Date   

Re: Fried my qcx+ finals...

 

i do agree, but i've found the menus at qrp labs complicated and have missed things, and due to memory problems also forgotten that some things exist there that i had not looked at in over a year. For example, i thought everything was in ONE PDF for each version. The firm ware PDF is also a little hidden away. it's not a criticism, far from it, the site is better than all the other QRP sites i've looked at and has very helpful information, but finding it or even knowing it is there, was sometimes a different thing, and i won't boast about what my profession was and my experience, and my skills, so for me to not find things though i am not at all in the shape i used to be by far i am not, many health issues, but i'm just saying that is going to also surely be the reason for some others who don't RTFM because they didn't know nor find the fucking manual :)  i don't take offense at the term RTFM as you can see, and it is useful and often relevant, even if only understood in some minority cultures (perhaps majority in here though). 

it's easy to misunderstand when communication is just via a keyboard, and all these other lack of knowledge of people's circumstances. no one here except perhaps Hans himself has a clue about my real circumstances, nor do i about others and i never make assumptions, life has taught me it is not a good idea to do so. This is why i attempt to write details about the problems not worry about looking foolish and give as much info as possible, and repeat again and again that i only know some very basic electronics from parrot memory more than anything else, i am a professional amateur not an amateur professional, and my only expertise is propagation, antennae to a degree, and Gerke Code aka CW, also some other CW. Repairs, probes, scopes, and other than resistors and diodes, i know let's presume NOTHING.

Such people should be able to ask for assistance in this group becuase this kit CAN be built by Mr Magoo and the Mad Professor provided we RTFM and i did that very, very precisely, only it was the construction manual. Every QCX i ever build (aside from a faulty SMD one) has always worked from the moment of completion. Every one i ever broke has a known reason for why it broke (except perhaps one mystery). When reaching for help, yes it is an opportunity to learn all about how and why but some of us are not able to due to age, health, memory, handicaps, etc, or just a lack of interest and other priorities like learning languages, or whatever, that don't allow much time for what some consider to be great fun and have great skill at it: REPAIRS. in such cases, if help is offered, it needs to be as good as Hans RTM.

Or, it needs to be very patient. Labouring this point probably makes this group look unhelpful to anyone totally new and just arrived, far from it, this group is very patient, very helpful, has amazing people in it offering wonderful help, but for the few who do not consider these possibilities, please do. 

For personal and privacy reasons we cannot always say much about ourselves nor should we have to but let's not make assumptions.  i have not seen a single message from anyone during the periods i have been active in and out over the years, more out than in, of one single instance of anyone taing offense or finding an instruction negative or argumentative to be pointed to a specific page in a manual, this is not what occured. it was one or two people saying they have been very patient with me, which i found offensive and uncalled for. i don't want anyone to have too much patience to the point it affect you negatively!

i hope this could be laid to rest as it is a storm in a tea cup. also keeping in mind that i can type faster than i can think, and few can block read, and for some, reading through my emails will just create a headache, as there are so many words. And that's ok, but those who can RTFM but cannot read such emails should simply skip over, or block, delete, move on, and not worry. There is help in abundance here!

You're one i've learned quite a few things from Arv, if only i can remember them, but i save snippets and useful information in a system that i hope will be more easily accessible to me when needed. Thank you!

On 26 Jul 2021, at 16:08, Arv Evans <arvid.evans@...> wrote:

Alan

You are correct.  RTFM has been ingrained into ham radio since before 
I was licensed in 1956.  As you say, pointing a builder to a specific page 
in the instruction is not argumentative or negative, it is just a simple way 
to help someone find the information they need.  There are those who 
find it easier to ask on-line for specific information rather than digging it 
out for themselves, but this is human nature.  Maybe as we evolve that 
will become the accepted way of doing things?

Arv
_._


On Mon, Jul 26, 2021 at 1:03 AM Alan G4ZFQ <alan4alan@...> wrote:
> That said I have read on a few other sites some nasty responses, like 
> READ THE MANUAL

  Harry,

But it is a response. It is a pity if it is interpreted as nasty.
I may know the answer is in the manual.
But if I see a post that has not been answered and have not the time to 
search for the manual and the page number what is wrong in suggesting 
the questioner looks?

And reading the manual is often better than a reply from someone who 
does not really know the answer.

