Date   

Re: Trying to understand differential heating issues of PA Fets in QCX40

ajparent1/KB1GMX
 

The key elements of class E is saturated switching of the devices
and the Phase relationship of the load matching network so that
switching happens at the zero voltage point.

IF the Keying device is getting hot it may be that the class E stage is
not running right and sucking down an inordinate amount of current
to produce power.

Allison
-------------------------------
Please reply on list so we can share.
No private email, it goes to a bit bucket due address harvesting


Re: WARNING: Promotional Material...sort of

Eike Lantzsch
 

On Monday, 8 February 2021 15:26:56 -03 ON7DQ Luc wrote:
Only a shame that shipping that book to Europe will be *35$*... yes
that's thirty five dollars ... OMG (and I don't see it in the eBook
section)

Luc ON7DQ *
Isn't there a Kindle edition on Amazon?

Eike ZP6CGE


Re: Yet another article (with pics) on QCX Mini regulator replacement #qcxmini #magicsmoke

PA1DMG
 

Hi Luc,

Brilliant enclosure that :)
BTW I totally agree on the fuse. Trust nothing, and leave nothing to chance. I mean, these same cells are also used in DIY spot welding devices. That should tell you something about the currents they are able to deliver. They're quite capable of burning a hole in your desk (or worse).


Re: Trying to understand differential heating issues of PA Fets in QCX40

Evan Hand
 

I do not believe that it is a "matching" issue.  Matching is required for push-pull amplifiers.  The QCX is not that type of amp.  It is a class E (though some have debated is it truly class e) with the three FETs in parallel.  The characteristics of these MOSFETs is that the Drain to Source resistance goes up as temperature increases.  They are also very much a resistance type of device, so putting resistors in parallel will distribute the load that each resistance dictates.  This distributes the load between the devices, regardless of how closely the device characteristics match.

There are also only 3 MOSFETs under the washer and one bipolar transistor.  I suspect the device getting hot is the bipolar key shaping device, as it is running more in the linear range to do the key shaping (still for only part of the key down cycle until it is past the shaping part).


Want to be sure of the understanding of how the QCX series of class E amps work.

If I got something wrong, please comment back so we all have the correct information.
73
Evan
AC9TU


Re: WARNING: Promotional Material...sort of

ON7DQ Luc
 

Only a shame that shipping that book to Europe will be  35$ ... yes that's thirty five dollars ... OMG
(and I don't see it in the eBook section)

Luc ON7DQ


Re: QCX for 2m VHF? #mods #vhf #qcx #chat

Jim Mcilroy
 

Beat me to it, Jim  :)

On 8 Feb 2021, at 18:14, Jim Allyn - N7JA <jim@...> wrote:

Well there, now we have a list of the difficulties in building a 2 meter QCX.  All that remains is to tackle and overcome those difficulties.

I don't think building 2 meter gear is terribly difficult, lots of hams have done it.  Heath made several VHF/UHF kits, probably the best known of which was the HW-2036/2036A.  I have designed and built receivers operating up to 929 MHz; it does start getting difficult there.


Re: QCX for 2m VHF? #mods #vhf #qcx #chat

Jim Mcilroy
 

To follow from my last post, and my errors for ambiguity as I see and agree with Allison’s comments.

I guess the original question could be read as ‘wouldn’t it be great if….’  and the answer is that it wouldn’t be possible for the QCX design to stretch that far.





On 8 Feb 2021, at 17:51, Jim Mcilroy via groups.io <gts53@...> wrote:

Fair comment.

I must confess that I assumed that most hams would know the distinct differences between HF and VHF design and construction.

On 8 Feb 2021, at 16:29, Jim Allyn - N7JA <jim@...> wrote:

On Sun, Feb 7, 2021 at 03:27 PM, Jim Mcilroy wrote:
I worked on the basis that the 17m band was the top limit of what Hans offered and went from there, which introduced the term transverter.

Not what the original poster asked.  He asked if the QCX design could be used at VHF. I commented on one aspect of doing so.



Re: Yet another article (with pics) on QCX Mini regulator replacement #qcxmini #magicsmoke

ON7DQ Luc
 

Marc,

like minds have similar ideas it seems ;-)

I reported about my powerpack in this thread on the SOTA reflector ...
https://reflector.sota.org.uk/t/qcx-mini-part-2/24564/57

Only 3 cells in my case, but also a FUSE and an ON/OFF switch.

73,
Luc ON7DQ


Re: QCX for 2m VHF? #mods #vhf #qcx #chat

Jim Allyn - N7JA
 

Well there, now we have a list of the difficulties in building a 2 meter QCX.  All that remains is to tackle and overcome those difficulties.

I don't think building 2 meter gear is terribly difficult, lots of hams have done it.  Heath made several VHF/UHF kits, probably the best known of which was the HW-2036/2036A.  I have designed and built receivers operating up to 929 MHz; it does start getting difficult there.


Re: Yet another article (with pics) on QCX Mini regulator replacement #qcxmini #magicsmoke

ON7DQ Luc
 

Well done Marc!

and @G0UPL
You see Hans, I'm not the only one with a FUSE fetish, haha.

73,
Luc ON7DQ


Re: QCX for 2m VHF? #mods #vhf #qcx #chat

Jim Mcilroy
 

Fair comment.

I must confess that I assumed that most hams would know the distinct differences between HF and VHF design and construction.

On 8 Feb 2021, at 16:29, Jim Allyn - N7JA <jim@...> wrote:

On Sun, Feb 7, 2021 at 03:27 PM, Jim Mcilroy wrote:
I worked on the basis that the 17m band was the top limit of what Hans offered and went from there, which introduced the term transverter.

