I need a good text on low noise amplifier design ASAP

;)

Besides, they only have supermarket tonic water. (In New Mexico, it's supermarket gin, too--thus it's a self-limiting process.)

Cheers

Phil Hobbs

Ok, a bit more adjustment, and things are looking a little better. Turns out that the good old 2N3904 is a better fit for this application than the BC850 I was using to start with. (AofE third edition just arrived with good info on page 501)

The circuit is a common base amplifier, biased to 100uA, with a parallel resonant tank in the collector path.

So with a 300nV input, LtSpice tells me I have 12uV on the output, but 99nV/Hz^1/2 noise on the output.

I'm new to using LTSpice noise functions, so do I have 60uV noise (horrible) on the output or 1.5uV ?

Noise sim setup: Output V(Out) Input V2 (300nV AC signal) Linear sweep, 1000 points, 425000 to 525000 hz.

Suggestions on where to post a spice file?

I bet the antenna doesn't have a secondary and those few turns are for a (hope I remember the type correctly) super regen type receiver

You wouldn't be having these problems if you drank rum like (slightly) more sensible people. Tonic water is a medication for malaria, isn't it?

Oh just slap the ltspice *.asc file into your text reader (notepad?) cut and paste it here.

George H.

No and no.

Ok, I wasn't sure that was appropriate. I've applied some info from the copy of AofE that arrived this morning. It turns out that the BC847 is not that great a device for this, and the lo wly 2N3904 is better, so I re-biased for that.

I have a gain of 40 on the front end, but I think spice is telling me I sti ll have a lot of noise. Should I divide the noise number (99nV/Hz^1/2) by the amplifier gain?

Version 4 SHEET 1 4508 2096 WIRE -1424 -448 -1616 -448 WIRE -1248 -448 -1344 -448 WIRE -1136 -448 -1248 -448 WIRE -64 -448 -1136 -448 WIRE 112 -448 -64 -448 WIRE 288 -448 112 -448 WIRE -1616 -400 -1616 -448 WIRE 112 -400 112 -448 WIRE 288 -400 288 -448 WIRE -1616 -304 -1616 -320 WIRE 112 -272 112 -320 WIRE 112 -272 -16 -272 WIRE -16 -240 -16 -272 WIRE 112 -240 112 -272 WIRE 288 -208 288 -320 WIRE 480 -208 288 -208 WIRE 608 -208 480 -208 WIRE 480 -176 480 -208 WIRE -16 -128 -16 -176 WIRE 112 -128 112 -160 WIRE 112 -128 -16 -128 WIRE 432 -112 416 -112 WIRE 480 -112 432 -112 WIRE 480 -96 480 -112 WIRE -1456 -48 -1552 -48 WIRE -1152 -48 -1280 -48 WIRE 288 -48 288 -208 WIRE -1456 -16 -1456 -48 WIRE -1280 -16 -1280 -48 WIRE 112 0 112 -128 WIRE 224 0 112 0 WIRE 480 0 480 -16 WIRE -1552 48 -1552 -48 WIRE 288 64 288 48 WIRE 480 64 288 64 WIRE 608 64 480 64 WIRE -608 80 -688 80 WIRE -496 80 -528 80 WIRE -320 80 -416 80 WIRE -144 80 -256 80 WIRE -32 80 -144 80 WIRE 112 80 112 0 WIRE 112 80 64 80 WIRE 288 80 288 64 WIRE -1456 96 -1456 64 WIRE -1280 96 -1280 64 WIRE -1280 96 -1456 96 WIRE -1456 128 -1456 96 WIRE -1280 128 -1280 96 WIRE 432 128 416 128 WIRE 480 128 432 128 WIRE 480 144 480 128 WIRE 176 160 112 160 WIRE 208 160 176 160 WIRE 112 176 112 160 WIRE 288 176 288 160 WIRE -688 192 -688 80 WIRE -1552 224 -1552 112 WIRE -1456 224 -1456 208 WIRE -1456 224 -1552 224 WIRE 480 240 480 224 WIRE -144 272 -144 80 WIRE 16 288 16 144 WIRE 112 288 112 256 WIRE 112 288 16 288 WIRE 112 304 112 288 WIRE 16 320 16 288 WIRE -688 400 -688 272 WIRE -144 400 -144 352 WIRE -144 400 -688 400 WIRE 0 400 -144 400 WIRE 16 400 16 384 WIRE 16 400 0 400 WIRE 112 400 112 384 WIRE 112 400 16 400 WIRE 0 416 0 400 FLAG -1136 -384 0 FLAG -1616 -304 0 FLAG -1248 -384 0 FLAG 288 176 0 FLAG 0 416 0 FLAG 480 240 0 FLAG 432 128 Xsig FLAG 480 0 0 FLAG 432 -112 Xinvsig FLAG -64 -448 2V8 FLAG 176 160 2V8 FLAG -1280 128 0 SYMBOL cap -1152 -448 R0 SYMATTR InstName C7

