Re: #qmx PCB Rev 2 #qmx

#112359

Hans,

Thanks for these improvements. Small but very useful.
ad 2: Have you in mind using another half for additional bands?

73 Bojan S53DZ

#qmx PCB Rev 2 #qmx

#112358

Hi all

We now began shipping QMX Rev 2 PCBs. The PCBs have the following minor changes, which are mostly about making the construction easier. There is no major performance or functionality enhancement so nobody?needs to upgrade from a Rev 1 PCB to a Rev 2 PCB. These are just minor improvements, not a major revision! There are always some people who get all panicked about PCB revision changes - please don't, these changes are so minor. I often make these kinds of minor changes each time I do a new PCB batch, just to incorporate customer feedback and observations on improving it. But they're usually MINOR. Just ongoing product improvement.?

1. Move trace that's under the screw for the PA heatsink

There's a?+5V trace that goes under the bolt of the PA heatsink. I have never heard of even one instance where someone over-tightened the bolt, causing the solder mask to be scratched off and a short from the?+5V trace to ground. However it seemed wise to move the trace a little just to sleep better.?

2. Use only half the BPF MUX switch

As per Steve K1RF's suggestion and discussion on this forum; now only half the IC402 74CBT3253 MUX switch is used for BPF switching. See attachment. This slightly reduces losses (due to the non-zero ON resistance of the switches) and reduces parasitic capacitance due to the capacitance of the switch inputs. As I mentioned yesterday, I have not seen any particular benefit from this change yet, I wasn't able to see any difference at all; but I will keep looking.?

3. Add 5.1K resistors at USB CC pins

People using a USB-C to USB-C cable reported this issue where the host computer also uses a USB-C connection, and in some cases the host computer tries to test the attached equipment, to find out whether it should supply power or not; it also decides to determine whether the thing at the other end is going to be the bus master (host) or bus slave (device). The USB specification is not very easy to understand but this does appear to be a correct interpretation. So on the Rev 2 board there are now 5.1K pull-down resistors on the USB-C connectors CC pins.?

4. Move capacitor C204 that is near the mounting hole for the left SMPS board

Just a small distance move to reduce the risk of damaging that capacitor C204 which is near the mounting hole (though I have never heard of an instance of this).?

5. Make 0.6mm wire holes larger on binoculars

On Rev 1, the wire holes are a little tight on the 0.6mm sized wire; so I made the holes a little larger to make it easier to install the 0.6mm sized wire.?

6. Enlarge some 3.5mm jack holes very slightly to make it easier to adjust their position

The 3.5mm jack connectors don't want to sit square with the board edge naturally, something about the Eagle CAD library component differing very slightly from the actual device; so I enlarged the pad holes slightly just to make it easier to precisely align these connectors square with the board edge.?

7. Move part of the power connector +12V pad to avoid nearby short

The 12V power connector center pin (+12V) is rather close to the 2x4-pin SMPS connector, and it is possible to create a short between the supply connector pin and one of the SMPS connector pins. I can't now move the SMPS at all and neither is there space to do so anyway, but I was able to shape the copper pad of the 12V pin to increase the distance and make a solder bridge much more unlikely.?

8. Make the ground connections of the rotary encoder pins easier to solder

Soldering the rotary encoder pins, particularly the ground pins, has been rather tricky; the reason for this is that although the ground pads in the design all use the normal good practice "thermals". which means there isn't a solid connection from the pin pad to the ground plane, there's a gap except for four traces at 0, 90, 180 and 270-degrees; the problem with the 6-layer board is that 6 x 4 = 24 and those 24 traces still conduct plenty of heat away. That combined with how close the rotary encoder pins are to the board edge, and to the body of the rotary encoder itself, made this part of the assembly a bit tricky. So I disconnected the ground pins from the ground plane except for one single trace; it makes them easier to solder now.?

9. BS170 holes slightly larger

BS170 pin pad holes were enlarged slightly, just to make it easier to clean out the holes if you do find yourself needing to replace BS170 transistors.?

10. Possible SMPS short area: moved transistors further apart slightly to avoid risk

The famous Q103/Q104 transistor short on the 5V SMPS board! This is not a practical issue for a few months now, not for you, because once the problem was known, we started testing all the boards here for the short, and when found, I corrected them all to remove the solder bridge. But on Rev 2 the transistors are a little further apart so there isn't a chance for the solder bridge to accidentally form during the assembly process. Well anyway we are still testing for the Q103/Q104 short too, just to make double sure.

11. C204 was 100n now 1n

This is in case it becomes desirable to make faster measurements of the supply voltage.?

