Date   
Re: Laser cut QO-100 patch parts

David Pearson
 

Hi Mike,

Ok, I have placed order but am not taking any money until I have possession. Delivery to me should be in the next week or so, so I’ll be in touch when I’ve got them

Best regards
Dave

On 8 Jun 2019, at 09:01, Mike Willis <willis.mj@...> wrote:

I'm interested too as I need another one. Why not bring some along to the RAL RT next weekend?
--
Mike G0MJW

Re: Laser cut QO-100 patch parts

David Pearson
 

Ouch,

Apologies for clogging up the forum. I intended to reply individually to people who are interested 

Regards
Dave


On 8 Jun 2019, at 09:18, David Pearson via Groups.Io <Davidfpearson@...> wrote:

Hi Mike,

Ok, I have placed order but am not taking any money until I have possession. Delivery to me should be in the next week or so, so I’ll be in touch when I’ve got them

Best regards
Dave

On 8 Jun 2019, at 09:01, Mike Willis <willis.mj@...> wrote:

I'm interested too as I need another one. Why not bring some along to the RAL RT next weekend?
--
Mike G0MJW

2.4G antenna

g3zqu
 

I would like one set of parts as well please.
Martin G3ZQU

Re: Laser cut QO-100 patch parts

Mike Willis
 

Dubus? That's news to me! Any chance of a scan?
--
Mike G0MJW

Re: Laser cut QO-100 patch parts

Andy G4JNT
 


On Sat, 8 Jun 2019 at 09:36, Mike Willis <willis.mj@...> wrote:
Dubus? That's news to me! Any chance of a scan?
--
Mike G0MJW

Re: patch antenna?

militaryoperator
 

Is a single band patch any better than the dual band patch?

I have no problem with 2 dish's and have built the dual patch, just wondering if a standard 2.4ghz patch without the 22mm hole in it, would be better. 

Ben.


Re: patch antenna?

Mike Willis
 

It depends, but probably. You might do better matching the feed to the focal length more accurately,especially if its f/d far from 0.5-0.6.
--
Mike G0MJW

Re: Radar Interference on 23cm

alwyn.seeds1
 

Dear All,

One other (unlikely) possibility for the very strong interference in JO01 is some new radar surveillance for migrant boats given recent activity- however I thought radars for this were in S-band not L-band, though not sure what the patrol vessels have- need to look up in Janes.

Whatever, as G6TRM says, the interference is very loud when it pops up. 

A recording of the pulses in freq./time should enable identification by those “in the business”.

In IO92FA I have yet to hear it; here there are just the occasional aircraft reflections from the ATC radars.

Regards,

Alwyn


_____________________________________________________

Alwyn Seeds, Director
SynOptika Ltd.,
114 Beaufort Street,
London,
SW3 6BU,
England.

Tel.: +44 (0) 20 7376 4110


SynOptika Ltd., Registered in England and Wales: No. 04606737
Registered Office: 114 Beaufort Street, London, SW3 6BU, United Kingdom.
_____________________________________________________

Re: Radar Interference on 23cm

SAM JEWELL
 

Alwyn. This radar interference has been apparent for some years. Its level is dependent on conditions across the NS. I would be really surprised if it turned out to be anything other than the en-route air traffic control radar. Most likely Belgian or possibly southern PA0.

Sam, G4DDK





On Saturday, June 8, 2019, 11:41 am, alwyn.seeds1 <a.seeds@...> wrote:

Dear All,

One other (unlikely) possibility for the very strong interference in JO01 is some new radar surveillance for migrant boats given recent activity- however I thought radars for this were in S-band not L-band, though not sure what the patrol vessels have- need to look up in Janes.

Whatever, as G6TRM says, the interference is very loud when it pops up. 

A recording of the pulses in freq./time should enable identification by those “in the business”.

In IO92FA I have yet to hear it; here there are just the occasional aircraft reflections from the ATC radars.

