Elecraft

QRO - KPA500 Failure

Classic

List

Threaded

38 messages Options

Doug Turnbull

Re: QRO - KPA500 Failure

Andy,
Yes, you are right and this applies to both KPA 500 and KPA 1500. The
conditions for maximum efficiency are not necessarily those for minimum
dissipation or reduced stress. I would like to know if running 2/3, 1/2
or even 1/4 maximum power output increases reliability and MTBF. It would
be a considerable surprise to me if this were not the case and it would be
good to know the numbers. Improvement in MTBF can be several fold for
some devices for a reduction in voltage, current or power of as little as
10%. Good design practice normally means that this is part of the
engineering design process. The proof of derating is looked for in design
reviews. Too many hams run their amplifiers flat out for very little
benefit. Signal cleanliness is still another issue; higher voltage but
lower power seems to help here.

Using FT8 or WSPR on dead band (160M in daytime or 12M), I sometimes
run 20 Watts plus. This can be stressful on the K3 especially for WSPR.
I crank the K3 down to a watt or so and drive the KPA 1500. Yes, my total
energy consumption is no doubt up but stress on either K3 or KPA is minimal.
This however is not my concern, I would like to know if running the KPA at
lower power increases reliability. In EI we can run 1500 Watts for
contests but outside of contests can only run 400W max.

73 Doug EI2CN

-----Original Message-----
From: [hidden email] <[hidden email]> On
Behalf Of Andy Durbin
Sent: Monday 10 February 2020 01:45
To: [hidden email]
Subject: [Elecraft] QRO - KPA500 Failure

"I am seeing a lot of confusion and erroneous conclusions based on the terms
'efficiency', 'dissipation' coupled with 'power output'."

Yes, very true. Only a few have actually contributed test data. All the
rest is assumption and speculation.

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]

______________________________________________________________
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: QRO - KPA500 Failure

Doug,

"I would like to know if running 2/3, 1/2 or even 1/4 maximum power output increases reliability and MTBF."

I have no experience of amplifier design but I would expect reliability of the finals to depend on temperature. Finals temperature will depend on the power dissipated in the finals (PA dissipation) and the effectiveness of the cooling design.

To understands finals dissipation one would need to produce one's own test data or rely on the work of others. My test data for the KPA500 shows PA dissipation peaks at about half power output. So far no one has disputed the validity of my test data.

It is my understanding that running an amplifier at reduced power only reduces PA dissipation if the finals voltage is adjusted give a good load match for that reduced power condition. The KPA500 has no provision for adjusting PA voltage for the reduced power case. For reduced power the PA voltage is actually higher than for full power because the high voltage supply is not regulated.

As I said, I have no experience in amplifier design and would appreciate comments from those who do.

73,
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]

Doug Turnbull

Re: QRO - KPA500 Failure

Hello Andy,

I do not have the numbers and also would like them. I understand
that efficiency decreases as one derates from the maximum of 1500W output.
It surprises me that the power dissipated would peak at half power output or
750 Watts. This may be so. If so then almost certainly the Mean Time
Between Failures or MTBF is worsened. I am not in a position to
determine any of this but maybe some guidance from Elecraft itself would be
helpful for both KPA 500 and 1500.

73 Doug EI2CN

From: Andy Durbin <[hidden email]>
Sent: Monday 10 February 2020 14:04
To: Doug Turnbull <[hidden email]>; [hidden email]
Subject: Re: [Elecraft] QRO - KPA500 Failure

Doug,

"I would like to know if running 2/3, 1/2 or even 1/4 maximum power output
increases reliability and MTBF."

I have no experience of amplifier design but I would expect reliability of
the finals to depend on temperature. Finals temperature will depend on the
power dissipated in the finals (PA dissipation) and the effectiveness of the
cooling design.

To understands finals dissipation one would need to produce one's own test
data or rely on the work of others. My test data for the KPA500 shows PA
dissipation peaks at about half power output. So far no one has disputed
the validity of my test data.

