My 12 volt high band QDX is outputting 2.6 watts into a 1.3 SWR antenna on 10 volts and 0.78 maps.
I've considered trying to wind a "Really Weird Twisted sisters Transformer" for it, then wondered if the 'factory' assembled kits are being constructed with the new improved Really Weird Twisted sisters Transformer?
Is this transformer winding only applicable to the QDX?
|
Re: #80m Turning the 80 meter winding of T1 from hell to heaven
#80m
I should add two reasons not to use 61 ferrite on 80m filter, where the loss is low enough.
First is temperature coefficient. The permeability changes with temperature with all cores, but carbonyl iron dust cores, particularly #17, #6 and #10 are more temperature stable. When you are making an analog VFO oscillator, you want to use nonmagnetic phenolic core (best), or one of those three, usually. Low Q filter is much less sensitive, but ferrites have coefficient on the order of a percent per Kelvin while carbonyl iron dust has 50-150ppm/K range.
Second is the material's manufacturing tolerance and design's reproducibility (pre-tweaking). Ferrite cores' permeabilities are much less tightly controlled compared to carbonyl iron dust cores.
These may or may not matter but each application have to weigh in these factors and decide.
|
Thanks all! I will stick with 12 V. Ron?
|
QMX Really Weird Twisted Sisters Transformer Outperforms everything so far
This is the third QMX I have built, each design improvement outperforming the last. Now on first powerup and testing RWTST on 12.00 volts I get the following output into a dummy load as per putty diagnostics:
40 meters 5.3 watts
30 meters 4.9 watts
20 meters 7.0 watts (amps went from .093 resting to 1.01 transmit)
80 meters 5.6 watts
60 meters 6.1 watts
So now what do I do???? Spread the coils on L513? I know Hans' advice is no power above 6 watts on the QMX. But the RWTST transformer is doing such a great job I hate to start degrading things. I have also heard that up to 1.1 amps is to be expected. ....So only 1.01 amps gives me 7 watts on 20 meters. That's a good thing, right? Sandy KB3EOF
|
After nearly blowing my ear drums, I'm successfully using two?reverse-parallel silicon switching diodes as described here.?
I assume that the implementation of the AGC in the firmware is a lot more difficult than assumed, otherwise we would surely have it already....
|
Exactly. If none of those factors is your concern, and 12V is more convenient, there you have the answer.
|
Ryuji Suzuki AB1WX wrote: I wrote a detailed answer to that exact same
question maybe one week ago. Please search the
archive. The detailed answer to which AB1WX refers might be this one: < /g/QRPLabs/message/112235> There's a "View All X Messages In Topic" button near the bottom of the page that will reveal other contributors' answers to the question. 73 --Neil Ormos N9NL
|
Re: #80m Turning the 80 meter winding of T1 from hell to heaven
#80m
Fair Rite 61 material may be usable on 80m for the BPF, but you'll have to use a different core for higher bands anyway (the inductor becomes lossy very quickly, and with that, the "filter" will behave more like an attenuator as the frequency goes up).
Note mu' and mu''. mu' is the value that contributes to the inductance and mu'' the ohmic loss component.
43 material would be worse, and 31 even more.
|
JZ, all that theory is not necessary to explain Tony's data and it does not even explain the phenomenon in question.
Tony's 10m closeup traces are clean (look at the new one with averaging). His observation that the impedance peaked a bit lower than where the frequencies lined up (with a spike) is also very clear.
An explanation that requires a minimum additional complexity is what I described: winding cap, leakage inductance, PCB trace stray cap or a combination thereof.
But none of that is important for the purpose of understanding QMX operation or guiding user mods. Once again, sidetracked.
|
I wrote a detailed answer to that exact same question maybe one week ago. Please search the archive.
|
New Chat: Short Circuit showing up Between Center Pin and Shell of the BNC
#chat-notice
A new chat has been created:
- Just after completing assembly point 1.15, encountered a short when I tested the BNC install.? The Center to Shell shows connectivity.
- I removed the BNC, and it had been very cleanly soldered.? No sloppy connections or bridges.? I will replace it when I solve the problem.
- The two internal PCB leads shown on page 12 of assembly instructions 1.00j , between the center connector of the BNC to the T507 show appripriate connectivity.? I was able to successfully unsolder the 4 T507 wire leads on the BNC side, and they remain in that state for now.
- There is no sign of any abrasion related connectivity between the two pairs of .6 and .3 mm wires in either side of T507.? ie, no apparent electrical connectivity.