73 Alan G4ZFQ







--
Love 30m and CW?
http://www.30cw.net


Re: QCX PA transistors - when blown

 

indeed you deserve a medal Julian :-)

how funny, i was just saying to OM Giuseppe off list, that i've ordered another mini main board to start again, as it is surely too fried, likely many if not all SMD ic's would have blown, smoke came up immediately from i think at least 3 parts of the board! Burn and smoke, almost fire, but no bang, though i'm deaf, so yes full 12V "everywhere". 

i was also saying that i have two other QCX that i need to fix, originals, one is mostly working but some odd problem which makes sending cw not practical, the other is complete no rx nor tx (futher info below). 

and that i was going to wait for things to settle in the list here, as i may have asked too many questions too frequently these past weeks, and well intentioned people misjudging my abilities, confused me some more, though i have been learning various things. 

it takes one to know one Julian and i very much appreciate it. thank you!

what i'll do i think is this: i will start building a different kit i have here, while awaiting qcx mini main board to arrive in about a week or so, then get that working first, by rebuilding. i don't mind the building, it is so easy, and 3rd time around gives me a change to perfect things even better than the first two times, and with a few tips i picked up along the way already in mind. 

then, for the other qcxs, if i had anything like fedex around here i'd take you up, but i have to travel a bit far i'm currently in a very remote place, may as well be in the middle of the desert. but, i think it could be helpful before then, for me to purchase a probe scope of sorts, that has been recommended by kind souls in here, firstly, so that i could do some fault tracing with your help perhaps remotely: in fact a few people have offered such remote assistance so when i am suitably equipped, perhaps that's the best way to go about it first, as it'll teach me something, and if that fails then take up that very kind offer of yours. 

i'm hoping i could fix both QCX but if not, then perhaps make one working one out of the two. Though for that, i'd need help to decide how to go about it as its a question which one would be the salvage one.
Again thanks for your kindness and in responding to this message. 

i do have something to listen to at the moment as the oldest original is working on rx, just on tx the screen flickers, and the keyer can't be used it doesn't key nicely and the side tone cuts in and out, even when used on the built in key in straight key mode, it goes silent, but there is no problems in tx output, just horrible to listen and worrying why it makes odd noises as if there is rf going into rx, so i do not want to risk damaging that one further without knowing what to be looking for. it had been several blown pa some years ago when it was used a lot, and some lifted tracks from repeated soldering, but was still working fine though at low power due to not having adjusted the coils properly, but then i decided to change it from 30m to 20m, and that's where this keying problem started since doing that. it's looking a bit ugly now, and has never had a case other than the free cardboard one. 

The other original is a beauty in its Baumatech case, but first i blew the PA with 19V and it had some problem already perhaps one of the PA was already gone first, then, with no idea what caused it the rx stopped working: one day all was absolutely fine with the RX, i switched it off, next day switched it on, block on top row, switched on/off several times (as normally it needs to be started twice as i did not make the mod on it for start up, it's 2017 model), still nothing, so i put in a new ic2, but, upside down (mislead by the sticker, absent mindedly, silly me), i don't know if that would damage anything having ic2 upside down, realised my error, turned it off, turned ic2 right way up, turned on, display all good again, but, no sound. Continuity throughout LPF to antenna. Nothing was touched other than switching it on, but it had been left connected to antenna but in off state, and there was a nearby storm that previous night, so don't know if static could have done anything, or, if something simply happened when i switched it on by bad luck, or, if putting in the new ic2 upside down did it, but then, why did ic2 (and perhaps other things) go in the first place? A real mystery that one.

i'm operating on a power supply that is not earthed, in fact there is no earth at all, so i'm also thinking this is probably very bad practice and need to sort out mains earth and other earth too. Antenna is ground planes single banders, but thinking to get a resistor and also put an earth leak on the outer of coax once i have an earth. 

i really think i need to obtain a few more tools first and also make my set up a bit safer, i have almost never used an earth in my stations but i've almost always used balanced antennas with no rf to worry about. But not even having a mains earth isn't a good idea i think for safety reasons. i think it was a lazy electrician who did the work but there is earth behind so just needs redoing the various wall sockets.

i'll contact you off list if it comes to that and thanks again, and also to the many others who tried to help!

On 26 Jul 2021, at 15:12, Julian N4JO <n4jo@...> wrote:

This is sad, and is why I've said off-forum to a couple of people, though perhaps more subtly: "if you get in too deep, stop digging". My time at Heathkit back in the UK gave me a particularly useful perspective on repairing "has never worked" electronics, and I enjoy resurrecting ailing QCXs, the Minis being a particular challenge because of their delicate nature.
One thing about the mini in particular, just to give an example of the trouble you can easily get yourself into, is that the display connector has 12V  on it, which is plenty enough to take out both the display and the processor in one swell foop with a misplaced probe or a solder whisker.
If your eyesight or dexterity are good enough to find and use your straight key, but not enough for soldering, then cut your losses, and send the ailing kit to me. For the price of USPS  both ways, and whatever parts we have to replace, you'll get it back, working.
As a radio beginner with four and a half decades of electronics experience, I consider it a privilege and an honor - almost a responsibility - to serve those of you who have given so much to the amateur radio community in terms of your radio experience.
And just think: you get to receive the same unit in the mail a second time ;-)
--
Julian, N4JO.


