Topics

QCX tuning

Matthias Zwoch
 

I used my 20m QCX until yet on my station antenna without any problems for numerous QSOs among them one to PY1AX with over 9900 km. Yesterday I had to prepare the QCX for a fieldday next week. I erected my outdoor inverted-V in my garden and hat a look at it with the antenna-analyzer. Yes, at 14030 SWR was 4,9 but a 1,1 minimum was 300 kHz nearby. No problem for my tiny elecraft T1 tuner. But after few seconds of tuning the display got dark. The fuse was blown and the finals including MPS751 were gone. The loss of some parts and the fix is not a big problem. But my question is -how to tune the QCX if it cannot resist few seconds of automatic tuning?

73 Matthias DD7NT

Alan de G1FXB
 

Hi Matthias,

Sad to hear of the failure,

1/.    out of interest what supply voltage and fuse value & was it (fast blow or time lag type) did you have?

It's strange the number of QSX PA failures,
yet Hans undertook rigorous testing and never encountered a failure?
A recent comment was  https://groups.io/g/QRPLabs/message/26421

2/.    What version PCB have you and any modifications made?

3/.    Are all the BS170's that failed the ones originally supplied or have you had reason to replace one or more before?

Alan

On 01/09/2018 18:04, Matthias Zwoch wrote:
I used my 20m QCX until yet on my station antenna without any problems for numerous QSOs among them one to PY1AX with over 9900 km. Yesterday I had to prepare the QCX for a fieldday next week. I erected my outdoor inverted-V in my garden and hat a look at it with the antenna-analyzer. Yes, at 14030 SWR was 4,9 but a 1,1 minimum was 300 kHz nearby. No problem for my tiny elecraft T1 tuner. But after few seconds of tuning the display got dark. The fuse was blown and the finals including MPS751 were gone. The loss of some parts and the fix is not a big problem. But my question is -how to tune the QCX if it cannot resist few seconds of automatic tuning?

73  DD7NT

Richard Harris
 

Hi Matthias,

When tuning, the T1 auto-ATU may present an impedance to the QCX that causes excessive current to flow through the PA FETs or causes an excessive voltage across them  One solution is to connect an attenuator between your QCX and T1 auto-ATU when it is tuning and disconnect it when a match has been found.   This will limit the range of impedance "seen" by the PA FETs.  A 3dB attenuator will ensure the QCX "sees" a load between about 17 and 150 Ohms.  A 6dB attenuator will ensure that the load is always between about 40 and 85 Ohms.  

When I built an auto-ATU for my 20m QCX, I thought that damage to the PA FETs was possible and so while the ATU is tuning an attenuator is switched into circuit by means of a relay.  

Richard G3OTK


On Saturday, 1 September 2018, 18:04:11 BST, Matthias Zwoch <zwoch@...> wrote:


I used my 20m QCX until yet on my station antenna without any problems
for numerous QSOs among them one to PY1AX with over 9900 km. Yesterday I
had to prepare the QCX for a fieldday next week. I erected my outdoor
inverted-V in my garden and hat a look at it with the antenna-analyzer.
Yes, at 14030 SWR was 4,9 but a 1,1 minimum was 300 kHz nearby. No
problem for my tiny elecraft T1 tuner. But after few seconds of tuning
the display got dark. The fuse was blown and the finals including MPS751
were gone. The loss of some parts and the fix is not a big problem. But
my question is -how to tune the QCX if it cannot resist few seconds of
automatic tuning?

73 Matthias DD7NT



Allen Poland
 

Interesting!  With a 3db attenuator installed in a rig that is putting out less than 5w,  the power output drops 50%.  And that is enough power to trigger an auto-tuner? 

 

From: QRPLabs@groups.io [mailto:QRPLabs@groups.io] On Behalf Of Richard Harris via Groups.Io
Sent: Saturday, September 1, 2018 2:07 PM
To: QRPLabs@groups.io
Subject: Re: [QRPLabs] QCX tuning

 

Hi Matthias,

 

When tuning, the T1 auto-ATU may present an impedance to the QCX that causes excessive current to flow through the PA FETs or causes an excessive voltage across them  One solution is to connect an attenuator between your QCX and T1 auto-ATU when it is tuning and disconnect it when a match has been found.   This will limit the range of impedance "seen" by the PA FETs.  A 3dB attenuator will ensure the QCX "sees" a load between about 17 and 150 Ohms.  A 6dB attenuator will ensure that the load is always between about 40 and 85 Ohms.  