I have completed the Power boards and connectors, and have built the LCD screen assembly plus the control button assembly.? I have successfully completed the full assembly, with all components fitting as they should.? I congratuate the design team on their techno-origami skills.? the QMX is an elegant victory of design. I have disassembled the whole unit back to the simple main PCB board - since the last (apparent) problem seems to reside there. Beyond that point, I get lost in the circuit diagrams, trying to figure out where the circuit from the center BNC pin gets to Ground. From my end I can go back and re-flow all of the solder joints to ensure that they are all solid.? I can see no solder bridges on the solder side of the PCB. Is it possible that a poor solder joint somewhere could induce a path to ground.? ?? My tools are simple - basically a DMM, solder iron and other standard bench tools. I would appreciate a pointer to my next step to figure out what to check.? Kevin Woods VE7KHI
By:
Kevin Woods VE7KHI <kevin.kc.woods@...>
View/Join This Chat
Do not reply to this message to post to the chat. You can participate in chats only through the group's website.
|
I am new to the forum. I have built the original QCX as well as 2QCX minis, one for 20 m and one for 40. Now in the process of building a QMX and wondering which voltage I should choose? Everything I own iworks on 12 V, so I guess what I¡¯m asking is why the 9 V option? Thanks for any input. Ron.?
?
|
Re: QMX PC Connect, Anyone?
|
Re: #80m Turning the 80 meter winding of T1 from hell to heaven
#80m
The T50-2 has a permeability (basically a magnetic concentration factor) of 10, while type 61 material is greater, at around 75. ?Yes, that means way fewer turns, but each turn will have a very large influence on inductance achieved, and it¡¯s hard to hit a desired value. ?Plus the material losses are greater, so lower Q and it will warm up with any power, shifting inductance. So not really a good idea.
toggle quoted message
Show quoted text
On Oct 22, 2023, at 15:10, Robert Madore <robertmadore@...> wrote:
?Sorry I'm late to the party. Would the FT50-61 work for an LC/detector circuit for a 3.5 mHz receiver? (instead of a T50-2)
Thanks, Robert? W7NER
|
Yes, got that. Thank you. If you look closely at Tony's 10m plot, you may notice that near the true LO frequency all the data looks "wobbly". That is why I suspected an instrument issue of some sort.
Here's what it might be. NanoVNA devices use a bridge circuit as the basic measuring tool, but then pick off a signal from it and, using a second synthesizer as a tracking LO, down-converts that signal to a beat note in the audio range for subsequent amplification and measurement.
In this case though. we have three signal sources going: two in the NanoVNA and one in QDX, and they all mix it up inside the NanoVNA.
Previously it was speculated that LO feed-through from QDX might be sufficient to interfere with the NanoVNA. We concluded that was probably not so, based on the strength of the signal source in the VNA. But we didn't fully take into account the inner workings of the NanoVNA and how a third signal in the mix might mess things up.
Alan reported that the QDX LO feed-through which he observed was small but definitely not negligible. That may indeed be what has corrupted both magnitude and phase data in the immediate vicinity of QDXs LO.
JZ
On Sun, Oct 22, 2023, 12:38 PM Ryuji Suzuki AB1WX < ab1wx@...> wrote: JZ, Tony is saying that the frequency where the peak is expected is off, not the impedance measurement is off.
|
Hi Wayne. You can expect between 500 and 800 mA or so on transmit depending on band, voltage and power output it's adjusted for. I find generally about 550 mA at 12v on 40 meters with roughly 4 watts out and 750 mA on 20 meter QCX Mini at the same voltage and power. Your results will vary, but something in that range is probably normal ... Ron
|
Re: #80m Turning the 80 meter winding of T1 from hell to heaven
#80m
You generally don't want to swap ferrite cores for powdered iron. Having said that, you'd have to recalculate the winding count to get the same inductance.
--John WB5YOO?
?
On Sun, 22 Oct 2023 10:09:25 -0700, "Robert Madore" <robertmadore@...> wrote:
Sorry I'm late to the party. Would the FT50-61 work for an LC/detector circuit for a 3.5 mHz receiver? (instead of a T50-2)
Thanks, Robert? W7NER
-- --John R. Strohm WB5YOO
|
Re: #80m Turning the 80 meter winding of T1 from hell to heaven
#80m
Sorry I'm late to the party. Would the FT50-61 work for an LC/detector circuit for a 3.5 mHz receiver? (instead of a T50-2)
Thanks, Robert? W7NER
|
Another member recommended qrp guys and I second that. This is a good one:
This one works well:
This one is ok but I cannot really recommend it, seems glitchy:
72
--
Wayne S.
KF7RCM
|
Re: QMX PC Connect, Anyone?
Pierre, ?You will need to download and install Beta v6.49b in order for this to work:? ??After installing it:
?
To disable audio and squelch setting enter this in the Terminal app (copy and paste) and press return.
?
defaults write com.dogparksoftware.MacLoggerDX radio_enable_volume_and_squelch "NO"
?
To enable it enter this into the terminal app and press return¡
?
defaults delete com.dogparksoftware.MacLoggerDX radio_enable_volume_and_squelch
?
Entering either will not take effect until you restart MacLoggerDX. ?This affects any radio that uses the TS-480 driver, NOT JUST THE QMX.
?
-- Jim / K7TXA Eagle, ID SKCC 10447C BUG 301
|
Ryuji Suzuki AB1WX
Pierre Robillard VE3HQZ
Ronald Taylor
