Elecraft

KPA-1500 issues

Classic

List

Threaded

28 messages Options

Elecraft mailing list

KPA-1500 issues

I've experienced some of these same issues with my KPA-1500. In addition, sometimes the amp will switch to stand-by for no apparent reason, on 6 meters the AT won't tune a 1.7 to 1 SWR under 1.5 to 1, and one of the exhaust fans got a good deal noisier after about 25 hours of operation. My amp is back at Elecraft for service as I write.

73,
Dick- K9OM

On Mon, Nov 26, 2018 at 5:15 PM K9MA <[hidden email]> wrote:

> This was my first serious contest effort with the new amplifier. It is
> paired with a K3. It survived about 24 hours of operation, though it
> faulted a few times, as described below. Clearly, the protection
> circuitry did its job. Firmware is 1.87.
>
> A couple glitches:
>
> The power gain drifts downward as the transistors heat up. If I start
> out at 1500 W, it will drift down to about 1100, without changing
> frequency or SWR. Conversely, if I turn the drive up after the amp is
> warm, and forget to turn it back down, the amp will fault when cold.
> Since the amp is talking to the K3, couldn't some clever software fix this?
>
> Despite setting tune power to 40 W, the ATU sometimes just won't get the
> SWR below 1.4:1 or so. This is with antennas below 2:1, and the ATU stop
> set to 1.0. This happens on 40 meters, so it's probably not an ATU L/C
> resolution limit. This 1.4:1 SWR is sometimes high enough to limit
> output to well below 1500 W due to high drive power or high current. I
> have not yet tried manually tuning the ATU to see if I can get the SWR
> lower.
>
> Finally, after "training" the ATU across the band, I sometimes find that
> the SWR has gone up significantly at some frequencies. I don't believe
> this is due to changing antenna SWR. It seems more like the ATU has
> "forgotten" the settings, or tried to retune itself without permission.
> (HiSWR RETUNE OFF)
>
> Has anyone else seen similar issues?
>
> 73,
>
> Scott K9MA
>
> --
> Scott K9MA
>
> [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]

Paul Baldock

Re: KPA-1500 issues

There is definitely room for improvements in the
tuning algorithm. Because of this I have tuned
all my segments manually. I find if the SWR is
not adjusted to be close to 1:1 in all segments
of a band then the drive requirements vary across
that band, which would be very annoying.

In addition the tuner may not always be in the
segment you think it is. This is due to the 8KHz
resolution of the internal counter, that wins
over you sending it a more accurate frequency,
and also the hysteresis algorithm that can
require a 2 segment frequency change before the
tuner changes. Gary from Elecraft says "if you
are off by one bin segment then it should not matter too much"

I also note that when into a perfect 50 Ohms, and
the tuner bypassed, on 6 meters, my KPA1500 says
the SWR is 1.4:1. On 10 meters it reads 1.2:1.
Rene from Elecraft says "this is the best they can do".

Other than the issues above I am very happy with the KPA1500.

- Paul

At 09:59 AM 11/27/2018, Dick via Elecraft wrote:

>I've experienced some of these same issues with
>my KPA-1500.Â In addition, sometimes the amp
>will switch to stand-by for no apparent reason,
>on 6 meters the AT won't tune a 1.7 to 1 SWR
>under 1.5 to 1, and one of the exhaust fans got
>a good deal noisier after about 25 hours of
>operation.Â My amp is back at Elecraft for
>service as I write. Â 73, Dick- K9OMÂ Â On
>Mon, Nov 26, 2018 at 5:15 PM K9MA
><[hidden email]> wrote: > This was my first
>serious contest effort with the new amplifier.
>It is > paired with a K3. It survived about 24
>hours of operation, though it > faulted a few
>times, as described below. Clearly, the
>protection > circuitry did its job. Firmware is
>1.87. > > A couple glitches: > > The power gain
>drifts downward as the transistors heat up. If I
>start > out at 1500 W, it will drift down to
>about 1100, without changing > frequency or SWR.
>Conversely, if I turn the drive up after the amp
>is > warm, and forget to turn it back down, the
>amp will fault when cold. > Since the amp is
>talking to the K3, couldn't some clever software
>fix this? > > Despite setting tune power to 40
>W, the ATU sometimes just won't get the > SWR
>below 1.4:1 or so. This is with antennas below
>2:1, and the ATU stop > set to 1.0. This happens
>on 40 meters, so it's probably not an ATU L/C >
>resolution limit. This 1.4:1 SWR is sometimes
>high enough to limit > output to well below 1500
>W due to high drive power or high current. I >
>have not yet tried manually tuning the ATU to
>see if I can get the SWR > lower. > > Finally,
>after "training" the ATU across the band, I
>sometimes find that > the SWR has gone up
>significantly at some frequencies. I don't
>believe > this is due to changing antenna SWR.
>It seems more like the ATU has > "forgotten" the
>settings, or tried to retune itself without
>permission. > (HiSWR RETUNE OFF) > > Has anyone
>else seen similar issues? > > 73, > > Scott
>K9MA > > -- > Scott K9MA > > [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]