12. Move MIC leads a little apart

The electret microphone on the controls board, is slightly different from its Eagle CAD library model, meaning during installation it was necessary to spread the mic legs a bit. On the Rev 2 board, I put the holes a little further apart just to make it easier to get the mic in.?

13. Tidy up spacing of C401/2/3/4

Just purely cosmetic, the four capacitors were unevenly spaced and I found it unsightly (as did Ross EX0AA who reported it!).?

73 Hans G0UPL

Re: SPAM Re: [QRPLabs] Trifilar toroid T401 continuity #qmx #toroid #troubleshoot #troubleshooting

#112357

I took a look at the datasheet for IC402 and IC403.

The voltage divider appears to be an RF termination network (suppress reflections and such) rather than a bias supply.? I kinda wonder why you need transmission line termination between the two ICs, and I'm still wondering about the frequency response of the network.

--John WB5YOO ?

On Tue, 17 Oct 2023 00:48:16 -0400, "John R. Strohm" <strohm@...> wrote:
?

Looking at the receiver board schematic, if T401 is on the board, you should see DC continuity from pin 7 on IC402 to pins 7 and 9 on IC403.

The time to do a DC continuity check on a trifilar winding is BEFORE you put it on the board.? At that point, you can verify that you have three independent windings.

Pin 7 on IC402 goes to the "hot" side (dotted) of the first winding.? The cold side of the first winding is AC grounded through C406, and gets 1/2 Vcc bias through the R401/R402 voltage divider network.

The hot side of the second winding is connected to the cold side of the third winding.? This is what produces the balanced output to pins 7 and 9 of IC403.? This common point is connected to the AC ground/DC 1/2 Vcc bias network described above, meaning you will have continuity on the board from the hot side of the first winding to the common point of the second and third windings, and from hence to the far ends of the second and third windings, giving you continuity from pin 7 of IC402 to pins 7 and 9 of IC403.

In short, the continuity you are seeing on the board is perfectly normal.

NOW, the Power Supplies schematic shows Vcc to be 5.6V.? 1/2 Vcc is 2.8V, so standing current in the voltage divider resistors, ignoring bias current draw, is 280 uA.? That's not a lot.? Without looking at the datasheets for the ICs, to see what they might actually draw on the bias leads, my first reaction (left over from bipolar transistor land) is that this is probably not stiff enough.? I'd've put a buffer on Vcc.? I'd probably at least glance at using a TI "rail splitter" IC to develop the 1/2 Vcc bias voltage, and, again, I'd worry about available current.

I'd also worry about frequency response of that RC network.? The tradeoff is frequency response vs. charge/discharge time, which controls regulation of the bias voltage.

--John WB5YOO
?

On Mon, 16 Oct 2023 21:12:59 -0700, "K3KDW via groups.io" <r.gammon@...> wrote:
?

I have continuity between the left A, B, C, and pin 7 as noted in the instructions. The right A, B, and C, share some continuity with the left. Is this correct? I expected these would be independent.

--
--John R. Strohm WB5YOO

[OT] High Efficiency class E/D amplifier

#112356

Hello everybody,

I'm looking after a design, kit or final project of a medium power (up to 50W but no more than 100W and no less than 30W) amplifier from 80 to 10m with high efficiency such as class E or D.

I saw few schematics and experiments in this way but either was only for narrow bands or does not reach (in reality) 28MHz or even 14MHz.

So, I'm asking the group to help with information about such a PA (schematics, project, kit, or final product).

Thank you.

73 de YO3FFF

Cristi

Re: BARB's FDT86256

#112355

DL2ARL the french 2CV was a fiscal calculation and not something you could easily convert into Watt. I Think the Citrióen 2CV had about 18 HP
But I like the virtual image of 2 sturdy horses pullig that car.

Martin
DK3UW

Re: QMX High swr into 50 ohm dummy load

#112354

Hi Mike,
I think you should make new windings. Secondaries have to be made very tight and turn by turn. You have to get maximum possible space between the primary and secondary in the total length of the hole.?

And, as Kees says, rigid coax would be the best, although it is more difficult to do.

GL 73 Bojan S53DZ

Re: QMX image rejection / rf sweep good enough?

#112353

On 17/10/2023 07:18, roebi@... wrote:
Should I try to get it better, and if so, how/what should I try to modify (BPF + LPF)?No, it is fundamentally a matter of matching phase and amplitude after the mixer.
All components are SM and fixed.
Hans has said one day it may be automatic, otherwise you need to be a dedicated tweaker.

73 Alan G4ZFQ

Re: QMX image rejection / rf sweep good enough?

#112352

Should I try to get it better, and if so, how/what should I try to modify (BPF + LPF)?

Thanks!

