Elecraft › [K2]

K2 / KPA ALC problem.

Classic

List

Threaded

20 messages Options

Chris Kimball

K2 / KPA ALC problem.

I'm getting reports that my digital mode Olivia 8/500 signals are distorted. Usually, the comments fall into two categories: 1) overdrive 2) audio compression on. I know that the audio compression is set 1:1 (off). And I've tried to prevent overdrive by the method suggested by Don Wilheim, W3FPR, in his response to my earlier posting of Apr 14, 2013. That is, adjust the input levels so that no bar appears in the K2 ALC bar graph.

But I've had a strange problem with the ALC. When the power level is 10 W or less (and the KPA is in-line, but off) I'm able to set the input levels so that the ALC bar indicator shows no bars AND a signal is output. When I increase the power level to 11 W (KPA is on), the exact same settings yield an unwanted, single ALC bar indication. If I then decrease input levels, the ALC bar disappears, but, alas, there's no signal output.

For power levels greater than 10 W, that is, with the KPA on, I can't get any reliable output unless at least one ALC bar is showing.

Listeners say the less than 10 W signal sounds fine.

As I understand it, the ALC stays entirely within the K2, that is, it doesn't go to the KPA. Control, DC, and RF lines are short and the system is well grounded.

Suggestions would be appreciated.

Chris
NQ8Z

Don Wilhelm-4

Re: K2 / KPA ALC problem.

Chris,

A few things come to mind. Does your KSB2 have the KI6WX increased RF
gain mod installed? If so, you may have to add some additional changes
to the KSB2 to control the ALC. Change R9 to 15k and cut the trace
between P1 pin 6 and Q1 base and insert a 1K resistor (you will have to
mount the resistor vertically).
Even without the KI6WX KSB2 mods, this change may help your situation.

The other thing is the value of R98 on the bottom of the RF Board. The
nominal value is 270 ohms, but the optimum value depends on the gain of
the RF transmit chain. With the KPA100 installed, the value may have to
be increased to maintain optimum ALC action.

There is a K2 app note describing the process to determine the optimum
value of R98 for any particular K2 (it will very from K2 to K2). In
short, temporarily replace R98 with a 2k (or greater) pot, and adjust
the value until the power begins to drop on 10 meters. Remove the pot
and measure its value. Replace R98 with a standard resistor value lower
than the resistance of the pot.

Hopefully one or both of those mods will solve your problem.

73,
Don W3FPR
On 12/2/2013 8:22 PM, Chris Kimball wrote:

> I'm getting reports that my digital mode Olivia 8/500 signals are distorted.
> Usually, the comments fall into two categories: 1) overdrive 2) audio
> compression on. I know that the audio compression is set 1:1 (off). And
> I've tried to prevent overdrive by the method suggested by Don Wilheim,
> W3FPR, in his response to my earlier posting of Apr 14, 2013. That is,
> adjust the input levels so that no bar appears in the K2 ALC bar graph.
>
> But I've had a strange problem with the ALC. When the power level is 10 W or
> less (and the KPA is in-line, but off) I'm able to set the input levels so
> that the ALC bar indicator shows no bars AND a signal is output. When I
> increase the power level to 11 W (KPA is on), the exact same settings yield
> an unwanted, single ALC bar indication. If I then decrease input levels,
> the ALC bar disappears, but, alas, there's no signal output.
>
> For power levels greater than 10 W, that is, with the KPA on, I can't get
> any reliable output unless at least one ALC bar is showing.
>
> Listeners say the less than 10 W signal sounds fine.
>
> As I understand it, the ALC stays entirely within the K2, that is, it
> doesn't go to the KPA. Control, DC, and RF lines are short and the system
> is well grounded.
>
> Suggestions would be appreciated.
>
> Chris
> NQ8Z
>
>
>
>
>
>
> --
> View this message in context: http://elecraft.365791.n2.nabble.com/K2-KPA-ALC-problem-tp7581353.html
> Sent from the Elecraft mailing list archive at 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
>

______________________________________________________________
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

Chris Kimball

Re: K2 / KPA ALC problem.

Don,

My KSB2 is revision E, July 20, 2004, with the "PSK31 ALC time constant change" of KI6WX built-in. It doesn't have the increased RF gain mod. Is there a reference to the mod that you suggest with the R9 -> 15K and a new 1K in the Q1 base?