Bill K9YEQ

Re: KPA-1500 issues

If one thinks about it, no auto tuner can be expected to get an exact match unless it is adjusted manually (that would be a PITA). And if done manually, that doesn't mean that the match is perfect. It just appears that way. Consider what goes on with feedline, connections and the antenna itself. Grounding, RF and impedance issues, etc., all can contribute to inconsistent behavior. The Elecraft tuning is among the best for price for in rig performance IMO. I prefer remote auto tuning myself which requires a different operation method.

Now if Elecraft could design a remotely controlled, wireless preferred, tuner, I would buy one yesterday!

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

-----Original Message-----
From: [hidden email] <[hidden email]> On Behalf Of Paul Baldock
Sent: Tuesday, November 27, 2018 1:39 PM

There is definitely room for improvements in the tuning algorithm. Because of this I have tuned all my segments manually. I find if the SWR is not adjusted to be close to 1:1 in all segments of a band then the drive requirements vary across that band, which would be very annoying.

In addition the tuner may not always be in the segment you think it is. This is due to the 8KHz resolution of the internal counter, that wins over you sending it a more accurate frequency, and also the hysteresis algorithm that can require a 2 segment frequency change before the tuner changes. Gary from Elecraft says "if you are off by one bin segment then it should not matter too much"

I also note that when into a perfect 50 Ohms, and the tuner bypassed, on 6 meters, my KPA1500 says the SWR is 1.4:1. On 10 meters it reads 1.2:1.
Rene from Elecraft says "this is the best they can do".

Other than the issues above I am very happy with the KPA1500.

- Paul

______________________________________________________________
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]

Wes Stewart-2

Re: KPA-1500 issues

In reply to this post by Paul Baldock

A few months ago I wrote about the potential errors in measuring SWR with
amateur grade equipment. See my comments in this thread:

http://elecraft.365791.n2.nabble.com/K3-SWR-Numerical-Indication-td7643839.html

I imagine "the best they can do" is accurate.

Wes N7WS

ps. I also said that if I was wrong, I was sure I would hear about it. Crickets.

On 11/27/2018 12:39 PM, Paul Baldock wrote:
> [snip]
>
> I also note that when into a perfect 50 Ohms, and the tuner bypassed, on 6
> meters, my KPA1500 says the SWR is 1.4:1. On 10 meters it reads 1.2:1. Rene
> from Elecraft says "this is the best they can do".
>
> Other than the issues above I am very happy with the KPA1500.
>
> - Paul

______________________________________________________________
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]

Michael Walker

Re: KPA-1500 issues

If you truly want to understand SWR, I recommend you pick a frequency and
make up a 1/4, 1/2 and 1 wavelength sections of 50ohm coax.

Then measure the SWR at those points into various different resistive
loads, like 10, 25, 50, 75, 100, 150 and 200 ohms.

I think you will find the results very interesting.

Mike va3mw

On Tue, Nov 27, 2018 at 5:23 PM Wes Stewart <[hidden email]> wrote:

> A few months ago I wrote about the potential errors in measuring SWR with
> amateur grade equipment. See my comments in this thread:
>
>
> http://elecraft.365791.n2.nabble.com/K3-SWR-Numerical-Indication-td7643839.html
>
> I imagine "the best they can do" is accurate.
>
> Wes N7WS
>
> ps. I also said that if I was wrong, I was sure I would hear about it.
> Crickets.
>
> On 11/27/2018 12:39 PM, Paul Baldock wrote:
> > [snip]
> >
> > I also note that when into a perfect 50 Ohms, and the tuner bypassed, on
> 6
> > meters, my KPA1500 says the SWR is 1.4:1. On 10 meters it reads 1.2:1.
> Rene
> > from Elecraft says "this is the best they can do".
> >
> > Other than the issues above I am very happy with the KPA1500.
> >
> > - Paul
>
> ______________________________________________________________
> 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]

Don Wilhelm

Re: KPA-1500 issues

Mike,

I don't know if that really explains SWR, but with the different loads,
I "get" what you are alluding to.
One comment is that those feedline lengths should be electrical lengths
- they will be shorter than the physical length by the amount
contributed by the feedline velocity factor.