Regards,

Alwyn


_____________________________________________________

Alwyn Seeds, Director
SynOptika Ltd.,
114 Beaufort Street,
London,
SW3 6BU,
England.

Tel.: +44 (0) 20 7376 4110


SynOptika Ltd., Registered in England and Wales: No. 04606737
Registered Office: 114 Beaufort Street, London, SW3 6BU, United Kingdom.
_____________________________________________________

Re: patch antenna?

SAM JEWELL
 

I will bring some of Kent’s antennas to RAL next weekend.
The circular polarisation patches have almost sold out. I’ll have more after the following weekend.....
Although not as rugged or able to handle extreme power, as the excellent patch feed from Mike et al, the PCB patches are just as effective and their position can be adjusted to suit your dish (within the constraints of your dish F/d).
Sam, G4DDK

On Saturday, June 8, 2019, 11:04 am, Mike Willis <willis.mj@...> wrote:

It depends, but probably. You might do better matching the feed to the focal length more accurately,especially if its f/d far from 0.5-0.6.
--
Mike G0MJW

Re: Laser cut QO-100 patch parts

PAUL NICKALLS
 

Hi David,

A bit late to the party but we would like a set if you go ahead.

Paul G8AQA and Heather.

On 25/05/2019 14:38, Davidfpearson via Groups.Io wrote:
 

I was working on a patch antenna for qo-100 and thought ‘there has to be a better way’...

if we get enough, could probably get these for about £12. 


Virus-free. www.avg.com

Re: Radar Interference on 23cm

Nicholas Shaxted
 

It is very likely to be from the radar system sitting by the motorway in Belgium, just outside Bruges in a small town called Jabbeke (51°11'30.01"N   3° 3'50.76"E)

Take a look on GiggleEarth, or Bing, or OpenMap, or even a paper atlas.

 

Don’t forget these systems have a Cosec beam null at the horizon so if you hear pulses outside ducting then it is likely to be from forward scatter.

 

Somebody with a wideband receiver covering the frequency band should be able to accurately identify the centre frequency and then you should be able to figure out if your filter is doing its job.

Remember that some fft algorithms will stretch (widen) the pulses so caution is advised when using additional signal processing on SDRs.

 

Nick – g4ogi

 

From: UKMicrowaves@groups.io <UKMicrowaves@groups.io> On Behalf Of SAM JEWELL via Groups.Io
Sent: Saturday, June 8, 2019 1:19 PM
To: UKMicrowaves@groups.io
Subject: Re: [UKMicrowaves] Radar Interference on 23cm

 

Alwyn. This radar interference has been apparent for some years. Its level is dependent on conditions across the NS. I would be really surprised if it turned out to be anything other than the en-route air traffic control radar. Most likely Belgian or possibly southern PA0.

 

Sam, G4DDK

 




On Saturday, June 8, 2019, 11:41 am, alwyn.seeds1 <a.seeds@...> wrote:

Dear All,

 

One other (unlikely) possibility for the very strong interference in JO01 is some new radar surveillance for migrant boats given recent activity- however I thought radars for this were in S-band not L-band, though not sure what the patrol vessels have- need to look up in Janes.

 

Whatever, as G6TRM says, the interference is very loud when it pops up. 

 

A recording of the pulses in freq./time should enable identification by those “in the business”.

 

In IO92FA I have yet to hear it; here there are just the occasional aircraft reflections from the ATC radars.

 

Regards,

 

Alwyn

 

 

_____________________________________________________

 

Alwyn Seeds, Director
SynOptika Ltd.,
114 Beaufort Street,
London,
SW3 6BU,
England.

Tel.: +44 (0) 20 7376 4110


SynOptika Ltd., Registered in England and Wales: No. 04606737
Registered Office: 114 Beaufort Street, London, SW3 6BU, United Kingdom.