I was amazed at the instability at the 5W level under tuning to a dummy load. The measured power levels alternate between 4.3 and 6.4 W with no points in between. Seems like too much. Since I don't have a pot in this range, I'm tempted to double R98 and see what happens. (I currently get 6.8 W on 10 meters for a 10 W setting and it doesn't bother me.)

Please let me know about the "1K in Q1 base" mod.

Thanks,

Chris
NQ8X

Don Wilhelm-4

Re: K2 / KPA ALC problem.

Chris,

That actually is 1.5k at R9, sorry for the slipped decimal point.
I believe you already have the information sent to you. For others who
are interested, it may be found near the end of the mod document found
at http://www.qrpproject.de/Media/pdf/SSBModsUK.pdf.

73,
Don W3FPR

On 12/3/2013 12:14 PM, Chris Kimball wrote:

>
> Don,
>
> My KSB2 is revision E, July 20, 2004, with the "PSK31 ALC time constant
> change" of KI6WX built-in. It doesn't have the increased RF gain mod. Is
> there a reference to the mod that you suggest with the R9 -> 15K and a new
> 1K in the Q1 base?
>
> I was amazed at the instability at the 5W level under tuning to a dummy
> load. The measured power levels alternate between 4.3 and 6.4 W with no
> points in between. Seems like too much. Since I don't have a pot in this
> range, I'm tempted to double R98 and see what happens. (I currently get 6.8
> W on 10 meters for a 10 W setting and it doesn't bother me.)
>
> Please let me know about the "1K in Q1 base" mod.
>
> Thanks,
>
> Chris
> NQ8X
>
>
>
> --
> View this message in context: http://elecraft.365791.n2.nabble.com/K2-KPA-ALC-problem-tp7581353p7581385.html
> Sent from the Elecraft mailing list archive at 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
>

Don Wilhelm-4

Re: K2 / KPA ALC problem.

In reply to this post by Chris Kimball

Chris,

Is there any chance you have T4 wound for the "better efficiency at 5
watts" ratio of 2:2:1:1 turns?
If so, be aware that this winding ratio is *only* for a QRP CW only K2.
If the KSB2 is added, it must be changed to the standard 2:3:1:1 ratio.
With the 'better efficiency at 5 watts" mod applied, the efficiency
above the 5 watt level becomes very poor, and may generate HiCur
messages. It should never be used if the KPA100 is added.

73,
Don W3FPR

On 12/3/2013 12:14 PM, Chris Kimball wrote:

Chris Kimball

Re: K2 / KPA ALC problem.

I bought the K2 (SN 4913) assembled. I don't see any special notes in the T4 section of the manual, but I'll check later today when the K2 is disassembled.

On-the-air checks with K3CC and VE3FMC show that the signal is good at 10W (KPA not active) and poor at 11 W (KPA on). SSB operation AT THE SAME GAIN SETTINGS is fine.

???
Would an impedance mismatch between the K2 and the KPA (when on) change the ALC response?
???

Thanks,

Chris
NQ8Z

Chris Kimball

Re: K2 / KPA ALC problem. Are power levels sensed at the same point?

Critical question: When I adjust the power level of the K2/KPA combination between low (10W) and high (100W) power is the power measurement taken at the same point for both levels?

Clearly, the RF level scale changes between the two settings. 10W is full scale below 10W, 100W is full scale above 10W. When operating below 10W is the power level measurement taken from the K2 or from the KPA?

Assuming the ALC level depends on the power level measurement, a difference in the point at which the power level is measured would explain the ALC changes around the 10W level. My K2 and KPA were built and aligned by different builders and there's no reason to believe they were set exactly the same.

Don Wilhelm-4

Re: K2 / KPA ALC problem. Are power levels sensed at the same point?

Chris,

The power should be measured by the wattmeter in the KPA100 in both
cases. Note that you must use the KPA100 SO-239 jack (not the base K2
BNC antenna jack) anytime the KPA100 is installed.

The KPA100 contains a scaling circuit for the wattmeter (Q10 and Q11
with resistors R24 and R25) to change the wattmeter range between low
power and high power.
Other than the range scaling, there is no difference in the power
measurement.
To do that correctly, the wattmeter must be balanced (nulled) and then
calibrated with pot R26 for forward power and R27 for reverse power
(SWR). Use a good dummy load for that calibration and an external
wattmeter that you can trust for accuracy.

The ALC and the power level are not directly related, but the ALC does
depend on the RF Gain through the transmit stages, and the KPA100 adds
additional gain.