After (or before) that experiment, grab most any good book on antennas
and transmission lines, looking with particular interest at the sections
dealing with mismatched lines which are directly associated with SWR.

73,
Don W3FPR

On 11/27/2018 5:51 PM, Michael Walker wrote:
> If you truly want to understand SWR, I recommend you pick a frequency and
> make up a 1/4, 1/2 and 1 wavelength sections of 50ohm coax.
>
> Then measure the SWR at those points into various different resistive
> loads, like 10, 25, 50, 75, 100, 150 and 200 ohms.
>
> I think you will find the results very interesting.
______________________________________________________________
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]

K9MA

Re: KPA-1500 issues

In reply to this post by Bill K9YEQ

The ATU in the K3, which should work exactly like the one in the
KPA1500, seems to consistently find a very good match, at least on the
lower bands where the L/C resolution is very good. Why the KPA1500 often
fails to find such a good match is a mystery to me. The only difference
should be the power handling capacity of the components. This would
affect strays, but I wouldn't expect them to be significant on 20 meters
and below.

73,
Scott K9MA

On 11/27/2018 14:02, Bill Johnson wrote:

> If one thinks about it, no auto tuner can be expected to get an exact match unless it is adjusted manually (that would be a PITA). And if done manually, that doesn't mean that the match is perfect. It just appears that way. Consider what goes on with feedline, connections and the antenna itself. Grounding, RF and impedance issues, etc., all can contribute to inconsistent behavior. The Elecraft tuning is among the best for price for in rig performance IMO. I prefer remote auto tuning myself which requires a different operation method.
>
> Now if Elecraft could design a remotely controlled, wireless preferred, tuner, I would buy one yesterday!
>
> 72 & 73,
> Bill
> K9YEQ
> FT'er for K2, KX1, KX3, KXPA100, KAT500, W2, etc.
>
> -----Original Message-----
> From: [hidden email] <[hidden email]> On Behalf Of Paul Baldock
> Sent: Tuesday, November 27, 2018 1:39 PM
>
> There is definitely room for improvements in the tuning algorithm. Because of this I have tuned all my segments manually. I find if the SWR is not adjusted to be close to 1:1 in all segments of a band then the drive requirements vary across that band, which would be very annoying.
>
> In addition the tuner may not always be in the segment you think it is. This is due to the 8KHz resolution of the internal counter, that wins over you sending it a more accurate frequency, and also the hysteresis algorithm that can require a 2 segment frequency change before the tuner changes. Gary from Elecraft says "if you are off by one bin segment then it should not matter too much"
>
> I also note that when into a perfect 50 Ohms, and the tuner bypassed, on 6 meters, my KPA1500 says the SWR is 1.4:1. On 10 meters it reads 1.2:1.
> Rene from Elecraft says "this is the best they can do".
>
> Other than the issues above I am very happy with the KPA1500.
>
> - Paul
>
> ______________________________________________________________
> 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]

--
Scott K9MA

[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]

Roger D Johnson

Re: KPA-1500 issues

In reply to this post by Wes Stewart-2

Interesting! I just ran some tests on my station into Bird load; tuner bypassed:

At 50.1 mHz:

LP-700 1.06
K3 1.5
KPA1500 1.0

At 9.56 mHz (geometric mean freq)

LP-700 1.05
K3 1.1
KPA1500 1.0

At 1.825 mHz:

LP-700 1.05
K3 1.1
KPA1500 1.0

My KPA1500 looks good but K3 is off on 6m.

73, Roger

On 11/27/2018 5:22 PM, Wes Stewart wrote:

> A few months ago I wrote about the potential errors in measuring SWR with
> amateur grade equipment. See my comments in this thread:
>
> http://elecraft.365791.n2.nabble.com/K3-SWR-Numerical-Indication-td7643839.html
>
> I imagine "the best they can do" is accurate.
>
> Wes N7WS
>
> ps. I also said that if I was wrong, I was sure I would hear about it. Crickets.
>
> On 11/27/2018 12:39 PM, Paul Baldock wrote:
>> [snip]
>>
>> I also note that when into a perfect 50 Ohms, and the tuner bypassed, on 6
>> meters, my KPA1500 says the SWR is 1.4:1. On 10 meters it reads 1.2:1. Rene
>> from Elecraft says "this is the best they can do".
>>
>> Other than the issues above I am very happy with the KPA1500.
>>
>> - Paul
>
> ______________________________________________________________
> 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]

Wes Stewart-2

Re: KPA-1500 issues

In reply to this post by Michael Walker

I fail to see what that will prove. The only length that will do anything
exciting is the 1/4 wavelength line and even that doesn't affect the SWR. The
other two will just repeat (less loss) on the input side, what terminates the
load side.

