Topics

2320 PA Modifications?


Anthony Coldman
 

Looking for information on modifications for this pallet to use on 13cm?

 

Any help/information gratefully received

 

Anthony G7LRQ

 


Robin Szemeti - G1YFG
 

I have a very similar one,  basically, not a lot is needed.

I removed the circulator, replaced the input connector with an SMA, coupled to the output to some hardline and that was about it. No real "tuning" seemed to be needed, it swept OK  and showed decent gain at 2320.  

Just put 0V on the "outer" sets of pins and 28V on the inner sets of pins.   If you look carefully you can see that on each side there are 4 pins connected to the main device and one to a "driver stage" ... I would check the current draw on the 4 left main device, right main device,  drier stage and input stage and record the values for future reference.

On Thu, 22 Oct 2020 at 00:28, Anthony Coldman via groups.io <anthony.coldman=ntlworld.com@groups.io> wrote:

Looking for information on modifications for this pallet to use on 13cm?

 

Any help/information gratefully received

 

Anthony G7LRQ

 


--
Robin Szemeti - G1YFG


Robin Szemeti - G1YFG
 

This may help

I would ignore the pictures showing added capacitors ... anyone adding capacitors with those length leads at 2320 is having a laugh.  There are 2 types of this slab, one on page one, the type I have is shown on page 27.  The type on P27 draws about 1.9A  idle.


On Thu, 22 Oct 2020 at 00:35, Robin Szemeti <robin@...> wrote:
I have a very similar one,  basically, not a lot is needed.

I removed the circulator, replaced the input connector with an SMA, coupled to the output to some hardline and that was about it. No real "tuning" seemed to be needed, it swept OK  and showed decent gain at 2320.  

Just put 0V on the "outer" sets of pins and 28V on the inner sets of pins.   If you look carefully you can see that on each side there are 4 pins connected to the main device and one to a "driver stage" ... I would check the current draw on the 4 left main device, right main device,  drier stage and input stage and record the values for future reference.

On Thu, 22 Oct 2020 at 00:28, Anthony Coldman via groups.io <anthony.coldman=ntlworld.com@groups.io> wrote:

Looking for information on modifications for this pallet to use on 13cm?

 

Any help/information gratefully received

 

Anthony G7LRQ

 


--
Robin Szemeti - G1YFG


geoffrey pike
 

But perhaps they are adding parallel inductance
Geoff
GI0GDP

On Thursday, 22 October 2020, 00:45:52 BST, Robin Szemeti - G1YFG <robin@...> wrote:


This may help

I would ignore the pictures showing added capacitors ... anyone adding capacitors with those length leads at 2320 is having a laugh.  There are 2 types of this slab, one on page one, the type I have is shown on page 27.  The type on P27 draws about 1.9A  idle.

On Thu, 22 Oct 2020 at 00:35, Robin Szemeti <robin@...> wrote:
I have a very similar one,  basically, not a lot is needed.

I removed the circulator, replaced the input connector with an SMA, coupled to the output to some hardline and that was about it. No real "tuning" seemed to be needed, it swept OK  and showed decent gain at 2320.  

Just put 0V on the "outer" sets of pins and 28V on the inner sets of pins.   If you look carefully you can see that on each side there are 4 pins connected to the main device and one to a "driver stage" ... I would check the current draw on the 4 left main device, right main device,  drier stage and input stage and record the values for future reference.

On Thu, 22 Oct 2020 at 00:28, Anthony Coldman via groups.io <anthony.coldman=ntlworld.com@groups.io> wrote:

Looking for information on modifications for this pallet to use on 13cm?

 

Any help/information gratefully received

 

Anthony G7LRQ

 


--
Robin Szemeti - G1YFG


militaryoperator
 

Looking for information on modifications for this pallet to use on 13cm?
Any help/information gratefully received
Anthony G7LRQ
-------------------

About to start on mine as well Anthony. Nice to work you on Tuesday. 