_____________________________________________________

 

Fan Cooling at the Masthead

Andy G4JNT
 

I'm putting together a 24GHz personal beacon using white box modules and direct upconversion.  The Tx module is modified for direct drive to the mixer diodes, using modulated baseband generated  in a DDS.
Tested the whole assembly yesterday and it works OK, but the baseplate the whole assembly is mounted on gets far too warm - it is dissipating something like  18 Watts.
Testing with a small 12V fan blowing at the baseplate cooled it perfectly.  So if I want to run this in a box at the masthead for extended periods, the fan will be mandatory

Where and how is the best place to mount a fan on an external box?  and how do you keep out rain?   And where's the other port need to be to keep the air flow moving. 
 I think I want to blow air onto/past the baseplate rather than suck it past, simply because that works in practice sitting on the bench.  But don't have much experience of air cooling, especially in a closed environment.    Intuitively, sucking is better for keeping rain out, as the inlet port could be placed at the bottom of the box.  Or it could blow from the bottom.

Ideas, thoughts ...
(But beware - it could end up in a future RadCom column :-)
 

Virus-free. www.avg.com

Re: Fan Cooling at the Masthead

Richard <richard@...>
 

why not use a heat pipe to an external fan , then you don't need to worry about moisture ingression
Also a lot easier to implement

Richard

On Sun, 2019-06-09 at 09:50 +0100, Andy G4JNT wrote:
I'm putting together a 24GHz personal beacon using white box modules and direct upconversion.  The Tx module is modified for direct drive to the mixer diodes, using modulated baseband generated  in a DDS.
Tested the whole assembly yesterday and it works OK, but the baseplate the whole assembly is mounted on gets far too warm - it is dissipating something like  18 Watts.
Testing with a small 12V fan blowing at the baseplate cooled it perfectly.  So if I want to run this in a box at the masthead for extended periods, the fan will be mandatory

Where and how is the best place to mount a fan on an external box?  and how do you keep out rain?   And where's the other port need to be to keep the air flow moving. 
 I think I want to blow air onto/past the baseplate rather than suck it past, simply because that works in practice sitting on the bench.  But don't have much experience of air cooling, especially in a closed environment.    Intuitively, sucking is better for keeping rain out, as the inlet port could be placed at the bottom of the box.  Or it could blow from the bottom.

Ideas, thoughts ...
(But beware - it could end up in a future RadCom column :-)
 

Virus-free. www.avg.com
-- 
 Best wishes /73 
 Richard Bown
 
 Email : richard@...

 HTTP  :  http://www.g8jvm.com
 
 ######################################################################

 Ham Call: G8JVM . QRV: 50-432 MHz + Microwave 
 Maidenhead QRA: IO82SP38, LAT. 52 39.720' N LONG. 2 28.171 W
 QRV 6mtrs 200W, 4mtrs 150W, 2mtrs 300W, 70cms 200W, 
Microwave 1296MHz 140W, 2320MHz 100W & 10368MHz 5W
 OS: Linux Mint 19.1 x86_64 Tessa, on a Dell Inspiron N5030 laptop
 ######################################################################

Re: Fan Cooling at the Masthead

Chris Bartram G4DGU
 

I thought about this subject quite a lot about this when working on ideas for a serious masthead PA at 1296. 'Tain't easy! The best solution I came up with, based on KISS principles, was to use external heatsinking with the 'fins' in a duct, and to blow air past them. That can separate the internal electronics from the cooling system, but the fan(s) need to be kept out of the direct line of water ingress. Don't forget that air doesn't like turning corners, so baffles aren't a good idea, and it's important to allow for that if you do some flow calculations. Some fans are available which can be operated, apparently reliably, in an an outdoor environment. Most now have solid-state motor commutation, and sealed electronics, others have sealed bearings. For long term operation you need to monitor the cooling system remotely so that you can see when the fan has become unusable.

Back to exam revision ...

