开云体育

ctrl + shift + ? for shortcuts
© 2025 开云体育
  1. QRPLabs
  2. Messages
Expanded Topics Messages Polls
Date
Re: Automatic Antenna Tuner, Internal Battery with a charger and built-in speaker for QMX+. Open Source.

 

Put me down as wanting a kit as well.?
--
73, Dan? NM3A
Re: QMX+ V2 - I've killed it

 

In my experience it’s very difficult to fully protect low voltage rails against overvoltage without an active sensing circuit that (depending on the nature of the risk ) interrupts the supply or crowbars the rail. ??

TVS diodes often fail after the most sensitive connected devices, ?
Zeners can be a bit laid-back, leaking early ?or letting too high a voltage pass for too long. ?
Their main benefit is their tendency to fail short which at least leaves a helpful hint as to the problem?
?
There’s definitely a market for a small fast and cheap 3 or 4 terminal protector for the common supply rail voltages and hopefully they arrive soon?
Re: how to set presets

 

Steve,
?I agree, the Super CMOS Keyers from Logitech (and before) really have a neat way to correct errors. I am sure Hans can add that to his list of things to do. Unfortunately, this isn't a high priority and he's got a long list to go through first.?
--
73, Dan? NM3A
Re: QMX Dead - 3.3v Rail Shorted

 

Bob,
?
Worth looking for your Vdd short, odds are pretty good you can find it.
Start with everything unplugged from the board.
I assume you have measured the Vdd rail with an ohmmeter using the same?
method and meter scale on a good board, and are sure there is a short.
?
First inspect the area around all Vdd power pins for shorts using good light and magnification.
It could be a speck of loose solder lodged between a couple pins somewhere,
so give the board a good wash and scrub with water, or better with 99% alcohol.
Give it plenty of time to dry.
?
If it's a blown part due to high SWR, hopefully the short is back on page 5.
Put a soldering iron on each Vdd power pin and lift it with an x-acto knife or dental pick,
clear any remaining solder with flux and solder wick.?
Check Vdd rail for short after lifting each one.
The only Vdd pins I see back there are IC501P-5, IC502P-5? and C501.
?
If not on page 5, perhaps page 4 with IC407 bottom left, or caps C429, C433.
No Vdd on page 2, and page 1 is just the power supply boards.
All that's left is page 3, lots of Vdd into the bottom of the STM32F,
but otherwise mostly into resistors and L201 and L202.
?
Should be some good parts on it, or maybe somebody else wants to try debugging it.
I'll start the bidding at $20 plus cheapest shipping if you just want to chuck it.
?
Jerry, KE7ER
?
?
On Mon, Oct 21, 2024 at 07:23 AM, Bob WV7W wrote:

With both PS boards out, I checked the 3.3v rail and it is shorted to ground. I suspect that something died on the main board but figured I would check if there is anything else I can do before scrapping this one and starting on a new kit.
Re: Automatic Antenna Tuner, Internal Battery with a charger and built-in speaker for QMX+. Open Source.

 

I likewise would purchase a kit or extra PCBs.
Stan KC7XE in Utah
Re: QMX Dead - 3.3v Rail Shorted

 

Thanks Ian,
?
Mine reads 0.5 ohms. I wish it could be as simple as a solder bridge but it was working and suddenly stopped so that is probably unlikely.?
?
I assume that either the CPU failed or one of the other SMD ICs.
?
I don't currently have the ability to work SMD chips such as a hot air station so I will probably have to put it aside for now.
?
Bob
WV7W
Re: QMX Dead - 3.3v Rail Shorted

 

