My understanding is that there is no place for active filter simply because opamps aren't fast enough. Isnt that kind of outdated?
--- End quote ---
No, active filters are still impractical for RF.
rwgast_lowlevellogicdesin:
Thank you short and to the point. The transistor options make a little more sense now. For anyone else PARD0T schematic the site i linked the pic from is http://radioaficion.com/cms/pa0rdt-mini-whip/ the schematic is the first pick on the page.
Can you maybe explain why a low noise 1.7ghz op amp isn't practical for RF filtering applications? I built this recently
It is a 3.3v 28.8 TCXO with a 5 pole tunable low pass filter, this is to feed an RTL SDR and cut down on the internal harmonic noise from the LO. It works well I have brought the spike down 20db or more, but I want to get rid of the harmonics totally. My problem is the SDR wont work if the VPeakToPeak of the oscillator is under about 90~95mv this varies between sticks. I would have gone much farther and kept adding poles until I was content but there is just enough loss with 5 poles to tune from 60mv to 112mv. Adding another 2 or 4 poles would definately attenuate the voltage far to low to drive the R80T2 tuner chip. If we could use active filters even if it requires expensive op amps we could get much closer to perfect and almost eliminate most or all birdies from the LO to the RF Front end and the USB clock. So im very interested to know why hi speed opamps are still a bad choice :/ im sad to get that answer? Also what if your hetrodyning down to 455Kc are active filters acceptable on the 455khz IF?
edavid:
I think the idea of sine wave drive to the RTL SDR is pretty questionable... doesn't it just square up the clock internally? Maybe you just need to control the slew rate.
But if you do want a sine wave, I think you would be better off with a bandpass filter instead of a lowpass filter.
Or forget the TCXO and use a sine wave XO in the first place.
At 28MHz, an LC filter will give you higher Q than an active one.
rwgast_lowlevellogicdesin:
Hmm my LO filter isnt really where i wanted to take this but ill definately take good advice. First off why bandpass, the dongle only tunes to 25mhz so any undertone harmonics dont effect anything. I know that the tuner chip probably wants a square wave but it doesnt seem to dislike a sign and it clears up a lot of harmonics! I got the idea from a guy using a 3 pole on a si5351 his results look good i jist had a tcxo on hand and decided to use elsie to help me design a 5 pole. I would definatly like a 28.8mhz ovenized oscillator if you know where i can get one ill snatch it up! Please tell me!
What i really want to know is why a 1.7gig opamp cant be used in a filter up to 850mhz especially if a wide q is ok? In a broadband radio it would be nice to stack bp filters for instance on every tv broadcast band, theres no wat to do this without suffering lots of insertion loss. I can accept a fast opamp cant be used as an active filter for rf. I just want to understand why not? Obviously if it could be done it would have already.
Navigation
[0] Message Index
[#] Next page
Go to full version