--
Love 30m and CW?
http://www.30cw.net


Re: SINAD

Albert Tatlock's Greatest Hits - Vol 1
 

On Mon, Jul 26, 2021 at 03:58 PM, Gerald Lemay - VA2GLU wrote:
Would like to delve further into receiver sensitivity measurements.
Spectrum Lab by DL2YHF can perform SINAD tests.
It's a Windows piece of software, but runs perfectly with WINE / Linux.
I've used the software for many of it's task, but cannot comment on the SINAD stuff a s I've never used that part of it.

https://www.qsl.net/dl4yhf/spectra1.html


Re: Fried my qcx+ finals...

Arv Evans <arvid.evans@...>
 

Alan

You are correct.  RTFM has been ingrained into ham radio since before
I was licensed in 1956.  As you say, pointing a builder to a specific page
in the instruction is not argumentative or negative, it is just a simple way
to help someone find the information they need.  There are those who
find it easier to ask on-line for specific information rather than digging it
out for themselves, but this is human nature.  Maybe as we evolve that
will become the accepted way of doing things?

Arv
_._


On Mon, Jul 26, 2021 at 1:03 AM Alan G4ZFQ <alan4alan@...> wrote:
> That said I have read on a few other sites some nasty responses, like
> READ THE MANUAL

  Harry,

But it is a response. It is a pity if it is interpreted as nasty.
I may know the answer is in the manual.
But if I see a post that has not been answered and have not the time to
search for the manual and the page number what is wrong in suggesting
the questioner looks?

And reading the manual is often better than a reply from someone who
does not really know the answer.

73 Alan G4ZFQ






Re: Fried my qcx+ finals...

Arv Evans <arvid.evans@...>
 


This URLs show more info on Class-E, for those who want to play with
the designs.




Arv
_._


On Mon, Jul 26, 2021 at 12:10 AM Hans Summers <hans.summers@...> wrote:
Hi Harry

Thanks for your comments and I hope you'll enjoy the radio. 

Just another word about the BS170s...

According to the BS170 datasheet the maximum allowable power dissipation of a BS170 is 830mW see 
Yes there are many other factors such as the ambient temperature etc. 

Anyway consider 830mW for the moment. There are three of them. 3 x 830mW = 2,490mW. So together those three transistors are supposed to be capable of dissipating around 2.5W (into a perfect heatsink).

The QCX PA is Class-E. Not highly optimized and of course every constructor is different too, there are lots of variations in component tolerances etc. So let's say we have 80% efficiency rather than the >90% that Class-E is well capable of. 

So for a PA outputting 5W, if it is 80% efficient, the power in will be 6.25W and the power dissipated in the three BS170 transistors will be 1.25W. 

In the case of CW operation the power dissipation is approximately half that because you key up key down merrily to send your message, and key-downs account for maybe half the total on average. 

But we basically have, for full duty 100% duty cycle operation, conservatively 1.25W of heat dissipation in transistors capable of dissipating 2.5W. In other words we are operating at 50% of the device rating or in yet other words, we have a 100% safety margin. That safety margin is for erosion by all sorts of real world factors like, imperfect antenna match, the transistors may not share the heat equally, imperfect heatsink, etc. Of course our Class-E may be more than 80% efficient too, that was a harsh judgment. 

Either way I don't think the design is terribly marginal. 

We did discuss protection methods on the group some time back but there are many ways to fail and protection against them all would be complex and expensive if possible at all. During development (and since) I never saw failures and concluded that the theoretical design safety margin was sufficient also in practice. 

73 Hans G0UPL 

On Mon, Jul 26, 2021, 00:38 Harry Latterman via groups.io <harrylatterman=yahoo.com@groups.io> wrote:
Hi Hans,

First of all thank you for getting my unit out when it became possible. Another lost week was not a big deal for me since I am working on a half a dozen electronic projects and twice that many house projects. For some reason being retired is harder than working 60-80 hr work weeks, like I did most of the 40+ yrs I was working.

I am painfully aware of heatsinking and the fact that it is more an art than science in many cases. I have not used the BS170 so it will be investigated just like any other new part I am exposed to. As far as some people pumping out 5 watts of WSPR and not blowing the parts and some people frying them, I am not surprised. Some of it I will grant you is SWR issues. Some coupling to the PCB and heat transfer.  The other is where is the radio mounted and was the room hot, warm, cool or cold and was a wind/breeze going over the case. The list can be expanded, but the bottom line is if you have a 100 QCX, QCX+, QCX-mini's all running WSJT-X you will have some that will die unless the heatsink is over rated for the need and the part is over rated for power requirements.  I have seen this in not only RF designs, but also analog designs.  I tend to be a bit over kill. If I have a spec for say 750 mw a typical engineer would put in a 1 watt part. Me... Nope, 2 watts or maybe more. I hate field failures and as one of my ex-employers who I ran into at a store once told me, all 500 of my units had run for 6 yrs and not one failure, but the designs by his other engineers all has some sort of failure in the same period of time and he was stumped as too why. When I told him he was paying a few percentage more per unit do to a overkill design and the added cost was nothing compared to field failures.