 

When I built an auto-ATU for my 20m QCX, I thought that damage to the PA FETs was possible and so while the ATU is tuning an attenuator is switched into circuit by means of a relay.  

 

Richard G3OTK

 

 

On Saturday, 1 September 2018, 18:04:11 BST, Matthias Zwoch <zwoch@...> wrote:

 

 

I used my 20m QCX until yet on my station antenna without any problems
for numerous QSOs among them one to PY1AX with over 9900 km. Yesterday I
had to prepare the QCX for a fieldday next week. I erected my outdoor
inverted-V in my garden and hat a look at it with the antenna-analyzer.
Yes, at 14030 SWR was 4,9 but a 1,1 minimum was 300 kHz nearby. No
problem for my tiny elecraft T1 tuner. But after few seconds of tuning
the display got dark. The fuse was blown and the finals including MPS751
were gone. The loss of some parts and the fix is not a big problem. But
my question is -how to tune the QCX if it cannot resist few seconds of
automatic tuning?

73 Matthias DD7NT





Avast logo

This email has been checked for viruses by Avast antivirus software.
www.avast.com


Matthias Zwoch
 

Am 01.09.2018 um 19:40 schrieb Alan de G1FXB via Groups.Io:
Hi Matthias,
Sad to hear of the failure,
1/.    out of interest what supply voltage and fuse value & was it (fast blow or time lag type) did you have?
It was a 1A fuse fast type. Supply voltage was at 15V via stepup converter fed from a 12V Pb akku.
It's strange the number of QSX PA failures,
yet Hans undertook rigorous testing and never encountered a failure?
A recent comment was  https://groups.io/g/QRPLabs/message/26421
2/.    What version PCB have you and any modifications made?
I assume its a PCB 3, it came already with the MPS751. Not much modified. Just the modifications from Hans' website and I set the display separately in the case. Have a stepup regulator for the PA and an extra 5V regulator for display and backlight.


3/.    Are all the BS170's that failed the ones originally supplied or have you had reason to replace one or more before?
All part are originally supplied. No reason to change something until yet...

73 Matthias, DD7NT

Charles W. Powell
 

The T1 is reported to tune with as little as 0.5 watts, and 2-5 watts recommended. Mine will tune even with repeated dashes.  I haven’t run into problems with the QCX-20, but I am using an antenna that is maybe 2:1 at its worst.  I do know that the radio has a slight mismatch from 50 ohms.  Into a dummy load, the radio puts out 3 watts, but with a manual tuner, I can tweak that to 5 watts with a low SWR.  My QCX-40 puts out 4 watts into a 50 ohm dummy load but 5 watts through the tuner.

72,

Charles - NK8O

On Sep 1, 2018, at 2:21 PM, Allen Poland <atpoland@...> wrote:

Interesting!  With a 3db attenuator installed in a rig that is putting out less than 5w,  the power output drops 50%.  And that is enough power to trigger an auto-tuner?  
 
From: QRPLabs@groups.io [mailto:QRPLabs@groups.io] On Behalf Of Richard Harris via Groups.Io
Sent: Saturday, September 1, 2018 2:07 PM
To: QRPLabs@groups.io
Subject: Re: [QRPLabs] QCX tuning
 
Hi Matthias,
 
When tuning, the T1 auto-ATU may present an impedance to the QCX that causes excessive current to flow through the PA FETs or causes an excessive voltage across them  One solution is to connect an attenuator between your QCX and T1 auto-ATU when it is tuning and disconnect it when a match has been found.   This will limit the range of impedance "seen" by the PA FETs.  A 3dB attenuator will ensure the QCX "sees" a load between about 17 and 150 Ohms.  A 6dB attenuator will ensure that the load is always between about 40 and 85 Ohms.  
 