It is my understanding that running an amplifier at reduced power only
reduces PA dissipation if the finals voltage is adjusted give a good load
match for that reduced power condition. The KPA500 has no provision for
adjusting PA voltage for the reduced power case. For reduced power the PA
voltage is actually higher than for full power because the high voltage
supply is not regulated.

As I said, I have no experience in amplifier design and would appreciate
comments from those who do.

73,

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]

Bob McGraw - K4TAX

Re: QRO - KPA500 Failure

One point to consider is the output network that is used to match the output device impedance to 50 ohms. It is a fixed value or ratio. Thus if designed for rated power then at 1/2 or 1/4 power the value would not seem to provide a proper match for effective transfer of energy. This can attribute to overall efficiency.

Bob, K4TAX

Sent from my iPhone

> On Feb 10, 2020, at 9:38 AM, Doug Turnbull <[hidden email]> wrote:
>
> Hello Andy,
>
> I do not have the numbers and also would like them. I understand
> that efficiency decreases as one derates from the maximum of 1500W output.
> It surprises me that the power dissipated would peak at half power output or
> 750 Watts. This may be so. If so then almost certainly the Mean Time
> Between Failures or MTBF is worsened. I am not in a position to
> determine any of this but maybe some guidance from Elecraft itself would be
> helpful for both KPA 500 and 1500.
>
> 73 Doug EI2CN
>
>
>
> From: Andy Durbin <[hidden email]>
> Sent: Monday 10 February 2020 14:04
> To: Doug Turnbull <[hidden email]>; [hidden email]
> Subject: Re: [Elecraft] QRO - KPA500 Failure
>
>
>
> Doug,
>
>
>
> "I would like to know if running 2/3, 1/2 or even 1/4 maximum power output
> increases reliability and MTBF."
>
>
>
> I have no experience of amplifier design but I would expect reliability of
> the finals to depend on temperature. Finals temperature will depend on the
> power dissipated in the finals (PA dissipation) and the effectiveness of the
> cooling design.
>
>
>
> To understands finals dissipation one would need to produce one's own test
> data or rely on the work of others. My test data for the KPA500 shows PA
> dissipation peaks at about half power output. So far no one has disputed
> the validity of my test data.
>
>
>
> It is my understanding that running an amplifier at reduced power only
> reduces PA dissipation if the finals voltage is adjusted give a good load
> match for that reduced power condition. The KPA500 has no provision for
> adjusting PA voltage for the reduced power case. For reduced power the PA
> voltage is actually higher than for full power because the high voltage
> supply is not regulated.
>
>
>
> As I said, I have no experience in amplifier design and would appreciate
> comments from those who do.
>
>
>
> 73,
>
> 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: QRO - KPA500 Failure

In reply to this post by Doug Turnbull

Several others have posted test data for the KPA500 here recently. As I
recall, it shows that dissipated power is fairly constant from about 50
percent to full output, at about 500 W, dropping a bit at lower power.

If it were an ideal class B amplifier, which no solid state RF amplifier
even approaches, the efficiency at full output would be 78.5 percent.
It would then dissipate only 141 W at full output, and maximum
dissipation would be at some lower output power level. This is probably
what folks are thinking of when they assume maximum dissipation occurs
at less than full output. REAL RF amplifiers don't behave that way, as
the test results show.

https://en.wikipedia.org/wiki/Power_amplifier_classes#Class_B

73,
Scott K9MA

On 2/10/2020 09:23, Doug Turnbull wrote:

> Hello Andy,
>
> I do not have the numbers and also would like them. I understand
> that efficiency decreases as one derates from the maximum of 1500W output.
> It surprises me that the power dissipated would peak at half power output or
> 750 Watts. This may be so. If so then almost certainly the Mean Time
> Between Failures or MTBF is worsened. I am not in a position to
> determine any of this but maybe some guidance from Elecraft itself would be
> helpful for both KPA 500 and 1500.
>
> 73 Doug EI2CN
>
>
>
> From: Andy Durbin <[hidden email]>
> Sent: Monday 10 February 2020 14:04
> To: Doug Turnbull <[hidden email]>; [hidden email]
> Subject: Re: [Elecraft] QRO - KPA500 Failure
>
>
>
> Doug,
>
>
>
> "I would like to know if running 2/3, 1/2 or even 1/4 maximum power output
> increases reliability and MTBF."
>
>
>
> I have no experience of amplifier design but I would expect reliability of
> the finals to depend on temperature. Finals temperature will depend on the
> power dissipated in the finals (PA dissipation) and the effectiveness of the
> cooling design.
>
>
>
> To understands finals dissipation one would need to produce one's own test
> data or rely on the work of others. My test data for the KPA500 shows PA
> dissipation peaks at about half power output. So far no one has disputed
> the validity of my test data.
>
>
>
> It is my understanding that running an amplifier at reduced power only
> reduces PA dissipation if the finals voltage is adjusted give a good load
> match for that reduced power condition. The KPA500 has no provision for
> adjusting PA voltage for the reduced power case. For reduced power the PA
> voltage is actually higher than for full power because the high voltage
> supply is not regulated.
>
>
>
> As I said, I have no experience in amplifier design and would appreciate
> comments from those who do.
>
>
>
> 73,
>
> 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]

ANDY DURBIN

Re: QRO - KPA500 Failure

In reply to this post by ANDY DURBIN

"I understand that efficiency decreases as one derates from the maximum of 1500W output.It surprises me that the power dissipated would peak at half power output or 750 Watts."

I have no test data for the KPA1500 and will not speculate on the shape of the PA dissipation curves. All I have is data for one KPA500. For that one KPA500, and under the conditions tested, it's clear that PA dissipation is greater at half power than at full power.

It would be easy to get a pair of data points for the KPA1500.

Connect the amplifier to a known good 50 ohm load.
Drive the amplifier to steady state 1500 W output
Record PA current and voltage
Drive the amplifier to steady state 750 W output
Record the PA current and voltage

Repeat on multiple bands if desired and report the results here.

73,
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]

Doug Turnbull

Re: QRO - KPA500 Failure

In reply to this post by K9MA

Dear OMs, Maybe this is an advantage of the SPE 2KFA amp which has three power levels 2000, 1200 and 600 W output. I have both amps and prefer the KPA 1500. Which I find quieter both acoustically and RFI wise on RX. The KPA 1500 is also happier handling Data modes. The SPE will drop down to 600W for RTTY. 73 Doug EI2CNSent from my Samsung Galaxy smartphone.
-------- Original message --------From: K9MA <[hidden email]> Date: 10/02/2020 16:29 (GMT+00:00) To: [hidden email] Subject: Re: [Elecraft] QRO - KPA500 Failure Several others have posted test data for the KPA500 here recently. As I recall, it shows that dissipated power is fairly constant from about 50 percent to full output, at about 500 W, dropping a bit at lower power.If it were an ideal class B amplifier, which no solid state RF amplifier even approaches, the efficiency at full output would be 78.5 percent. It would then dissipate only 141 W at full output, and maximum dissipation would be at some lower output power level. This is probably what folks are thinking of when they assume maximum dissipation occurs at less than full output. REAL RF amplifiers don't behave that way, as the test results show.https://en.wikipedia.org/wiki/Power_amplifier_classes#Class_B73,Scott K9MAOn 2/10/2020 09:23, Doug Turnbull wrote:> Hello Andy,>> I do not have the numbers and also would like them. I understand> that efficiency decreases as one derates from the maximum of 1500W output.> It surprises me that the power dissipated would peak at half power output or> 750 Watts. This may be so. If so then almost certainly the Mean Time> Between Failures or MTBF is worsened. I am not in a position to> determine any of this but maybe some guidance from Elecraft itself would be> helpful for both KPA 500 and 1500.>> 73 Doug EI2CN>> >> From: Andy Durbin <[hidden email]>> Sent: Monday 10 February 2020 14:04> To: Doug Turnbull <[hidden email]>; [hidden email]> Subject: Re: [Elecraft] QRO - KPA500 Failure>> >> Doug,>> >> "I would like to know if running 2/3, 1/2 or even 1/4 maximum power output> increases reliability and MTBF.">> >> I have no experience of amplifier design but I would expect reliability of> the finals to depend on temperature. Finals temperature will depend on the> power dissipated in the finals (PA dissipation) and the effectiveness of the> cooling design.>> >> To understands finals dissipation one would need to produce one's own test> data or rely on the work of others. My test data for the KPA500 shows PA> dissipation peaks at about half power output. So far no one has disputed> the validity of my test data.>> >> It is my understanding that running an amplifier at reduced power only> reduces PA dissipation if the finals voltage is adjusted give a good load> match for that reduced power condition. The KPA500 has no provision for> adjusting PA voltage for the reduced power case. For reduced power the PA> voltage is actually higher than for full power because the high voltage> supply is not regulated.>> >> As I said, I have no experience in amplifier design and would appreciate> comments from those who do.>> >> 73,>> 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]-- Scott K9MAk9ma@sdellington.us______________________________________________________________Elecraft mailing listHome: http://mailman.qth.net/mailman/listinfo/elecraftHelp: http://mailman.qth.net/mmfaq.htmPost: mailto:[hidden email] list hosted by: http://www.qsl.netPlease help support this email list: http://www.qsl.net/donate.htmlMessage 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]