I know all about hands on building and testing and being swamped with orders. You have  my deepest respect for what you are doing and the stress it sometimes puts on you and in your life. I am looking forward to getting my hands on your handy work.  The only possible issue I might have is how much audio output drive you have. I do know a sidetone at 700 hz for me is almost impossible to hear in my right ear and a bit better in my left so I use 600 hz and as low as 550 hz. Even then fast CW is a disaster so I will be a QRS CW person at 10-18 WPM. From having many radios I know sound is subjective and not a exacting science, thus hands on use is the only way I know how well I will be able to copy CW.  From what I can see your design is much better than most and I don't expect any issues.

If I have any issues I will be posting them here for help and not bugging you... You have too much to deal with already. If I have a suggestion that is simple to make the design better I will bug you with that information.  I don't expect that to happen, but hey I have a history of finding simple ways to improve designs when I was doing engineering consulting.

Thank you again for your prompt response to my email and explaining why my radio was hung up in the system and shipping it at the first chance you had.  My WSPR work with new compromise antennas will ALWAYS be at 1 watt. By leaving a baseline of 1 watt for all the antennas I will have a better way of determining a good design from a not so good one.

Have a great week... Get some rest and I hope you can get to play more with some new ideas.

73 Harry K7ZOV

On Sunday, July 25, 2021, 1:39:01 PM MST, Hans Summers <hans.summers@...> wrote:


Hi Harry, all

QCX transistors had no heatsinks and were fine for 5W (or a bit more) of CW but I did recommend people who want to run WSPR either lower the power or fit heatsinks to the three BS170 PA transistors. 

QCX+ and QCX-mini *do* have heatsinking on the BS170 transistors (and the Q6 keying transistor). Contrary to some misconceptions the 12mm diameter steel washer is not the heatsink. It touches the curved part of the transistors, a line of contact with theoretically zero cross-sectional area, that transfers a very very low amount of heat away from the transistor body. The steel washer is therefore almost irrelevant from the point of view of heat dissipation. It's purpose is to push the transistors with their flat sides flush against the PCB surface. The PCB under the transistors is solid groundplane, with no solder mask; the transistor bodies are pressed against the tinned copper directly. There is copper ground plane on both sides of the PCB and densely packed vias between the two sides under the transistors. Both sides of the entire PCB are ground-plane wherever possible and with vias at frequent intervals, never more than 0.1 inches apart. Therefore the entire copper plated PCB itself, is the heatsink for the BS170s. 

Heatsink compound does no harm. I don't recommend it however because it is also of negligible benefit. The thermal resistance between the flat faces of the transistors and the PCB "heatsink" is a very small fraction of the total thermal resistance, which is dominated by the plastic enclosure of the transistor itself, or in other words all the plastic the heat has to travel through to get from the BS170 silicon to the flat face surface of the transistor housing. Reducing the surface-to-PCB thermal resistance by putting in some thermal paste is therefore of no significant benefit. 

Incidentally the same applies to IRF510s. You have to go to much larger devices and lower thermal resistance packages than TO92 (BS170) or even TO220 (IRF510) before thermal paste (heatsink compound) becomes of much use. This is why I don't mention it for the 5W, 10W or 50W PA kits that use IRF510s.

On WSPR, I've heard of quite a lot of people who run 5W WSPR from their QCX+/QCX-mini without incident. Clearly there's a limit somewhere and high SWR, or high power, are two of the factors that are going to get you closer to that limit. Nevertheless a properly set up QCX+/QCX-mini and properly set up antenna system shouldn't have difficulties with 5W WSPR transmissions at sensible transmission rates. 

Regarding testing of assembled units, Harry... in the last 12 months my team have assembled almost 600 QCX+ and QCX-mini kits. All of those have been tested, adjusted and calibrated (incl GPS calibration of oscillators) by me personally. In some cases, as you'd expect, fault-finding is also required. Assembled radioa now account for around 15% of QCX-mini sales and 8% of QCX+ sales. It has been a significant drain on my time, which along with email support, admin, component sourcing, managing production and people and problems etc etc, leaves far too little time for precious R&D (i.e. new kits development). 

2 weeks ago I built a QRP Labs dummy load and a power meter consisting of AD8307 log amp detector and ATmega328 and an OLED display. I calibrated it per band to exactly match the power readings I had previously been obtaining from my oscilloscope. Then I wrote a test procedure script and trained our team supervisor to do the testing/adjustment/calibration. It now comes to me only for a final visual check of the completed radio and sign-off of the test/calibration results. Furthermore we have a summer intern, EE undergraduate student, who is now doing fault-finding. If he can't fix it, then it comes to me. Hopefully this will create a bit more R&D time for me. 