When I built an auto-ATU for my 20m QCX, I thought that damage to the PA FETs was possible and so while the ATU is tuning an attenuator is switched into circuit by means of a relay.  
 
Richard G3OTK
 

Alan G4ZFQ
 

It was a 1A fuse fast type. Supply voltage was at 15V via stepup converter fed from a 12V Pb akku.
That might confirm the saying that semiconductors blow fast in order to protect fuses from blowing...

73 Alan G4ZFQ

Matthias Zwoch
 

Am 01.09.2018 um 20:07 schrieb Richard Harris via Groups.Io:
Hi Matthias,
When tuning, the T1 auto-ATU may present an impedance to the QCX that causes excessive current to flow through the PA FETs or causes an excessive voltage across them  One solution is to connect an attenuator between your QCX and T1 auto-ATU when it is tuning and disconnect it when a match has been found.   This will limit the range of impedance "seen" by the PA FETs.  A 3dB attenuator will ensure the QCX "sees" a load between about 17 and 150 Ohms.  A 6dB attenuator will ensure that the load is always between about 40 and 85 Ohms.
When I built an auto-ATU for my 20m QCX, I thought that damage to the PA FETs was possible and so while the ATU is tuning an attenuator is switched into circuit by means of a relay.
Richard G3OTK
Hi Richard,

Thats a good idea. T1 should do it too with 3dB att - 2,5W. Using a 6db attenuator may cause the problem, that the tuner stops at a level what later is not good for the finals. I have a homemade Z-Match which does the same in the tune mode. But dont like the fiddling with the Z-Match.
At my 40m QRP trx (Mrs. Mosquita) I had blown a driver transistor years ago also using the autotuner. The autotuner brought the driver stage to oscillate, the output got at abt 30W for a sceond (made for 5w) and the fuse was blown out. However the RD16HHF1 survived and only one transistor had to be changed. Made a modification in the driver stage and now I can tune Mrs. Mosquita even with a long tuning sequence with the T1.

73 Matthias, DD7NT

Charles W. Powell
 

So, in the midst of this discussion, if my finals blow is there a “drop in” replacement with similar characteristics but that will withstand considerably more abuse?

72,

Charles - NK8O

On Sep 1, 2018, at 2:56 PM, Matthias Zwoch <zwoch@...> wrote:

Am 01.09.2018 um 20:07 schrieb Richard Harris via Groups.Io:
Hi Matthias,
When tuning, the T1 auto-ATU may present an impedance to the QCX that causes excessive current to flow through the PA FETs or causes an excessive voltage across them One solution is to connect an attenuator between your QCX and T1 auto-ATU when it is tuning and disconnect it when a match has been found. This will limit the range of impedance "seen" by the PA FETs. A 3dB attenuator will ensure the QCX "sees" a load between about 17 and 150 Ohms. A 6dB attenuator will ensure that the load is always between about 40 and 85 Ohms.
When I built an auto-ATU for my 20m QCX, I thought that damage to the PA FETs was possible and so while the ATU is tuning an attenuator is switched into circuit by means of a relay.
Richard G3OTK

ajparent1/KB1GMX
 

>>>So, in the midst of this discussion, if my finals blow is there a “drop in” replacement with similar characteristics but that will withstand considerably more abuse?<<<

No, a more robust device will reequire a stronger driver device and likely a 
revised DC feed.   The class E has fairly narrow bounds before it gets to
a bad state.  

To protect it use a switchable attenuator before the tuner 3 or 6DB is
not all that much but easily leaves enough power for a T1 to auto tune.
The T1 tunes well even at 1W. 

Or use a manual tuner with a resistive bridge SWR detector (tayloe).  

Better suggestion, tune the antenna to not need the tuner.  IF a tuner is
needed you not going from total mis-match to 1:1

All tuners will go through a state of high SWR to the radio, its unavoidable.
Try it by hand and see.
.
Allison

Alan de G1FXB
 

Never a truer word said in jest,
transistors being the fastest fuses on three legs.....

Q1-3 & Q6 blown while protected by a F1A fuse,
It does put things in perspective what a task it appears to be?
It would be good to get to the bottom of
what brings on the perfect storm while Hans testing didn't highlight anything.