Doug Turnbull

Re: QRO - KPA500 Failure

In reply to this post by ANDY DURBIN

We should also know the power supply efficiency though as a switcher, I expect 90 plus percent.73 Doug EI2CNSent from my Samsung Galaxy smartphone.
-------- Original message --------From: Andy Durbin <[hidden email]> Date: 10/02/2020 16:58 (GMT+00:00) To: Doug Turnbull <[hidden email]>, [hidden email] Subject: Re: [Elecraft] QRO - KPA500 Failure

"I understand that efficiency decreases as one derates from
the maximum of 1500W output.It surprises me that the power dissipated would peak at half power output or 750
Watts."

I have no test data for the KPA1500 and will not speculate on the shape of the PA dissipation curves. All I have is data for one KPA500. For that one KPA500, and under the conditions tested, it's clear that PA dissipation is greater at half power than at
full power.

It would be easy to get a pair of data points for the KPA1500.

Connect the amplifier to a known good 50 ohm load.

Drive the amplifier to steady state 1500 W output

Record PA current and voltage

Drive the amplifier to steady state 750 W output

Record the PA current and voltage

Repeat on multiple bands if desired and report the results here.

73,

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]

ANDY DURBIN

QRO - KPA500 Failure

In reply to this post by KK1W

"This is probably what folks are thinking of when they assume maximum dissipation occurs at less than full output."

A possible source of error in my data is that the dummy load is cool at the start of the power ramp and hot by the time peak power is reached. I recorded LP-100A data during each test run and can see some change in the load characteristics. I did not attempt to compensate for this when plotting the data. I simply reported values of R and X that were about average for each test run.

One way to find out if the change in load is significant would be to start the power ramp at maximum power rather than minimum power.

Some time after I ran those tests I expanded the capability of my Arduino station controller to add a page that displays PA voltage, PA current, PA DC input, PA dissipation, RF Output, and efficiency. A snapshot of the displayed data can be sent to the controller's logger output.

I just ran a test which had a brief transmission close to full power and another brief transmission at half power. This test was performed at 14.01 MHz using a dummy load. Here are the data:

26:36:36.672 KPA Mon frozen
PA voltage=60.6
PA current=14.1
PA input=854
RF out=495
PA diss=359
PA effic=57

26:37:32.437 KPA Mon frozen
PA voltage=64.6
PA current=10.0
PA input=646
RF out=250
PA diss=396
PA effic=38

Dummy load heating would not have been significant factor in this test but the results still show higher PA dissipation at 250 W out than at 495 W out.

It's really simple to produce a similar data set and there are lots of KPA500 owners here. If you think my data are invalid please share your test data. I don't mind being wrong as long as I learn something new.

73,
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]

Bob McGraw - K4TAX

Re: QRO - KPA500 Failure

Unless one is concerned about power consumption, the efficiency factor should be really of no concern. Total heat output will always be greater at maximum output and less at lower power. Nice for warming the shack on a Winter day.