You might want to check the KPA100 bias adjustment before proceeding
with anything else. It sounds to me like your KPA100 may have too much
bias on the PA transistors which can drive them into distortion and
increase the gain of the KPA100.

The instructions for setting the bias as well as balancing and
calibrating the wattmeter can be found in the manual. When setting the
wattmeter balance, it is important to use a non-metallic screwdriver
because both ends of that capacitor are above ground and have some small
bit of RF on the cap. Using a metallic screwdriver will give you an
erroneous indication.

To make those adjustments, lay the KPA100 'on its back' to the right of
the K2 - place it on a book or other support about half the height of
the K2. You will have to remove the shield - remove only the zinc
screws, not the black ones that hold the shield clips.

73,
Don W3FPR

On 12/4/2013 3:54 PM, Chris Kimball wrote:

> Critical question: When I adjust the power level of the K2/KPA combination
> between low (10W) and high (100W) power is the power measurement taken at
> the same point for both levels?
>
> Clearly, the RF level scale changes between the two settings. 10W is full
> scale below 10W, 100W is full scale above 10W. When operating below 10W is
> the power level measurement taken from the K2 or from the KPA?
>
> Assuming the ALC level depends on the power level measurement, a difference
> in the point at which the power level is measured would explain the ALC
> changes around the 10W level. My K2 and KPA were built and aligned by
> different builders and there's no reason to believe they were set exactly
> the same.
>
>

Chris Kimball

Re: K2 / KPA ALC problem. Are power levels sensed at the same point?

Thanks for the reply.

The setup is a K2 with a KAT2 tuner feeding a KPA-100/KAT-100 in a separate enclosure. The connection is from Ant 1 of the K2 KAT2 to the BNC input of the KAT-100. The internal wiring of the KPA-100/KAT-100 routes this to the amplifier input in the usual way for this conglomeration("Remoting the KPA100 into the EC2 enclosure", kk7p.com).

The KPA-100 is fully up-to-date as of Rev A-2 Nov 13, 2006. The upgrades were made in March, 2012, and I remember that the bias setting went as planned. Since I don't have a wattmeter, no adjustment of the wattmeter balance or calibration was made with the upgrade. Since the upgrade involved a lot of component changes, this could be the problem!

Would an incorrectly set KPA-100 power meter lead to the observed differences in ALC response between low and high power settings?

Thanks,

Chris

Rob KK4R

Re: K2 / KPA ALC problem. Are power levels sensed at the same point?

Chris,

I have been monitoring this thread, since I have had some similar problems with my K2. My recent experience may help you.

In the K2, amplifier gain is in the ALC loop. When you switch from low power to high (go from less than 10W to greater than 10W), the KPA100 gain changes the ALC behavior and may be the source of many of the differences you see between low and high power operation. If the KPA100 needs some adjustment, that could also be a factor. In my case both calibration/alignment and gain were causing problems.

I recalibrated the KPA100 first and noticed some improvements immediately in output power on the high bands and in the agreement between set power and output power. Note that I did this even though I believed that I had taken care in the original calibration. Maybe things change over time, or perhaps I did a better job this time, but the effort was rewarded.

Once the recalibration was done, I followed Don's suggestion and adjusted R98 to reduce the ALC loop gain. Prior to adjusting R98, my K2 would hunt endlessly at some power levels, jumping up and down by significant amounts and never converging to the set power level. At other power levels, it might settle down at a value that wasn't as close to the set value as you might like. After adjusting R98, the K2 always converges to the set value with a relatively small error, because the ALC steps are smaller.

You mentioned that you don't have a power meter to use for adjustments to the KPA100. I have two, but after a comparison, decided instead to use a voltage measurement across a good dummy load to calculate power. This method seemed to give me the most reliable result across the range of interest. If you have a good dummy load, you can use a scope to measure peak-to-peak voltage or use a diode detector to get a DC voltage off the dummy load. The Elecraft 20W load is very good and has a built in provision to take a DC voltage measurement for conversion to power. With that taken care of, you can adjust your amp with little trouble.

Rob

Chris Kimball

Re: K2 / KPA ALC problem. Are power levels sensed at the same point?

Rob,

Thanks for your helpful comments.

My setup has a KAT-100 in it and the KAT-100 is, apparently, is the source of the power level measurement, not the KPA-100. I've got the Elecraft dummy load and your suggestion was great.