Alan

On 01/09/2018 19:54, Alan G4ZFQ wrote:
It was a 1A fuse fast type. Supply voltage was at 15V via stepup converter fed from a 12V Pb akku.
That might confirm the saying that semiconductors blow fast in order to protect fuses from blowing...

73 Alan G4ZFQ

Lee
 

The LDG Z-100Plus requires only 0.1 watt to tune according to LDG and is rated for 125 watts which leads me to think that it has lower losses than some of the "QRP" tuners with much lower ratings.  Dunno for sure about that.  It uses latching relays so power consumption is almost zero once tuned. 
 
No relation to the company...just a satisfied user.
 
73,
 
-Lee-    WA3FIY
 

------ Original Message ------
From: "Richard Harris via Groups.Io" <g3otk@...>
Sent: 9/1/2018 2:07:18 PM
Subject: Re: [QRPLabs] QCX tuning
 
Hi Matthias,

When tuning, the T1 auto-ATU may present an impedance to the QCX that causes excessive current to flow through the PA FETs or causes an excessive voltage across them  One solution is to connect an attenuator between your QCX and T1 auto-ATU when it is tuning and disconnect it when a match has been found.   This will limit the range of impedance "seen" by the PA FETs.  A 3dB attenuator will ensure the QCX "sees" a load between about 17 and 150 Ohms.  A 6dB attenuator will ensure that the load is always between about 40 and 85 Ohms.  

When I built an auto-ATU for my 20m QCX, I thought that damage to the PA FETs was possible and so while the ATU is tuning an attenuator is switched into circuit by means of a relay.  

Richard G3OTK

Alan G4ZFQ
 

It would be good to get to the bottom of
what brings on the perfect storm while Hans testing didn't highlight anything.
Alan,

I don't know how Hans had no trouble although it could be that most QCX do survive abuse. Maybe only a few actually have problems?
The design here http://www.norcalqrp.org/files/NC2030/NC2030_v5.pdf was mentioned by someone, it shows how to protect 3 little BS170s from some class E hazards. Difficult to spot the actual PA components!

73 Alan G4ZFQ

Glen Leinweber
 

In defense of Hans,
I don't recall seeing a brag about QCX able to handle infinite-SWR conditions. (ie: antenna short or open-circuit).

 I can see a dangerous scenario....
if you're using an open-circuit type outdoor antenna - one that measures open-circuit resistance between driven
element and ground. An extreme example might be a kite-elevated long wire. This type of antenna can attain
a considerable charge, and rise to very large DC voltage levels.
Then you attach this antenna to your radio. Could even be an un-powered radio....
Ka-blooy!
The QCX has an output filter that nearly floats as far as DC is concerned. It has a path through R43, a 120k resistor
to one of the four 74ACT00 logic gates - most likely in a logic "high" state. If you measure the DC voltage at the
antenna connection terminals with a high-Z DC voltmeter, you'll measure close to +5V.
I would ensure that any antenna has a DC path to ground, either through a RF choke, or a resistor, to bleed off any
static charge. The QCX's 120k resistor path is too feeble to serve as a reliable discharge. A charged antenna can
blow the finals nearly instantly when connected. There is no over-voltage antenna protection in the QCX.

A loop-type antenna is safer from this scenario, but I'd still include a discharge path from the loop to earth.

Alan de G1FXB
 

Hi Glenn,

I'm in defence of all Hans views as well.  https://groups.io/g/QRPLabs/message/26421
(I keep pointing out his statement that he has never experienced a failure of any prototypes and beyond to production testing.
Last was my Post https://groups.io/g/QRPLabs/message/26498)

What I'm trying to understand is the conditions that lead upto what appears to be a one or two a week report of PA failure,
Is it always user abuse, and / or is there any easy solutions ?.
My posts:-
https://groups.io/g/QRPLabs/message/26420
 https://groups.io/g/QRPLabs/message/25533
https://groups.io/g/QRPLabs/message/25822

Putting it into perspective the QCX has celebrated it's first birthday
Best case, its perhaps 52 out of 5500 units built, equates to just over 1% failure(?)
of the most heavily stressed and abused components of any transmitter.
(it's probably fair to assume that failure is likely to be under rather than over reported,
many would just replace the components rather than post a question about it here?)