Wes N7WS

On 11/27/2018 3:51 PM, Michael Walker wrote:

> If you truly want to understand SWR, I recommend you pick a frequency and
> make up a 1/4, 1/2 and 1 wavelength sections of 50ohm coax.
>
> Then measure the SWR at those points into various different resistive
> loads, like 10, 25, 50, 75, 100, 150 and 200 ohms.
>
> I think you will find the results very interesting.
>
> Mike va3mw
>

K9MA

Re: KPA-1500 issues

I think the point Mike is trying to make is that many SWR meters don't
actually measure SWR all that accurately as impedance changes. While the
length of transmission line doesn't affect SWR (other than the effect of
loss), the impedance at a given SWR does change with line length, and
that may cause the indicated SWR to change with some instruments. I
wonder if an antenna analyzer, which actually measures impedance, and
calculates SWR from that, would do any better. If it measures impedance
accurately, it should.

73,
Scott K9MA

On 11/27/2018 21:34, Wes Stewart wrote:

> I fail to see what that will prove. The only length that will do
> anything exciting is the 1/4 wavelength line and even that doesn't
> affect the SWR. The other two will just repeat (less loss) on the
> input side, what terminates the load side.
>
> Wes N7WS
>
> On 11/27/2018 3:51 PM, Michael Walker wrote:
>> If you truly want to understand SWR, I recommend you pick a frequency
>> and
>> make up a 1/4, 1/2 and 1 wavelength sections of 50ohm coax.
>>
>> Then measure the SWR at those points into various different resistive
>> loads, like 10, 25, 50, 75, 100, 150 and 200 ohms.
>>
>> I think you will find the results very interesting.
>>
>> Mike va3mw
>>
>
> ______________________________________________________________
> 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]

Wes Stewart-2

Re: KPA-1500 issues

In the link I provided earlier (including some references) I point out some of
the pitfalls and error sources so that is a possibility. But I still contend
that, with a lossless line anyway, the impedance of a load, resistive or
otherwise, is the same at either end of a 1/2 wavelength line, or integer
multiple thereof. A quarter-wavelength line has the "magical" property of
inverting impedances, best seen on a Smith chart. But the rotation is around a
circle of constant SWR, with radius=rho.

Of course better instruments as a rule make better (more accurate)
measurements. I have several vector analyzers, an N2PK that I built years ago,
an SDR-Kits by DG8SAQ, an FA-VA5 by DG5MK, an AA-55 Zoom by Rig Expert and if it
ever clears US Customs, an FAA-450 by EU1KY. Any of these, used with
appropriate care, is more than accurate enough for any normal transmission
line/antenna evaluation and remarkably inexpensive.

Wes N7WS

On 11/27/2018 9:33 PM, K9MA wrote:

> I think the point Mike is trying to make is that many SWR meters don't
> actually measure SWR all that accurately as impedance changes. While the
> length of transmission line doesn't affect SWR (other than the effect of
> loss), the impedance at a given SWR does change with line length, and that may
> cause the indicated SWR to change with some instruments. I wonder if an
> antenna analyzer, which actually measures impedance, and calculates SWR from
> that, would do any better. If it measures impedance accurately, it should.
>
>
> 73,
> Scott K9MA
>
>
> On 11/27/2018 21:34, Wes Stewart wrote:
>> I fail to see what that will prove. The only length that will do anything
>> exciting is the 1/4 wavelength line and even that doesn't affect the SWR.
>> The other two will just repeat (less loss) on the input side, what terminates
>> the load side.
>>
>> Wes N7WS
>>
>> On 11/27/2018 3:51 PM, Michael Walker wrote:
>>> If you truly want to understand SWR, I recommend you pick a frequency and
>>> make up a 1/4, 1/2 and 1 wavelength sections of 50ohm coax.
>>>
>>> Then measure the SWR at those points into various different resistive
>>> loads, like 10, 25, 50, 75, 100, 150 and 200 ohms.
>>>
>>> I think you will find the results very interesting.
>>>
>>> Mike va3mw
>>>
>>
>> ______________________________________________________________
>> 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]
>
>

ANDY DURBIN

KPA-1500 issues

In reply to this post by Elecraft mailing list

"In addition the tuner may not always be in the segment you think it is. This is due to the 8KHz resolution of the internal counter, that wins over you sending it a more accurate frequency, and also the hysteresis algorithm that can require a 2 segment frequency change before the tuner changes. Gary from Elecraft says "if you are off by one bin segment then it should not matter too much"

I have the same frustration with my KAT500 tuner. I send it a frequency with 1 kHz resolution ( I was using 1 Hz resolution but found that was futile) but it wants to override that frequency with a much less accurate RF derived frequency. I suggested to Elecraft that the frequency received over the serial interface should have priority unless the RF detected frequency differed by more that a user set threshold. That proposal was rejected on the basis that here cannot be two master frequency sources. That argument was lost on me since I have worked in the development of avionics systems where master swaps are common and dynamic data source selection is also common.

