Hi, all,
I'm up in the Adirondacks sitting on a deck above a lake, so for R&R I'm de signing a transistor noise tester to stand in for the late lamented HP 4470 A that I can't find for sale.
I'd like it to have a current range of 100:1 or so, ideally without range s witching. That sort of range isn't hard to do with a pot, and I can progra m two BJT current sources (for the source and drain) from the same pot by u sing an ordinary phase splitter circuit. No problems there.
However, digging a bit deeper, it's actually pretty interesting. The main sticking point is just how quiet the bias current sources have to be when y ou're testing JFETs.
The output current noise of a JFET is the Johnson noise of a resistance of
2/(3 g_M). To look at the relative influence of the bias current source in the output noise, we can start with full shot noise and calculate the resu lting noise figure asNF = 1+(i_Nshot / i_Nfet)**2.
After a few lines of algebra this comes out to be
NF = 1 + I_d/(g_M * 33 mV)
If you plug in the performance of a BF862 at 10 mA or so, you get a NF of a bout 13 (11 dB), which is the pits.
Sooo, we need to apply some feedback to quieten down the current source. B ut how much?
The small-signal emitter circuit of a BJT looks like a DC current source in parallel with a shot noise current source, in parallel with a resistance 1 /g_M = 26mV/I_C.
If you add another 26 mV worth of drop in an external resistor, half the no ise current loops around the g_M path and half goes into the external circu it. Of course you get the Johnson noise of the external resistor, so a bit more algebra leads to the conclusion that the resulting current noise goes down by 3 dB when you drop 52 mV across the external resistor. It's down to a quarter at 200 mV, and so on. 200 mV thus gets us within 1 dB of the Johnson noise. For a measuring instrument, one would like to be within 0.5 dB at worst, which needs 400 mV. Not awful.
Since with the JFET the current has to be 13 times quieter than shot noise even to get to a 3 dB NF, we need like 13 x 400 mV, or a bit over 5V. Stil l not horrible.
However, combining that with wanting 100:1 in a single range leads to the r equirement of 500V of supply headroom, and some gross amount of power dissi pation.
I don't want to add breakpoints to the current source, because that's going to mess up the tracking of the source and drain, and probably cause a lot of voltage swing at the drain.
So, I'm looking at ways of putting a bandaid on the current source to make it quieter without needing such a high voltage supply.
The obvious one is to use a simulated inductor, which is just a cap multipl ier used back to front (going in the emitter and coming out the collector). It'll need breakpoints too, but that has to be done without adding excess noise from the breakpoints themselves.
Any wisdom?
Thanks
Phil Hobbs