1. QRPLabs
  2. Topics

While poking around on the QRP=Labs website this morning ...

Jim Strohm
 

... I noticed a suggestion to glue a coin to one of the
heat-generating semiconductors. I don't remember where I saw it.

However, this would be excellent for replacing the steel washer on the
BS170s in my QDX. A plain old US all-copper penny with a hole drilled
in it.

A newer-than-1982 penny would work almost as well, since those are
made from copper-clad aluminum.

73
Jim N6OTQ
Da Amazin' man G0FTD
 

On Sat, Jul 30, 2022 at 02:09 PM, Jim Strohm wrote:
However, this would be excellent for replacing the steel washer on the
BS170s in my QDX. A plain old US all-copper penny with a hole drilled
in it.
Sorry to  be a pest, but not quite true.

The washer holds the BS170's *to* the heatsink (the PCB conductive copper area).

The washer can't actually act as a heatsink because it is only in contact with a tangent
of the transistor, which can be thought of of an infinitesimal contact area between the
washer and the hemispherical plastic casing of the BS170.

73 de Andy
Jim Strohm
 

Hi Andy

I had planned to sand or file small flats on the round side of the
finals. Reducing the amount of epoxy between the die and the clamping
washer, and increasing the contact area, will increase heat transfer
into the washer. Changing from a steel washer to a copper washer will
also improve heat transfer.

How much? I have no idea. That's why I'm experimenting.

Yes, I realize that my efforts may provide only minimal improvements at best.

Since I'm going to socket the finals and add a copper heat transfer
pad between the PCB and the flats on the finals to take up the spacing
difference, that'll change things too. That effort may yield only
minimal improvements -- and since the finals leads won't be directly
in contact with the PCB, it may even make things worse, since thermal
conduction through the finals leads into the PCB traces delivers a lot
of cooling for these flea-power FETs.

But once I start blowing finals, the sockets will make my life much better.

I may have to buy another QDX once Rev 4 boards and kits become
available. I hope that doesn't become as problematic as buying a
Raspberry Pi 4 is now. However, I trust that Hans will have another
batch of QDX by Christmas. RPi4s are back-ordered until Christmas
next year, and probably will still sell at scalper prices then.

My premium for the RPi4 I think I've sourced will only cost me maybe
USD$100 in parts and shipping, above the agreed-on price that reflects
fair market prices from last year. Assuming I could buy an RPi4
today, they're running at about a USD$250 premium to the scalpers.

I attribute that to the Bruce Springsteen concert ticket effect -- a
million suckers are born every minute, and they all have money just
falling out of their pockets.

That's why this is a hobby, and I can afford to spend a little more
time and a little less money to get mission-critical hardware.

If you go to the official Raspberry website, you can see all the
disadvantaged people complaining about how they can't get RPi4s for
their mission-critical projects. And some of them are trying to do
work with social and educational relevance, and CAN'T because of all
the scalpers who've programmed their bots to snatch up every available
RPi4 for resale at that USD$250 Springsteen ticket premium.

I have something else to add to my rant here, but I'll keep it to
myself for now. This is a hobby for me, or at least it's supposed to
be.

73
Jim N6OTQ

On Sat, Jul 30, 2022 at 10:44 AM Andy G0FTD via groups.io
<punkbiscuit@...> wrote:

On Sat, Jul 30, 2022 at 02:09 PM, Jim Strohm wrote:

However, this would be excellent for replacing the steel washer on the
BS170s in my QDX. A plain old US all-copper penny with a hole drilled
in it.

Sorry to be a pest, but not quite true.

The washer holds the BS170's *to* the heatsink (the PCB conductive copper area).

The washer can't actually act as a heatsink because it is only in contact with a tangent
of the transistor, which can be thought of of an infinitesimal contact area between the
washer and the hemispherical plastic casing of the BS170.

73 de Andy
Larry Acklin
 

I bought a RPi400 for $100, pretty happy with it. Pi 4 inside a keyboard, mouse, usb. Wsjt-x is happy. 

Larry 
KB3CUF 


On Sat, Jul 30, 2022 at 12:23 PM Jim Strohm <jim.strohm@...> wrote:
Hi Andy

I had planned to sand or file small flats on the round side of the
finals.  Reducing the amount of epoxy between the die and the clamping
washer, and increasing the contact area, will increase heat transfer
into the washer.  Changing from a steel washer to a copper washer will
also improve heat transfer.