Bob, K4TAX

Sent from my iPhone

> On Feb 10, 2020, at 12:18 PM, Andy Durbin <[hidden email]> wrote:
>
>
> "This is probably what folks are thinking of when they assume maximum dissipation occurs at less than full output."
>
> A possible source of error in my data is that the dummy load is cool at the start of the power ramp and hot by the time peak power is reached. I recorded LP-100A data during each test run and can see some change in the load characteristics. I did not attempt to compensate for this when plotting the data. I simply reported values of R and X that were about average for each test run.
>
> One way to find out if the change in load is significant would be to start the power ramp at maximum power rather than minimum power.
>
> Some time after I ran those tests I expanded the capability of my Arduino station controller to add a page that displays PA voltage, PA current, PA DC input, PA dissipation, RF Output, and efficiency. A snapshot of the displayed data can be sent to the controller's logger output.
>
> I just ran a test which had a brief transmission close to full power and another brief transmission at half power. This test was performed at 14.01 MHz using a dummy load. Here are the data:
>
> 26:36:36.672 KPA Mon frozen
> PA voltage=60.6
> PA current=14.1
> PA input=854
> RF out=495
> PA diss=359
> PA effic=57
>
> 26:37:32.437 KPA Mon frozen
> PA voltage=64.6
> PA current=10.0
> PA input=646
> RF out=250
> PA diss=396
> PA effic=38
>
> Dummy load heating would not have been significant factor in this test but the results still show higher PA dissipation at 250 W out than at 495 W out.
>
> It's really simple to produce a similar data set and there are lots of KPA500 owners here. If you think my data are invalid please share your test data. I don't mind being wrong as long as I learn something new.
>
> 73,
> 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]
>

ke9uw

Re: QRO - KPA500 Failure

Yes, I was wondering why this was so interesting.

Chuck Jack Hawley
KE9UW

Sent from my iPhone, cjack

> On Feb 10, 2020, at 12:47 PM, Bob McGraw K4TAX <[hidden email]> wrote:
>
> Unless one is concerned about power consumption, the efficiency factor should be really of no concern. Total heat output will always be greater at maximum output and less at lower power. Nice for warming the shack on a Winter day.
>
>
> Bob, K4TAX
>
>
> Sent from my iPhone
>
>> On Feb 10, 2020, at 12:18 PM, Andy Durbin <[hidden email]> wrote:
>>
>>
>> "This is probably what folks are thinking of when they assume maximum dissipation occurs at less than full output."
>>
>> A possible source of error in my data is that the dummy load is cool at the start of the power ramp and hot by the time peak power is reached. I recorded LP-100A data during each test run and can see some change in the load characteristics. I did not attempt to compensate for this when plotting the data. I simply reported values of R and X that were about average for each test run.
>>
>> One way to find out if the change in load is significant would be to start the power ramp at maximum power rather than minimum power.
>>
>> Some time after I ran those tests I expanded the capability of my Arduino station controller to add a page that displays PA voltage, PA current, PA DC input, PA dissipation, RF Output, and efficiency. A snapshot of the displayed data can be sent to the controller's logger output.
>>
>> I just ran a test which had a brief transmission close to full power and another brief transmission at half power. This test was performed at 14.01 MHz using a dummy load. Here are the data:
>>
>> 26:36:36.672 KPA Mon frozen
>> PA voltage=60.6
>> PA current=14.1
>> PA input=854
>> RF out=495
>> PA diss=359
>> PA effic=57
>>
>> 26:37:32.437 KPA Mon frozen
>> PA voltage=64.6
>> PA current=10.0
>> PA input=646
>> RF out=250
>> PA diss=396
>> PA effic=38
>>
>> Dummy load heating would not have been significant factor in this test but the results still show higher PA dissipation at 250 W out than at 495 W out.
>>
>> It's really simple to produce a similar data set and there are lots of KPA500 owners here. If you think my data are invalid please share your test data. I don't mind being wrong as long as I learn something new.
>>
>> 73,
>> 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]
>>
>
>
> ______________________________________________________________
> 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]

Chuck, KE9UW

Doug Turnbull

Re: QRO - KPA500 Failure

In reply to this post by Bob McGraw - K4TAX

Thank you Andy,
Bob, if maximum power dissipation occurs at 50% max output power then the greatest amount of heat should be generated for the shack at half max output power. Power dissipation is in the form of heat not RF. The subject is interesting for a number of reasons perhaps most importantly to me is that of reliability. I believe the chance of failure is reduced as one lowers dissipation while staying within all other component ratings.