73

Chris G4DGU

On 09/06/2019 09:50, Andy G4JNT wrote:
I'm putting together a 24GHz personal beacon using white box modules and direct upconversion.  The Tx module is modified for direct drive to the mixer diodes, using modulated baseband generated  in a DDS.
Tested the whole assembly yesterday and it works OK, but the baseplate the whole assembly is mounted on gets far too warm - it is dissipating something like  18 Watts.
Testing with a small 12V fan blowing at the baseplate cooled it perfectly.  So if I want to run this in a box at the masthead for extended periods, the fan will be mandatory

Where and how is the best place to mount a fan on an external box?  and how do you keep out rain?   And where's the other port need to be to keep the air flow moving. 
 I think I want to blow air onto/past the baseplate rather than suck it past, simply because that works in practice sitting on the bench.  But don't have much experience of air cooling, especially in a closed environment.    Intuitively, sucking is better for keeping rain out, as the inlet port could be placed at the bottom of the box.  Or it could blow from the bottom.

Ideas, thoughts ...
(But beware - it could end up in a future RadCom column :-)
 

Re: Fan Cooling at the Masthead

Neil Smith G4DBN
 

If the box is thinnish plastic, simply moving the air around inside will dissipate a fair amount of heat over the whole surface, with a temperature gradient of only a few degrees at 20W, but you need to keep it really well stirred inside. The thermal conductivity of typical plant case materials, along with the area of the case might mean you can manage OK just with some external finned heatsinks on the case. Polypropylene and similar plastics have a thermal conductivity around 0.2 W/m K, so a box with an area of 0.1 m^2 and 3mm wall thickness dissipating 20W will have a gradient of only 3 degrees, so if the boundary layer inside the case is disrupted, and there is external finning or any significant air movement, 20 W will not cook anything, but you need a decent fan to prevent any hotspots. 

For higher powers, I cut a porthole through the case wall and make a stepped plate for the inside so when I clamp the outside plate on, there is a contact area with Arctic White and an area round the edge with silicone sealant. http://www.g4dbn.uk/?p=917 

Then I use either a finned passive heatsink, or fluid cooling.  I tried a ducted external fan using a hollow extrusion, but it was a swine to keep insects and dust and water out.

With the porthole solution, there is a problem on cold days that you have a VERY cold surface inside the case, so humidity control is important.  I use a GoreTex breather and a bag of silica gel, and flush the case with dry Argon during assembly and maintenance.  With a beacon though, the heat flux will be pretty constant, so you won’t get the diurnal pumping and heat cycle pumping effects that I get with the masthead PAs.

If the area of your box is less than about 0.03m^2 or thicker than 3mm, the porthole approach might help with a plastic case.  A metal case wouldn’t have any significant thermal gradient assuming good internal stirring, so some external natural convection fins rated at 0.5C/W would keep the box cool enough. My masthead 5.7GHz masthead box takes around 100W on transmit and it has no fans, just the porthole plates and some external finned heatsinks. 30 minute full carrier transmissions only raise the temp of the plate by 15 degrees.

I’l see if I can find a pic of the ducted fan arrangement. It used a hollow box section mounted vertically, with a fan at the top, sucking air up the tube, and a wide cap to cover the fan and allow the exhaust out.  There was a mesh grid at the bottom of the tube to prevent insects and leaves being sucked in.

Neil G4DBN



On 9 Jun 2019, at 09:50, Andy G4JNT <andy.g4jnt@...> wrote:

I'm putting together a 24GHz personal beacon using white box modules and direct upconversion.  The Tx module is modified for direct drive to the mixer diodes, using modulated baseband generated  in a DDS.
Tested the whole assembly yesterday and it works OK, but the baseplate the whole assembly is mounted on gets far too warm - it is dissipating something like  18 Watts.
Testing with a small 12V fan blowing at the baseplate cooled it perfectly.  So if I want to run this in a box at the masthead for extended periods, the fan will be mandatory

Where and how is the best place to mount a fan on an external box?  and how do you keep out rain?   And where's the other port need to be to keep the air flow moving. 
 I think I want to blow air onto/past the baseplate rather than suck it past, simply because that works in practice sitting on the bench.  But don't have much experience of air cooling, especially in a closed environment.    Intuitively, sucking is better for keeping rain out, as the inlet port could be placed at the bottom of the box.  Or it could blow from the bottom.