How much?  I have no idea.   That's why I'm experimenting.

Yes, I realize that my efforts may provide only minimal improvements at best.

Since I'm going to socket the finals and add a copper heat transfer
pad between the PCB and the flats on the finals to take up the spacing
difference, that'll change things too.  That effort may yield only
minimal improvements -- and since the finals leads won't be directly
in contact with the PCB, it may even make things worse, since thermal
conduction through the finals leads into the PCB traces delivers a lot
of cooling for these flea-power FETs.

But once I start blowing finals, the sockets will make my life much better.

I may have to buy another QDX once Rev 4 boards and kits become
available.  I hope that doesn't become as problematic as buying a
Raspberry Pi 4 is now.  However, I trust that Hans will have another
batch of QDX by Christmas.  RPi4s are back-ordered until Christmas
next year, and probably will still sell at scalper prices then.

My premium for the RPi4 I think I've sourced will only cost me maybe
USD$100 in parts and shipping, above the agreed-on price that reflects
fair market prices from last year.  Assuming I could buy an RPi4
today, they're running at about a USD$250 premium to the scalpers.

I attribute that to the Bruce Springsteen concert ticket effect -- a
million suckers are born every minute, and they all have money just
falling out of their pockets.

That's why this is a hobby, and I can afford to spend a little more
time and a little less money to get mission-critical hardware.

If you go to the official Raspberry website, you can see all the
disadvantaged people complaining about how they can't get RPi4s for
their mission-critical projects.  And some of them are trying to do
work with social and educational relevance, and CAN'T because of all
the scalpers who've programmed their bots to snatch up every available
RPi4 for resale at that USD$250 Springsteen ticket premium.

I have something else to add to my rant here, but I'll keep it to
myself for now.  This is a hobby for me, or at least it's supposed to
be.

73
Jim N6OTQ

On Sat, Jul 30, 2022 at 10:44 AM Andy G0FTD via groups.io
<punkbiscuit=googlemail.com@groups.io> wrote:
>
> On Sat, Jul 30, 2022 at 02:09 PM, Jim Strohm wrote:
>
> However, this would be excellent for replacing the steel washer on the
> BS170s in my QDX. A plain old US all-copper penny with a hole drilled
> in it.
>
> Sorry to  be a pest, but not quite true.
>
> The washer holds the BS170's *to* the heatsink (the PCB conductive copper area).
>
> The washer can't actually act as a heatsink because it is only in contact with a tangent
> of the transistor, which can be thought of of an infinitesimal contact area between the
> washer and the hemispherical plastic casing of the BS170.
>
> 73 de Andy
>





Da Amazin' man G0FTD
 

On Sat, Jul 30, 2022 at 05:23 PM, Jim Strohm wrote:
I had planned to sand or file small flats on the round side of the
finals. Reducing the amount of epoxy between the die and the clamping
washer, and increasing the contact area, will increase heat transfer
into the washer. Changing from a steel washer to a copper washer will
also improve heat transfer.

How much? I have no idea. That's why I'm experimenting.

Well I suppose you have a point. It might be a minor incremental stage, but who the
heck would know if that 5% did save your finals. It might just do it, but it's a bit ugly.

But ugly vs useful in experimenters world is valid ;-)



Since I'm going to socket the finals and add a copper heat transfer pad between the PCB and the flats on the finals to take up the spacing difference, that'll change things too.

Socketing is often a good idea, but you might be better off considering:

a)Hans has already done extensive testing of the product when used correctly.
So hows about using the product correctly.

The QDX QCX range uses Class E = max efficiency = lack of dissipated heat.

If a device is efficiently used then the power input is used properly and not wasted as heat
and that helps keep the device within it's rated dissipation rating when used properly.

But what happens when it's not used properly ?

Well you have to take into consideration P(input) and P(output) *and* what might get "reflected"
back into the device because of a poor SWR + suitable temperature deration.

Put simply, if you want to push the device then make sure the darn SWR is low and stop trying to
build in excuses for end user operation.

As for socketing per se, well that can also be influenced by the  physical construction of a given product.