Not what the original poster asked.  He asked if the QCX design could be used at VHF. I commented on one aspect of doing so.


Re: QCX+ SSB mod - display blank

Guido PE1NNZ
 

Hi Mike,

Great build!

It looks like the fuse settings are incorrect, resulting in setting XCK clock output on enable pin 6.
If you program with Arduino IDE, simply run a "Burn bootloader", or program it manually: avrdude -c avrisp -b 19200 -P /dev/ttyACM0 (or: /dev/ttyUSB0) -p m328p -e -U efuse:w:0xFD:m -U hfuse:w:0xD6:m -U lfuse:w:0xFF:m -U flash:w:R1.02j.hex

Hope this helps. Would you like to document your QCX+ modification when you get it working?  Thanks.

73, Guido
PE1NNZ


Re: Trying to understand differential heating issues of PA Fets in QCX40

tony.volpe.1951@...
 

Thanks again.

I now understand the very hot PA and low power out (Cheers Alan).

And I will now remove the hot Fet to avoid killing Q6 (again) and remove the added capacitance. I will also run wspr at 7.5 volts (because I happen to have a nice linear power supply that can manage that. It is in fact an adjustable 5 volt supply that will ho to 7.5 volts which will give three amps peak and happens to be available. Thanks Marc for that info.

At 7.5 volts the QCX seems to make about 800 milliwatts out which really is plenty for wspr.

Cheers all.


Re: Trying to understand differential heating issues of PA Fets in QCX40

Brad Noblet
 



On Mon, Feb 8, 2021 at 12:03 PM PA1DMG <info@...> wrote:
Hi Tony,
I'd definitely reduce the supply voltage for WSPR use. A couple of volts less makes a big difference in heat and will definitely improve reliability. 7.5 Volts is on the low side though; maybe try 10 - 11 Volts and see what happens?
Adding more output transistors is not guaranteed to work as you are also increasing the load on the driver. The added capacitance will change the circuit characteristics, and not necessarily for the better.

In the end this is a 5Watt CW transmitter. As an experiment, I've added a bit of extra cooling to my QCX Mini but there's limits to what you can expect from TO92 PA transistors...

73,
Marc PA1DMG


Re: QCX+ SSB mod - display blank

Mike G8GYW
 

OK I've narrowed it down a bit because the LCD module works fine with an Arduino.

I've checked the LCD signals at the processor with a x10 scope probe and get this:

Read/Write: Low
D4 to D7: High
Read/Write Select: see below (1 volt per division)



Enable: see below (200mV per division)



So Enable looks wrong. What should it look like?

Mike G8GYW


Re: Trying to understand differential heating issues of PA Fets in QCX40

PA1DMG
 

Hi Tony,
I'd definitely reduce the supply voltage for WSPR use. A couple of volts less makes a big difference in heat and will definitely improve reliability. 7.5 Volts is on the low side though; maybe try 10 - 11 Volts and see what happens?
Adding more output transistors is not guaranteed to work as you are also increasing the load on the driver. The added capacitance will change the circuit characteristics, and not necessarily for the better.

In the end this is a 5Watt CW transmitter. As an experiment, I've added a bit of extra cooling to my QCX Mini but there's limits to what you can expect from TO92 PA transistors...

73,
Marc PA1DMG


Re: Trying to understand differential heating issues of PA Fets in QCX40

Alan G4ZFQ
 

Is there a failure mode where one bad one would impact on the good PA fets like this?
Tony,

I can only guess that one (the hot one?) sometimes fails short circuit - but not quite. Still has some resistance that drags the PA voltage down but not to zero. It would also have the effect of partly shorting the output of the functioning BS170s.

73 Alan G4ZFQ


Re: #u3s #troubleshooting No Spots / not sure transmitting #u3s #troubleshooting #help #lpf

Alan G4ZFQ
 

you need to enter a frequency for park mode, I believe this is used in the cal process,
Andrew,

This is not essential. It derived from the time drift was very common. The park frequency was initiated to maintain a more stable generator output, the frequency determined by trial and error.
It may still be used if drift is seen, it might help.
Full details are on the web pages.

Calibration frequency is fixed (around 6MHz?) again full details are published.

73 Alan G4ZFQ


Re: QCX for 2m VHF? #mods #vhf #qcx #chat

Jim Allyn - N7JA
 

On Sun, Feb 7, 2021 at 03:27 PM, Jim Mcilroy wrote:
I worked on the basis that the 17m band was the top limit of what Hans offered and went from there, which introduced the term transverter.

Not what the original poster asked.  He asked if the QCX design could be used at VHF. I commented on one aspect of doing so.


Re: Trying to understand differential heating issues of PA Fets in QCX40

tony.volpe.1951@...
 

Thanks Mark and Alan for excellent info.

I can now understand why I have one 'hottie' in my batch of four PA fets. I am thinking the hot one isn't well matched with the others. 

One more thing though; A couple of times I have got up in the morning and looked at my WSPR beacon QCX and noticed that power output as judged by forward power as judged on my twin meter swr meter is only about 20% of what it was when I went to bed and all the PA transistors were steaming hot. I replaced them one at a time and suddenly (obviously hitting on a bad one and replacing it) the PA went straight back to normal, with full power. 

Is there a failure mode where one bad one would impact on the good PA fets like this?

I'm sort of thinking that when running wspr on it, reducing the PA stress by swapping power supplies from 13v to 7.5v is probably a sensible idea. 2 minutes key down with a PA which has no heat sinking is probably a bit of a harsh regime to out the radio through.

Thanks

6401 - 6420 of 68287