K3 increasing SWR with increasing Power


dl9hda@...
 

Hello, K3 #827. Checked with dummy load. If power is less than 13 W SWR is perfekt. Increasing the output power results in an increasing of swr. So SWR at 20 is 2 and maximum at 100 W.

Any hints
73,
Holger, DL9HDA


Yngvi Hardarson
 

Hi Holger
What is the power rating of the dummy load?
73 Yngvi TF3Y

On Fri, Sep 24, 2021 at 4:45 PM <dl9hda@...> wrote:
Hello, K3 #827. Checked with dummy load. If power is less than 13 W SWR is perfekt. Increasing the output power results in an increasing of swr. So SWR at 20 is 2 and maximum at 100 W.

Any hints
73,
Holger, DL9HDA




dl9hda@...
 

Hello, it is rated for 1,5 KW. But at a resonant it is same issue.

73, Holger DL9HDA


dl9hda@...
 

Hello, I realized, that only SWR is rising. I have a P3 with connected Power-Sensor and also a KPA500. Both are showing (KPA500 in standy-by of course) 100 W of output power. So the internal pa seems to be ok but swr is rising to maximum. So SWR is max, power is max and not reduced due to high SWR. I'm a little bit irritated ...

73, Holger DL9HDA


Bob McGraw - K4TAX
 

Sounds as though the SWR circuit in the K3 has a defective diode on the REF side.  That would likely be D37 on the RF Board.  Reference schematic E850651

73
Bob, K4TAX


Wes Stewart
 

If the diode is bad the indicated SWR would be 1:1 at all levels.

Wes  N7WS


On Sat, Sep 25, 2021 at 05:48 AM, Bob McGraw - K4TAX wrote:
Sounds as though the SWR circuit in the K3 has a defective diode on the REF side.  That would likely be D37 on the RF Board.  Reference schematic E850651

73
Bob, K4TAX


dl9hda@...
 
Edited

Hello, additional tests.

Additional to my P3 (2 KW sensor) I checked with LP-100A (1.5 KW senor) and a cross needle meter DAIWA CN-101L.

When I increase the power, P3 and LP-100A show a maximum power of !!!! 170 watts !!! The needle is at the absolute limit! All three show an SWR of less than 1.1-1.

1. I have checked all cables and tried with different dummy loads.
2. If the internal ATU is in bypass, the results are identical.
3. There is the same behavior with both antenna connections.
4. Additional to my P3  (2 KW sensor) I checked with LP-100A (1.5 KW senor).
5. The problem occurs during tuning as well as during SSB/CW operation.
6. The problem is independent of the band used. Only the values change then.

Here some measurements at 7165 KHz:

Tune-Power 5 W:
K3: 5.0 W and SWR 1.1-1
LP100A: 7.5 W and SWR 1.02-1

Tune-Power 12 W:
K3: 11.1 W and SWR 1.3-1
LP100A: 18.2 W and SWR 1.02-1

Tune-Power 13 W:
K3: 13 W and SWR 1.6-1
LP100A: 25.7 W and SWR 1.02-1

Tune-Power 20 W:
K3: 20 W and SWR 2.5-1
LP100A: 56.4 W and SWR 1.02-1

Tune-Power 30 W:
K3: 26 W and SWR 5.6-1
LP100A: 137 W and SWR 1.02-1

This can't just be related to the KPA3, I think.

73, Holger DL9HDA


Paul G4KZY
 

Hi Holger,

The measured VSWR is nothing to do with the PA.  A defective PA can't change the VSWR which is a measure of the antenna match.

I'm assuming that you have bypassed your KAT3 if you have one, and that the bypassing is actually working.  If your antenna tuner has, for example, a stuck relay you might have a poor match at the PA output whilst having a dummy load connected the the ATU output.  I'm not sure how you would check this.  Maybe remove your KAT3 as a start point?