If it were me, I would adopt a more surface mount approach, and that would mean only soldering
the leads on the top side of the PCB for easy removal (assuming thru plated holes are good / used
and that their integrity is good).

For something like a U3S I'd recommend socketing, for others I'd *consider* top blobbling.


 RPi4s are back-ordered until Christmas next year, and probably will still sell at scalper prices then.
4's might have more horsepower, but user reports indicate a poorer rfi performance.
That can effect in shack problems, and radiation from interconnects = crap local noise floors.
Lower models like the 3's appear to have few issues so discount them.

They might have *only* 1Gb of RAM, but then does it matter ?

You can run a whole OS + Libre Office + Browser on Youtube and other utils and not even break
the 450Mb limit.  Unused RAM is operationally useless.

No point buying a 1000 BHP Lamborghini when the speed limit is 30 MPH (or 20MPH on many British town roads).

Go faster stripes anyone ?
I attribute that to the Bruce Springsteen concert ticket effect -- a million suckers are born

I'd rather pay $1 admission to see https://www.youtube.com/watch?v=QOHOM1hVM-M
than waste time with a corporate whore ;-)

- Andy -
Hans Summers
 

Hi Jim

I did manage to buy another 2,000 chip sets so will be able to produce a QDX batch 5 soon. Currently 150 batch 4 kits are still available http://shop.qrp-labs.com/qdx 

73 Hans G0UPL
http://qrp-labs.com
https://www.buymeacoffee.com/g0upl


-------- Original message --------
From: Jim Strohm <jim.strohm@...>
Date: Sat, Jul 30, 2022, 7:23 PM
To: QRPLabs@groups.io
Subject: Re: [QRPLabs] While poking around on the QRP=Labs website this morning ...
Hi Andy

I had planned to sand or file small flats on the round side of the
finals.  Reducing the amount of epoxy between the die and the clamping
washer, and increasing the contact area, will increase heat transfer
into the washer.  Changing from a steel washer to a copper washer will
also improve heat transfer.

How much?  I have no idea.   That's why I'm experimenting.

Yes, I realize that my efforts may provide only minimal improvements at best.

Since I'm going to socket the finals and add a copper heat transfer
pad between the PCB and the flats on the finals to take up the spacing
difference, that'll change things too.  That effort may yield only
minimal improvements -- and since the finals leads won't be directly
in contact with the PCB, it may even make things worse, since thermal
conduction through the finals leads into the PCB traces delivers a lot
of cooling for these flea-power FETs.

But once I start blowing finals, the sockets will make my life much better.

I may have to buy another QDX once Rev 4 boards and kits become
available.  I hope that doesn't become as problematic as buying a
Raspberry Pi 4 is now.  However, I trust that Hans will have another
batch of QDX by Christmas.  RPi4s are back-ordered until Christmas
next year, and probably will still sell at scalper prices then.

My premium for the RPi4 I think I've sourced will only cost me maybe
USD$100 in parts and shipping, above the agreed-on price that reflects
fair market prices from last year.  Assuming I could buy an RPi4
today, they're running at about a USD$250 premium to the scalpers.

I attribute that to the Bruce Springsteen concert ticket effect -- a
million suckers are born every minute, and they all have money just
falling out of their pockets.

That's why this is a hobby, and I can afford to spend a little more
time and a little less money to get mission-critical hardware.

If you go to the official Raspberry website, you can see all the
disadvantaged people complaining about how they can't get RPi4s for
their mission-critical projects.  And some of them are trying to do
work with social and educational relevance, and CAN'T because of all
the scalpers who've programmed their bots to snatch up every available
RPi4 for resale at that USD$250 Springsteen ticket premium.

I have something else to add to my rant here, but I'll keep it to
myself for now.  This is a hobby for me, or at least it's supposed to
be.

