Elecraft

KPA-1500 ATU not working well

Classic

List

Threaded

27 messages Options

K9MA

Re: KPA-1500 ATU not working well

My observation , also.

Another approach would be for automatic drive adjustment, or ALC that worked. The latter, I know, is very difficult to do with an amp.

Scott K9MA

----------

Scott Ellington

--- via iPad

> On Jun 22, 2018, at 2:36 PM, Paul Baldock <[hidden email]> wrote:
>
> I have found that if the KPA1500 is feeding a largish SWR (like 1.5:1 as you suggest) then the drive power required to maintain a constant output will vary significantly as you tune across a band. This means you have to keep adjusting the power control within a single band. This appears not to occur if the SWR is 1.1:1 or better.

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]

Bill K9YEQ

Re: KPA-1500 ATU not working well

In reply to this post by Wes Stewart-2

Wes, so well put. The expectations expressed don't include a whole new set of variables when dealing with full power. The good thing: this list may help those challenged to get help to figure out their issues. I can't wait to get mine on the air!

72 & 73,
Bill
K9YEQ
FT’er for K2, KX1, KX3, KXPA100, KAT500, W2, etc.

-----Original Message-----
From: [hidden email] <[hidden email]> On Behalf Of Wes Stewart
Sent: Friday, June 22, 2018 3:45 PM
To: Paul Baldock <[hidden email]>; [hidden email]
Subject: Re: [Elecraft] KPA-1500 ATU not working well

OK, note my qualifier, "If". If there are other detrimental issues with greater load mismatch then they might need to be addressed. Unfortunately, Elecraft doesn't specify a load requirement for rated output. Gain ripple that you observe is possibly a reflection (no pun intended) of the LPFs rather than the intrinsic performance of the transistors, although as I have reported elsewhere K3S IMD is frequency sensitive, but not overly so within a band.

One point I would like to make is that guys who wring their hands because the SWR meters on the K3, KAT500 and KPAs all read differently should lighten up. There are reasons why this can be without there being product defects.

Wes N7WS

.On 6/22/2018 12:36 PM, Paul Baldock wrote:

