[Elecraft] LDMOS for QRO

Sun Dec 14 10:12:25 EST 2014

> Your issue with heat sinks and copper spreaders is already solved
> and one can run simple fans instead of blowers. And combining two
> devices is more complicated than combining 8 or 16? Huh?

I'm not saying it isn't possible - nor that is hasn't been solved.
Only that is it not less expensive than an 8877 and that those who
say a single 1.25 KW LDMOS is a suitable replacement for an 8877
are not considering other performance issues like IMD, etc.

The multiple module with combiner configurations have been around for
a long time.  The LDMOS modules were designed for television broadcast
service in addition to ISM purposes.  The broadcaster transmitters use
a large number of combined modules to reach quite high power levels
but each module is running less than 1 KW and the transmitters have
significant amounts of adaptive precorrection to maintain linearity.

If you want to step back and design precorrection capability into
the transceiver DSP (e.g., an input for a directional coupler/RF
tap at the output of the *system*) and can maintain linearity of
multiple modules "close enough" to use a single composite correction,
be my guest.  However, the cost and complexity are an order of
magnitude greater than a good cathode driven triode - but then if
you add phase modulation to the precorrection, perhaps you can get
away saturated (pulse) amplifiers.

73,

    ... Joe, W4TV

On 2014-12-14 4:10 AM, Edward R Cole wrote:
> Joe,
>
> You are correct that the specs are for pulse and one can run CW/JT65 at
> this level since there is no IMD produced by a single sine wave.  So
> combining two running 750w each gets you 1500w in linear operation with
> 2400w dissipation.
>
> 1500w mobile!!!  Are you serious?  I'd guess 200-250w would be adequate
> for mobile.  But I do not do serious HF'ing.
>
> Your issue with heat sinks and copper spreaders is already solved and
> one can run simple fans instead of blowers.  And combining two devices
> is more complicated than combining 8 or 16?  Huh?  If you think this is
> all theoretical you are mistaken as these are out there as kits and
> assembled amps on VHF to 1296.  Several companies have them for sale (M2
> is one).  The kits come with the LDMOS already installed on a copper
> spreader designed for the needed heat conduction.  They haven't hit the
> HF market as yet but no reason why not (be interesting to re-examine
> this topic in dec. 2015).
>
> I sure would consider a 50v PS a lot simpler and safer than a 4kV PS.
> In fact I have one that I bought on e-bay for $31 made by HP (it will
> run my 1100w 6m PA).  I also have a HB 4kV - 1.5A PS for my 8877...cost
> me a lot more than $31.
>
> Once I use up my current 8877, I will likely sell my amp with the final
> 8877 "pull" that I have in hand and replace it with a LDMOS amp since I
> can install it next to the base off my tower instead of having it inside
> with all the noise and HV.  Then I can haul the 1-5/8 inch Hardline to
> the metal recycle'r as I will only need to run RG213 to the amp with
> 4w.  Yes, 240vac will be needed but running that is not rocket science
> as every home well pump is wired with buried 240vac.
>
> 73, Ed - KL7UW
>
> From: "Joe Subich, W4TV" <lists at subich.com>
> To: elecraft at mailman.qth.net
> Subject: Re: [Elecraft] (no subject)
> Message-ID: <548CEE2E.7050400 at subich.com>
> Content-Type: text/plain; charset=utf-8; format=flowed
>
>
>   > But no longer do you need to run a dozen "pills" (man I hate that
>   > term).  Here is all you need for 1250w 1.8-600 MHz:
>
> Unfortunately, the LDMOS device data sheets provide *no* IMD spec's
> for linear operation.  The only data is for CW and pulse service.  If
> one is to extrapolate from similar LDMOS devices (single FET vs. two
> FETs on a common die), one would need to derate to 800-900 W PEP in
> order to achieve reasonable IMD levels.  That conclusion is further
> corroborated by the compression spec's (actual vs. ideal output power)
> which show the onset of output compression above 59 to 60 dBm (59 dBm
> => ~800W).
>
> The devices work at 1200 W CW (or JT65) because those modes are single
> tone and work with saturated operation (class C amplifiers) where IMD
> performance is not "tested".
>
> At full output these devices would be as dirty in SSB operation as the
> old FM "brick" amplifiers were when run in SSB service - perhaps like
> the RMA Italia solid state amps <G>.
>
> BTW, since these are 50V parts they are not suited for mobile use at
> 12V (13.8 V nominal) service.
>
>   > $241.50 compared with the new price of a 8877 is pretty cheap!
>
> To even approach the IMD performance and reliability of an 8877 you
> would need two devices and the cooling problems (cost of heat sinks,
> heat spreaders, etc.) are much more difficult with two of these devices
> than with a single 8877.  Any cost advantage for even two of the LDMOS
> devices over an 8877 will be more than offset by cooling system (in
> addition to splitter/combiner and protection system) costs.
>
> 73,
>
>     ... Joe, W4TV
>
>
> 73, Ed - KL7UW
> http://www.kl7uw.com
>      "Kits made by KL7UW"
> Dubus Mag business:
>      dubususa at gmail.com
>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:Elecraft at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
> Message delivered to lists at subich.com
>