73
Jim N6OTQ

On Sat, Jul 30, 2022 at 10:44 AM Andy G0FTD via groups.io
<punkbiscuit@...> wrote:
>
> On Sat, Jul 30, 2022 at 02:09 PM, Jim Strohm wrote:
>
> However, this would be excellent for replacing the steel washer on the
> BS170s in my QDX. A plain old US all-copper penny with a hole drilled
> in it.
>
> Sorry to  be a pest, but not quite true.
>
> The washer holds the BS170's *to* the heatsink (the PCB conductive copper area).
>
> The washer can't actually act as a heatsink because it is only in contact with a tangent
> of the transistor, which can be thought of of an infinitesimal contact area between the
> washer and the hemispherical plastic casing of the BS170.
>
> 73 de Andy
>





Jim Strohm
 

Andy,

The QDX is Class D. The QCX is Class E.

The difference is somewhat minimal, but it makes a difference. It's
why the QCX is frequency-locked to a single band and the QDX can be
used on multiple bands.

I don't have much more to add for now, except that a copper penny with
a hole in it will make a nice clamp on the top of the finals, but
won't be feasible on the bottom because of all the machining necessary
to fit on the PCB heatsink pad and to be flat enough to work.

I'm still looking for my scrap copper.

I'm also looking for more money to get a Rev 4 QDX (when available) so
I can easily attach a set of sockets for the finals, with the
semi-promised 0.1" spacing on the finals pins.

I am seeing that developing a product like the QDX is a much greater
challenge than most of us realize, and Hans is doing excellent work.

73
Jim N6OTQ
Da Amazin' man G0FTD
 

On Sat, Jul 30, 2022 at 07:41 PM, Jim Strohm wrote:
The QDX is Class D. The QCX is Class E.
I'm can't claim too much on this one, and will happily stand corrected.

Too many Q's these days to keep up.

Over to Hans ;-)

- Andy -
Jim Strohm
 

Hi Andy

Class D and Class E are more like high-speed switches than amps. If
you're familiar with Class A, B, and C amps -- the biggest difference
is that when you get to Class C, you start to lose some linearity but
you gain much more efficiency. Past Class C, we really don't have a
body of anecdotal data.

Hans' methods of massaging signal phasing allows achieving much
greater economy of transmitted signals -- when you don't have to strip
off a carrier and one of the sideband phases, life is good. However,
since Class D and Class E "amps" are basically RF switches, stuff can
and will blow up more easily.

I don't know if my heatsink experimentation will let me run my QDX
above 4 watts but I want to see -- not because I _need_ more power
(amps are easy), but because it seems that most of the finals failures
appear to be caused by cranking the power too high. So -- make the
finals easier to replace, and improve the thermal controls for them.

I also have a bunch of adapter boards that should let me socket a DIP
chip where the SMD ICs go, because sometimes those ICs like to go to
radio heaven with the finals.

Yeah, this is likely to turn into a Frankenstein radio in a few weeks.

73
Jim N6OTQ

On Sat, Jul 30, 2022 at 1:56 PM Andy G0FTD via groups.io
<punkbiscuit@...> wrote:

On Sat, Jul 30, 2022 at 07:41 PM, Jim Strohm wrote:

The QDX is Class D. The QCX is Class E.

I'm can't claim too much on this one, and will happily stand corrected.

Too many Q's these days to keep up.

Over to Hans ;-)

- Andy -
Jack Kibelbek
 

I wouldn't use glue. Depending upon which glue, iI could act as a thermal insulator.

I made a heat sink for a friend's mini whose transistors were getting too hot very quickly.  After 4 seconds of key down, they were too hot to hold a finger on.

I made my sink from two layers of flattened copper pipe. It turned out a little bit thicker than the supplied washer. I made it 0.5" square with a hole in the center for the screw. That would be 1,25 cm. It covered all the transistors and fit with plenty of room to clear everything, Later I was able to activate with that unit for over an hour with no over heating.



FWIW... The old wheat pennies were an alloy of copper 95% and zinc 5%. I would not use current production copper clad pennies. Current clad pennies are 97.5% zinc clad with copper 2.5%. They do not contain aluminum.
Barry VA7GEM
 

On Sat, Jul 30, 2022 at 11:41 AM, Jim Strohm wrote:
Andy,

The QDX is Class D. The QCX is Class E.

The difference is somewhat minimal, but it makes a difference. It's
why the QCX is frequency-locked to a single band and the QDX can be
used on multiple bands.

I don't have much more to add for now, except that a copper penny with
a hole in it will make a nice clamp on the top of the finals, but
won't be feasible on the bottom because of all the machining necessary
to fit on the PCB heatsink pad and to be flat enough to work.