Ben G4BXD

 


Richard White
 

Not the neatest of modifications but this worked for me - if I were doing it again I would probably just wire round the circulator- I removed in in slices with a Dremel. But perfectly functional.
Other mods got a bit more oomph out of it.
73 Richard GI4DOH


Anthony Coldman
 

Many thanks for the replies, I have now powered it up and getting rf out.  7mw in 18w out at 2320, but at 2220 its 60w for same drive.  I guess its need a little more modification and drive?  I see from varies pictures that adding copper section to some tracks has been done, fettling I believe, Is there any information available on this? 

 

Has anyone modified it so that it doesn’t draw 2amps all the time, I noticed that the heatsink starts to get rather warm after its been on for a while and even producing any O/P.

 

Robin- Thanks for the document, spent some time translating important area, shame it doesn’t give more details on the mods.

 

Richard, Thanks for the pictures, could you please email me the originals so I can zoom in and have a proper look at what you done. And also can you please confirm your drive level and resulting O/P at 2320?

 

Ben – thanks for contact and let’s keep in touch as we travel the modification road together  😊

 

Any other help/information appreciated, there seems to be a few trying this and some already done it.

 

Regards Anthony

 

From: UKMicrowaves@groups.io <UKMicrowaves@groups.io> On Behalf Of Richard White
Sent: 22 October 2020 12:02
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] 2320 PA Modifications?

 

Not the neatest of modifications but this worked for me - if I were doing it again I would probably just wire round the circulator- I removed in in slices with a Dremel. But perfectly functional.

Other mods got a bit more oomph out of it.

73 Richard GI4DOH


Robin Szemeti - G1YFG
 

The "other" type of board as per P27 of the PDF is easier to remove the circulator, 4 bolts.   I took a piece of brass bar, 10mm square, drilled a 0.141" hole through it at just the right height and soldered a piece it UT141 hardline through it with the centre conductor at board height.  Bolted down, it removes any possible strain on the junction.


On Thu, 22 Oct 2020 at 12:02, Richard White <gi4doh@...> wrote:
Not the neatest of modifications but this worked for me - if I were doing it again I would probably just wire round the circulator- I removed in in slices with a Dremel. But perfectly functional.
Other mods got a bit more oomph out of it.
73 Richard GI4DOH


--
Robin Szemeti - G1YFG


Clint Sharp
 

As these boards have 'sense' FETs would that suggest they're using an automatic biasing scheme like the one shown in Ampleon's Doherty Amplifier App notes?

Should make it more simple to modify the biasing no?

On Thu, 22 Oct 2020 at 13:23, Robin Szemeti - G1YFG <robin@...> wrote:
The "other" type of board as per P27 of the PDF is easier to remove the circulator, 4 bolts.   I took a piece of brass bar, 10mm square, drilled a 0.141" hole through it at just the right height and soldered a piece it UT141 hardline through it with the centre conductor at board height.  Bolted down, it removes any possible strain on the junction.


On Thu, 22 Oct 2020 at 12:02, Richard White <gi4doh@...> wrote:
Not the neatest of modifications but this worked for me - if I were doing it again I would probably just wire round the circulator- I removed in in slices with a Dremel. But perfectly functional.
Other mods got a bit more oomph out of it.
73 Richard GI4DOH


--
Robin Szemeti - G1YFG



--
Clint. M0UAW IO83

No trees were harmed in the sending of this mail. However, a large number of electrons were greatly inconvenienced.


Robin Szemeti - G1YFG
 

I'm not sure you'll get more power out, you may get more gain.   Remember that in "Doherty" mode, the Class A/B side can be driven hard into clipping, the peaking side then picks up the slack .... its very similar to a "Class H" audio amplifier in some respects.  So while you could re-bias it into Class A/B both sides and they will still sum correctly due to the phasing, you'd have to back the drive off so the devices did not clip .. you, get more gain, but probably the same power, with less efficiency.    That's my take on it anyway.  

