8761	2013-06-27 08:24:23	mdgolfbum	Active antenna amplifier
Message	Date	From	Subject
This is to discuss the active antenna amplifier mentioned in the distortion thread earlier. I've posted two schematics in my AB3CV files area. The first pass file is my initial attempt at capturing Figure 5 from the article by Chris Trask (http://home.earthlink.net/~christrask/Complementary%20Push-Pull%20Amplifiers.pdf) modified by adding the differential buffer stage he mentions in the closing section. Since I didn't have the right transistor models immediately at hand I just stuck in what was easily obtained with LTSpice. AC simulation shows it works from about 10Khz to 10Mhz fairly well. DC simulation shows very low current in the output stage which I suspect will show a lot of crossover distortion as a result. The Output only file lops off the fet stage and inputs directly into the differential buffer for analysis purposes. I could have started anywhere but I chose the output for simplicity. This stage showed output loss of about 1db which I suspected was due to the low idle current so I added the 470ohm resistors to raise the idle current a bit and bypassed them with the caps shown. This improved the output loss to about 0.17db. I then figured out how to do the FFT analysis using the options shown in the schematic. 2nd harmonic is -90db and third is -84db which isn't a bad start. Next step is to replace the final stage with BFQ19 and BFQ149. I'll then play with idle currents and see what effect it has on output distortion products. Once I'm satisfied with the output stage I'll add back in the fet input and also try playing with BF994 as the input device. This is for my upconverter for sub 500Khz to 7Mhz to listen to the emerging LF MF ham bands and whatever else I can hear down there. No plans right now to do a transmitter. Right now I'm assembling the Jackson Harbor converter. This amplifier will be used at my TBD external antenna and run to the shack on 75ohm coax. The converter will be fed into my K3 transverter input. The display on the P3 should be interesting. I'm a LTSpice novice so please feel free to add suggestions. 73 jim ab3cv
8762	2013-06-27 08:44:51	Paul Daulton	Re: Active antenna amplifier
I've been a weaksignal chaser and beacon operator for about 13 years, I operate a beacon on 185.3khz under part 15 rules. Farcircuits sells the boards for the AMRAD see (April 2002 QST) up converter. I built a modified M0AYF loop preamp and 8 ft dia 14 turn loop with this antenna I have copoied beacons from NH and CT at a distance of 1250 miles. These beacons like mine run one watt into a 50ft antenna, 80 mw erp. Join us on the lowfer forum on qth.net. Paul k5wms Quoting mdgolfbum : This is to discuss the active antenna amplifier mentioned in the distortion thread earlier. I've posted two schematics in my AB3CV files area. The first pass file is my initial attempt at capturing Figure 5 from the article by Chris Trask (http://home.earthlink.net/~christrask/Complementary%20Push-Pull%20Amplifiers.pdf) modified by adding the differential buffer stage he mentions in the closing section. Since I didn't have the right transistor models immediately at hand I just stuck in what was easily obtained with LTSpice. AC simulation shows it works from about 10Khz to 10Mhz fairly well. DC simulation shows very low current in the output stage which I suspect will show a lot of crossover distortion as a result. The Output only file lops off the fet stage and inputs directly into the differential buffer for analysis purposes. I could have started anywhere but I chose the output for simplicity. This stage showed output loss of about 1db which I suspected was due to the low idle current so I added the 470ohm resistors to raise the idle current a bit and bypassed them with the caps shown. This improved the output loss to about 0.17db. I then figured out how to do the FFT analysis using the options shown in the schematic. 2nd harmonic is -90db and third is -84db which isn't a bad start. Next step is to replace the final stage with BFQ19 and BFQ149. I'll then play with idle currents and see what effect it has on output distortion products. Once I'm satisfied with the output stage I'll add back in the fet input and also try playing with BF994 as the input device. This is for my upconverter for sub 500Khz to 7Mhz to listen to the emerging LF MF ham bands and whatever else I can hear down there. No plans right now to do a transmitter. Right now I'm assembling the Jackson Harbor converter. This amplifier will be used at my TBD external antenna and run to the shack on 75ohm coax. The converter will be fed into my K3 transverter input. The display on the P3 should be interesting. I'm a LTSpice novice so please feel free to add suggestions. 73 jim ab3cv Paul Daulton K5WMS beacon WMS 185.302 khz qrss30/slow 24/7 Jacksonville,Ar 72076 em34wu [Non-text portions of this message have been removed]