I'm still looking for my scrap copper.

I'm also looking for more money to get a Rev 4 QDX (when available) so
I can easily attach a set of sockets for the finals, with the
semi-promised 0.1" spacing on the finals pins.

I am seeing that developing a product like the QDX is a much greater
challenge than most of us realize, and Hans is doing excellent work.

73
Jim N6OTQ
Jim
Not sure how that penny will work as it is not flat. Also curious about the law down there
regarding defacing money. Not supposed to do it here. Copper flashing is cheap at the
hardware store and nice and flat, read not prep needed. If you were fortunate enough
to find soft copper you may not have to remove as much epoxy but then again you may
be trying to get the die closer to the heat sink.
GL de Barry
Bill Allen
 

Copper clad zinc.
Bob - K2KI
 

Hi Jim,
I am not sure if someone touched base with you regarding the '82 or newer cents pieces. But, The cents newer than 82 are not Copper Clad Aluminum but Copper Clad Zinc. If you ever drill into a newer than '82 and find it is all copper or any other metals through & through, you just killed a cent (usually '82 small date or '83 and maybe '84) worth some serious $$$ (Depending on Condx). I only know this because I only collect Lincoln Cents looking for errors (usually called Varieties). I am not a full blooded Numismatist.

Good luck with whichever way you roll!

--
"Of all the things I've lost in life, I miss my mind the most!"
Very best 72/73 from Vermont. .
Bob de k2ki
k2ki@... (Primary Email)
k2ki.bb@... (Secondary: Goes to my cell)
Grid: FN34LW
NEQRP#: 911
4SQRP#: 2292
FP#: 4214
SKCC#: 12195
NAQCC#: 2337
Sasquatch#: 769
30CW#: 190
100WattsID#: 4675
SOC#: 1312
Zombie# 1567

On 7/30/2022 9:08 AM, Jim Strohm wrote:
... I noticed a suggestion to glue a coin to one of the
heat-generating semiconductors. I don't remember where I saw it.

However, this would be excellent for replacing the steel washer on the
BS170s in my QDX. A plain old US all-copper penny with a hole drilled
in it.

A newer-than-1982 penny would work almost as well, since those are
made from copper-clad aluminum.

73
Jim N6OTQ
Jim Strohm
 

Hi

I misspoke when I said "aluminum". I should have said "zinc."

There's another interesting problem with the new zinc pennies. When
swallowed, the old copper pennies would merely dissolve after some
time. The copper-clad zinc pennies would shed their copper, and then
present a jagged, razor-sharp zinc disk requiring surgical
intervention.

Back in college, I'd photocopy texts when copying was cheaper than
buying the text itself. To bind them, three-hole punch them, I'd
locate some matching machine screws and nuts, and then drill out some
pennies because at one cent each, they were cheaper than washers.

Yes, these were before the clad zinc cents were common, so all were
solid copper but none were wheat cents.

73
Jim N6OTQ

On Sun, Jul 31, 2022 at 7:24 PM Bob - K2KI <k2ki@...> wrote:

Hi Jim,
I am not sure if someone touched base with you regarding the '82 or
newer cents pieces. But, The cents newer than 82 are not Copper Clad
Aluminum but Copper Clad Zinc. If you ever drill into a newer than '82
and find it is all copper or any other metals through & through, you
just killed a cent (usually '82 small date or '83 and maybe '84) worth
some serious $$$ (Depending on Condx). I only know this because I only
collect Lincoln Cents looking for errors (usually called Varieties). I
am not a full blooded Numismatist.

Good luck with whichever way you roll!

--
"Of all the things I've lost in life, I miss my mind the most!"
Very best 72/73 from Vermont. .
Bob de k2ki
k2ki@... (Primary Email)
k2ki.bb@... (Secondary: Goes to my cell)
Grid: FN34LW
NEQRP#: 911
4SQRP#: 2292
FP#: 4214
SKCC#: 12195
NAQCC#: 2337
Sasquatch#: 769
30CW#: 190
100WattsID#: 4675
SOC#: 1312
Zombie# 1567

On 7/30/2022 9:08 AM, Jim Strohm wrote:
... I noticed a suggestion to glue a coin to one of the
heat-generating semiconductors. I don't remember where I saw it.