I settled for sending my KAT500 a new frequency any time the actual transmitted frequency differs from the KAT500 reported frequency by more than 8 kHz. I have also tweaked FCCS and FCMD (not easy because they don't seem to work as documented) to make is less likely that RF detected frequency will usurp my serial data frequency at least for CW transmissions.

I spent a fair bit of time manually selecting L and C to get a near perfect match in the band segments I use. I find it frustrating that the KAT500 firmware thinks it can do better than what I tell it to do.

I wish Elecraft would give serial data frequency priority over RF detected frequency until a difference threshold is exceed. It would not be difficult to do.

Andy, k3wyc

______________________________________________________________
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]

Don Wilhelm

Re: KPA-1500 issues

Andy and all,

The KAT500 and KAT1500 tuner band segments are divided into frequency
band segments - 25kHz wide on the lower bands and 50kHz wide on the
upper bands if I recall correctly, so an 8kHz shift will not be
significant unless the frequency shift moves the tuning into the
adjacent segment.

There is also a limit to how far the firmware will go to reducing SWR.
In the interest of minimizing tuning time (and relay wear and tear). If
the match is less than 1.3:1 or so, tuning will be declared adequate and
will be stopped.

Since the typical driving transceiver can tolerate an SWR of at least
2:1 without power foldback or damage, setting that 1.3:1 limit is
reasonable - unless you are a 'purest' and mentally can only tolerate a
1:1 SWR (big grin).

73,
Don W3FPR

On 11/29/2018 5:42 PM, ANDY DURBIN wrote:
>
> "In addition the tuner may not always be in the segment you think it is. This is due to the 8KHz resolution of the internal counter, that wins over you sending it a more accurate frequency, and also the hysteresis algorithm that can require a 2 segment frequency change before the tuner changes. Gary from Elecraft says "if you are off by one bin segment then it should not matter too much"
>
>
> I have the same frustration with my KAT500 tuner. I send it a frequency with 1 kHz resolution ( I was using 1 Hz resolution but found that was futile) but it wants to override that frequency with a much less accurate RF derived frequency. I suggested to Elecraft that the frequency received over the serial interface should have priority unless the RF detected frequency differed by more that a user set threshold. That proposal was rejected on the basis that here cannot be two master frequency sources. That argument was lost on me since I have worked in the development of avionics systems where master swaps are common and dynamic data source selection is also common.
>
> I settled for sending my KAT500 a new frequency any time the actual transmitted frequency differs from the KAT500 reported frequency by more than 8 kHz. I have also tweaked FCCS and FCMD (not easy because they don't seem to work as documented) to make is less likely that RF detected frequency will usurp my serial data frequency at least for CW transmissions.
>
______________________________________________________________
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]

ANDY DURBIN

Re: KPA-1500 issues

In reply to this post by ANDY DURBIN

"There is also a limit to how far the firmware will go to reducing SWR. In the interest of minimizing tuning time (and relay wear and tear). If the match is less than 1.3:1 or so, tuning will be declared adequate and will be stopped. Since the typical driving transceiver can tolerate an SWR of at least 2:1 without power foldback or damage, setting that 1.3:1 limit is reasonable - unless you are a 'purest' and mentally can only tolerate a 1:1 SWR (big grin)."

I have no interest in how good any Elecraft tuner is at finding a minimum SWR solution. I use my LP-100A which will give good complex load readings with far lower power output that is required for the KAT500 to give a much less useful SWR reading. None of my tuning solutions was derived by the KAT500. It can't find a solution with the 5 W signal I like to tune with.

As to your assertion that an SWR of 2:1 should be tolerable I call "bollocks!" The KPA500 will not even accept 1.5:1 if the R is on the low side of 50 ohm. SWR is a poor indicator of the load presented to an amplifier. I was perfectly happy with an SWR of 1.5:1 until I started using my KPA500 and realized the very significant difference the complex load made to PA dissipation and the probability of fault trips.

