Hi Jay,
Well, i have further questions to you but first yours...
Am 31.05.2012 15:59, schrieb [email protected]:
Stefan
Thanks for the positive comments.
Have to admit I haven't been following what you're up to ... curious
what problems you're running into with VMCD at 500 kHz?
I didn't spend much time on the MF PA yet.
First i expected it is very uncomplicated at this frequency. I tried
different ferrite cores starting with a FT140-77 (2x6 turns primary,
later 2x10) and then a EPCOS core (B64290L0022).
Later i thought i can use a much smaller core it this frequency and so
i'm using a smaller Epcos core now, size about like a T106.
All PAs are Voltage Mode Class D PAs so far! The driver is a single ICL7667CPA
device. It becomes warm but the gate voltage looks very good. If there
is only the primary windong on the transformer then the current
consumption is about 0.4A at 12VDC. Don't know if this value would be
lower if a IXDD614CI would be used but 0.4 A is still "OK". This
current rises when the secondary winding is applied and the output band
filter is applied (still without a output load) to about 0.7A. Note
that my output filter configuration is similar than in my LF PA (
http://dl.dropbox.com/u/19882028/LF/schematics/LF%20H%20bridge%20PA%2C%20DK7FC.png),
here at MF starting with a series LC, then parallel, serial and a final
parallel filter on the BNC output (6.8nF/16.5uH each).
However there is a heavy oscillation on the Drain-Source during the OFF
time, beeing clipped by the FET at arround 60V (losses!) and then
falling to 0V. That's the first problem. This is why i thought i should
use a smaller core, not so store so much energy in it which oscillates
with the output C of the FET then. Indeed, the smaller core allowed to
better damp this oscillation with a 220pF and serial 100 Ohm resistor
across Drain-Source. Anyway the efficiency is poor.
BUT at least this PA has no problem with a missing load. I think it
behaves like a voltage source and also the output filter config should
be less critical to mismatch. This is what i like, a PA that is as
robust as possible against mismatch.
This leads me to my question about the CMCD config: How critical does
it act on a mismatch? What happens if i forget to apply the antenna or
so? Regarding mismatch and a robust design, which design will behave
better, what do you think?
Also i would prefer to apply the first parallel LC circuit after the
transformer first, because the impedance is higher there and it is
easier to modify from the current configuration.
Your advice is apprechiated :-)
73, Stefan/DK7FC
Jay
----- Original Message ----- From: "Stefan Schäfer"
<[email protected]>
To: <[email protected]>
Sent: Thursday, May 31, 2012 9:14 AM
Subject: Re: LF: Re: Current mode class D PA for 475 kHz
Hi Jay, LF,
Tnx! I already found your paper in a forum :-)
Really, your work looks perfect. My compliments! Actually it is no
amateur radio any more but looks more than proffessional. Amazing!
I'm still going through that IEEE paper. It is worth to mention that
the
explanation of the working principle is excellent! Most papers i found
are very complex and related to very special UHF stuff.
Will try the CMCD today!
73, Stefan
Am 31.05.2012 15:01, schrieb [email protected]:
Stefan
CMCD working well up through 40 meters here on AM ...
http://www.w1vd.com/375wattclassD.html .
Jay W1VD WD2XNS WE2XGR/2
----- Original Message ----- From: "Stefan Schäfer"
<[email protected]>
To: <[email protected]>
Sent: Thursday, May 31, 2012 7:12 AM
Subject: LF: Current mode class D PA for 475 kHz
Am 31.05.2012 12:46, schrieb Stefan Schäfer:
[...] Looks all very interesting. I will
take the idea of the "current mode class D PA" and play arround with
it...
This paper looks most interesting:
http://sigma.ucsd.edu/research/articles/2001/2001_2.pdf
Maybe just the solution for my problem in my current MF PA
construction.
73, Stefan
|
|