At this point it looks like the R98 replacement is the way to go. I'm getting a bipolar distribution of output power under "Tune" at a power set level of 5w, with a display of either 4.4 or 6.4 W and nothing in between.

Chris

Don Wilhelm-4

Re: K2 / KPA ALC problem. Are power levels sensed at the same point?

Chris,

If you have both the KPA100 and the KAT100 in the system, the power
output is measured with the KPA100 wattmeter.
If the KAT100 is used on a K2/10 than the KAT100 wattmeter is used for
power output measurement.
Take both the KPA100 and KAT100 away and the base K2 uses a simple diode
detector to measure power. That detector will only be accurate when
working into a 50 ohm non-inductive load (dummy load or antenna with
some properly tuned antenna tuner.

Try to break the problem down into pieces to be certain what is
happening. First thing is to get the wattmeter in the KPA100 properly
calibrated (no KAT100 at that point),
Then work on the value of R98 to stabilize the power control loop. Once
that is accomplished, then look at the ALC action - if you still have
excessive ALC on SSB with the KPA100 in-circuit (11 watts or above),
then look to the ALC changes for the KSB2 board.

Attempting to do things in any other order may lead to erroneous
conclusions.

73,
Don W3FPR

On 12/6/2013 1:36 PM, Chris Kimball wrote:

> Rob,
>
> Thanks for your helpful comments.
>
> My setup has a KAT-100 in it and the KAT-100 is, apparently, is the source
> of the power level measurement, not the KPA-100. I've got the Elecraft
> dummy load and your suggestion was great.
>
> At this point it looks like the R98 replacement is the way to go. I'm
> getting a bipolar distribution of output power under "Tune" at a power set
> level of 5w, with a display of either 4.4 or 6.4 W and nothing in between.
>
>

Chris Kimball

Re: K2 / KPA ALC problem. Progress, but ...

Some progress..

1. The BFO setting for the R bar mode was set incorrectly, resulting in a dsb-like transmission. The 't' in the filter settings really does mean that the FL1 BFO is used for transmitting.

2. The power levels were within 1 watt at 20 or 30 watts via a DL2 measurement scheme. No change made.

3. R98 was adjusted going up to 1400 ohm before 10 meter levels reduced a bit. R98 was set to 1100 ohms. The variance at 7.073 MHz for a 5 W power setting was improved.

But ..

K2 at 5 W and KAT-100 perform nicely in digital modes. With the KPA and a power level of 40W, with the KAT into a dummy load the rig works nicely, too. With the antenna tuned by the KAT automatically to 1:1 SWR, the ALC level display changes to 1 bar and the RF output level shows more motion. The signal, by reports, still distorted.

During transmission (Olivia 8/500) I can hear a weak version of the transmitted signal on the speaker or headphones.

Suppose the KAT isn't really matching the antenna and the 1:1 SWR reading is false. Would that cause the ALC/RF output effects?

Chris Kimball

Re: K2 / KPA ALC problem. Disconnect VRFDET?

The KAT-100 manual, page 33, suggests disconnecting the VRFDET line at the K2 in the event of "Erratic power control with K2/100".

Since may KPA/KAT are in the same box, it's unlikely I'll use the KAT without the KPA.

Does this seem reasonable?

Thanks,

Chris
NQ8Z

Don Wilhelm-4

Re: K2 / KPA ALC problem. Disconnect VRFDET?

Chris,

That is *not* in the newer manual that I have, and it is *not*
reasonable to do disconnect VRFDET.
Secondly, removing the VRFDET signal from the K2 MCU will cause the K2
(and KPA100) to go to full output at all times - sometimes in excess of
160 watts. If that continues, the base K2 PA transistors will be
stressed and will fail in a short time.

Bottom line - don't disconnect VRFDET.

It may be that your KPA100 wattmeter is not properly calibrated. Put an
external wattmeter in the coax link between the KPA100 and the KAT100
and see if the wattmeter closely follows the power requested setting on
the K2. If it goes to full power, then the wattmeter diodes may be
damaged. OTOH if the power is quite different (more than 15%) than the
requested power, you will need to calibrate the KPA100 wattmeter.

73,
Don W3FPR

On 12/13/2013 12:55 PM, Chris Kimball wrote:
> The KAT-100 manual, page 33, suggests disconnecting the VRFDET line at the
> K2 in the event of "Erratic power control with K2/100".
>
> Since may KPA/KAT are in the same box, it's unlikely I'll use the KAT
> without the KPA.
>
> Does this seem reasonable?
>
>

______________________________________________________________
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

Chris Kimball

Re: K2 / KPA ALC problem. Ancient manual, recent suggests bypass C.