8769	2013-06-27 23:47:48	mdgolfbum	Re: Active antenna amplifier
hi paul thanks for the lowfer list info. just submitted to join jim ab3cv
8771	2013-06-29 17:35:48	mdgolfbum	Re: Active antenna amplifier
I replaced the 2n2222 and 2n2907 with the BFQ19 and BFQ149 respectively as suggested in the Trask paper. You can see them in the BFQ Outonly file I uploaded. Their models were added to the standard.bjt file that I also placed in my files folder. The simulations appear a bit worse than the 2n2222/2n2907 in 2nd and 3rd output. Not really sure why that's occurring. I'm wondering if the 3904/3906 pair might be a better fit especially considering the LF/MF intended use. Comments encouraged. jim ab3cv
8772	2013-06-30 06:21:35	Chris Trask	Re: Active antenna amplifier
> >I replaced the 2n2222 and 2n2907 with the BFQ19 and BFQ149 respectively as suggested in the Trask paper. >You can see them in the BFQ Outonly file I uploaded. Their models were added to the standard.bjt file >that I also placed in my files folder. > >The simulations appear a bit worse than the 2n2222/2n2907 in 2nd and 3rd output. Not really sure why >that's occurring. > >I'm wondering if the 3904/3906 pair might be a better fit especially considering the LF/MF intended use. > >Comments encouraged. > The BFQ19/BFQ149 pair will outperform the 2N2222/2N2907, but at a higher collector current. In my complementary amplifier design the collector current for the output stage is around 100mA to get the performance that I quoted. The value for resistor R5 in figure 5 is a starting value, and it needs to be adjusted so as to obtain the desired operating current. This was mentioned in the text, and it might have been more obvious if I had put a potentiometer in the schematic rather than a fixed resistor. The KAA1000 that my design was based on had a potentiometer for that adjustment. Chris
8773	2013-06-30 06:46:32	mdgolfbum	Re: Active antenna amplifier
Hi Chris Thanks for the info! My datasheet from NXP shows Icc Max to be 100mA. You mention running the DC operating point there. Is there any issue with running on the edge at that point? Also I'm looking at 2nd and 3rd harmonic generation in the FFT as a proxy for linearity. Does that make sense? Thanks Jim ab3cv
8774	2013-06-30 07:42:47	Chris Trask	Re: Active antenna amplifier
> >Thanks for the info! My datasheet from NXP shows Icc Max to be 100mA. You mention running the DC >operating point there. > >Is there any issue with running on the edge at that point? > I should have read my own paper first before making that comment. The collector currents are around 50mA or more so that a comparison could be made with the 2N2222/2N2907 that I started with. BTW: The prices for the BFQ devices have decreased significantly since I wrote that. They can now be bought for less than $0.50 each in 100pc quantities from Mouser. > >Also I'm looking at 2nd and 3rd harmonic generation in the FFT as a proxy for linearity. Does that make sense? > Yes, it does as the harmonics and IMD are the result of the same device nonlinearities. Chris
8775	2013-06-30 11:56:58	mdgolfbum	Re: Active antenna amplifier
I've played around a bit more with both the BFQ and the 390x designs. Both work better when driven so that each transistor is working well below its rated hfe. I expect that is because it leaves lots of excess gain to ensure linearity. I've posted the best 390x file I've got working so far. 2nd is now -103db and 3rd is now -93db. I need to play around more with the BFQ design to see if I can get enough drive to bring it that low. 390x design posted. Next I'll start adding the fet front end. Comments encouraged. jim ab3cv
8776	2013-06-30 12:04:30	mdgolfbum	Re: Active antenna amplifier
Just a note about NF. I don't have a good feel for the noise figure of the 390x output at all. I'm making an assumption that since this design is targeted at <500Khz that with a proper antenna the natural background and manmade noise will completely swamp any reasonable preamp noise. 73 jim ab3cv
8780	2013-07-05 06:28:30	mdgolfbum	Re: Active antenna amplifier
I managed to get the combination of the front end and the 390x backend working with low simulated distortion numbers. I'll post them later today. I'll probably get this laid out for PCB sometime this month. Of course I can just connect it to an antenna to see how it works but I was wondering if there is someone with time and equipment who could do proper testing of NF and IMD of a unit. If so I'd build up a couple of units rather than just one for myself. 73 jim ab3cv