On Thu, 22 Oct 2020 at 13:38, Clint Sharp <cjaysharp@...> wrote:
As these boards have 'sense' FETs would that suggest they're using an automatic biasing scheme like the one shown in Ampleon's Doherty Amplifier App notes?

Should make it more simple to modify the biasing no?

On Thu, 22 Oct 2020 at 13:23, Robin Szemeti - G1YFG <robin@...> wrote:
The "other" type of board as per P27 of the PDF is easier to remove the circulator, 4 bolts.   I took a piece of brass bar, 10mm square, drilled a 0.141" hole through it at just the right height and soldered a piece it UT141 hardline through it with the centre conductor at board height.  Bolted down, it removes any possible strain on the junction.


On Thu, 22 Oct 2020 at 12:02, Richard White <gi4doh@...> wrote:
Not the neatest of modifications but this worked for me - if I were doing it again I would probably just wire round the circulator- I removed in in slices with a Dremel. But perfectly functional.
Other mods got a bit more oomph out of it.
73 Richard GI4DOH


--
Robin Szemeti - G1YFG



--
Clint. M0UAW IO83

No trees were harmed in the sending of this mail. However, a large number of electrons were greatly inconvenienced.


--
Robin Szemeti - G1YFG


Anthony Coldman
 

I am not necessarily looking for a way to change the bias just stop the PA drawing 2 amps when it’s not actually producing and O/P i.e. when the system is in RX. I am keen to keep the heat down as I mast mount my system on a large heatsink but have no fans, I just started looking at ways to read the on board temperature sensor to see what the temperature is actually doing.

 

I did think that completely removing the 28v when on RX would be one way, anyone think this might cause issues?

 

An update:  With an unmodified pallet,  I have now driven it with 30mW and the O/P is just over the 100w drawing 15amps@27v (max voltage for this PSU) , I need to find a way to read above 100w, (I have a directional coupler but it doesn’t seem to work correctly.)

 

Thanks again for help and any information.

 

Anthony

 

 

 

From: UKMicrowaves@groups.io <UKMicrowaves@groups.io> On Behalf Of Robin Szemeti - G1YFG
Sent: 22 October 2020 16:04
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] 2320 PA Modifications?

 

I'm not sure you'll get more power out, you may get more gain.   Remember that in "Doherty" mode, the Class A/B side can be driven hard into clipping, the peaking side then picks up the slack .... its very similar to a "Class H" audio amplifier in some respects.  So while you could re-bias it into Class A/B both sides and they will still sum correctly due to the phasing, you'd have to back the drive off so the devices did not clip .. you, get more gain, but probably the same power, with less efficiency.    That's my take on it anyway.  

 

On Thu, 22 Oct 2020 at 13:38, Clint Sharp <cjaysharp@...> wrote:

As these boards have 'sense' FETs would that suggest they're using an automatic biasing scheme like the one shown in Ampleon's Doherty Amplifier App notes?

 

Should make it more simple to modify the biasing no?

 

On Thu, 22 Oct 2020 at 13:23, Robin Szemeti - G1YFG <robin@...> wrote:

The "other" type of board as per P27 of the PDF is easier to remove the circulator, 4 bolts.   I took a piece of brass bar, 10mm square, drilled a 0.141" hole through it at just the right height and soldered a piece it UT141 hardline through it with the centre conductor at board height.  Bolted down, it removes any possible strain on the junction.

 

 

On Thu, 22 Oct 2020 at 12:02, Richard White <gi4doh@...> wrote:

Not the neatest of modifications but this worked for me - if I were doing it again I would probably just wire round the circulator- I removed in in slices with a Dremel. But perfectly functional.

Other mods got a bit more oomph out of it.