Assuming the dummy load is actually connected to the PA output, it would seem that the VSWR detection circuit on your K3 is defective.  It is of a well understood design (called a Tandem Bridge I think) so its operation should be easy to figure out.  T3 measures the antenna current and T4 measures the antenna voltage.  You might find that, for example, there is an open on one of the windings of T3 or T4, or maybe there is a defective C230, which is supposed to provide an RF ground to one side of T4's secondary.  These faults would put the bridge out of balance and an out of balance bridge would cause the kind of trouble you are seeing.

The bridge seems to be reading significantly less power than is actually flowing, as the K3 is pumping more power out than it thinks, which would also suggest a simple issue with the bridge circuit.  If you put a DC voltmeter on the cathodes of D36 and D37 you should be able to measure the raw output from the bridge and see if the problem lies with the bridge or with the following electronics.  If you have an oscilloscope then try looking at the RF voltage appearing on the anodes of D36 and D37.  With a good match you should see a small (RF) voltage on D37 and a larger voltage on D36.

Looking at the schematic it looks like R34 is intended to add a small DC bias to the diodes which would help with measurements at low power levels.  With no power output (ie in Rx mode) you would expect to measure about 0.14V across R44 and R35 (ie on the anodes of D36 and D37).  If you don't measure this voltage, look at the voltage across R34 which ought to be about 12V.  Anything amiss here would suggest that there is a disconnect on the secondary circuit of T3 somewhere.

Finally, I would keep at low power settings because your K3 seems to be significantly overdriving your PA at the moment (assuming your LPA100 meter is accurate) and you could easily damage it.

73,

Paul G4KZY


dl9hda@...
 

Hi Paul, thank you very much.

The voltage of my power-supply is 14.1V. It is a SS30-DV.

Rx-Mode:
Voltage at the anodes of D36 and D37 is 0,127V. At both.
Voltage across R34 is 13,47V

Tune with 5W.
Voltage at the anodes of D36 and D37 is 0,12V. At both.

73, Holger DL9HDA


Paul G4KZY
 

Hi Holger,

The RF voltage when transmitting is on the anodes of the diodes (junctions with R44 and R35).  You need a n oscilloscope to measure here. The small DC voltage on the anodes of the diodes comes via R34 and the secondary of T3. and I think the low voltage here shows that T3 is connected OK. 

The DC voltage when transmitting comes up on the diode cathodes (junctions with R46 and R38).  With a low VSWR you expect the voltage on D36 cathode / R46 to increase (forward power) but the voltage on D37 cathode / R38 to stay close to 0V (reflected power), or at least low compared to D36 cathode voltage.  If you see the voltage on D37 cathode significantly rising when transmitting into a dummy load, try to put a (temporary) short across C230 to see if this improves the situation.

73,  Paul


Mel Farrer, K6KBE <farrermesa@...>
 

Please  remember that SWR is a reflection of all of the energy going to the antenna.  In the past, I have seen incorrect bias on the finals causing an increase in the harmonic content of the signal.  I usually look for harmonic output along with everything else.  Harmonics can really screw up the SWR reading.

Mel, K6KBE

On Tue, Sep 28, 2021 at 9:49 AM Paul G4KZY <paul.g4kzy@...> wrote:
Hi Holger,

The RF voltage when transmitting is on the anodes of the diodes (junctions with R44 and R35).  You need a n oscilloscope to measure here. The small DC voltage on the anodes of the diodes comes via R34 and the secondary of T3. and I think the low voltage here shows that T3 is connected OK. 

The DC voltage when transmitting comes up on the diode cathodes (junctions with R46 and R38).  With a low VSWR you expect the voltage on D36 cathode / R46 to increase (forward power) but the voltage on D37 cathode / R38 to stay close to 0V (reflected power), or at least low compared to D36 cathode voltage.  If you see the voltage on D37 cathode significantly rising when transmitting into a dummy load, try to put a (temporary) short across C230 to see if this improves the situation.

73,  Paul


dl9hda@...
 

Hi Paul, thanks for your tips and information. Possibly I come tomorrow to measure with an oscilloscope.

I have now measured again the diodes (soldered, I know not very well to measure), with the diode measurement function of the Fluke 177. Each are 0.37V.

73, Holger DL9HDA


dl9hda@...
 