SYMATTR SpiceLine V=4 Irms=1.62 Rser=0.00502825 Lser=0 mfg="KEMET " pn="C0201C104K7PAC" type="X5R" SYMBOL voltage -1616 -416 R0 WINDOW 123 0 0 Left 2 WINDOW 39 24 44 Left 2 SYMATTR SpiceLine Rser=0.01 SYMATTR InstName V1 SYMATTR Value 2.8V SYMBOL cap -1264 -448 R0 SYMATTR InstName C17

SYMBOL voltage -688 176 R0 WINDOW 123 24 124 Left 2 WINDOW 39 0 0 Left 2 SYMATTR Value2 AC 300nV SYMATTR InstName V2 SYMATTR Value SINE(0 300nV 475000) SYMBOL res 96 -416 R0 SYMATTR InstName R13 SYMATTR Value 14k SYMATTR SpiceLine tol=1% SYMBOL npn 224 -48 R0 SYMATTR InstName Q5 SYMATTR Value BC847C SYMBOL res 304 176 R180 WINDOW 0 36 76 Left 2 WINDOW 3 36 40 Left 2 SYMATTR InstName R24 SYMATTR Value 1k SYMATTR SpiceLine tol=5% SYMBOL res -1328 -464 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R6 SYMATTR Value 1 SYMBOL ind 96 -256 R0 SYMATTR InstName L1

SYMATTR SpiceLine Rser=12 SYMBOL cap -32 -240 R0 SYMATTR InstName C1

SYMBOL npn 64 144 M270 SYMATTR InstName Q1 SYMATTR Value 2N3904 SYMBOL res -160 256 R0 SYMATTR InstName R1 SYMATTR Value 1k SYMATTR SpiceLine tol=5% pwr=0.1 SYMBOL res 96 160 R0 SYMATTR InstName R5 SYMATTR Value 25.5k SYMATTR SpiceLine tol=1% pwr=0.1 SYMBOL cap -256 64 R90 WINDOW 0 0 32 VBottom 2 WINDOW 3 32 32 VTop 2 SYMATTR InstName C4

SYMATTR SpiceLine V=4 Irms=162m Rser=0.0322889 Lser=0 SYMBOL ind -400 64 R90 WINDOW 0 5 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName L7

SYMBOL res 96 288 R0 SYMATTR InstName R9 SYMATTR Value 8.45k SYMATTR SpiceLine tol=1% pwr=0.1 SYMBOL res -512 64 R90 WINDOW 0 0 56 VBottom 2 WINDOW 3 32 56 VTop 2 SYMATTR InstName R20 SYMATTR Value 129 SYMBOL cap 0 320 R0 SYMATTR InstName C16

SYMBOL res 304 -304 R180 WINDOW 0 36 76 Left 2 WINDOW 3 36 40 Left 2 SYMATTR InstName R26 SYMATTR Value 1k SYMATTR SpiceLine tol=5% SYMBOL cap 464 64 R0 SYMATTR InstName C39

SYMBOL res 464 128 R0 SYMATTR InstName R53 SYMATTR Value 100k SYMBOL cap 464 -176 R0 SYMATTR InstName C33

SYMBOL res 464 -112 R0 SYMATTR InstName R44 SYMATTR Value 100k SYMBOL ind -1472 -32 R0 SYMATTR InstName L2