73 Richard GI4DOH


--
Robin Szemeti - G1YFG


 

--

Clint. M0UAW IO83

No trees were harmed in the sending of this mail. However, a large number of electrons were greatly inconvenienced.


--
Robin Szemeti - G1YFG


SAM JEWELL
 


Just remove the (5v) bias. That's how I and many others have been switching our 13cm 200W- 270W RFMOS PAs when on receive.
A 24V truck auto relay (few pounds from Wilco or Halfords) or P channel MOS power FeT to switch the 28V on and off. Relay gives complete isolation.

73 de Sam, G4DDK




------ Original Message ------
From: "Anthony Coldman via groups.io" <anthony.coldman@...>
To: UKMicrowaves@groups.io
Sent: Thursday, 22 Oct, 20 At 17:45
Subject: Re: [UKMicrowaves] 2320 PA Modifications?

I am not necessarily looking for a way to change the bias just stop the PA drawing 2 amps when it’s not actually producing and O/P i.e. when the system is in RX. I am keen to keep the heat down as I mast mount my system on a large heatsink but have no fans, I just started looking at ways to read the on board temperature sensor to see what the temperature is actually doing.

I did think that completely removing the 28v when on RX would be one way, anyone think this might cause issues?

An update: With an unmodified pallet, I have now driven it with 30mW and the O/P is just over the 100w drawing 15amps@27v (max voltage for this PSU) , I need to find a way to read above 100w, (I have a directional coupler but it doesn’t seem to work correctly.)

Thanks again for help and any information.

Anthony

From: UKMicrowaves@groups.io <UKMicrowaves@groups.io> On Behalf Of Robin Szemeti - G1YFG
Sent: 22 October 2020 16:04
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] 2320 PA Modifications?

I'm not sure you'll get more power out, you may get more gain. Remember that in "Doherty" mode, the Class A/B side can be driven hard into clipping, the peaking side then picks up the slack .... its very similar to a "Class H" audio amplifier in some respects. So while you could re-bias it into Class A/B both sides and they will still sum correctly due to the phasing, you'd have to back the drive off so the devices did not clip .. you, get more gain, but probably the same power, with less efficiency. That's my take on it anyway.

On Thu, 22 Oct 2020 at 13:38, Clint Sharp <cjaysharp@...> wrote:

As these boards have 'sense' FETs would that suggest they're using an automatic biasing scheme like the one shown in Ampleon's Doherty Amplifier App notes?

Should make it more simple to modify the biasing no?

On Thu, 22 Oct 2020 at 13:23, Robin Szemeti - G1YFG <robin@...> wrote:

The "other" type of board as per P27 of the PDF is easier to remove the circulator, 4 bolts. I took a piece of brass bar, 10mm square, drilled a 0.141" hole through it at just the right height and soldered a piece it UT141 hardline through it with the centre conductor at board height. Bolted down, it removes any possible strain on the junction.

On Thu, 22 Oct 2020 at 12:02, Richard White <gi4doh@...> wrote:

Not the neatest of modifications but this worked for me - if I were doing it again I would probably just wire round the circulator- I removed in in slices with a Dremel. But perfectly functional.

Other mods got a bit more oomph out of it.

73 Richard GI4DOH


--
Robin Szemeti - G1YFG


--

Clint. M0UAW IO83

No trees were harmed in the sending of this mail. However, a large number of electrons were greatly inconvenienced.


--
Robin Szemeti - G1YFG


Andy G4JNT
 

Is there no shutdown or standby line into them.
(Only speculating, never been anywhere near one of those amps).

Dive inside, see if there is a way to shut off the bias.



Virus-free. www.avg.com

On Thu, 22 Oct 2020 at 17:45, Anthony Coldman via groups.io <anthony.coldman=ntlworld.com@groups.io> wrote:

I am not necessarily looking for a way to change the bias just stop the PA drawing 2 amps when it’s not actually producing and O/P i.e. when the system is in RX. I am keen to keep the heat down as I mast mount my system on a large heatsink but have no fans, I just started looking at ways to read the on board temperature sensor to see what the temperature is actually doing.