I use my KAT500 as a manual tuner and just wish it would do what I tell it. (I found a solution for this frequency, I gave it the frequency, use the solution I told it to use for that frequency.)

Andy k3wyc
______________________________________________________________
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]

K9MA

Re: KPA-1500 issues

In reply to this post by Don Wilhelm

On 11/29/2018 16:57, Don Wilhelm wrote:
> Since the typical driving transceiver can tolerate an SWR of at least
> 2:1 without power foldback or damage, setting that 1.3:1 limit is
> reasonable -

That certainly is not the case with the KPA1500. An SWR of 1.4:1 can
cause it to fault well below 1500 W due to either excess current or
excessive drive power, or to greatly exceed 1500 W output.

I don't think the K3 will deliver anywhere near 100 W at 2:1, either.

73,

Scott K9MA

--
Scott K9MA

[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]

Christopher Hoover

Re: KPA-1500 issues

In reply to this post by ANDY DURBIN

>
> I wish Elecraft would give serial data frequency priority over RF detected
> frequency until a difference threshold is exceed.

+1

73 de AI6KG

On Thu, Nov 29, 2018 at 2:43 PM ANDY DURBIN <[hidden email]> wrote:

>
> "In addition the tuner may not always be in the segment you think it is.
> This is due to the 8KHz resolution of the internal counter, that wins over
> you sending it a more accurate frequency, and also the hysteresis algorithm
> that can require a 2 segment frequency change before the tuner changes.
> Gary from Elecraft says "if you are off by one bin segment then it should
> not matter too much"
>
>
> I have the same frustration with my KAT500 tuner. I send it a frequency
> with 1 kHz resolution ( I was using 1 Hz resolution but found that was
> futile) but it wants to override that frequency with a much less accurate
> RF derived frequency. I suggested to Elecraft that the frequency received
> over the serial interface should have priority unless the RF detected
> frequency differed by more that a user set threshold. That proposal was
> rejected on the basis that here cannot be two master frequency sources.
> That argument was lost on me since I have worked in the development of
> avionics systems where master swaps are common and dynamic data source
> selection is also common.
>
> I settled for sending my KAT500 a new frequency any time the actual
> transmitted frequency differs from the KAT500 reported frequency by more
> than 8 kHz. I have also tweaked FCCS and FCMD (not easy because they
> don't seem to work as documented) to make is less likely that RF detected
> frequency will usurp my serial data frequency at least for CW transmissions.
>
> I spent a fair bit of time manually selecting L and C to get a near
> perfect match in the band segments I use. I find it frustrating that the
> KAT500 firmware thinks it can do better than what I tell it to do.
>
> I wish Elecraft would give serial data frequency priority over RF detected
> frequency until a difference threshold is exceed. It would not be
> difficult to do.
>
> Andy, k3wyc
>
>
>
>
>
> ______________________________________________________________
> 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]
>

Randy Farmer

Re: KPA-1500 issues

In reply to this post by Don Wilhelm

I've had difficulties with my KAT500 / KPA500 combo (only) on 160
meters, apparently due to the tuner relying on RF sensing. I've
carefully "trained" my KAT500 all across the 160 meter band. Since the
manual states that the tuning bins are 10 kHz wide below 3 MHz, I do a
tune at 1805, 1815 ... up to about 1895. My antenna is an impossible
load above 1900 kHz, so I don't try to tune up there. Since I don't have
exactly a great antenna system, I almost always stick to S&P operation
in 160 meter contests. As I tune across the band I can hear the relays
selecting the appropriate tuning solutions. But... every once in awhile,
as soon as I key the transmitter I hear the relays in the KAT500 switch,
presumably out of the commanded tuning bin due to frequency measurement
error in the RF detection circuit. This shouldn't be a problem, since
the tuning doesn't change significantly between adjacent bins. However,
since the amp is being keyed while the tuner is changing presets, the
amp immediately faults and goes off line. This is very annoying, and
probably not so good for either the amp or the tuner. If the RF
measurement resolution is really as poor as 8kHz, I don't see why
there's any point in providing for 10 kHz wide tuning bins on the low
bands. I would also strongly agree that if good frequency information is
available to the tuner via the serial bus that's what it should use to
select tuning bins. The RF sensing information should be a backup
source, to be relied on only in the absence of such data.