The manual I was looking at was KAT100 Revision A, Nov 5, 2002. (A vintage manual!)

The current manual at the Elecraft site is KAT100 is Revision C, March 12, 2003, which on page 34 suggests putting "a 0.01 uF capacitor from pin 7 to ground (inside the backshell)." for "Erratic power control with K2/100".

Should I try this?

Thanks,

Chris
NQ8Z

Don Wilhelm-4

Re: K2 / KPA ALC problem. Ancient manual, recent suggests bypass C.

Chris,

There is no harm in that at all.
If you have a very long cable between your EC2 enclosure and the K2, you
may try shortening it.

I suggest you check the actual power output vs. the requested power
setting before attempting anything else.

73,
Don W3FPR

On 12/13/2013 2:52 PM, Chris Kimball wrote:

> The manual I was looking at was KAT100 Revision A, Nov 5, 2002. (A vintage
> manual!)
>
> The current manual at the Elecraft site is KAT100 is Revision C, March 12,
> 2003, which on page 34 suggests putting "a 0.01 uF capacitor from pin 7 to
> ground (inside the backshell)." for "Erratic power control with K2/100".
>
> Should I try this?
>
> Thanks,
>
> Chris
> NQ8Z
>

Chris Kimball

Re: K2 / KPA ALC problem. Ancient manual, recent suggests bypass C.

OK, I'll give it try.

I've done that, using the DL2 and a DC voltmeter. Given the max is 30W, how much of a difference between requested and measured power can there be (at 30W)?

Thanks,

Chris
NQ8Z

Chris Kimball

Re: K2 / KPA ALC problem. * Solved *

I added 19.5 ft of RG-58 coax formed into a 6 inch diameter coil between the rig and the LDG 1:1 balun. Meter results on ALC or RF for the dummy load and for the antenna are now the same.

Steps to the solution:

1. Get the transmitting BFO aligned with the built-in SSB filter in R and R-reversed modes.

2. Fix the R98 gain problems with the standard mod .

3. Added 0.01 uf capacitor on the VRFDET line in the shell of the connector between the K2 and the KPA/KAT.

4. Changed the RF path length from the K2 to the balun, providing a choke balun at the same time.

Thanks to all that assisted in this venture!

Chris
NQ8Z

Don Wilhelm-4

Re: K2 / KPA ALC problem. * Solved *

Chris,

I am glad you have things restored to sanity.

73,
Don W3FPR

On 12/14/2013 8:45 PM, Chris Kimball wrote:
> I added 19.5 ft of RG-58 coax formed into a 6 inch diameter coil between the
> rig and the LDG 1:1 balun. Meter results on ALC or RF for the dummy load
> and for the antenna are now the same.
>
>

______________________________________________________________
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

K2 / KPA ALC problem.

K2 / KPA ALC problem.

Re: K2 / KPA ALC problem.

Re: K2 / KPA ALC problem.

Re: K2 / KPA ALC problem.

Re: K2 / KPA ALC problem.

Re: K2 / KPA ALC problem.

Re: K2 / KPA ALC problem. Are power levels sensed at the same point?

Re: K2 / KPA ALC problem. Are power levels sensed at the same point?

Re: K2 / KPA ALC problem. Are power levels sensed at the same point?

Re: K2 / KPA ALC problem. Are power levels sensed at the same point?

Re: K2 / KPA ALC problem. Are power levels sensed at the same point?

Re: K2 / KPA ALC problem. Are power levels sensed at the same point?

Re: K2 / KPA ALC problem. Progress, but ...

Re: K2 / KPA ALC problem. Disconnect VRFDET?

Re: K2 / KPA ALC problem. Disconnect VRFDET?

Re: K2 / KPA ALC problem. Ancient manual, recent suggests bypass C.

Re: K2 / KPA ALC problem. Ancient manual, recent suggests bypass C.

Re: K2 / KPA ALC problem. Ancient manual, recent suggests bypass C.

Re: K2 / KPA ALC problem. *** Solved ***

Re: K2 / KPA ALC problem. *** Solved ***

Re: K2 / KPA ALC problem. * Solved *

Re: K2 / KPA ALC problem. * Solved *