 

I did think that completely removing the 28v when on RX would be one way, anyone think this might cause issues?

 

An update:  With an unmodified pallet,  I have now driven it with 30mW and the O/P is just over the 100w drawing 15amps@27v (max voltage for this PSU) , I need to find a way to read above 100w, (I have a directional coupler but it doesn’t seem to work correctly.)

 

Thanks again for help and any information.

 

Anthony

 

 

 

From: UKMicrowaves@groups.io <UKMicrowaves@groups.io> On Behalf Of Robin Szemeti - G1YFG
Sent: 22 October 2020 16:04
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] 2320 PA Modifications?

 

I'm not sure you'll get more power out, you may get more gain.   Remember that in "Doherty" mode, the Class A/B side can be driven hard into clipping, the peaking side then picks up the slack .... its very similar to a "Class H" audio amplifier in some respects.  So while you could re-bias it into Class A/B both sides and they will still sum correctly due to the phasing, you'd have to back the drive off so the devices did not clip .. you, get more gain, but probably the same power, with less efficiency.    That's my take on it anyway.  

 

On Thu, 22 Oct 2020 at 13:38, Clint Sharp <cjaysharp@...> wrote:

As these boards have 'sense' FETs would that suggest they're using an automatic biasing scheme like the one shown in Ampleon's Doherty Amplifier App notes?

 

Should make it more simple to modify the biasing no?

 

On Thu, 22 Oct 2020 at 13:23, Robin Szemeti - G1YFG <robin@...> wrote:

The "other" type of board as per P27 of the PDF is easier to remove the circulator, 4 bolts.   I took a piece of brass bar, 10mm square, drilled a 0.141" hole through it at just the right height and soldered a piece it UT141 hardline through it with the centre conductor at board height.  Bolted down, it removes any possible strain on the junction.

 

 

On Thu, 22 Oct 2020 at 12:02, Richard White <gi4doh@...> wrote:

Not the neatest of modifications but this worked for me - if I were doing it again I would probably just wire round the circulator- I removed in in slices with a Dremel. But perfectly functional.

Other mods got a bit more oomph out of it.

73 Richard GI4DOH


--
Robin Szemeti - G1YFG


 

--

Clint. M0UAW IO83

No trees were harmed in the sending of this mail. However, a large number of electrons were greatly inconvenienced.


--
Robin Szemeti - G1YFG


Virus-free. www.avg.com


Robin Szemeti - G1YFG
 

Certainly on mine there is no shutdown,  it was not needed with "always on" operation in its original use.

There are 3 separate bias circuits, one each for the main devices, one for the driver, each is fed with a (very small) SMD resistor off the relevant 28V line.     I'd suggest a FET or a "car type relay"  as the best solution.   You might want to leave 28V on the first stage, as that is fed from some miniature switch-mode PSU on the board which may not respond well to being bounced on and off.  The feed to that is on a separate 28V pin anyway, so it is easy to do.


On Thu, 22 Oct 2020 at 18:07, Andy G4JNT <andy.g4jnt@...> wrote:
Is there no shutdown or standby line into them.
(Only speculating, never been anywhere near one of those amps).

Dive inside, see if there is a way to shut off the bias.



Virus-free. www.avg.com

On Thu, 22 Oct 2020 at 17:45, Anthony Coldman via groups.io <anthony.coldman=ntlworld.com@groups.io> wrote:

I am not necessarily looking for a way to change the bias just stop the PA drawing 2 amps when it’s not actually producing and O/P i.e. when the system is in RX. I am keen to keep the heat down as I mast mount my system on a large heatsink but have no fans, I just started looking at ways to read the on board temperature sensor to see what the temperature is actually doing.

 

I did think that completely removing the 28v when on RX would be one way, anyone think this might cause issues?

