Message-ID: <4D2F736C.6010001@iup.uni-heidelberg.de>
Date: Thu, 13 Jan 2011 22:49:32 +0100
From: =?ISO-8859-1?Q?Stefan_Sch=E4fer?= <schaefer@iup.uni-heidelberg.de>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.1; de; rv:1.9.1.11) Gecko/20100711 Thunderbird/3.0.6
MIME-Version: 1.0
To: rsgb_lf_group@blacksheep.org
References: <003801cbb33e$8b59e620$0401a8c0@xphd97xgq27nyf> <4929CA3383DD45FCBC44A80A2733B139@PcMinto> <004801cbb34a$5b8ab580$0401a8c0@xphd97xgq27nyf> <E65C1779B3B94AAB83235F0C4D0DF677@PcMinto>
In-Reply-To: <E65C1779B3B94AAB83235F0C4D0DF677@PcMinto>
Subject: Re: LF: FET RDS
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org


Minto,

14 nF gate capacity can easily be driven by a IXDD414, at 500 kHz and 
even at 1800 kHz. I've built a 600W class E for 160m running at 36 VDC 
(for /p, 3 car batteries in series ;-) ), using 4xIRFP260N. Efficiency 
is above 90% but mostly due to losses in the coils ( i like to built it 
compact :-) ).

73, Stefan/DK7FC

Am 13.01.2011 19:08, schrieb Minto Witteveen:
> Mal,
>
> Yes, putting several in parallel means that the total capacitance that 
> the driver stage sees (sum of gate capacitances and the reverse 
> transfer capacitances) increases so more power is needed to 
> hard-switch the fets. I use a TIP31/32 totem-pole to accomplish this.
> And the capacitance becomes quit big, Ciss = 4200 pF,  Crss= 135 pF 
> (translates roughly to approx 135 * 20 = 2700 pF) so total capacitance 
> the driver sees is somewhere near 14 nF!!
> Cx @ 500 KHz = 22.7 Ohms.
> The IRFP360 is even worse.
> Now image 4 of these in parallel.
>
>
> Regards,
> Minto pa3bca.
>
>
>
> --------------------------------------------------
> From: "mal hamilton" <g3kevmal@talktalk.net>
> Sent: Thursday, January 13, 2011 18:50
> To: <rsgb_lf_group@blacksheep.org>
> Subject: Re: LF: FET RDS
>
>> Minto
>> One approach is to use a few in parallel like they do in plasma tv's but
>> there must then be other considerations to hinder the application.
>> mal/g3kev
>>
>> ----- Original Message -----
>> From: "Minto Witteveen" <minto.witteveen@gmail.com>
>> To: <rsgb_lf_group@blacksheep.org>
>> Sent: Thursday, January 13, 2011 5:42 PM
>> Subject: Re: LF: FET RDS
>>
>>
>>> Yup that is correct. High(er) voltage fets usually have the higher 
>>> RDS-on
>>> values... Tradeoff based on physics...
>>> I started my 500-600 Watts 500 KHz transverter with two IRFP360's in
>>> parallel.
>>> Later I switched to IXFH26N50 (cheaper at EUR 1 a piece and slightly
>> better
>>> than the IRFP360).
>>> The IXFH26N50 has a VDSS of 500 Volts, and a RDS-on of 0.23 Ohms and 
>>> an Id
>>> of 25A.
>>> With two of these in parallel the efficiency is > 90%. DC supply is 
>>> (max)
>> 54
>>> Volts.
>>> Peak voltage on the drains is somewhere around max 250 Volts. So I 
>>> might
>>> search for Fets with a somewhat lower RDSon and a lower max voltage, 
>>> but
>>> these fets are indestructible in my setup, they survive open and
>>> short-circuited antennas without a problem for several minutes until 
>>> heat
>>> becomes a problem.
>>>
>>>
>>> For more info wrt my setup see www.pa3bca.nl
>>>
>>> Regards,
>>> Minto pa3bca (500 KHz in PA idle at the moment, alas...)
>>>
>>>
>>>
>>>
>>> From: mal hamilton
>>> Sent: Thursday, January 13, 2011 17:25
>>> To: rsgb
>>> Subject: LF: FET RDS
>>>
>>>
>>> LF/MF
>>> It seems to me if you are working with low V high current FETS the RDS
>> seems
>>> reasonable 0.02 for example but when a High V low current device is 
>>> need
>> the
>>> RDS of these devices seem to be around 0.4 considerably higher.
>>> therefore the efficiency of the amplifier will never reach the 90% plus
>> that
>>> some claim.
>>> I stripped a plasma tv recently and found banks of FETS (6 per bank) 
>>> and
>>> wondered why the application neederd so many and have come to the
>> conclusion
>>> that because of the high RDS lots were required in parallel to 
>>> reduce the
>>> losses.
>>> Maybe there are low RDS fets about that will handle several hundred 
>>> volts
>> at
>>> modest currents ie 10A at 1000 volts
>>> Room for thought
>>>
>>> de mal/g3kev
>>>
>>>
>>>
>>
>>