Edited

Hi, got my scope.

With 5W tuning power sending into a dummy load I got a nice sinus with 750 mV peak to peak at D36 anode and at D37 anode 82 mV peak to peak. The signal at D37 doesn't look like a sinus at all.

At both cathodes I get mor or less haunted signals with 5 mV pp or so. I think it is not significantly. It is a SDS 1202X-E DSO.

When I bypass the KPA3, the waveforms do not change.

73, Holger DL9HDA


Paul G4KZY
 

Hi Holger,

You are expecting close to zero on D37 and a nice signal on D36 (anodes).

If you increase the power to, say, 20W tuning (that will be 55W on your LP100 according to your earlier measurements), if all was well with your bridge  would I expect you to see 2V on D36 anode.  But I think you might see about 1.5V instead, with the voltage on D37 anode increasing to maybe 0.5V or so.

While you are there, put your scope on C230.  You should not see much RF voltage on either side of C230.

73,

Paul G4KZY


dl9hda@...
 

Hi Paul,

at 20 W my LP-100A now shows 25.5 W. I don't know.

SWR on K3 rises up to 1.4-1. Voltage at anode of D36 is 4.20 V peak to peak.

D37 anode shows 1V peak to peak.

At C230 on the side which shows to the case wall 40 mV and on the other side 16 mV.



73, Holger DL9HDA


Paul G4KZY
 

Hi Holger,

I've more-or-less run out of ideas now. The voltage on D36 anode ought to be proportional to the square root of the power.  If you run 5 watts and get 1V then you should get 2V by running 20W. You seem to be going from 0.75V to 4V which would be about 25x power increase.

There should not be a voltage on D37 anode if you have a 50 ohm load.  And the SWR should not change with power level.  I was wondering if the two resistors R44 and R35 are still 51 ohms each.  I think that if, for example, R35 was higher than 51 ohms you might get this kind of problem.  If you try to measure them you probably will get about 26 ohms because they are DC shorted by the secondary of T4.

Another possibility is that there is something shorting to R35 and injecting a voltage there which ought not to be there.  Especially as you mention it is an ugly kind of waveform.  Perhaps there is some kind of problem with T4?

Or perhaps your connection from the VSWR bridge output to the antenna socket has an intermittent connection or something there like an ATU component which ought not to be connected?  Maybe the VSWR at the bridge is really changing even though the VSWR at the antenna socket is not changing?

Hope you manage to find the problem.

73,

Paul


dl9hda@...
 

Hi Paul, my last measurements for now. I am then on vacation and will only take my K2/100.

Both resistors show 25.3 ohms when measured soldered in place.

Before November I can not continue.

Many greetings and thanks!

73, Holger DL9HDA


Charlie T, k3ICH
 

If driving a resonant antenna, excessive/increasing distortion products, harmonics etc. in the output will cause an apparent increase in SWR since they would be off resonance and result in higher reflected energy even if the antenna were a perfect match at resonance.

 

This of course is a moot point if the load is a purely resistive 50Ω.

 

73, Charlie k3ICH

 

 

 

 

From: Elecraft-K3@groups.io <Elecraft-K3@groups.io> On Behalf Of dl9hda@...
Sent: Thursday, September 30, 2021 9:14 AM
To: Elecraft-K3@groups.io
Subject: Re: [Elecraft-K3] K3 increasing SWR with increasing Power

 

Hi Paul, my last measurements for now. I am then on vacation and will only take my K2/100.

Both resistors show 25.3 ohms when measured soldered in place.

Before November I can not continue.

Many greetings and thanks!

73, Holger DL9HDA


Paul G4KZY
 

OK Holger,

Have a great holiday.

73,

Paul


dl9hda@...
 

Hello, to be really sure once again, I made a few measurements with the oscilloscope. Anode of D36.

Set tune power:
2.5W, 5W and 10W.

And this peak-to-peak voltage I measured:
1.86V, 2.56V and 3.36V.

That should be correct so far - right?

KPA3 and KAT3 were in operation respectively.

73, Holger DL9HDA