It's academic, we will move on and soon have a new unit to talk about.  :-)


Alan


On 02/09/2018 16:44, Glen Leinweber wrote:
In defense of Hans,
I don't recall seeing a brag about QCX able to handle infinite-SWR conditions. (ie: antenna short or open-circuit).

 I can see a dangerous scenario....
if you're using an open-circuit type outdoor antenna - one that measures open-circuit resistance between driven
element and ground. An extreme example might be a kite-elevated long wire. This type of antenna can attain
a considerable charge, and rise to very large DC voltage levels.
Then you attach this antenna to your radio. Could even be an un-powered radio....
Ka-blooy!
The QCX has an output filter that nearly floats as far as DC is concerned. It has a path through R43, a 120k resistor
to one of the four 74ACT00 logic gates - most likely in a logic "high" state. If you measure the DC voltage at the
antenna connection terminals with a high-Z DC voltmeter, you'll measure close to +5V.
I would ensure that any antenna has a DC path to ground, either through a RF choke, or a resistor, to bleed off any
static charge. The QCX's 120k resistor path is too feeble to serve as a reliable discharge. A charged antenna can
blow the finals nearly instantly when connected. There is no over-voltage antenna protection in the QCX.

A loop-type antenna is safer from this scenario, but I'd still include a discharge path from the loop to earth.

jmh6@...
 

Hi Alan,

Academic??

ALL failures are an opportunity to learn and make a product more robust.

I am surprised no one has talked about bringing up the power in steps? Measuring things as we go up to full power.

From where I look, this would give up real data to look at in the event of failure.

Lots of fun :).

John Concord, NH

On Sun, 2 Sep 2018, Alan de G1FXB via Groups.Io wrote:

Hi Glenn,
I'm in defence of all Hans views as well. 
https://groups.io/g/QRPLabs/message/26421
(I keep pointing out his statement that he has never experienced a failure
of any prototypes and beyond to production testing.
Last was my Post https://groups.io/g/QRPLabs/message/26498)
What I'm trying to understand is the conditions that lead upto what appears
to be a one or two a week report of PA failure,
Is it always user abuse, and / or is there any easy solutions ?.
My posts:-
https://groups.io/g/QRPLabs/message/26420
 https://groups.io/g/QRPLabs/message/25533
https://groups.io/g/QRPLabs/message/25822
Putting it into perspective the QCX has celebrated it's first birthday
Best case, its perhaps 52 out of 5500 units built, equates to just over 1%
failure(?)
of the most heavily stressed and abused components of any transmitter.
(it's probably fair to assume that failure is likely to be under rather than
over reported,
many would just replace the components rather than post a question about it
here?)
It's academic, we will move on and soon have a new unit to talk about.  :-)
Alan
On 02/09/2018 16:44, Glen Leinweber wrote:
In defense of Hans,
I don't recall seeing a brag about QCX able to handle
infinite-SWR conditions. (ie: antenna short or open-circuit).

 I can see a dangerous scenario....
if you're using an open-circuit type outdoor antenna - one that
measures open-circuit resistance between driven
element and ground. An extreme example might be a kite-elevated
long wire. This type of antenna can attain
a considerable charge, and rise to very large DC voltage levels.
Then you attach this antenna to your radio. Could even be an
un-powered radio....
Ka-blooy!
The QCX has an output filter that nearly floats as far as DC is
concerned. It has a path through R43, a 120k resistor
to one of the four 74ACT00 logic gates - most likely in a logic
"high" state. If you measure the DC voltage at the
antenna connection terminals with a high-Z DC voltmeter, you'll
measure close to +5V.
I would ensure that any antenna has a DC path to ground, either
through a RF choke, or a resistor, to bleed off any
static charge. The QCX's 120k resistor path is too feeble to
serve as a reliable discharge. A charged antenna can
blow the finals nearly instantly when connected. There is no
over-voltage antenna protection in the QCX.

A loop-type antenna is safer from this scenario, but I'd still
include a discharge path from the loop to earth.