Ideas, thoughts ...
(But beware - it could end up in a future RadCom column :-)
 

Virus-free. www.avg.com

Re: Fan Cooling at the Masthead

Mike Willis
 

Not about fans. Its always marginal with blowers outside. If its not water it's insects and birds. It is surprising how much heat a think aluminium pole takes away and also the dissipation of a large metal areas in free air. Failing that, water cooling is efficient and lightweight with a pair of hose pipes (or H100 coax) and a small pump.
--
Mike G0MJW

Re: Fan Cooling at the Masthead

Ian White
 

Definitely *blow* air directly at the baseplate from close range. A finned heatsink would be better, of course - and best of all, a ducted cooling system.

Blowing from close range gives a vast improvement in cooling compared with natural convection.  Left on its own, the surface of a heatsink becomes blanketed by a stagnant boundary layer of warm air, so the main purpose of the fan is to scrub away that insulating blanket. To do this efficiently needs turbulent air, which only exists immediately downstream of the fan. That is why it's so important to *blow* the air, and also to mount the fan close to the heatsink. Everywhere else (even immediately upstream of the fan) the airflow is much smoother and much less efficient for cooling.

So, the mistakes to *avoid* include:

= Sucking instead of blowing. The smooth airflow upstream of the fan is much less efficient for cooling, and meanwhile that useful turbulence goes somewhere else where it isn't needed.

= Mounting the fan too far away, just aiming hopefully in the general direction. Airflow always takes the easiest route, so it will not go between the fins of a heatsink until you have closed off all the easier options.

= Cooling the hottest components (usually the PA transistors) but then letting the hot air swirl around all the rest of the unit.

Attached is a photo of "one I made earlier" for a PA application where space was very limited. 50mm fans are the largest that would fit, so two fans are mounted side by side (rear of picture) and all the air  is ducted through the heatsink so none of the turbulent airflow is wasted. The hot air exits directly through the rear panel of the enclosure. Even with those two very small fans the cooling is very efficient, and also very quiet because the fans are buried deep inside the enclosure.

73 from Ian GM3SEK

On 09/06/2019 09:50, Andy G4JNT wrote:

I'm putting together a 24GHz personal beacon using white box modules and direct upconversion.  The Tx module is modified for direct drive to the mixer diodes, using modulated baseband generated  in a DDS.
Tested the whole assembly yesterday and it works OK, but the baseplate the whole assembly is mounted on gets far too warm - it is dissipating something like  18 Watts.
Testing with a small 12V fan blowing at the baseplate cooled it perfectly.  So if I want to run this in a box at the masthead for extended periods, the fan will be mandatory

Where and how is the best place to mount a fan on an external box?  and how do you keep out rain?   And where's the other port need to be to keep the air flow moving. 
 I think I want to blow air onto/past the baseplate rather than suck it past, simply because that works in practice sitting on the bench.  But don't have much experience of air cooling, especially in a closed environment.    Intuitively, sucking is better for keeping rain out, as the inlet port could be placed at the bottom of the box.  Or it could blow from the bottom.

Ideas, thoughts ...
(But beware - it could end up in a future RadCom column :-)
 

Virus-free. www.avg.com

Re: Fan Cooling at the Masthead

mike G6TRM
 


My experience was to seal the electronics. fit a heatsink on the outside. fit it all inside a larger box with a fan blowing air on to the heatsink. Airflow is in through a hole in the bottom and blown out the top.
the bottom entry has a drip lip of 2-3 inches (tube) with a grill inside to prevent birds etc. the top has a similar tube with a inverted bowl on the top of that.( snorkel effect)
This prevents most rain/snow moisture form entering.
 