73 Doug EI2CN

-----Original Message-----
From: [hidden email] <[hidden email]> On Behalf Of Bob McGraw K4TAX
Sent: Monday 10 February 2020 18:47
To: Andy Durbin <[hidden email]>
Cc: [hidden email]
Subject: Re: [Elecraft] QRO - KPA500 Failure

Unless one is concerned about power consumption, the efficiency factor should be really of no concern. Total heat output will always be greater at maximum output and less at lower power. Nice for warming the shack on a Winter day.

Bob, K4TAX

Sent from my iPhone

> On Feb 10, 2020, at 12:18 PM, Andy Durbin <[hidden email]> wrote:
>
>
> "This is probably what folks are thinking of when they assume maximum dissipation occurs at less than full output."
>
> A possible source of error in my data is that the dummy load is cool at the start of the power ramp and hot by the time peak power is reached. I recorded LP-100A data during each test run and can see some change in the load characteristics. I did not attempt to compensate for this when plotting the data. I simply reported values of R and X that were about average for each test run.
>
> One way to find out if the change in load is significant would be to start the power ramp at maximum power rather than minimum power.
>
> Some time after I ran those tests I expanded the capability of my Arduino station controller to add a page that displays PA voltage, PA current, PA DC input, PA dissipation, RF Output, and efficiency. A snapshot of the displayed data can be sent to the controller's logger output.
>
> I just ran a test which had a brief transmission close to full power and another brief transmission at half power. This test was performed at 14.01 MHz using a dummy load. Here are the data:
>
> 26:36:36.672 KPA Mon frozen
> PA voltage=60.6
> PA current=14.1
> PA input=854
> RF out=495
> PA diss=359
> PA effic=57
>
> 26:37:32.437 KPA Mon frozen
> PA voltage=64.6
> PA current=10.0
> PA input=646
> RF out=250
> PA diss=396
> PA effic=38
>
> Dummy load heating would not have been significant factor in this test but the results still show higher PA dissipation at 250 W out than at 495 W out.
>
> It's really simple to produce a similar data set and there are lots of KPA500 owners here. If you think my data are invalid please share your test data. I don't mind being wrong as long as I learn something new.
>
> 73,
> 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]
>

______________________________________________________________
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]

Bob McGraw - K4TAX

Re: QRO - KPA500 Failure

Oh but contraire. I agree that lower dissipation typically reduces
MTBF. Therefore, never plug in your KPA500, leave it in the box and it
should last a very long time.

Sorry folks, I just couldn't resist.

I did not do extensive measurements, but the ones I did show the following:

450 watts out, 819 watts in, for a delta of 369 watts, I presume for
heat or 55% efficiency

210 watts out, 576 watts in, for a delta of 366 watts, I presume for
heat or 37% efficiency {I do work 30M that has a power limit of 200 watts}

Note the difference in heat dissipation between 450 watts out and 210
watts out is only 33 watts but is 240 watts difference in RF output.
Based on these numbers, there is not a lot of difference in heat to be
dissipated in the upper half of the power rating.

100 watts out, 393 watts in, for a delta of 293 watts, I presume for
heat or 25% efficiency.

73

Bob, K4TAX

On 2/10/2020 1:16 PM, Doug Turnbull wrote:

> Thank you Andy,
> Bob, if maximum power dissipation occurs at 50% max output power then the greatest amount of heat should be generated for the shack at half max output power. Power dissipation is in the form of heat not RF. The subject is interesting for a number of reasons perhaps most importantly to me is that of reliability. I believe the chance of failure is reduced as one lowers dissipation while staying within all other component ratings.
>
> 73 Doug EI2CN
>
> -----Original Message-----
> From: [hidden email] <[hidden email]> On Behalf Of Bob McGraw K4TAX
> Sent: Monday 10 February 2020 18:47
> To: Andy Durbin <[hidden email]>
> Cc: [hidden email]
> Subject: Re: [Elecraft] QRO - KPA500 Failure
>
> Unless one is concerned about power consumption, the efficiency factor should be really of no concern. Total heat output will always be greater at maximum output and less at lower power. Nice for warming the shack on a Winter day.
>
>
> Bob, K4TAX
>
>
> Sent from my iPhone
>
>> On Feb 10, 2020, at 12:18 PM, Andy Durbin <[hidden email]> wrote:
>>
>>
>> "This is probably what folks are thinking of when they assume maximum dissipation occurs at less than full output."
>>
>> A possible source of error in my data is that the dummy load is cool at the start of the power ramp and hot by the time peak power is reached. I recorded LP-100A data during each test run and can see some change in the load characteristics. I did not attempt to compensate for this when plotting the data. I simply reported values of R and X that were about average for each test run.
>>
>> One way to find out if the change in load is significant would be to start the power ramp at maximum power rather than minimum power.
>>
>> Some time after I ran those tests I expanded the capability of my Arduino station controller to add a page that displays PA voltage, PA current, PA DC input, PA dissipation, RF Output, and efficiency. A snapshot of the displayed data can be sent to the controller's logger output.
>>
>> I just ran a test which had a brief transmission close to full power and another brief transmission at half power. This test was performed at 14.01 MHz using a dummy load. Here are the data:
>>
>> 26:36:36.672 KPA Mon frozen
>> PA voltage=60.6
>> PA current=14.1
>> PA input=854
>> RF out=495
>> PA diss=359
>> PA effic=57
>>
>> 26:37:32.437 KPA Mon frozen
>> PA voltage=64.6
>> PA current=10.0
>> PA input=646
>> RF out=250
>> PA diss=396
>> PA effic=38
>>
>> Dummy load heating would not have been significant factor in this test but the results still show higher PA dissipation at 250 W out than at 495 W out.
>>
>> It's really simple to produce a similar data set and there are lots of KPA500 owners here. If you think my data are invalid please share your test data. I don't mind being wrong as long as I learn something new.
>>
>> 73,
>> 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]
>>
>
> ______________________________________________________________
> 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]
>
>

KK1W

Re: QRO - KPA500 Failure

In reply to this post by KK1W

UPDATE:

Jack/W6FB contacted me directly shortly after posting. I sent him my fault
log to analyze and a detailed description of my equipment setup. Best guess
is a cable or antenna failure causing high SWR sometime in the recent past
that stressed the finals. They gave up the ghost when I was running 17m FT8.
I'm leaning toward a bad patch cable I found between the KPA500 and KAT500 a
few months previous to the failure. A cheap cable that became very costly!

His guess was right on, when I was pulling the power transformer out of the
amp to make it easier to ship I found the ceramic part of a VRF2933 MOSFET
rattling around in the cabinet. The problem is now obvious :(

The amp left today to Elecraft for repair. I'll let the group know the
results once it gets there.

Thanks everyone for your constructive comments and to Jack for reaching out
and helping me find the problem.

'73...

Jim

KK1W wrote

> Was running about 150W on 18 MHz this afternoon. My KPA 500 made a loud
> pop. No more output. Any pointers of where to start looking?
>
> *Details:*
> KPA500 (use with KAT500 and Icom 7610)
> Antenna on this band is a SteppIR
> Amp has been running FB
> KPA Utility 1.13.7.16
> 240V supply
> A single loud pop was heard.
> Mode remained in OPER
> No fault light displayed, nor can I find one in the Fault Table
>
> *After the failure:*
>
> *In STBY:*
> Radio and antennas work fine. SWR shown to exciter is low, as expected.
> Exciter power is displayed on KPA500 power meter.
> HV = 78.2
> Works correctly on all bands
>
> *In OPER:*
> HV = 78.2 key up
> On key down (input of 15W) the asterisk briefly displays, then display
> changes to 18.1, back to HV display on key up.
> Zero output, SWR shown to exciter is very high (over 3 to 1)
> No power shown on output meter.
> Fails *on all bands* with high SWR to the exciter.
> No fault indication.
>
> *Other data:*
> Here's the last three entries from the fault table:
> 00000000 00 PWR ON -- 0007 Power was turned on
> 00000039 09 REFL HI 80 00F0 Excessive reflected power (high SWR)
> 00000000 00 PWR ON -- Power was turned on
> .............
>
> Apparently only faults get logged, subsequent power cycles have not
> changed
> the table.
>
> I power cycled the amp, including removing the mains power. - no change.
> I reloaded the last configuration - no change.
> I removed the cover and looked for any visible damage or bad odors.
> Nothing
> stands out.
>
> Amp is about 4 years old and was a kit model. Before I start to
> disassemble
> if anyone can offer a likely culprit I'm all ears.
>
> Thanks,
>
> Jim/KK1W
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:

> Elecraft@.qth

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

> lists+1215531472858-365791@.nabble

--
Sent from: http://elecraft.365791.n2.nabble.com/
______________________________________________________________
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]

Edward R Cole

Re: QRO - KPA500 Failure

In reply to this post by KK1W

I checked the operating specs on two of my QRO amps. Not a KPA500.

2m-1500w PA using a MRF1K50 in class AB:
945w 1398w dc 67% eff 453w disp
1511w 1888w dc 80% eff 377w disp

2m-1200w PA using a MRFE6VP61K25H in class AB:
566w 1200w dc 47% eff 634w disp
1133w 1542w dc 73% eff 409w disp

less heat dissipated when running full power RF ouput.
Both are W6PQL amps.

73, Ed - KL7UW
http://www.kl7uw.com
Dubus-NA Business mail:
[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]

K9MA

Re: QRO - KPA500 Failure

That's very high efficiency for a solid-state amplifier. Are they
operating in class C? Are there any IMD specs at those output levels?

73,
Scott K9MA

On 2/11/2020 03:18, Edward R Cole wrote:

> I checked the operating specs on two of my QRO amps. Not a KPA500.
>
> 2m-1500w PA using a MRF1K50 in class AB:
> 945w 1398w dc 67% eff 453w disp
> 1511w 1888w dc 80% eff 377w disp
>
> 2m-1200w PA using a MRFE6VP61K25H in class AB:
> 566w 1200w dc 47% eff 634w disp
> 1133w 1542w dc 73% eff 409w disp
>
> less heat dissipated when running full power RF ouput.
> Both are W6PQL amps.
>
> 73, Ed - KL7UW
> http://www.kl7uw.com
> Dubus-NA Business mail:
> [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]

Conrad PA5Y

Re: QRO - KPA500 Failure

Hello Ed, I don't think that you are in class AB with 80% efficiency, in fact you cannot be. Class B can theoretically achieve 78.5% but this is seldom achieved in practice, especially on VHF. Most likely there is a problem with instrumentation.

________________________________
From: [hidden email] <[hidden email]> on behalf of K9MA <[hidden email]>
Sent: 11 February 2020 17:11
To: [hidden email] <[hidden email]>
Subject: Re: [Elecraft] QRO - KPA500 Failure

That's very high efficiency for a solid-state amplifier. Are they
operating in class C? Are there any IMD specs at those output levels?

73,
Scott K9MA

On 2/11/2020 03:18, Edward R Cole wrote:

--
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]
______________________________________________________________
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]

Petr, OK1RP/M0SIS

Re: QRO - KPA500 Failure

In reply to this post by KK1W

Hi Jim,

thanks for your KPA500 issue update and returning the thread to its original
point there.
Let us know the results once you gets there.

Ahoj,
Petr

-----
73 - Petr, OK1RP
"Apple & Elecraft freak"
B:http://ok1rp.blogspot.com
MeWe: https://bit.ly/2HGPoDx
MeWe: https://bit.ly/2FmwvDt
--
Sent from: http://elecraft.365791.n2.nabble.com/
______________________________________________________________
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]

73 - Petr, OK1RP
"Apple & Elecraft freak"
B:http://ok1rp.blogspot.com
MeWe: https://bit.ly/2HGPoDx
MeWe: https://bit.ly/2FmwvDt