 

An update:  With an unmodified pallet,  I have now driven it with 30mW and the O/P is just over the 100w drawing 15amps@27v (max voltage for this PSU) , I need to find a way to read above 100w, (I have a directional coupler but it doesn’t seem to work correctly.)

 

Thanks again for help and any information.

 

Anthony

 

 

 

From: UKMicrowaves@groups.io <UKMicrowaves@groups.io> On Behalf Of Robin Szemeti - G1YFG
Sent: 22 October 2020 16:04
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] 2320 PA Modifications?

 

I'm not sure you'll get more power out, you may get more gain.   Remember that in "Doherty" mode, the Class A/B side can be driven hard into clipping, the peaking side then picks up the slack .... its very similar to a "Class H" audio amplifier in some respects.  So while you could re-bias it into Class A/B both sides and they will still sum correctly due to the phasing, you'd have to back the drive off so the devices did not clip .. you, get more gain, but probably the same power, with less efficiency.    That's my take on it anyway.  

 

On Thu, 22 Oct 2020 at 13:38, Clint Sharp <cjaysharp@...> wrote:

As these boards have 'sense' FETs would that suggest they're using an automatic biasing scheme like the one shown in Ampleon's Doherty Amplifier App notes?

 

Should make it more simple to modify the biasing no?

 

On Thu, 22 Oct 2020 at 13:23, Robin Szemeti - G1YFG <robin@...> wrote:

The "other" type of board as per P27 of the PDF is easier to remove the circulator, 4 bolts.   I took a piece of brass bar, 10mm square, drilled a 0.141" hole through it at just the right height and soldered a piece it UT141 hardline through it with the centre conductor at board height.  Bolted down, it removes any possible strain on the junction.

 

 

On Thu, 22 Oct 2020 at 12:02, Richard White <gi4doh@...> wrote:

Not the neatest of modifications but this worked for me - if I were doing it again I would probably just wire round the circulator- I removed in in slices with a Dremel. But perfectly functional.

Other mods got a bit more oomph out of it.

73 Richard GI4DOH


--
Robin Szemeti - G1YFG


 

--

Clint. M0UAW IO83

No trees were harmed in the sending of this mail. However, a large number of electrons were greatly inconvenienced.


--
Robin Szemeti - G1YFG


Virus-free. www.avg.com


--
Robin Szemeti - G1YFG


Conrad, PA5Y
 

Use a FET switch and turn the supply off. Very simple. W6PQL sells suitable FET switches. I do it on my 70cms Class A driver PA.

 

Conrad

 

From: UKMicrowaves@groups.io <UKMicrowaves@groups.io> On Behalf Of Anthony Coldman via groups.io
Sent: 22 October 2020 18:46
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] 2320 PA Modifications?

 

I am not necessarily looking for a way to change the bias just stop the PA drawing 2 amps when it’s not actually producing and O/P i.e. when the system is in RX. I am keen to keep the heat down as I mast mount my system on a large heatsink but have no fans, I just started looking at ways to read the on board temperature sensor to see what the temperature is actually doing.

 

I did think that completely removing the 28v when on RX would be one way, anyone think this might cause issues?

 

An update:  With an unmodified pallet,  I have now driven it with 30mW and the O/P is just over the 100w drawing 15amps@27v (max voltage for this PSU) , I need to find a way to read above 100w, (I have a directional coupler but it doesn’t seem to work correctly.)

 

Thanks again for help and any information.

 

Anthony

 

 

 

From: UKMicrowaves@groups.io <UKMicrowaves@groups.io> On Behalf Of Robin Szemeti - G1YFG
Sent: 22 October 2020 16:04
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] 2320 PA Modifications?