Mike G6TRM

----- Original Message -----
From: Ian White
Sent: Sunday, June 09, 2019 11:22 AM
Subject: Re: [UKMicrowaves] Fan Cooling at the Masthead

Definitely *blow* air directly at the baseplate from close range. A finned heatsink would be better, of course - and best of all, a ducted cooling system.

Blowing from close range gives a vast improvement in cooling compared with natural convection.  Left on its own, the surface of a heatsink becomes blanketed by a stagnant boundary layer of warm air, so the main purpose of the fan is to scrub away that insulating blanket. To do this efficiently needs turbulent air, which only exists immediately downstream of the fan. That is why it's so important to *blow* the air, and also to mount the fan close to the heatsink. Everywhere else (even immediately upstream of the fan) the airflow is much smoother and much less efficient for cooling.

So, the mistakes to *avoid* include:

= Sucking instead of blowing. The smooth airflow upstream of the fan is much less efficient for cooling, and meanwhile that useful turbulence goes somewhere else where it isn't needed.

= Mounting the fan too far away, just aiming hopefully in the general direction. Airflow always takes the easiest route, so it will not go between the fins of a heatsink until you have closed off all the easier options.

= Cooling the hottest components (usually the PA transistors) but then letting the hot air swirl around all the rest of the unit.

Attached is a photo of "one I made earlier" for a PA application where space was very limited. 50mm fans are the largest that would fit, so two fans are mounted side by side (rear of picture) and all the air  is ducted through the heatsink so none of the turbulent airflow is wasted. The hot air exits directly through the rear panel of the enclosure. Even with those two very small fans the cooling is very efficient, and also very quiet because the fans are buried deep inside the enclosure.

73 from Ian GM3SEK

On 09/06/2019 09:50, Andy G4JNT wrote:
I'm putting together a 24GHz personal beacon using white box modules and direct upconversion.  The Tx module is modified for direct drive to the mixer diodes, using modulated baseband generated  in a DDS.
Tested the whole assembly yesterday and it works OK, but the baseplate the whole assembly is mounted on gets far too warm - it is dissipating something like  18 Watts.
Testing with a small 12V fan blowing at the baseplate cooled it perfectly.  So if I want to run this in a box at the masthead for extended periods, the fan will be mandatory

Where and how is the best place to mount a fan on an external box?  and how do you keep out rain?   And where's the other port need to be to keep the air flow moving. 
 I think I want to blow air onto/past the baseplate rather than suck it past, simply because that works in practice sitting on the bench.  But don't have much experience of air cooling, especially in a closed environment.    Intuitively, sucking is better for keeping rain out, as the inlet port could be placed at the bottom of the box.  Or it could blow from the bottom.

Ideas, thoughts ...
(But beware - it could end up in a future RadCom column :-)

 

Virus-free. www.avg.com

Re: Fan Cooling at the Masthead

Andy G4JNT
 

Yes, that's what I've found on the bench many times - even a small fan directly blowing a sheet of aluminium from a couple of cm has a huge effect.  

Neil's comments about the heat conductivity of plastic made me think - I've never even considered that before, other than for PCB material and SMT device cooling 
Heat pipes are too esoteric, Richard.   I met them in spacecraft design, and also after dismantling a laptop - but for this application, I think not
Plastic right angle elbow joints, pointing down to keep out rain as suggested by G0API (on Skype !) have an appeal.

For info, here's a picture of the baseplate.    http://www.g4jnt.com/DropF/imgj0401.JPG

The rear side is all countersunk and flat, so bolting against a box face is possible, it was designed to fit snugly into a plastic box, before I realised it would need fan cooling.  Now it can go in a bigger one to allow the fan some space.   

