[600MRG] Power Supply Discussion

Mon Feb 27 13:44:04 CST 2017

Referencing my previous email, it seems the email system truncated the
hyperlinking for two links, rendering them null and void.  You may have to
wrap the linking on the following two referenced links:

<goog_1809950462>

http://www.gw3uep.ukfsn.org/100W_QTX/100WTX/PA_pa.gif

http://njdtechnologies.net/wp-content/uploads/2015/01/capture-112016-e1479690533994.jpg

73,

jl

On Mon, Feb 27, 2017 at 1:24 PM, John Langridge <kb5njd at gmail.com> wrote:

> Time out - lets play the semantics game for just a moment - then I will
> comment on a few of Frank's questions from my personal experience.
>
> Frank said: "... of my KB5NJD / WG2XIQ designed amplifier."
>
> To be clear and pick nits, that amplifier is not my design.  It was
> developed by GW3UEP and I become so enamored with the simplicity vs.
> performance that I built a few for operators at one time when I had time.
> Now I am happy to give away all of the info and improvements that I have
> learned about through the use so that anyone who can follow directions
> carefully can replicate the very good, simple  design.  All of the design
> work was GW3UEP!  Please don't attribute that to me.
>
> Now all that said, to address a couple of Franks questions from my
> experience:
>
> Frank asked :"How would you sufficiently suppress the key clicks for such
> a power supply approach if there might be a power supply droop of 20
> percent or more? "
>
> if you are seeing that much droop, there may be a major supply problem but
> in many cases, keep adding capacitance and that will help buffer the beast.
>
> Frank said:  "Does power supply droop instigate spurious oscillations?"
>
>  A sagging power supply definitely does create spurious signals.  When I
> started on 472 back in 2012 I had a computer PS that lost a lot of its
> capacity and it was a mess on the air...You could watch it AM on a scope
> even on steady carriers...  I probably could have propped it up with a lot
> of capacitance but I chose to replace the PS and problem was solved.
>
> Referencing the key click question and assuming that your PS is NOT
> sagging, the only solution I have found to quieten down clicks in that
> amplifier design is the one specified by GW3UEP where he controls the drain
> rise and fall time shown in this schematic:
> http://www.gw3uep.ukfsn.org/100W_QTX/100WTX/PA_pa.gif
>
> If you are using the above approach to keying that amplifier and have not
> changed the ratio of Tr/Tr (5ms / 20ms) you should be in good shape.  If
> the PS is sagging, it doesn't matter what you do, you are going to see
> spurious as I reported two paragraphs up.
>
> The performance using good power supplies with and without external keying
> can be seen in the below links.  You can figure out which CW signal is
> using waveform shaping and which isn't:
>
> http://njdtechnologies.net/wp-content/uploads/2015/06/XKA-at-474R5.jpeg
>
> and
>
> http://njdtechnologies.net/wp-content/uploads/2015/01/captur
> e-112016-e1479690533994.jpg
>
> Keep in mind that in both captures, the poor SDR receiving my signal is in
> the near field and getting beat to death but you can at least see the
> relative difference.  If you are using a cruddy power supply both will be
> much worse.
>
> There is more here that Frank asked that I am not qualified to answer but
> I felt I needed to address the design attribution statement, any perceived
> click situation on that amp design and my experiences with sagging supplies
> and how it might impact all of that.
>
> 73,
>
> John KB5NJD / WG2XIQ..
>
>
>
> On Mon, Feb 27, 2017 at 12:01 PM, Frank Lotito <k3dz at live.com> wrote:
>
>> A recent post regarding power supply designs for nonlinear MF and LF
>> power amplifiers suggested a brute force unregulated design.  Most
>> definitely, I agree, for continuous carrier type emissions.  What about key
>> clicks when the carrier has to be keyed, such as regular old fashion CW?
>> Or for that matter QRSS keying, if you would, the start / stop of a WSPR or
>> JT9 transmission, the start / stop of a dit or dah in QRSS transmission?  I
>> submit that if the brute force supply's internal resistance is not
>> "very-very low," there will be a substantial click on make / break.  How
>> would you sufficiently suppress the key clicks for such a power supply
>> approach if there might be a power supply droop of 20 percent or more?  How
>> do you get a very-very low power supply droop?  It might be a little more
>> involved than using "large bore" components for modest power delivery.
>>
>>
>> I think the key question is "Does power supply droop instigate spurious
>> oscillations?"  If yes, can the power supply droop be combatted with driver
>> keying?  As a means to minimize keying clicks my gut feeling says driver
>> keying may not do it.  As I understand, if the driver keying spends to much
>> time making the transition from hard turn on to hard turn off (and visa
>> versa) there is an increased risk of spurious oscillations while the
>> amplifier is making this transition.
>>
>>
>> I use a regulated DC power supply, but key the Vcc line going to the my
>> non-linear RF power amplifier using a series P-channel power mosfet.  It
>> took a while for me to work out a means to key the mosfet's gate to get an
>> acceptable rise / fall time of my KB5NJD / WG2XIQ designed amplifier.   I
>> am still not 100% sure that in the on / off keying transition I am not
>> generating spurious oscillations.
>>
>>
>> Suggestions?  Comments?
>>
>>
>> 73 Frank K3DZ / WH2XHA
>> _______________________________________________
>> 600MRG mailing list
>> 600MRG at w7ekb.com
>> http://w7ekb.com/mailman/listinfo/600mrg_w7ekb.com
>>
>
>