However, this would be excellent for replacing the steel washer on the
BS170s in my QDX. A plain old US all-copper penny with a hole drilled
in it.

A newer-than-1982 penny would work almost as well, since those are
made from copper-clad aluminum.

73
Jim N6OTQ
@BruceUK
 

My experience is on QCX+ rather than QDX but for what its worth :

Having put a very very light smear of heatsink compound (standard white silicone/zinc stuff) on the flat faces that get screwed to the PCB there was still quite a bit of heat in the top sides /steel washer on prolonged continuous 5W+ TX with the case closed. 

Yes, I know that nowhere does Hans suggest that prolonged 5W+ TX is what the units are designed for.  In fact the opposite 

As I'm interested in WSPR more than CW  the usual duty cycle saviours may not apply to my case either.  However Ham radio is experimentation and I have a drawer of BS170's handy..so

 

I drilled and bent a heatsink from a significant area of 1mm flat copper sheet to nicely fit the round upper surfaces of the transistors and carefully set up ultra-thin compound along the ridges and a thicker band either side of each package so most of the curved surface has compound between it and the heatsink.   The central 5mm x 1mm of each transistor case is in full contact and will have a reasonably good thermal path  to the heatsink, The same again on either side will be poorer but still useful.  Call it a good 5x 2mm per transistor, 5x 8 mm for all 4.  This seemed a reasonable contact area and I was able to run 5.8W out key down for 20 minutes without any problem and the temperature somewhere in the midst of the transistor group  and ALSO the temperature in the bottom of the new copper heatsink was around 51c in each case with the top of my 30mm high sink fin still at around 40-45C.   If I assume the 'worst temperature in the midst of the transistors was 58C  I think that means junction temperatures for the BS170s of around 120C worst case which is allowable and sustainable.  

 

From this I concluded that it is possible to extract significant heat out of the curved tops of the transistors even though the transistors junction- to -case thermal resistance and / or the contact method used is not perfect for higher powers.


I would add that no one has suggested that running continuous 5.8W for 20 minutes is a very good idea, indeed my motivation is to be very sure a 3.5Watt beacon is bomb-proof in a heat wave.

I think the existing advice on limiting voltage and power in all the literature is good advice but if you wish to push the limits or raise the room temperature you can extract heat from above the packages

 
@BruceUK
 

Steve WB3LGC
 

Great...

You answered a question I was thinking relative to my (on order) QDX as after reading the manual, I wondered if thermal compound could be used.  Your answer, YES.

Now I am think about using one of copper heatsinks from my Rpi and doing a little machining to replicate what you did.

Thanks

73, steve WB3LGC


On 8/1/22 11:38 AM, bagoog@... wrote:

My experience is on QCX+ rather than QDX but for what its worth :

Having put a very very light smear of heatsink compound (standard white silicone/zinc stuff) on the flat faces that get screwed to the PCB there was still quite a bit of heat in the top sides /steel washer on prolonged continuous 5W+ TX with the case closed. 

Yes, I know that nowhere does Hans suggest that prolonged 5W+ TX is what the units are designed for.  In fact the opposite 

As I'm interested in WSPR more than CW  the usual duty cycle saviours may not apply to my case either.  However Ham radio is experimentation and I have a drawer of BS170's handy..so

 

I drilled and bent a heatsink from a significant area of 1mm flat copper sheet to nicely fit the round upper surfaces of the transistors and carefully set up ultra-thin compound along the ridges and a thicker band either side of each package so most of the curved surface has compound between it and the heatsink.   The central 5mm x 1mm of each transistor case is in full contact and will have a reasonably good thermal path  to the heatsink, The same again on either side will be poorer but still useful.  Call it a good 5x 2mm per transistor, 5x 8 mm for all 4.  This seemed a reasonable contact area and I was able to run 5.8W out key down for 20 minutes without any problem and the temperature in the midst of the transistor group  and ALSO the temperature in the bottom of the new copper heatsink was around 51c in each case with the top of my 30mm high sink fin still at around 40-45C.    I think that means junction temperatures for the BS170s of around 100C which is allowable and sustainable.  

