Preamp, which way is best?

[brandon7861]

I was looking at building the preamp on this page using a gear-driven capacitor from a TS-130s VFO in the tank. My first thought was to model it in simsmith using a tank with a reasonable Q directly into the Jfet biased as shown in the schematic, but that doesn't seem to be working in simsmith (could be my error)...

So I decided to try something different in simsmith to achieve better power transfer, couple to 50Ω output (where the jfet would be) using a capacitor and got this

and then I got to thinking, the JFET can have any resistor value at the gate, so why not get rid of the coupling cap and got this

Truth is, I have no clue what I am doing. Is it better to stay with the original schematic, couple the tank to 50Ω using a capacitor and use a 50Ω preamp, or do the last one and put the required 1.84kΩ resistor at the gate? I always thought lower resistances had less noise voltage, so the 1MΩ gate bias resistors is throwing me off. What should I do?

Thanks!!

Edit: just to clarify, L is the antenna, I forgot to rename that. Pigtail is the inch of wire where the coax splits off and gets soldered to the coil, about 20nH per inch, TankBelowTap and TankAboveTap are representing the tank inductor with a tap point, TankCap and Variable are both the tank capacitor, my variable only goes from 10pF to 40pF and it needed more. C1 is the coupling cap I used to prevent the tank from being loaded down and OutputLead is the same as the pigtail, the trace or wire going to the jfet or output connector.

 

[brandon7861]

I am going to guess that my simsmith confusion is coming from my assumption that I need good power transfer. I failed to consider that JFET inputs don't consume power. Since the JFET only cares about voltage and have essentially no current draw, I am now thinking the original schematic in the link may be best since the voltage is all that matters.

I guess I will go back into simsmith, set the load on the tank back to 5MΩ or so, then play with the tap point looking at bandwidth (tank loading) and overall tank voltage only.

this thing is getting built this weekend whether it works or not lol

 

[unit_399]

Brandon -

Not many of our members can solve their own problems, especially in just two posts !! Good job. . . hope that the preamp works.

- J.J. 399

 

[brandon7861]

It ended up being a pointless exercise. I spent all night looking for those two fairly expensive low-noise J310's I have been saving and couldn't find them. I am now planning to use a dual gate mosfet since I have them from my active probe projects and they are also fairly low noise. The cool part is that I can have a fairly wide control of the gain by changing the bias voltage on gate 1. I also have a decent idea what to expect for gate input impedance at that frequency thanks to others blazing the trail, so it should work out quite well.

ok, not pointless, just frustrating

 

[brandon7861]

I think my reasoning in post 2 was also incorrect. The idea that the JFET has an extremely high input impedance may apply to low frequencies, but probably not at RF. I was told on another forum that I need to look at power, something I cannot do if there is total reflection, so there has to be some finite input impedance at RF just like mosfets have.

My plan now is to build the amp without the antenna/tank part, then drive it with a signal generator using a multi-tap xfmr to try to figure out the actual input impedance based on the power that comes out as I change taps. Once I know the approximate input impedance, I can then match the known tank impedance to that value. Like I said before, there is reference to the BF998 having a 2kΩ input, but I'd like to know for sure.

What a trip lol

I will probably let this thread rest until I have a working amp, and then report on what I did to get there. I am way out of my depth on this one, could have just followed someone else's schematic and been happy with the result lol.

Edit: and I just had my koolaid pissed in again. Read a forum post on eev that states the mosfet I am using has the lowest NF at the highest gain and reducing the gain using gate 2 screws up the NF. Same post mentioned not using tanks because the insertion loss adds directly to NF, something I hadn't considered. Maybe I can use litz wire on ceramic or borosilicate as a form to lower the insertion loss. I don't know. The more I dig, the worse my idea looks, but this is more for learning than anything else, so whatever happens, happens. Starting to realize the best front ends don't come easy!