Close

Even faster !

A project log for YGREC-ECL

At the opposite site of the spectrum, this project is incredibly smaller and faster than the relay implementation of YGREC.

Yann Guidon / YGDESYann Guidon / YGDES 01/17/2019 at 21:3316 Comments

I found even faster transistors than the BFS480...
The BFP420 is 3x faster, at 25GHz gain-bandwidth product ! This means a gain of 25 at 1GHz, for example.

I received a few hundreds for tests. They are of course more expensive and are packaged alone in a small package (4 pins ??) so the density is lower but it could be helpful for the critical datapath, like the ALU.

I doubt I could measure the frequency of a ring oscillator made with it so I'll have to implement at least 9 stages ! :-P

One important limitation is the operating voltage : 4.5V so this limits the possible topologies.

OTOH it could also be used to build a very good picosecond pulse generator at low voltage :-P

Discussions

Tim wrote 01/26/2021 at 19:22 point

I also bought some of these and actually designed a PCB with an LNA, but never dared to test it. The application note is seriously scary (to me at least, maybe not to those well versed in HF black magic) - they use inductive emitter degeneration by some strategically placed vias. Also, the 4.5V limit is apparently quite strict and you can easily kill the device.


"For oscillators up to 10 GHz". Well, I guess building an oscillator is always easier than building an amplifier with devices like these.

  Are you sure? yes | no

Yann Guidon / YGDES wrote 01/26/2021 at 19:26 point

"inductive emitter degeneration by some strategically placed vias."
wut

"the 4.5V limit is apparently quite strict and you can easily kill the device."
I suppose so...

"building an oscillator is always easier than building an amplifier with devices like these."
And not only those...

  Are you sure? yes | no

Yann Guidon / YGDES wrote 01/26/2021 at 20:32 point

Would you share the appnote with us ?
Is this LNA a project on HaD as well ?

  Are you sure? yes | no

Tim wrote 01/26/2021 at 22:31 point

I think it was this appnote:

https://www.infineon.com/dgdl/Infineon-AN_1805_PL32_1806_113119_design%20guide%20for%20low%20noise%20transistors%20-AN-v01_00-EN.pdf?fileId=5546d46265f064ff016643d954321028

I notice now that it does not explicitly mention the BFP420, but it is generally about LNA design with SOT343 packages. I also got some BF776 in SOT343.

The LNA is not on HaD. I built up one version with BFU520W and got to about 350MHZ BW (and no oscillation or overshoot) at 19.4 dB gain. The design used shunt feedback. I would do it differently now.

  Are you sure? yes | no

alice crush wrote 01/26/2021 at 17:21 point

look for the pink light at bends in the traces j/k

  Are you sure? yes | no

Yann Guidon / YGDES wrote 01/26/2021 at 17:56 point

I don't get it...

  Are you sure? yes | no

Tim wrote 01/26/2021 at 19:17 point

cherenkov radiation? :)

  Are you sure? yes | no

Yann Guidon / YGDES wrote 01/26/2021 at 22:32 point

muahahaha

  Are you sure? yes | no

alice crush wrote 01/27/2021 at 06:30 point

the traces will leak photons pink in color from forcing the electrons around the corner

  Are you sure? yes | no

Yann Guidon / YGDES wrote 01/27/2021 at 11:53 point

You mean "synchrotron effect" ? :-D

  Are you sure? yes | no

matseng wrote 01/18/2019 at 00:13 point

At 11 Eurocents each in 1K qty (Farnell) I wouldn't call them particularly expensive - just a little bit more than double the price of plain old BC547's.

  Are you sure? yes | no

Yann Guidon / YGDES wrote 01/18/2019 at 00:56 point

BC547 can be sourced as low as .2 to .015€...

https://www.ebay.de/itm/1000-x-BC546B-Si-Transistor-NPN-65V-0-1A-0-5W-TO-92/312004936430 : 14€, or 20€ incl. shipping and taxes.

BFS480 has 2 transistors per package, which is pretty price-effective and dense on the board.

  Are you sure? yes | no

matseng wrote 01/18/2019 at 08:35 point

Well, you really can't compare prices on eBay with a proper supplier like Farnell. (Which had the prices I quoted). Oranges and apples you know ;-)

  Are you sure? yes | no

Yann Guidon / YGDES wrote 01/18/2019 at 10:47 point

@matseng in fact... I can ;-) and dollars are dollars ;-)

  Are you sure? yes | no

Ted Yapo wrote 01/17/2019 at 23:11 point

I wonder what kinds of steps you need to take to discourage parasitic oscillations at some ridiculously high frequency with these things. How would you even know if it were oscillating at 10+ GHz?

  Are you sure? yes | no

Yann Guidon / YGDES wrote 01/17/2019 at 23:21 point

to prevent resonance, you have to use the https://en.wikipedia.org/wiki/Barkhausen_stability_criterion : add phase or have gain <1 at the unwanted frequencies...

  Are you sure? yes | no