Bit of thought needed, but some ideas to be going with now.

It won't be mounted long term up a mast - especially as my QTH is not that high, and has too many trees around for that band, but I do plan on taking it to remote hilltops and leaving alone for periods - so needs rain proofing.

Mike. 'TRM's email arrived as I type this
Birds, yes.   I attract them to my garden, so air holes will need to be bird proofed.

BTW, before anyone queries things, the open PCBs in the photo are running at approx  8.4MHz.   A very old DDS module with the DIL (!) PIC is the controller and keyer, delivering whatever mod I choose to use.   Beauty of this DDS chip is that it has amplitude control, so linear modulations are possible -with an upconverter approach it would be a pity not to use this possibility.   The two small PCBs are a  BGA6489 amp raising the signal to around +15dBm, then the one with the with bifilar wound toroid is a quadrature network directly feeding into the I/Q mixer diodes in the white box module Tx source.    I'm getting 25dB opposite sideband suppression (with no I/Q amplitude trimming), and about 35 - 40dB carrier rejection.  As the image is still in the amateur primary location at 24031MHz, I don't feel any necessity to lower the image any further.



Virus-free. www.avg.com


On Sun, 9 Jun 2019 at 11:23, Ian White <gm3sek@...> wrote:

Definitely *blow* air directly at the baseplate from close range. A finned heatsink would be better, of course - and best of all, a ducted cooling system.

Blowing from close range gives a vast improvement in cooling compared with natural convection.  Left on its own, the surface of a heatsink becomes blanketed by a stagnant boundary layer of warm air, so the main purpose of the fan is to scrub away that insulating blanket. To do this efficiently needs turbulent air, which only exists immediately downstream of the fan. That is why it's so important to *blow* the air, and also to mount the fan close to the heatsink. Everywhere else (even immediately upstream of the fan) the airflow is much smoother and much less efficient for cooling.

So, the mistakes to *avoid* include:

= Sucking instead of blowing. The smooth airflow upstream of the fan is much less efficient for cooling, and meanwhile that useful turbulence goes somewhere else where it isn't needed.

= Mounting the fan too far away, just aiming hopefully in the general direction. Airflow always takes the easiest route, so it will not go between the fins of a heatsink until you have closed off all the easier options.

= Cooling the hottest components (usually the PA transistors) but then letting the hot air swirl around all the rest of the unit.

Attached is a photo of "one I made earlier" for a PA application where space was very limited. 50mm fans are the largest that would fit, so two fans are mounted side by side (rear of picture) and all the air  is ducted through the heatsink so none of the turbulent airflow is wasted. The hot air exits directly through the rear panel of the enclosure. Even with those two very small fans the cooling is very efficient, and also very quiet because the fans are buried deep inside the enclosure.

73 from Ian GM3SEK

On 09/06/2019 09:50, Andy G4JNT wrote:
I'm putting together a 24GHz personal beacon using white box modules and direct upconversion.  The Tx module is modified for direct drive to the mixer diodes, using modulated baseband generated  in a DDS.
Tested the whole assembly yesterday and it works OK, but the baseplate the whole assembly is mounted on gets far too warm - it is dissipating something like  18 Watts.
Testing with a small 12V fan blowing at the baseplate cooled it perfectly.  So if I want to run this in a box at the masthead for extended periods, the fan will be mandatory

Where and how is the best place to mount a fan on an external box?  and how do you keep out rain?   And where's the other port need to be to keep the air flow moving. 
 I think I want to blow air onto/past the baseplate rather than suck it past, simply because that works in practice sitting on the bench.  But don't have much experience of air cooling, especially in a closed environment.    Intuitively, sucking is better for keeping rain out, as the inlet port could be placed at the bottom of the box.  Or it could blow from the bottom.

Ideas, thoughts ...
(But beware - it could end up in a future RadCom column :-)
 

Virus-free. www.avg.com