73 Hans G0UPL 


On Sun, Jul 25, 2021, 19:42 Harry Latterman via groups.io <harrylatterman=yahoo.com@groups.io> wrote:
Actually John you are partly correct about 100% duty cycle and modern radios. Most Yeasu radios are not happy with 100% duty cycle. Most Icom have no issue with 100% duty cycle. I am running my IC-7300 for many hours running FT8 and RTTY at 100 watts and the radio barely gets warm. Same was true with my IC-756PORII and K3/100.  All were designed to take the heat and get rid of it. I have run RTTY, Olivia, PSK31/63, JT-65 and now FT8 flat out and never blew a final nor had heating problem... With ICOM.. With the Yeasu F-450, FT-991A, and even the FTdx5000MPL beast you had to cut the power back on or have problems.  The OCX line does not have even close to safe heatsinking, thus they will be prone to final failure it not taken care of with low SWR and keeping high duty cycles low.

I am new here and my mini was sent from Turkey yesterday and I should see it by Friday. It was built factory built one and I know for a fact that Hans tested it.  With only a screw and washer and I think with no silicon heat sink added heating will always be a problem with the mini when doing WSPR. I plan on treating the radio as though it could blow at anytime. I do plan on doing a lot of WSPR for QRP testing, so 1 to 2 watts will be all I will ever use... I also plan on keeping the final transistors in stock in my radio shack, just in case.

73 Harry K7ZOV


Re: QCX Challenge - July 2021

wa8yan.radio
 

Hi Erich.  Sorry to hear of your hospitalization.  I've recently come through some medical issues of my own and I sincerely feel for you.  We will keep you in our prayers !  Wishing you well and back on-the-air soon.
Keep us all informed of your progress.
73 Phil WA8YAN



Sent from my Verizon, Samsung Galaxy smartphone


-------- Original message --------
From: HB9FIH <erich.rieder@...>
Date: 7/26/21 2:20 AM (GMT-05:00)
To: QRPLabs@groups.io
Subject: Re: [QRPLabs] QCX Challenge - July 2021

Sri not qrv..

I am (after health disaster - heard/sugar - on Samos) back in Switzerland in the hospital of Canton Schaffhausen.
(For repair my body ...hihi)
Takes time > 1 month.

--
---
73 de Erich

HB9FIH

HS0ZLS


Re: Fried my qcx+ finals...

wa8yan.radio
 

Hi Hans.  Just a quick thank you for continuing my education.  This explanation corrected a few points in my thinking.  I appreciate you taking the time to give all of us the advantage of your knowledge.
73 Phil WA8YAN



Sent from my Verizon, Samsung Galaxy smartphone


-------- Original message --------
From: Hans Summers <hans.summers@...>
Date: 7/26/21 2:10 AM (GMT-05:00)
To: QRPLabs@groups.io
Subject: Re: [QRPLabs] Fried my qcx+ finals...

Hi Harry

Thanks for your comments and I hope you'll enjoy the radio. 

Just another word about the BS170s...

According to the BS170 datasheet the maximum allowable power dissipation of a BS170 is 830mW see 
Yes there are many other factors such as the ambient temperature etc. 

Anyway consider 830mW for the moment. There are three of them. 3 x 830mW = 2,490mW. So together those three transistors are supposed to be capable of dissipating around 2.5W (into a perfect heatsink).

The QCX PA is Class-E. Not highly optimized and of course every constructor is different too, there are lots of variations in component tolerances etc. So let's say we have 80% efficiency rather than the >90% that Class-E is well capable of. 

So for a PA outputting 5W, if it is 80% efficient, the power in will be 6.25W and the power dissipated in the three BS170 transistors will be 1.25W. 

In the case of CW operation the power dissipation is approximately half that because you key up key down merrily to send your message, and key-downs account for maybe half the total on average. 

But we basically have, for full duty 100% duty cycle operation, conservatively 1.25W of heat dissipation in transistors capable of dissipating 2.5W. In other words we are operating at 50% of the device rating or in yet other words, we have a 100% safety margin. That safety margin is for erosion by all sorts of real world factors like, imperfect antenna match, the transistors may not share the heat equally, imperfect heatsink, etc. Of course our Class-E may be more than 80% efficient too, that was a harsh judgment. 

Either way I don't think the design is terribly marginal. 

We did discuss protection methods on the group some time back but there are many ways to fail and protection against them all would be complex and expensive if possible at all. During development (and since) I never saw failures and concluded that the theoretical design safety margin was sufficient also in practice. 

73 Hans G0UPL 

On Mon, Jul 26, 2021, 00:38 Harry Latterman via groups.io <harrylatterman=yahoo.com@groups.io> wrote:
Hi Hans,

First of all thank you for getting my unit out when it became possible. Another lost week was not a big deal for me since I am working on a half a dozen electronic projects and twice that many house projects. For some reason being retired is harder than working 60-80 hr work weeks, like I did most of the 40+ yrs I was working.

