On Sat, Sep 21, 2024 at 04:37 PM, AndyH wrote:
Terminations as farhan says is important!
?
Also Rf (feedback resistors) vs bias...?
Sometimes that resistor can be juggled to be the bias value.? Often
its too low for that so series resistance bypassed with suitable value
cap gives a valid DC bias with correct feedback.
?
Additionally many devices the FT and Beta vary with emitter current.
Some like the 5551A show curves that improve with increased current
also same for 2n2219A.? Others like the frequently used 2n3904 actually
decreases above about 12-15ma (best peak is 10ma).? Searching for
detailed datasheets helps with getting that information. Some vendors
have notably sparse datasheets making that harder to know.? Looking
at the same device from a different vendor often results in a more
detailed datasheet.
?
One thing seen with very high FT devices is oscillation at vhf or uhf
(usually well above frequency of interest) and usually in multi stage
amplifiers that often shows as reduced gain or less than expected
power.? For example the FT for the 2sc5551A 3.5ghz!? ?With a gain
bandwidth product layout and construction is a factor.? Sloppy long
leads and poor grounds will bite back.
?
At the other extreme the usual 2n3904 the FT is 300mhz which varies
from 250 to 380mhz or more.? The problem with that is; build a gain
of 17DB for use at 28mhz...? it does not net the correct gain in practice
due to low FT, it has more than enough beta.? The test is
FT/frequency-max. so for a gain of 20 (13db) at 28mhz the FT must
be more than 560mhz (500mhz will likely do)!? ?Conversely the best
a 300mhz part can do?across many samples is about 10 (10db!).
?
Resistive feedback can also be used with jfets and mosfets as well
The calculations are different due to the gain nature of the devices.
FETs are voltage to current so the calculations need some math
translation.? Generally it works the same and FETs (mos or junction)
do not for the most part exhibit FT.
?
Building wideband amplifies is actually more difficult than the
basic calculations from W1ZOI suggest as they hold well for
small signal at below VHF.? ?As we go up the reactive properties
and unintended feedback become a more significant factor.
?
--
Allison
------------------
Please use the forum, offline and private will go to bit bucket.