Sounds like bad luck - sorry.? While it could be a failed component like Ian notes, based on your description a more likely reason is a solder whisker or a tiny piece of cut-off that has migrated to a bad place, causing a short.? This especially if your measurement is closer to 0 ohms.? This happens occasionally even to those of us who are very experienced and careful with our builds.? I would firmly rap each corner of the main board on a table top to hopefully dislodge any such thing, and also use some air to clean it off.? And then I would re-inspect carefully each side of the board, paying close attention to any areas near where I soldered, and to the pins of the CPU and PCM1804 chips, since they have adjacent pins with power and ground that could easily get bridged.
Good luck -
And if you decide to 'scrap' it, send it my way and I will take a look.? I kind-of enjoy debugging this sort of thing.
Stan KC7XE
Re: QMX Dead - 3.3v Rail Shorted

 

开云体育

Sorry to hear that Bob, what is the short circuit reading? If it is around the 4 Ohm mark it could well be the CPU failed. Only way to isolate that is to isolate all the VDD paths until you find the short, unfortunately there are several pins connected to VDD at the CPU, you would have to lift each pin in turn, not a nice job. I had this problem as has many others. I had it on a QMX +, managed to remove the dead CPU and got hold of a replacement from Hans. So don't scrap it I am sure it can be sorted.
?
Of course it could be a simple short somewhere like a flash of soder!!
?
73
?
Ian G4GIR
?

----- Original Message -----
From: Bob WV7W via groups.io <bcreigh@...>
Reply-To: <[email protected]>
Sent: 21/10/2024 15:14:34
Subject: [QRPLabs] QMX Dead - 3.3v Rail Shorted
This is my third QMX Build so not completely new to this.
?
I was building this one for a friend and?I fired up newly built QMX mid-band yesterday and loaded the firmware with no issues. Everything came up fine. I was able to navigate menus and change settings. All the stuff it should normally do.
The issue I was noticing was that into a dummy load, I was getting high SWR and no power out. Sometimes protection would kick in and other times not.?
?
I turned the power off to go do some errands and came back to hook up to terminal to do an RF sweep and nothing. No display nothing.
?
I thought that maybe one of my SMPS boards may have died so swapped them out of a working QMX and still nothing.
?
With both PS boards out, I checked the 3.3v rail and it is shorted to ground. I suspect that something died on the main board but figured I would check if there is anything else I can do before scrapping this one and starting on a new kit. I do have another mid-band kit that was supposed to be mine, but won't leave my friend hanging.
?
Any thoughts/suggestions?
?
73
Bob
WV7W
Re: Allow 28v on 50w QCX PA ?

 

What lead to my answer and comments..
?
Some 18 years ago I built a copy of the 'EBY amp.? During testing?
I pushed it... just to see.? ?At 85W the IRF510s died from excessive
heating.? That was near 160W DC input!!!? Also despite a copper
plate and?large finned heatsink the thermal resistance allowed
too much heating.? It was however running with efficiency of
about 50% which for a linear push-pull amp is decent.? Tamed
down to 25V it was driven to ~50W on all bands (37W on 10M)
into various (specialized mismatches for testing) loads and
proved durable if it didn't get too hot.
?
That amp is still part of the station on the same set of IRF510s?
and does a good job of turning 4W (with input attenuation on)
into about 45W. The attenuation was required to keep the input
below 2W or max power was exceeded with all the unwanted
benefits (splatter).? Does a nice job of converting my QRP radios
to respectable power.
?
There are limits for IRF510:
*4A continuous, more than that and heat and lead bond issues arise.
*~40W dissipation max, assuming large heatsink.
*100V is its max rating but never go near it.
*At 32V Vdrain that 100V rating is not stressed unless the load is reactive... boom!
*Gate voltage plus bias should be kept under 12V, at 20V it dies.
? Note the reason is that the feed back capacitance to the gate can be additive
? and exceed limits.
*Also never forget The IRF510 is a VHF capable device.? I use 8 of them (4x4 push pull)
? for over 200W at 50mhz.
?
Class C also has a full off state so the Drain voltage rings higher depending on load
and that is very hard on devices of not managed.
?
As to the diodes the forward on resistance (at RF) is a factor for power dissipated.
The reverse capacitance for the amp use is at 70 to 100V not 4V.? ?Less capacitance
is desired or you need to minimize RF going down the wrong path (the off switches)
and doing damage or creating a working feedback path.? While familiar with the
1n5408 and how W6JL used it one has to pay attention to things done.? That includes
added switching (to protect the RX) and higher (550V!) reverse bias voltages.? There
are multiple switches (3 series and one shunt for the RX path) and only one is the?
1n5408 (tx path).? That design is for a TR switch to the antenna not a amp level
switch that has to isolate the input from the output.? Really nice design for that
use.
?
A better example would be KPA500 Elecraft 500W) as that is a complete amp
with switching like the QRPL 50W amp.? As a result the switching there would
be similar.? ?Or W6JL has a 1500W amp with switching and that show the efforts
needed to switch high power while making sure the potential feedback paths do
not happen.? Note the use of real pin diodes and 4007s to supplement the 5408s
and the 900V reverse bias.
?
?
As they say, the devil is in the details.
?
--
Allison
------------------
Post online only,?
direct email will go to a bit bucket.
Re: Automatic Antenna Tuner, Internal Battery with a charger and built-in speaker for QMX+. Open Source.