I am painfully aware of heatsinking and the fact that it is more an art than science in many cases. I have not used the BS170 so it will be investigated just like any other new part I am exposed to. As far as some people pumping out 5 watts of WSPR and not blowing the parts and some people frying them, I am not surprised. Some of it I will grant you is SWR issues. Some coupling to the PCB and heat transfer.  The other is where is the radio mounted and was the room hot, warm, cool or cold and was a wind/breeze going over the case. The list can be expanded, but the bottom line is if you have a 100 QCX, QCX+, QCX-mini's all running WSJT-X you will have some that will die unless the heatsink is over rated for the need and the part is over rated for power requirements.  I have seen this in not only RF designs, but also analog designs.  I tend to be a bit over kill. If I have a spec for say 750 mw a typical engineer would put in a 1 watt part. Me... Nope, 2 watts or maybe more. I hate field failures and as one of my ex-employers who I ran into at a store once told me, all 500 of my units had run for 6 yrs and not one failure, but the designs by his other engineers all has some sort of failure in the same period of time and he was stumped as too why. When I told him he was paying a few percentage more per unit do to a overkill design and the added cost was nothing compared to field failures.

I know all about hands on building and testing and being swamped with orders. You have  my deepest respect for what you are doing and the stress it sometimes puts on you and in your life. I am looking forward to getting my hands on your handy work.  The only possible issue I might have is how much audio output drive you have. I do know a sidetone at 700 hz for me is almost impossible to hear in my right ear and a bit better in my left so I use 600 hz and as low as 550 hz. Even then fast CW is a disaster so I will be a QRS CW person at 10-18 WPM. From having many radios I know sound is subjective and not a exacting science, thus hands on use is the only way I know how well I will be able to copy CW.  From what I can see your design is much better than most and I don't expect any issues.

If I have any issues I will be posting them here for help and not bugging you... You have too much to deal with already. If I have a suggestion that is simple to make the design better I will bug you with that information.  I don't expect that to happen, but hey I have a history of finding simple ways to improve designs when I was doing engineering consulting.

Thank you again for your prompt response to my email and explaining why my radio was hung up in the system and shipping it at the first chance you had.  My WSPR work with new compromise antennas will ALWAYS be at 1 watt. By leaving a baseline of 1 watt for all the antennas I will have a better way of determining a good design from a not so good one.

Have a great week... Get some rest and I hope you can get to play more with some new ideas.

73 Harry K7ZOV

On Sunday, July 25, 2021, 1:39:01 PM MST, Hans Summers <hans.summers@...> wrote:


Hi Harry, all

QCX transistors had no heatsinks and were fine for 5W (or a bit more) of CW but I did recommend people who want to run WSPR either lower the power or fit heatsinks to the three BS170 PA transistors. 

QCX+ and QCX-mini *do* have heatsinking on the BS170 transistors (and the Q6 keying transistor). Contrary to some misconceptions the 12mm diameter steel washer is not the heatsink. It touches the curved part of the transistors, a line of contact with theoretically zero cross-sectional area, that transfers a very very low amount of heat away from the transistor body. The steel washer is therefore almost irrelevant from the point of view of heat dissipation. It's purpose is to push the transistors with their flat sides flush against the PCB surface. The PCB under the transistors is solid groundplane, with no solder mask; the transistor bodies are pressed against the tinned copper directly. There is copper ground plane on both sides of the PCB and densely packed vias between the two sides under the transistors. Both sides of the entire PCB are ground-plane wherever possible and with vias at frequent intervals, never more than 0.1 inches apart. Therefore the entire copper plated PCB itself, is the heatsink for the BS170s. 

Heatsink compound does no harm. I don't recommend it however because it is also of negligible benefit. The thermal resistance between the flat faces of the transistors and the PCB "heatsink" is a very small fraction of the total thermal resistance, which is dominated by the plastic enclosure of the transistor itself, or in other words all the plastic the heat has to travel through to get from the BS170 silicon to the flat face surface of the transistor housing. Reducing the surface-to-PCB thermal resistance by putting in some thermal paste is therefore of no significant benefit. 

Incidentally the same applies to IRF510s. You have to go to much larger devices and lower thermal resistance packages than TO92 (BS170) or even TO220 (IRF510) before thermal paste (heatsink compound) becomes of much use. This is why I don't mention it for the 5W, 10W or 50W PA kits that use IRF510s.

On WSPR, I've heard of quite a lot of people who run 5W WSPR from their QCX+/QCX-mini without incident. Clearly there's a limit somewhere and high SWR, or high power, are two of the factors that are going to get you closer to that limit. Nevertheless a properly set up QCX+/QCX-mini and properly set up antenna system shouldn't have difficulties with 5W WSPR transmissions at sensible transmission rates. 

Regarding testing of assembled units, Harry... in the last 12 months my team have assembled almost 600 QCX+ and QCX-mini kits. All of those have been tested, adjusted and calibrated (incl GPS calibration of oscillators) by me personally. In some cases, as you'd expect, fault-finding is also required. Assembled radioa now account for around 15% of QCX-mini sales and 8% of QCX+ sales. It has been a significant drain on my time, which along with email support, admin, component sourcing, managing production and people and problems etc etc, leaves far too little time for precious R&D (i.e. new kits development). 

