Envelope-to: dave@picks.force9.co.uk
Delivery-date: Wed, 18 May 2005 11:57:43 +0100
Message-ID: <428B1F97.7040007@tele2.se>
Date: Wed, 18 May 2005 10:57:27 +0000
From: Johan Bodin <jbodin@tele2.se>
User-Agent: Mozilla Thunderbird 1.0.2 (Windows/20050317)
To: rsgb_lf_group@blacksheep.org
References: <000c01c55b7c$8c384f20$14d0fc3e@l8p8y6>
In-Reply-To: <000c01c55b7c$8c384f20$14d0fc3e@l8p8y6>
Mime-Version: 1.0
Subject: Re: LF: CLASS D/E
Content-Transfer-Encoding: 8bit
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org

Mal wrote:

 > Hi All
 > Where  would I find a good book about class D and E amp design including
 > the mathematics?

Mal,

regarding class E, the best paper I have ever seen is
"Class-E RF Power Amplifiers" by Nathan O. Sokal, WA1HQC.
The author is one of the inventors of class E.
It was published in Jan/Feb QEX 2001. You will find it
in .pdf format here:

   http://www.classeradio.com/sokal2corrected.pdf

Another good source of information on class D/E/F switching amplifiers,
with some math, can be found in the chapter "High Efficiency Power
Amplifiers" in the book "Solid State Radio Engineering" by F. Raab et al.


Class E has lower "transistor utilization factor" than class D so
any given transistor will have less maximum output power in class E.
However, class E has some great advantages. For example, The nasty
(go-bang!) voltage spikes often seen in class D amplifiers does not
exist in a properly designed class E amplifier!
Another advantage is that the drive requirements are very relaxed
compared to class D, especially current switching class D, because
the majority of the current flows in the middle of the "ON period"
of the class E transistor, very little near the on/off edges.
Some HF designs even use sinusoidal drive with good results!
In my opinion, Class E is smooth, simple, elegant and (literally) cool.

There is a considerable amount of even harmonics in the output
of a single ended class E stage, especially if the series resonant
tank has relatively low Q. Instead of building a super-duper
lowpass filter, you may wish to build a push-pull class E amplifier,
where the even harmonics are cancelled by symmetry, and get away
with a less elaborate lowpass filter. It is simple. Design a
class E stage for half the desired output. Then build two such
stages, each with separate DC feeding chokes and separate
drain-source parallel capacitors. Drive the two stages with a
180-degree (push-pull) signal (50% duty cycle square wave) and let
both stages share a common series tank and output transformer.
Tank L = twice the L from the single stage design and tank
C = design C / 2. Transformer primary input impedance =
2 * single stage design load impedance.

The DC feeding choke(s) has to be quite large in order not to interfere
with the (non-sinusoidal) waveform at the transistor drain/collector.
In the "schoolbook example", the choke is there only to feed a constant
DC current into the stage and the choke simply doesn't exist at RF
frequencies... Values as high as XL = 30*XC of the drain/source capacitor
has been mentioned but I think it is possible to get away with a somewhat
smaller value.

A nice program for class E design is available at
http://tonnesoftware.com/
(Look for ClassE)

A handful of IRFP450 at 80V supply voltage (or so) will
deliver a respectable power (even in class E :-)...


73
Johan SM6LKM


-- 
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.308 / Virus Database: 266.11.12 - Release Date: 2005-05-17