 

I'm not sure you'll get more power out, you may get more gain.   Remember that in "Doherty" mode, the Class A/B side can be driven hard into clipping, the peaking side then picks up the slack .... its very similar to a "Class H" audio amplifier in some respects.  So while you could re-bias it into Class A/B both sides and they will still sum correctly due to the phasing, you'd have to back the drive off so the devices did not clip .. you, get more gain, but probably the same power, with less efficiency.    That's my take on it anyway.  

 

On Thu, 22 Oct 2020 at 13:38, Clint Sharp <cjaysharp@...> wrote:

As these boards have 'sense' FETs would that suggest they're using an automatic biasing scheme like the one shown in Ampleon's Doherty Amplifier App notes?

 

Should make it more simple to modify the biasing no?

 

On Thu, 22 Oct 2020 at 13:23, Robin Szemeti - G1YFG <robin@...> wrote:

The "other" type of board as per P27 of the PDF is easier to remove the circulator, 4 bolts.   I took a piece of brass bar, 10mm square, drilled a 0.141" hole through it at just the right height and soldered a piece it UT141 hardline through it with the centre conductor at board height.  Bolted down, it removes any possible strain on the junction.

 

 

On Thu, 22 Oct 2020 at 12:02, Richard White <gi4doh@...> wrote:

Not the neatest of modifications but this worked for me - if I were doing it again I would probably just wire round the circulator- I removed in in slices with a Dremel. But perfectly functional.

Other mods got a bit more oomph out of it.

73 Richard GI4DOH


--
Robin Szemeti - G1YFG


 

--

Clint. M0UAW IO83

No trees were harmed in the sending of this mail. However, a large number of electrons were greatly inconvenienced.


--
Robin Szemeti - G1YFG


Clint Sharp
 

Not necessarily looking for more power or gain, either or both would be nice of course but, primarily, I am looking for knowledge.

Reading the application notes for LDMOS devices in Doherty amplifiers is interesting but hasn't yet answered all my questions and I'm learning.


On Thu, 22 Oct 2020, 16:04 Robin Szemeti - G1YFG, <robin@...> wrote:
I'm not sure you'll get more power out, you may get more gain.   Remember that in "Doherty" mode, the Class A/B side can be driven hard into clipping, the peaking side then picks up the slack .... its very similar to a "Class H" audio amplifier in some respects.  So while you could re-bias it into Class A/B both sides and they will still sum correctly due to the phasing, you'd have to back the drive off so the devices did not clip .. you, get more gain, but probably the same power, with less efficiency.    That's my take on it anyway.  

On Thu, 22 Oct 2020 at 13:38, Clint Sharp <cjaysharp@...> wrote:
As these boards have 'sense' FETs would that suggest they're using an automatic biasing scheme like the one shown in Ampleon's Doherty Amplifier App notes?

Should make it more simple to modify the biasing no?

On Thu, 22 Oct 2020 at 13:23, Robin Szemeti - G1YFG <robin@...> wrote:
The "other" type of board as per P27 of the PDF is easier to remove the circulator, 4 bolts.   I took a piece of brass bar, 10mm square, drilled a 0.141" hole through it at just the right height and soldered a piece it UT141 hardline through it with the centre conductor at board height.  Bolted down, it removes any possible strain on the junction.


On Thu, 22 Oct 2020 at 12:02, Richard White <gi4doh@...> wrote:
Not the neatest of modifications but this worked for me - if I were doing it again I would probably just wire round the circulator- I removed in in slices with a Dremel. But perfectly functional.
Other mods got a bit more oomph out of it.
73 Richard GI4DOH


--
Robin Szemeti - G1YFG



--
Clint. M0UAW IO83

No trees were harmed in the sending of this mail. However, a large number of electrons were greatly inconvenienced.