2 weeks ago I built a QRP Labs dummy load and a power meter consisting of AD8307 log amp detector and ATmega328 and an OLED display. I calibrated it per band to exactly match the power readings I had previously been obtaining from my oscilloscope. Then I wrote a test procedure script and trained our team supervisor to do the testing/adjustment/calibration. It now comes to me only for a final visual check of the completed radio and sign-off of the test/calibration results. Furthermore we have a summer intern, EE undergraduate student, who is now doing fault-finding. If he can't fix it, then it comes to me. Hopefully this will create a bit more R&D time for me. 

73 Hans G0UPL 


On Sun, Jul 25, 2021, 19:42 Harry Latterman via groups.io <harrylatterman=yahoo.com@groups.io> wrote:
Actually John you are partly correct about 100% duty cycle and modern radios. Most Yeasu radios are not happy with 100% duty cycle. Most Icom have no issue with 100% duty cycle. I am running my IC-7300 for many hours running FT8 and RTTY at 100 watts and the radio barely gets warm. Same was true with my IC-756PORII and K3/100.  All were designed to take the heat and get rid of it. I have run RTTY, Olivia, PSK31/63, JT-65 and now FT8 flat out and never blew a final nor had heating problem... With ICOM.. With the Yeasu F-450, FT-991A, and even the FTdx5000MPL beast you had to cut the power back on or have problems.  The OCX line does not have even close to safe heatsinking, thus they will be prone to final failure it not taken care of with low SWR and keeping high duty cycles low.

I am new here and my mini was sent from Turkey yesterday and I should see it by Friday. It was built factory built one and I know for a fact that Hans tested it.  With only a screw and washer and I think with no silicon heat sink added heating will always be a problem with the mini when doing WSPR. I plan on treating the radio as though it could blow at anytime. I do plan on doing a lot of WSPR for QRP testing, so 1 to 2 watts will be all I will ever use... I also plan on keeping the final transistors in stock in my radio shack, just in case.

73 Harry K7ZOV


Re: QCX PA transistors - when blown

Michael.2E0IHW
 

Julian, if I had the power, your Congressional Medal would, proverbially speaking, be in the post!

UK : anyone offering this service?

Michael UK
A fellow beginner...

On 26/07/2021 16:12, Julian N4JO wrote:
This is sad, and is why I've said off-forum to a couple of people, though perhaps more subtly: "if you get in too deep, stop digging". My time at Heathkit back in the UK gave me a particularly useful perspective on repairing "has never worked" electronics, and I enjoy resurrecting ailing QCXs, the Minis being a particular challenge because of their delicate nature.
One thing about the mini in particular, just to give an example of the trouble you can easily get yourself into, is that the display connector has 12V  on it, which is plenty enough to take out both the display and the processor in one swell foop with a misplaced probe or a solder whisker.
If your eyesight or dexterity are good enough to find and use your straight key, but not enough for soldering, then cut your losses, and send the ailing kit to me. For the price of USPS  both ways, and whatever parts we have to replace, you'll get it back, working.
As a radio beginner with four and a half decades of electronics experience, I consider it a privilege and an honor - almost a responsibility - to serve those of you who have given so much to the amateur radio community in terms of your radio experience.
And just think: you get to receive the same unit in the mail a second time ;-)
--
Julian, N4JO.


Re: QCX PA transistors - when blown

 

This is sad, and is why I've said off-forum to a couple of people, though perhaps more subtly: "if you get in too deep, stop digging". My time at Heathkit back in the UK gave me a particularly useful perspective on repairing "has never worked" electronics, and I enjoy resurrecting ailing QCXs, the Minis being a particular challenge because of their delicate nature.
One thing about the mini in particular, just to give an example of the trouble you can easily get yourself into, is that the display connector has 12V  on it, which is plenty enough to take out both the display and the processor in one swell foop with a misplaced probe or a solder whisker.
If your eyesight or dexterity are good enough to find and use your straight key, but not enough for soldering, then cut your losses, and send the ailing kit to me. For the price of USPS  both ways, and whatever parts we have to replace, you'll get it back, working.
As a radio beginner with four and a half decades of electronics experience, I consider it a privilege and an honor - almost a responsibility - to serve those of you who have given so much to the amateur radio community in terms of your radio experience.
And just think: you get to receive the same unit in the mail a second time ;-)
--
Julian, N4JO.


SINAD

Gerald Lemay - VA2GLU
 

I teach a radio class here in Quebec province. Would like to delve further into receiver sensitivity measurements. Is there a company selling a 12 dB SINAD meter kit? Maybe someone has a Sinadder 3 for sale? Seems Motorola used to make a similar low-cost product. I have the block diagram and schematic for the Sinadder 3; thus one option would be to breadboard a functioning SINAD meter. Comments and suggestions are most welcomed. 73 de VA2GLU, Gerald


Re: Contest

Christopher Maness
 

I love this type of QRN.  It reminds me of sitting in my dark room by the dial light listening to 80m in the Summer.  All the lighting static in the background.  It is nostalgic for me.