 

From this I concluded that it is possible to extract significant heat out of the curved tops of the transistors even though the transistors junction- to -case thermal resistance and / or the contact method used is not perfect for higher powers.


I would add that no one has suggested that running continuous 5.8W for 20 minutes is a very good idea, indeed my motivation is to be very sure a 3.5Watt beacon is bomb-proof in a heat wave.

I think the existing advice on limiting voltage and power in all the literature is good advice but if you wish to push the limits or raise the room temperature then more heatsinking will help the junction temperatures out a bit and CAN be done from above
howard winwood G4GPF
 

If you are going to put a heatsink on  I would turn it 9o degree and mount it under the transistors.

On 1 Aug 2022, at 18:00, Steve WB3LGC <email@...> wrote:

Great...

You answered a question I was thinking relative to my (on order) QDX as after reading the manual, I wondered if thermal compound could be used.  Your answer, YES.

Now I am think about using one of copper heatsinks from my Rpi and doing a little machining to replicate what you did.

Thanks

73, steve WB3LGC


On 8/1/22 11:38 AM, bagoog@... wrote:

My experience is on QCX+ rather than QDX but for what its worth :

Having put a very very light smear of heatsink compound (standard white silicone/zinc stuff) on the flat faces that get screwed to the PCB there was still quite a bit of heat in the top sides /steel washer on prolonged continuous 5W+ TX with the case closed. 

Yes, I know that nowhere does Hans suggest that prolonged 5W+ TX is what the units are designed for.  In fact the opposite 

As I'm interested in WSPR more than CW  the usual duty cycle saviours may not apply to my case either.  However Ham radio is experimentation and I have a drawer of BS170's handy..so

 

I drilled and bent a heatsink from a significant area of 1mm flat copper sheet to nicely fit the round upper surfaces of the transistors and carefully set up ultra-thin compound along the ridges and a thicker band either side of each package so most of the curved surface has compound between it and the heatsink.   The central 5mm x 1mm of each transistor case is in full contact and will have a reasonably good thermal path  to the heatsink, The same again on either side will be poorer but still useful.  Call it a good 5x 2mm per transistor, 5x 8 mm for all 4.  This seemed a reasonable contact area and I was able to run 5.8W out key down for 20 minutes without any problem and the temperature in the midst of the transistor group  and ALSO the temperature in the bottom of the new copper heatsink was around 51c in each case with the top of my 30mm high sink fin still at around 40-45C.    I think that means junction temperatures for the BS170s of around 100C which is allowable and sustainable.  

 

From this I concluded that it is possible to extract significant heat out of the curved tops of the transistors even though the transistors junction- to -case thermal resistance and / or the contact method used is not perfect for higher powers.


I would add that no one has suggested that running continuous 5.8W for 20 minutes is a very good idea, indeed my motivation is to be very sure a 3.5Watt beacon is bomb-proof in a heat wave.

I think the existing advice on limiting voltage and power in all the literature is good advice but if you wish to push the limits or raise the room temperature <IMG_4849.JPG>then more heatsinking will help the junction temperatures out a bit and CAN be done from above
Dan Swanson
 

Hi Larry,
RPi400 looks really cool.  Have been looking for compact computer solutions for limited space on ham shack desktop.
Can you comment on what ham software is available?
73's Dan
William Smith
 

There are a couple of different packaged systems for the Pi computers that incorporate a _lot_ of ham software. A couple that come to mind are:

https://mawcg.org/ham-pi/
https://github.com/dslotter/HamPi (Yeah, I know, very similar names)
https://github.com/km4ack/pi-build

Get spare SD cards so you can play with different ones.

I tried the middle one above, but then ended up rolling my own QDXpi by building HamLib, WSJT-x, Gridtracker, and JS8Call from sources, which was harder than it should have been, and makes me leery of doing upgrades. Also got tsql and other such working, so I'm pretty happy with it. Will probably rebuild it from scratch (on a separate SD card) when I get my new QDX rebuilt.

73, Willie N1JBJ

On Aug 4, 2022, at 12:32 PM, Dan Swanson <dan@...> wrote:

RPi400 looks really cool. Have been looking for compact computer solutions for limited space on ham shack desktop.
Can you comment on what ham software is available?

1 - 20 of 21