SYMBOL ind -1472 112 R0 SYMATTR InstName L3

SYMBOL ind -1264 -32 M0 SYMATTR InstName L4

SYMBOL cap -1568 48 R0 SYMATTR InstName C2 SYMATTR Value 560pF TEXT 600 -312 Left 2 !;ac lin 100000 425000 525000 TEXT -1440 -392 Left 2 ;Long track TEXT 592 -344 Left 2 !.noise V(Xsig) V2 lin 10000 425000 525000 TEXT 600 -400 Left 2 !;tran 100uS TEXT -568 136 Left 2 ;X Antenna TEXT -720 -48 Left 2 ;Antenna Z=132 +13.7 deg TEXT -720 -24 Left 2 ;Measured on PCB with antenna tuned TEXT -720 -72 Left 2 ;Approximation of tuned ferrite bar antenna TEXT -184 -16 Left 2 ;Also tried BC847 here TEXT -1840 -144 Left 2 ;Antenna is 200uH nominal, in two series connected w indings with center tap grounded. TEXT -1232 -64 Left 2 ;To input at C4

Bit of a body mass issue. Rum is a medicine to relieve poverty, but too much Zacapa 25 has the opposite effect. ;)

Cheers

Phil Hobbs

Am 25.01.2017 um 20:53 schrieb snipped-for-privacy@gmail.com:

yes. The result "gain" is predeclared. V(onoise) / gain You can also plot 20*log10(gain) for gain in dB

and see the noise contribution of the R in the tuned circuit: V(R20)/gain

But, current starved and voltage miser, your problem won't be noise but IMD.

regards, Gerhard, DK4XP

Thank you!

Ok, it's starting to make sense now, I have parts on order, and I'll be building it ASAP.

Agreed on the additional issues, but I'm making progress.

I think you can now buy genuine Havana Club Rum (not the Bicardi imitation) in the USA, at least until The Donald shuts things down.

I got a bottle of Havana Club when I was in Moscow. It was excellent. But I couldn't bring it back into the USA, so we drank it.

Shiny! Now we can make sense of things. You can plot the input-referred noise by selecting the plot window, and then Plot settings -> Add trace -> V(inoise) and click OK. You'll find about 2.4nV/rtHz at 480kHz, which is not so bad at all. Good enough for a receiver at that frequency.

Something I find a real time saver is that you can click on resistors and transistors, and the contribution of each to the *output* noise appears in the plot window. So it's very easy to see that Q1 makes most of the noise, and R20 is the 2nd largest contributor. The other components are negligible. Keep in mind that noise *powers* add, not noise voltages. (Square Voltages, add, take square root.)

Finally, for AC analysis the circuit is linearized, so you can set the source amplitude to unity and then the output level equals the gain. You'll find about 32dB at 480kHz. If you prefer to think of the emitter of Q1 as the input, you could plot V(xsig)/V(n011). You'd get about 36dB in that case, independent of the AC level of V2.

Jeroen Belleman

Now that we finally know that the amplifier input related noise performance is about acceptable and the pick-up coil impedance is not very strange (110 ohms) but still no difference between connected and disconnected loop, the conclusion is that the ferrite antenna is simply not perfect for the application.

• The permeability of the rod might be wrong for this frequency range.

• The resonant circuit losses might be too high, are there proper Litz wires i the resonator wire and pick-up loop ?

With proper ferrite materials and coils even the DCF 77.5 kHz Mayflingen (Frankfurt) radio clocks works with a radius of 2000 km with a 5 cm ferrite rod.

To the OP, take a look at loop antenna theory, e.g.

Your antenna model is fishy: If you have two 100uH coils in series on common ferrite rod, the combined inductance is around 400uH. For LTSpice, you have to set coupling coefficients for the three coils on the antenna rod.

You're running the amplifier with far too small collector current to survive the mess of signals on the MF band. I'd look for something around 10mA instead of 0.1mA.

On medium frequencies, the band is so noisy that big-signal linearity and front-end selectivity are far more important than noise figure.

That wasn't intended as a circuit model, just as a representation of the antenna circuit. Note that it is not connected to the main simulation at all. The two coils measure 74uH individually, but I wanted to show the effective circuit of a center tapped 200uH coil.

Possibly so. But at the moment that is a different problem.

Ok, where can I read up on this?