 

Interested in a kit package.
if anyone here decides to put together a package kit of parts and PCBs I would like to purchase one.
Alternatively, I’d purchase a set of PCBs if someone here wants to order a few extra.
Glen, WB4KTF, Austin, Tx
--
73s,
Glen, WB4KTF, Austin, TX
Re: QMX died pulling out headphones

 

On Sat, Oct 19, 2024 at 09:02 AM, Daniel Conklin wrote:
Update:? Success!
Glad to hear it, Dan.
?
73, Mike KK7ER
?
Re: Automatic Antenna Tuner, Internal Battery with a charger and built-in speaker for QMX+. Open Source.

 

Hi,
?
I would be interested in buying 2 assembled, working :) units for a reasonable price.
?
Thanks ,
Sam - AC4OW
Re: Why The EER Method For SSB Generation Works

 

Thanks to Hans and all of you for clarifying this. Greatly appreciated!

Tony

On Mon, Oct 21, 2024 at 8:28?AM Scott Dupuie via <groupsio=[email protected]> wrote:
On Sun, Oct 20, 2024 at 12:21 PM, Hans Summers wrote:
It's true yes, that rhw phase noise when divided by 2 is lower. However the 5351 in quadrature mode is also dividing just the same way as you'd be doing if you were doing it externally. There's no difference in phase noise between the two cases, due to the division. The division is done somewhere anyway. It's either done in the '5351 or externally (such as by a 74AC74). Iss all the same, regarding total division ratios.
?
The reason why phase noise is likely to be lower if done internally in the Si5351A instead of externally is that generally speaking the more chips, PCB, wiring traces etc the signal goes through, the more noise it gets. Doing it all, under the control of SiLabs experts, in that tiny spec of '5351 silicon should be cleanest.?
?
If the total division ratio is the same (VCO runs at the same frequency in both cases), then I agree that doing it on-board the 5351 will result in the best performance.

FYI back in the days of the 2017 design when I developed the way of getting the '5351 to do quadrature LO, I did compare unwanted sideband level with the '5351 quadrature and the external 74AC74 Johnson counter way (x4 clock). The '5351 quadrature method produced considerably better performance. So that was an advantage as well as saving the cost and board area of a chip (74AC74). I didn't have capability at that time to measure phase noise.?
What you measured was phase offset accuracy. I can believe that the phase offset accuracy is better when internal generated in the 5351. The 74AC74 is designed for a maximum clock of about 50MHz and a propagation delay of up to 20ns. That is 3/5 of a period at 30MHz! Any mismatch in delay between the two paths will just kill you.
?
73,
Scott
AK5SD
QMX Dead - 3.3v Rail Shorted

 

