I took out the Locky Tracer and remove some dust, it was about time. One of the most popular lower FETs I use in the gyrator PCB is the now obsolete BF862. Here is the tracing output:
You can see that for a typical preamp stage where the DHT has an anode current from 5 to 15mA, the transconductance is 30mS to 40mS. This yield an output impedance of 30Ω down to 25Ω in a hybrid mu-follower stage.
Now, we can look at the enhancement MOSFETs which I introduced long time ago: BSH111BK or BSN20K:
Both devices behave similarly at low voltage and current. Despite the BSN20BK has higher transconductance at large current, when we look at both of them below 30mA, you can see that they are not far off. Even at 5mA they have higher transconductance than the BF862 (50mS versus 30mS). This will yield an even lower output impedance.
Now, let’s look at the top devices.
One of the reasons why I introduced the IXTP08N100D was that as a top device in a cascode pair can provide a higher VDS headroom to the lower device given the VGS behaviour compared to the DN2540. Below is the clear evidence. Let’s pick 20mA for example. The DN2540 will bias the lower device VDS at about 1.4V. On the contrary the IXT08N100D will provide 2.0V of VDS. This will yield a better frequency response of the hybrid mu-follower stage as the FET capacitances are minimised the higher the VDS voltage is.
Now, have a look at the IXTP3N100D, which can deliver 3A!. It has an even better VGS curve than the 08N100D. Adding this device as the top MOSFET in the gyrator PCB will deliver an even better frequency response.
Now, it’s time for matching FETs for my new RIAA project. 🙂
Stay tuned
Excellent measurements, thanks Ale.
I need it, especially for calculate driver stage Zout (from MU output of CCS or gyrator).
Rajko
Hi Ale,
How about looking at Walt Jung’s bias multiplier to further increase the VDS for the bottom device?
tim
Hi Tim, I’d tried the bias multiplier but don’t see the trouble of doing so with a device like the IXT3N100D available now. Even the IXT08N100D is good enough and I can get +1.5MHz bandwidth which is more than what we need in audio 🙂
Exceptional Ground Breaking effort from Yourself Ale, With exception to the IXTP3, I have listened to all of these devices & can say that I agree with those results totally on listening alone & no longer use BF862 or DN2540.
Yes Barry, the IXTP3 is a new discovery. Must try it in some boards soon! Thanks