Re: QMX remains dead

#112351

What is the exact name of the firmware file, including the extension, that you are copying onto the mass storage device?
--
Compton
VK2HRX
Sydney, Australia

Re: QMX remains dead

#112350

Hello Brian

If it shows up as a mass storage?device that means it is in firmware update mode. In that case it won't connect with PuTTY, until you copy in the firmware. My apologies if I missed anything you posted earlier on this problem.

73 Hans G0UPL

Show quoted text

On Tue, Oct 17, 2023 at 1:21?AM Brian Lewis <Dazeoff@...> wrote:

I’ve rechecked all my solder joints 4 times now, it appears to load the firmware but then nothing, no backlit display or characters on the display (I tried the contrast adjustment, nothing). I try to connect with Putty and it won’t connect. It shows up as a mass storage device but it does not show up on a com port on my windows10 computer. Checked the encoders and they are correctly installed and functioning properly. Any further things to check? ?

Re: Trifilar toroid T401 continuity #troubleshooting #qmx #toroid #troubleshoot

#112349

Glad I could help.

It was a good exercise for me, as well.

--John WB5YOO ?

On Mon, 16 Oct 2023 23:09:26 -0700, "K3KDW via groups.io" <r.gammon@...> wrote:

Thank you John! I've tested everything and it looks good so I will proceed.

?

--
--John R. Strohm WB5YOO

Re: QMX High swr into 50 ohm dummy load

#112348

Mike, the fact that your "after" SWR gets worse at lower frequency suggest to me that something probably broke or made an unwanted connection in the process.

Re: Trifilar toroid T401 continuity #troubleshooting #qmx #toroid #troubleshoot

#112347

Thank you John! I've tested everything and it looks good so I will proceed.

Re: WSPR in Mexico

#112346

Does anyone have information about WSPR in Mexico that is current?Paul

XE2ITY 	has been on 60m I think continuously for quite a while.
A quick search on all bands finds no other.

73 Alan G4ZFQ

Re: Trifilar toroid T401 continuity #troubleshooting #qmx #toroid #troubleshoot

#112345

Looking at the receiver board schematic, if T401 is on the board, you should see DC continuity from pin 7 on IC402 to pins 7 and 9 on IC403.

The time to do a DC continuity check on a trifilar winding is BEFORE you put it on the board.? At that point, you can verify that you have three independent windings.

Pin 7 on IC402 goes to the "hot" side (dotted) of the first winding.? The cold side of the first winding is AC grounded through C406, and gets 1/2 Vcc bias through the R401/R402 voltage divider network.

The hot side of the second winding is connected to the cold side of the third winding.? This is what produces the balanced output to pins 7 and 9 of IC403.? This common point is connected to the AC ground/DC 1/2 Vcc bias network described above, meaning you will have continuity on the board from the hot side of the first winding to the common point of the second and third windings, and from hence to the far ends of the second and third windings, giving you continuity from pin 7 of IC402 to pins 7 and 9 of IC403.

In short, the continuity you are seeing on the board is perfectly normal.

NOW, the Power Supplies schematic shows Vcc to be 5.6V.? 1/2 Vcc is 2.8V, so standing current in the voltage divider resistors, ignoring bias current draw, is 280 uA.? That's not a lot.? Without looking at the datasheets for the ICs, to see what they might actually draw on the bias leads, my first reaction (left over from bipolar transistor land) is that this is probably not stiff enough.? I'd've put a buffer on Vcc.? I'd probably at least glance at using a TI "rail splitter" IC to develop the 1/2 Vcc bias voltage, and, again, I'd worry about available current.

I'd also worry about frequency response of that RC network.? The tradeoff is frequency response vs. charge/discharge time, which controls regulation of the bias voltage.

--John WB5YOO ?

On Mon, 16 Oct 2023 21:12:59 -0700, "K3KDW via groups.io" <r.gammon@...> wrote:
?

I have continuity between the left A, B, C, and pin 7 as noted in the instructions. The right A, B, and C, share some continuity with the left. Is this correct? I expected these would be independent.

--
--John R. Strohm WB5YOO

QMX suffers zero negative impact from deleting L401 (with measurements)

#112344

In previous episodes of this saga -

* L401 has a habit of producing a null in the 20m band if you wind it wrong
* I wound it wrong
* For giggles, I deleted it from my QMX. Suffered no ill effects, posted about it.
* Was told I would surely awaken the wrath of the RF gods in doing so
* I pointed out that I was already operating on the air with it and encountering zero wrath
* Was told I must not be operating near any high power AM radio stations
* I decided to come back with data

Here's my data:
is the AM radio environment where I live. At night I have one 15kW station 11km away, a 25kW station 28km away, and a 50kW station 33km away. Whether you consider this a challenging RF environment is up to you, but I'm not exactly out in the sticks.