> I have found that if the KPA1500 is feeding a largish SWR (like 1.5:1
> as you
> suggest) then the drive power required to maintain a constant output
> will vary significantly as you tune across a band. This means you have
> to keep adjusting the power control within a single band. This appears
> not to occur if the SWR is 1.1:1 or better.
>
> I don't know, but I would guess a solid state untuned amplifier
> efficiency is better into a 1:1 SWR than a 1:5:1. This could lead to
> the dreaded more fan noise.
>
> By the way by suggesting 1.0:1. I did not mean exactly
> 1.0000000000000000000:1. I tired to infer 1 digit resolution/accuracy.
>
> - Paul KW7Y
>
>
> At 11:11 AM 6/22/2018, Wes Stewart wrote:
>> I do not, and probably never will, have a KPA1500.Â I do have a
>> KPA500 and
>> KAT500 so I think I can comment.Â In my opinion, the function of
>> these tuners is to provide a match into which the amp can deliver
>> full power.Â If
>> 1.5:1 does this, then that's good enough in my book.Â Furthermore, a
>> better match might entail higher tuner losses.Â So why worry about
>> it? Another thing; a 1.0:1 SWR infers infinite return loss.Â To
>> measure infinite return loss, you need a directional coupler with
>> infinite directivity or some mathematical correction derived from
>> very well known calibration standards and no other errors,Â So maybe
>> you say, 1.0:1 can't be measured (you would be correct), let's shoot
>> for 1.22:1.Â Â That's a return loss of ~20 dB.Â Now I have no idea
>> of what the directivity of the coupler is in the KAT500 or
>> KPA1500 but considering it has to work from 1.8 to 54 MHz and to keep
>> the numbers easy, 20 dB wouldn't be unreasonable. (Note: I'm only
>> considering the directivity error, there are several others, internal
>> mismatches, frequency tracking, detector non-linearities, etc.) If we
>> measure a load with 20 dB RL using a directional coupler with 20 dB
>> directivity the answer can be anywhere between infinity to 14 dB.Â
>> (Full cancellation of the two reflection coefficients to the sum of
>> the reflection coefficients)Â In SWR terms, a
>> 1.22:1 load can measure anywhere between 1.0:1 and 1.5:1.Â A
>> "perfect" load will measure 1.22:1. Folks, these things aren't
>> laboratory instruments and until Elecraft builds in vector network
>> analyzers with full error correction this is what we get. WesÂ N7WS
>> On 6/21/2018 11:17 PM, Paul Baldock wrote: > First let me say that
>> other than some issues with the ATU, I am after a week
>> > of use, very happy with the KPA1500. > > So here's the ATU issues:
>> > > > I
>> find that the ATU very rarely will learn a setting to get the SWR
>> below >
>> 1.2:1 on any band. I can usually tweak it down to 1.0:1 using the
>> Utility > Program.Â And yes, before anybody asks, I have ATU STOP
>> TUNE and the ATU > BYPASS set at 1.0:1. > > On 6M into a perfect load
>> (laboratory quality to
>> 8GHz) with the ATU bypassed, > the internal SWR meter reads 1.4:1Â I
>> would have expected it to be 1.0:1, but > Elecraft tell me it is
>> normal. So the ATU on 6M will have to tune out what is > left when
>> the 1.4:1 internal mismatch sees your external antenna mismatch. >
>> Maybe this is contributing to your problem. Try adding a quarter wave
>> of coax, > maybe it will achieve a better match. Fortunately my 6M
>> antenna is flat enough > that I can work CW, SSB and
>> FT-8 in BYPASS. > > For your interest, on 10M in to perfect load with
>> ATU bypassed the internal > SWR meter reads 1.2:1. Not 1.0:1. > >
>> Another issue I have with the ATU is that the internal frequency
>> counter has > 8KHz resolution which according to one of the Elecraft
>> techs can lead to a > 16KHz error in measurement. It appears that
>> during "learning" that the ATU is > based on the internal frequency
>> counter measurement, not the frequency you > transceiver sends to the
>> KPA1500.This means that on the bands that have 10KHz
>> > or 20KHz segments, you cannot be sure which segment you are using,
>> > and when you use the tuner it could be in a different segment.
>> > Being off by a
>> segment > could be a problem with a high Q antenna. > > Apparently
>> the counter originally had 1KHz resolution but was changed to fix >
>> some other problem. An Elecraft tech told me they currently have no
>> plan to > change it back to 1KHz. I am not suggesting that the tuner
>> should change > segments exactly on their edge. Clearly there has to
>> be some hysteresis to > stop hunting. 25% of a segment width might
>> width be a good number to choose, > but this would require the current frequency counter to have improved resolution.
>> > > > - PaulÂ KW7Y
>> ______________________________________________________________
>> Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help:
>> http://mailman.qth.net/mmfaq.htm Post:
>> mailto:[hidden email] This list hosted by: http://www.qsl.net Please help support this email list:
>> http://www.qsl.net/donate.html Message delivered to
>> [hidden email]
>
>

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html Message delivered to [hidden email]
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]

John Oppenheimer

Re: KPA-1500 ATU not working well

In reply to this post by Wes Stewart-2

Here is a picture and some plot images:

http://www.kn5l.net/Elecraft/CP1/

Not super accurate 60 and 100 Ohm loads, but provides an insight of the
coupler operation.

John KN5L

On 06/22/2018 04:15 PM, Wes Stewart wrote:

> I don't quite understand your methodology but there is one, apparently little
> known caveat with the use of the "Elecraft" coupler topology. AFAIK, this was
> first used by John Grebenkemper, KI6WX, in his Tandem Match, described in QST,
> January 1987.
>
> Paul Kiciak, N2PK, mentions in his paper, "An HF In-Line Return Loss And Power
> Meter", that this coupler suffers from poor input match at lower frequencies.
> So the SWR meter has high SWR!
>
> If the CP1 wasn't so expensive, I'd buy one just to measure it.

Wes Stewart-2

Re: KPA-1500 ATU not working well

In reply to this post by K9MA

Offhand, since they are adamantly opposed to using operational ALC, I could see
incorporating a "1500 W" gain calibration, as they do with K3s. They could read
the output power at several frequencies per band and save the gain settings.
This assumes they have enough memory in the K3 to save the values.

Of course fixing the root cause would be better.