This is my third QMX Build so not completely new to this.
?
I was building this one for a friend and?I fired up newly built QMX mid-band yesterday and loaded the firmware with no issues. Everything came up fine. I was able to navigate menus and change settings. All the stuff it should normally do.
The issue I was noticing was that into a dummy load, I was getting high SWR and no power out. Sometimes protection would kick in and other times not.?
?
I turned the power off to go do some errands and came back to hook up to terminal to do an RF sweep and nothing. No display nothing.
?
I thought that maybe one of my SMPS boards may have died so swapped them out of a working QMX and still nothing.
?
With both PS boards out, I checked the 3.3v rail and it is shorted to ground. I suspect that something died on the main board but figured I would check if there is anything else I can do before scrapping this one and starting on a new kit. I do have another mid-band kit that was supposed to be mine, but won't leave my friend hanging.
?
Any thoughts/suggestions?
?
73
Bob
WV7W
Re: Allow 28v on 50w QCX PA ?

 

Don’t do it. I blew the finals at 24v and they took the input of the QMX with them after a cascading failure of parts inside the amp. I had to replace an IC in the QMX after some sleuthing to figure out what broke.
?
If you do the math on it, 24 volts results in a DS voltage in excess of 100V, exceeding the rating of the transistor.?
?
?
Re: QMX reduced power function?

 

Thanks, Stan.? As you saw, I have been absent for a few months and have not kept up!
--
Peter Lee
G3SPL
Re: #QMX IC402 issue in combination with 50W PA? #qmx

 

This is a much likelier scenario than my speculation. I don’t know why I didn’t think of the simpler/more obvious explanation first.?

In any case, it’s relevant to reiterate that when I blew IC-402, it was when I blew the finals, and it certainly wasn’t during normal operation!?

I didn’t see any other posts in the forum about the 50W amp blowing the QMX. It’s nice that IC is easy to replace!?
Re: how to set presets

 

Really, what peter wrote here is the easiest way.
?
On Mon, Oct 21, 2024 at 12:36 AM, Peter HB9EBE wrote:
The easiest way for me is to log on to QMX terminal applications via putty - select configuration - select messages, and type or edit the messages on the keyboard.
?
73 Peter
?
--
73, Dan - W2DLC
Re: Why The EER Method For SSB Generation Works

 

On Sun, Oct 20, 2024 at 12:21 PM, Hans Summers wrote:
It's true yes, that rhw phase noise when divided by 2 is lower. However the 5351 in quadrature mode is also dividing just the same way as you'd be doing if you were doing it externally. There's no difference in phase noise between the two cases, due to the division. The division is done somewhere anyway. It's either done in the '5351 or externally (such as by a 74AC74). Iss all the same, regarding total division ratios.
?
The reason why phase noise is likely to be lower if done internally in the Si5351A instead of externally is that generally speaking the more chips, PCB, wiring traces etc the signal goes through, the more noise it gets. Doing it all, under the control of SiLabs experts, in that tiny spec of '5351 silicon should be cleanest.?
?
If the total division ratio is the same (VCO runs at the same frequency in both cases), then I agree that doing it on-board the 5351 will result in the best performance.

FYI back in the days of the 2017 design when I developed the way of getting the '5351 to do quadrature LO, I did compare unwanted sideband level with the '5351 quadrature and the external 74AC74 Johnson counter way (x4 clock). The '5351 quadrature method produced considerably better performance. So that was an advantage as well as saving the cost and board area of a chip (74AC74). I didn't have capability at that time to measure phase noise.?
What you measured was phase offset accuracy. I can believe that the phase offset accuracy is better when internal generated in the 5351. The 74AC74 is designed for a maximum clock of about 50MHz and a propagation delay of up to 20ns. That is 3/5 of a period at 30MHz! Any mismatch in delay between the two paths will just kill you.
?
73,
Scott
AK5SD
Re: L506 Winding Gap Improves Filter Sweeps

Robert AD6XJ
 

I used 30AWG wire.
11241 - 11260 of 141177