73...
Randy, W8FN

On 11/29/2018 4:57 PM, Don Wilhelm wrote:

> Andy and all,
>
> The KAT500 and KAT1500 tuner band segments are divided into frequency
> band segments - 25kHz wide on the lower bands and 50kHz wide on the
> upper bands if I recall correctly, so an 8kHz shift will not be
> significant unless the frequency shift moves the tuning into the
> adjacent segment.
>
> There is also a limit to how far the firmware will go to reducing SWR.
> In the interest of minimizing tuning time (and relay wear and tear).
> If the match is less than 1.3:1 or so, tuning will be declared
> adequate and will be stopped.
>
> Since the typical driving transceiver can tolerate an SWR of at least
> 2:1 without power foldback or damage, setting that 1.3:1 limit is
> reasonable - unless you are a 'purest' and mentally can only tolerate
> a 1:1 SWR (big grin).
>
> 73,
> Don W3FPR
>
> On 11/29/2018 5:42 PM, ANDY DURBIN wrote:
>>
>> "In addition the tuner may not always be in the segment you think it
>> is. This is due to the 8KHz resolution of the internal counter, that
>> wins over you sending it a more accurate frequency, and also the
>> hysteresis algorithm that can require a 2 segment frequency change
>> before the tuner changes. Gary from Elecraft says "if you are off by
>> one bin segment then it should not matter too much"
>>
>>
>> I have the same frustration with my KAT500 tuner. I send it a
>> frequency with 1 kHz resolution ( I was using 1 Hz resolution but
>> found that was futile) but it wants to override that frequency with a
>> much less accurate RF derived frequency. I suggested to Elecraft
>> that the frequency received over the serial interface should have
>> priority unless the RF detected frequency differed by more that a
>> user set threshold. That proposal was rejected on the basis that
>> here cannot be two master frequency sources. That argument was lost
>> on me since I have worked in the development of avionics systems
>> where master swaps are common and dynamic data source selection is
>> also common.
>>
>> I settled for sending my KAT500 a new frequency any time the actual
>> transmitted frequency differs from the KAT500 reported frequency by
>> more than 8 kHz. I have also tweaked FCCS and FCMD (not easy
>> because they don't seem to work as documented) to make is less likely
>> that RF detected frequency will usurp my serial data frequency at
>> least for CW transmissions.

Bob McGraw - K4TAX

Re: KPA-1500 issues

In reply to this post by K9MA

In checking my K3S into an antenna that presents a 2:1 SWR as indicated
on my external Power/SWR meter, my KPA500 in STBY, and my KAT500 in
bypass mode, it does deliver 100 watts.   Although, I prefer not to
operate in that configuration if not necessary.

I think in general we fret too much over values of SWR which are 1.5:1
or less at the station end i.e. between the tuner and the amp or
transceiver. This is without regard to the actual SWR on the feedline
between the tuner and the antenna load. That is where the real power
loss occurs. Also, in many cases, using a tuner in a feedline where the
SWR is 1.5:1 or less, most likely the tuner introduces more loss than if
the tuner was not in the path.    There is a nice XLS spreadsheet
http://www.dj0ip.de/antenna-matchboxes/matchbox-shoot-out/ showing the
actual loss of many different tuners, different bands, and different
matched and unmatched losses.   Rather eye opening.

Regarding the SWR fault at 1.4:1, is that at 35 ohms or at 70 ohms?
Significant difference in current demand of the PA between the two
values while SWR appears to remain the same.   This is largely the
reason I often state that SWR only tells 1/2 of the story of antenna
system performance. A PA correctly designed will monitor the current in
addition to SWR and not just SWR as most seem to do.

73

Bob, K4TAX

On 11/29/2018 6:16 PM, K9MA wrote:

>
> That certainly is not the case with the KPA1500. An SWR of 1.4:1 can
> cause it to fault well below 1500 W due to either excess current or
> excessive drive power, or to greatly exceed 1500 W output.
>
> I don't think the K3 will deliver anywhere near 100 W at 2:1, either.
>
> 73,
>
> Scott K9MA
>

K9MA

Re: KPA-1500 issues

It's true that the effect a given SWR has on an amplifier output depends
on the actual impedance presented. That 2:1 SWR can be anywhere from 25
to 100 Ohms, plus various reactive impedances in between. On
simple-minded way to look at it is to imagine an amplifier as a
low-impedance source. (A crude approximation, but a useful one.) With
the same drive level, it might deliver 200 W to a 25 Ohm load, or 50 W
to a 100 Ohm load. If you reduced the drive to the 25 Ohm load for 100
W output, the amplifier would be very inefficient, as current would be
high. On the other hand, if you increase drive with the 100 Ohm load, it
is likely the amplifier will saturate before reaching full power, due to
limited voltage output capability. The only way to make an amplifier
tolerate such big load differences is to make it capable of much higher
output, which increases cost. A good ATU is probably more cost effective.