Wes N7WS

On 6/22/2018 2:31 PM, K9MA wrote:

> My observation , also.
>
> Another approach would be for automatic drive adjustment, or ALC that worked. The latter, I know, is very difficult to do with an amp.
>
> Scott K9MA
>
> ----------
>
> Scott Ellington
>
> --- via iPad
>
>> On Jun 22, 2018, at 2:36 PM, Paul Baldock <[hidden email]> wrote:
>>
>> I have found that if the KPA1500 is feeding a largish SWR (like 1.5:1 as you suggest) then the drive power required to maintain a constant output will vary significantly as you tune across a band. This means you have to keep adjusting the power control within a single band. This appears not to occur if the SWR is 1.1:1 or better.
>

Elecraft mailing list

Re: KPA-1500 ATU not working well

In reply to this post by John Oppenheimer

Hi John,

Well if a picture is worth a thousand words a graph is worth at least
10,000.

THANKS for the posting!
73,
Bob
K2TK ex KN2TKR (1956) & K2TKR

On 6/22/2018 6:27 PM, John Oppenheimer wrote:
> Here is a picture and some plot images:
>
> http://www.kn5l.net/Elecraft/CP1/
>
> Not super accurate 60 and 100 Ohm loads, but provides an insight of the
> coupler operation.
>
> John KN5L

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]

Bill K9YEQ

Re: KPA-1500 ATU not working well

In reply to this post by John Oppenheimer

Can we change subject matter. Min is on the way and this is way to freakey!!!!!!!

73,
Bill
K9YEQ

https://wrj-tech.com/

-----Original Message-----
From: [hidden email] <[hidden email]> On Behalf Of John Oppenheimer
Sent: Friday, June 22, 2018 5:27 PM
To: [hidden email]
Subject: Re: [Elecraft] KPA-1500 ATU not working well

Here is a picture and some plot images:

http://www.kn5l.net/Elecraft/CP1/

Not super accurate 60 and 100 Ohm loads, but provides an insight of the coupler operation.

John KN5L

On 06/22/2018 04:15 PM, Wes Stewart wrote:

> I don't quite understand your methodology but there is one, apparently
> little known caveat with the use of the "Elecraft" coupler topology.
> AFAIK, this was first used by John Grebenkemper, KI6WX, in his Tandem
> Match, described in QST, January 1987.
>
> Paul Kiciak, N2PK, mentions in his paper, "An HF In-Line Return Loss
> And Power Meter", that this coupler suffers from poor input match at lower frequencies.
> So the SWR meter has high SWR!
>
> If the CP1 wasn't so expensive, I'd buy one just to measure it.

David Woolley (E.L)

Re: KPA-1500 ATU not working well

In reply to this post by Paul Baldock

I'm not clear whether you are talking about the SWR before or after the
tuner, but efficiency is determined by the load seen by the PA, which
has two degrees of freedom (reactance and resistance), both of which can
vary either side of the ideal. SWR reduces this to a single variable,
that doesn't even respect the sign of the deviation from the ideal load.
That means that efficiency may even be better than the nominal
efficiency at a particular SWR, whereas that same SWR can also produce a
lowered efficiency. (Similarly, an SWR may destroy a PA or be safe,
depending on the fine details.)

In particular, a higher load resistance may result in a higher
efficiency, at the expense of a lower maximum output. There is actually
a K2 build option, that winds the output transformer to impose a higher
load resistance, in order give better efficiencies at low powers.

Tuning across the band will change at least the reactance component of
the load on the PA.

Also, if input SWR isn't 1:1, I think you will find that the measured
forward power doesn't reflect what is actually available to go out
through the antenna. Whilst some of it will circulate back and forth
and eventually make its way out in the right direction, some will also
get absorbed by the PA.

On 22/06/18 20:36, Paul Baldock wrote:
> I have found that if the KPA1500 is feeding a largish SWR (like 1.5:1
> as you suggest) then the drive power required to maintain a constant
> output will vary significantly as you tune across a band. This means you
> have to keep adjusting the power control within a single band. This
> appears not to occur if the SWR is 1.1:1 or better.
>
> I don't know, but I would guess a solid state untuned amplifier
> efficiency is better into a 1:1 SWR than a 1:5:1. This could lead to
> the dreaded more fan noise.

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]