--
Robin Szemeti - G1YFG


Richard White
 

I have both types of Nokia board and the screwed down component one is my 'active' one at the moment - it came with the original heatsink so was easier to package. It benefited from a bit of snowflaking too.
I use an Arduino Nano to read the onboard temperature sensor. I then control a 24V fan based on the temperature. At idle 20% is fine and the noise is low - there is always the PSU on/off switch if I am not going to be transmitting.
From a cold start this morning it was 5 minutes before the board got up to 24 degrees at which point the fan came on. I draw about 1.5 amps at 28V when idle.
Above 40 degrees the fan goes to full speed - more noise but the temperature doesn't increase by much more after that.
For those plagued by the nominal 10 ohm resistors in the bias chain that don't like the power on surge, 33 ohm 0402 seems to be working for me at the moment with now noticeable reduction in power out.
Note the fan isn't happy with PWM settings between 30% and 90%. I assume that is because of its inbuilt brushless controller not liking the on/off ratios. But the fan was cheap not a problem functionally..
Richard GI4DOH


Andy G4JNT
 

Can't you just increase the PWM rate?



On Fri, 23 Oct 2020 at 10:32, Richard White <gi4doh@...> wrote:

Note the fan isn't happy with PWM settings between 30% and 90%. I assume that is because of its inbuilt brushless controller not liking the on/off ratios. But the fan was cheap not a problem functionally..


Virus-free. www.avg.com


Robin Szemeti - G1YFG
 

Is the Arduino code available?   I currently just turn my 12V fan on around  8V and just leave it on. Something more "intelligent" would be ideal.

On Fri, 23 Oct 2020 at 10:32, Richard White <gi4doh@...> wrote:
I have both types of Nokia board and the screwed down component one is my 'active' one at the moment - it came with the original heatsink so was easier to package. It benefited from a bit of snowflaking too.
I use an Arduino Nano to read the onboard temperature sensor. I then control a 24V fan based on the temperature. At idle 20% is fine and the noise is low - there is always the PSU on/off switch if I am not going to be transmitting.
From a cold start this morning it was 5 minutes before the board got up to 24 degrees at which point the fan came on. I draw about 1.5 amps at 28V when idle.
Above 40 degrees the fan goes to full speed - more noise but the temperature doesn't increase by much more after that.
For those plagued by the nominal 10 ohm resistors in the bias chain that don't like the power on surge, 33 ohm 0402 seems to be working for me at the moment with now noticeable reduction in power out.
Note the fan isn't happy with PWM settings between 30% and 90%. I assume that is because of its inbuilt brushless controller not liking the on/off ratios. But the fan was cheap not a problem functionally..
Richard GI4DOH


--
Robin Szemeti - G1YFG


Clint Sharp
 

You could beg, borrow, buy a PWM compatible 4 wire CPU fan, they're very cheapo or even free if you find a scrap/defunct PC and are easier to control than 2 wire fans which can be unpredictable from one brand to another but you'd need a source of 12V added into the mix.



 


On Fri, 23 Oct 2020 at 10:32, Richard White <gi4doh@...> wrote:
I have both types of Nokia board and the screwed down component one is my 'active' one at the moment - it came with the original heatsink so was easier to package. It benefited from a bit of snowflaking too.
I use an Arduino Nano to read the onboard temperature sensor. I then control a 24V fan based on the temperature. At idle 20% is fine and the noise is low - there is always the PSU on/off switch if I am not going to be transmitting.
From a cold start this morning it was 5 minutes before the board got up to 24 degrees at which point the fan came on. I draw about 1.5 amps at 28V when idle.
Above 40 degrees the fan goes to full speed - more noise but the temperature doesn't increase by much more after that.
For those plagued by the nominal 10 ohm resistors in the bias chain that don't like the power on surge, 33 ohm 0402 seems to be working for me at the moment with now noticeable reduction in power out.
Note the fan isn't happy with PWM settings between 30% and 90%. I assume that is because of its inbuilt brushless controller not liking the on/off ratios. But the fan was cheap not a problem functionally..
Richard GI4DOH



--
Clint. M0UAW IO83

No trees were harmed in the sending of this mail. However, a large number of electrons were greatly inconvenienced.