73,
Scott K9MA

On 11/29/2018 19:22, Bob McGraw K4TAX wrote:

> In checking my K3S into an antenna that presents a 2:1 SWR as
> indicated on my external Power/SWR meter, my KPA500 in STBY, and my
> KAT500 in bypass mode, it does deliver 100 watts.   Although, I prefer
> not to operate in that configuration if not necessary.
>
> I think in general we fret too much over values of SWR which are 1.5:1
> or less at the station end i.e. between the tuner and the amp or
> transceiver. This is without regard to the actual SWR on the
> feedline between the tuner and the antenna load. That is where the
> real power loss occurs. Also, in many cases, using a tuner in a
> feedline where the SWR is 1.5:1 or less, most likely the tuner
> introduces more loss than if the tuner was not in the path.    There
> is a nice XLS spreadsheet
> http://www.dj0ip.de/antenna-matchboxes/matchbox-shoot-out/ showing the
> actual loss of many different tuners, different bands, and different
> matched and unmatched losses.   Rather eye opening.
>
> Regarding the SWR fault at 1.4:1, is that at 35 ohms or at 70 ohms?
> Significant difference in current demand of the PA between the two
> values while SWR appears to remain the same.   This is largely the
> reason I often state that SWR only tells 1/2 of the story of antenna
> system performance. A PA correctly designed will monitor the current
> in addition to SWR and not just SWR as most seem to do.
>
> 73
>
> Bob, K4TAX
>
>
> On 11/29/2018 6:16 PM, K9MA wrote:
>>
>> That certainly is not the case with the KPA1500. An SWR of 1.4:1 can
>> cause it to fault well below 1500 W due to either excess current or
>> excessive drive power, or to greatly exceed 1500 W output.
>>
>> I don't think the K3 will deliver anywhere near 100 W at 2:1, either.
>>
>> 73,
>>
>> Scott K9MA
>>
>
>
> ______________________________________________________________
> 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]

Jim Brown-10

Re: KPA-1500 issues

In reply to this post by Bob McGraw - K4TAX

Lots of important stuff here, most of which I agree with. See comments
interspersed.

On 11/29/2018 5:22 PM, Bob McGraw K4TAX wrote:
> I think in general we fret too much over values of SWR which are 1.5:1
> or less at the station end i.e. between the tuner and the amp or
> transceiver.
I paid a lot less attention to the SWR that my Titan 425 amps saw
because their manually tuned Pi-Li output stage could make the tubes
happy working into much higher SWR than the auto-tuned Pi-L in the 87As
that replaced the Titans. And they both use the same tubes. The Alpha
will fault as the SWR approaches 2:1. And, has been noted, both KPA
amps introduce loss in the drive circuit to reduce power if the SWR gets
too high because it makes the output devices too warm.
> This is without regard to the actual SWR on the feedline between the
> tuner and the antenna load. That is where the real power loss occurs.

Yes and no -- most hams believe that excess loss due to SWR is much
higher than it is. I got into this many years ago when specifying low
loss 75 ohm CATV coax for remotely located wireless mic antennas for
sound systems in large spaces. Engineers for the major wireless system
mfrs told me that 1) they didn't know what the input Z of their
receivers were and 2) that was probably somewhere between 50 and 100
ohms. My logic was (and still is) pretty simple -- thanks to the very
high volume of coax for CATV systems, low loss 75 ohm coax is FAR
cheaper than low loss 50 ohm coax, and even assuming a 50 ohm RX input,
worst case excess loss for a 1.5:1 mismatch is 0.18 dB no matter how
long the line is.

That said, because the bandwidth of 80M is a rather percentage of its
operating frequency, SWR can get fairly high at the band edges of a
dipole cut to the center of the band. My dipoles are up about 120 ft at
the end of about 170 ft of coax with feedpoint Z are resonance about 88
ohms (measured), so they're fed with the lowest loss RG11 I can find
(currently Davis RF RG11, which measures like Belden 8213).

> Also, in many cases, using a tuner in a feedline where the SWR is
> 1.5:1 or less, most likely the tuner introduces more loss than if the
> tuner was not in the path. There is a nice XLS spreadsheet
> http://www.dj0ip.de/antenna-matchboxes/matchbox-shoot-out/ showing the
> actual loss of many different tuners, different bands, and different
> matched and unmatched losses. Rather eye opening.

Nice chart, which I've seen before. Over the years, ARRL has published
this sort of testing for a lot of tuners. I'd far rather see this in dB
-- 10% lost power is 0.46 dB, 20% is 0.97 dB.

73, Jim K9YC

______________________________________________________________
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]