I decided to use a NanoVNA as a signal generator. Its output is approximately -22 dBm. I set it to "CW" stimulus mode at 3.501 MHz, and plugged it directly into my QMX, tuned to 3.5 MHz:

1

The frontend is complaining quite visibly! Now let's tune the NanoVNA to 1700 khz, the highest frequency in the AM radio band:

The noise floor is now jet black. One should note that the output of the NanoVNA is a square wave, meaning there is also a strong second harmonic at 3.4MHz, 100khz away from the passband.

I then unplugged the NanoVNA and hooked up a 50 ohm termination to see if there was any change in the noise floor:

To my eyes the noise floor looks identical to when I was pushing -22dBm of out-of-band noise into it.

Finally, I hooked up my antenna. It's a busy night for FT8!

In summary, one or more of the following are true:
* The Tayloe detector is very good at rejecting out of band interference.
* The low frequency rolloff provided by C401/C402/C404 (whichever is currently selected by IC402) is more than adequate to prevent AM radio stations from causing out of band interference.

Further comments:
* I still have no idea what anyone thinks L401 is supposed to accomplish in terms of filtering out signals below the passband, considering that it is a series inductor, whose effect is to produce a high frequency rolloff. If anyone can show me a SPICE simulation which demonstrates how the addition of a series inductor can produce a low frequency rolloff, I would very much appreciate it.
* I believe that in practice, my antenna tuner also produces a band pass effect which further attenuates out of band signals. This is on top of the fact that the antenna itself is only resonant at certain frequencies, and is not large enough to be resonant in the AM radio band.

Re: New QMX has not transmit power #qmx

#112343

开云体育

You're good, for 1 & 2-- The windings of T401 and T507 do connect together on their respective component.

Paul -- AI7JR

On 10/16/23 16:08, Kevin Woods VE7KHI wrote:

Show quoted text

I'm still mid-build on the QMX.? I have the same symptom showing up - being a short circuit between the center of the BNC and the ground / outer shell.? Just checked it by accident, and drew some instant emotion!? I examined the center pin under the BNC fitting, and it goes through the PCB cleanly, so no chance of a solder bridge.? The Solder connection on the back is beautiful and clean.

Since I cant see inside the multi-layer PCB, it holds many mysteries.? 2 other (perhaps related) anomalies that arose were:

1) when I connected the Trifilar winding, all AA BB CC continuity checks were perfect, and solder joints were clean.? BUT, I not only got continuity between AA BB and CC, but all 6 connections showed reciprocal continuity.? Is this normal?!

2) Similarly, when I wound the T507 Binolcular, AA BB CC and DD all checked out prior to install, but after soldering (also cleanly) all 8 terminals showed mutual continuity -? one-and-all!

Can I proceed with 2.16 - Install Pin Header Connectors, or do I have to Down-Tools until the mystery is solved?

Does this point to a DEEP problem somewhere, or should I just proceed, and "it all comes clean in the wash"?? ?Does the SDR magic make sense of it all??

?Kevin Woods,
VE7KHI - BC

Re: New QMX has not transmit power #qmx

#112342

I'm still mid-build on the QMX.? I have the same symptom showing up - being a short circuit between the center of the BNC and the ground / outer shell.? Just checked it by accident, and drew some instant emotion!? I examined the center pin under the BNC fitting, and it goes through the PCB cleanly, so no chance of a solder bridge.? The Solder connection on the back is beautiful and clean.

Since I cant see inside the multi-layer PCB, it holds many mysteries.? 2 other (perhaps related) anomalies that arose were:

1) when I connected the Trifilar winding, all AA BB CC continuity checks were perfect, and solder joints were clean.? BUT, I not only got continuity between AA BB and CC, but all 6 connections showed reciprocal continuity.? Is this normal?!

2) Similarly, when I wound the T507 Binolcular, AA BB CC and DD all checked out prior to install, but after soldering (also cleanly) all 8 terminals showed mutual continuity -? one-and-all!

Can I proceed with 2.16 - Install Pin Header Connectors, or do I have to Down-Tools until the mystery is solved?

Does this point to a DEEP problem somewhere, or should I just proceed, and "it all comes clean in the wash"?? ?Does the SDR magic make sense of it all??

?Kevin Woods,
VE7KHI - BC

Trifilar toroid T401 continuity #troubleshooting #qmx #toroid #troubleshoot

#112341

I have continuity between the left A, B, C, and pin 7 as noted in the instructions. The right A, B, and C, share some continuity with the left. Is this correct? I expected these would be independent.

?

Re: Problems with QMX CW sidetone level

#112340

It is a good question OM and it has been discussed here more than once....

It doesn't actually work atm.

AND it is on the "to be fixed soon" list.

es vy 73 de "baab" w9ya