Chris KQ6UP 

On Mon, Jul 26, 2021 at 7:02 AM jjpurdum via groups.io <jjpurdum=yahoo.com@groups.io> wrote:
All:

Horrible QRN and QSB here, but manage contacts with N9GVP and K0FHG with my QCX on 40M and EFHW at about 20'. I even used the microswitch for the key! Fun time!!

Jack, W8TEE


Contest

jjpurdum
 

All:

Horrible QRN and QSB here, but manage contacts with N9GVP and K0FHG with my QCX on 40M and EFHW at about 20'. I even used the microswitch for the key! Fun time!!

Jack, W8TEE


Re: QCX-mini simple audio limiter #qcxmini

Nigel ZS6RN
 

Thank you Gunnar for sharing a very simple, easy to implement and effective solution to lack of an AGC system - APPRECIATED!

73

Nigel

ZS6RN
KG43eu


Re: Want to hire an experienced builder

VE7VXO
 

Hi Tom

I saw your earlier post about this and didn't read it because I haven't ever built a kit for someone else and figured for sure there would be someone on this list who would jump on that.  However I do enjoy building and am a very careful and often painstaking with attention to detail so I can assure you I can build something well.  One issue of course is I am outside the USA so that might be a show stopper for you due to cross border shipping, but if you are interested we can discuss it via private email.  I have never built a U3S but have built a few QCX's and many other designs and many things of my own design amateur radio and otherwise.

Best regards...Joe


Re: help on alignment

K9NUD-Steve
 

The necessary Ref Frequency for my Mini ended up at 27,400,300. 


Re: Fried my qcx+ finals...

Yannig - F4IUJ
 

On Mon, Jul 26, 2021 at 12:36 PM, Roelof Bakker wrote:
Hello Yannig,

Maybe you mentioned it already and have I missied it, just curious how much output power you
were using?

73,
Roelof, pa0rdt
Hi Roelof,

The QCX+ was outputting about 3.5-4 Watts in a dummy load, I didn't try to optimise the power output and I was feeding the kit with about 12.4V ( a small DC power supply resembling a phone charger I had lying around).

73

Yannig


Re: Fried my qcx+ finals...

Roelof Bakker
 

Hello Yannig,

Maybe you mentioned it already and have I missied it, just curious how much output power you
were using?

73,
Roelof, pa0rdt


Re: Fried my qcx+ finals...

Yannig - F4IUJ
 

Hello all,

I didn't realise that my topic would create so much controversy. I am a new Ham (licensed March 2021) and bought my QCX+ because I wanted to learn and decided that putting together a kit was the best thing to do to teach myself the practical aspects of building radios. It was also important to me that it was cheap enough so that I could afford to possibly "destroy it" by error as I knew that I would be more relaxed using it. Obviously I wasn't overly happy with myself when the finals failed nevertheless I see this as part of the learning experience and I just started this topic to try to learn as much as possible from my mistake.

Thank you Hans for the clarifications on the heat dissipation calculations, I wouldn't call 100% safety coefficient marginal either... In "real life", I am a structural engineer and this is roughly what we use for timber structures (30% for the material on top of 50% for variable loads -> 1.3 x 1.5 = 1.95 ) and it could be lower than 50% (Steel under permanent loads). Bearing in mind that the risks aren't the same, if we get it wrong we risk killing people (and thanks god, we very rarely do!), whereas the finals of my kit didn't risk harming anybody, 100% for an electronic kit is very comfortable!

After a more detailed inspection I realised that the cable I made to connect the Z match to the QCX had one BNC plug that was less than well connected (cable is a bit short and the coax shield got damaged from flexing it) while the TX had been in Beacon mode all night before the incident, so may be the unplugged antenna was just the last straw that broke the camel's back! Next time I will be more cautious...

Looks like I will get some BS170s from the local electronic shop tomorrow for a few Euros, it is a shame that I am out of action for tonight QCX challenge (this was going to be my first one as I missed the June one and wasn't QRV in May) but hopefully Tuesday night I will back!

73,

Yannig - F4IUJ


Re: Fried my qcx+ finals...

Alan G4ZFQ
 

That said I have read on a few other sites some nasty responses, like READ THE MANUAL
Harry,

But it is a response. It is a pity if it is interpreted as nasty.
I may know the answer is in the manual.
But if I see a post that has not been answered and have not the time to search for the manual and the page number what is wrong in suggesting the questioner looks?

And reading the manual is often better than a reply from someone who does not really know the answer.

73 Alan G4ZFQ


Re: QCX Challenge - July 2021

HB9FIH
 

Sri not qrv..

I am (after health disaster - heard/sugar - on Samos) back in Switzerland in the hospital of Canton Schaffhausen.
(For repair my body ...hihi)
Takes time > 1 month.

--
---
73 de Erich

HB9FIH

HS0ZLS

18681 - 18700 of 87009