Ham Radio - tr(u)SDX splatter problem, further testing.
Vložit
- čas přidán 11. 11. 2022
- Trying suggestions from DL2MAN and other viewers.
If you like my videos, please subscribe here:
czcams.com/users/loughkb?s...
If you want me to keep making more videos like this, consider becoming my patron at Patreon: / kb9rlw
Discuss this and other videos on Facebook:
/ kb9rlw - Věda a technologie
My issue with your Videos is: You´re just "pushing" and creating publicity in my opinion. The issue could have been resolved immediately if you just had accepted replacement. You insisted, you wanted to fix it yourself, and here we are. I´d have appreciated, if you just had worked with us, trying to resolve the issue, and only then make a Video about the solution and the results. From a content creators perspective I can understand, as "pushing" creates content and clickbait. From a ham perspective, I think you´re mis-representing the whole issue and doing even a dis-service. 1. There´s definately a different picture remote vs. local. So jumping immediately to the conclusion your local measurements are right is a little bit fast at least. It clearly shows mitigation remote, yet there´s an issue with your (tr)uSDX. 2. The (tr)uSDX is designed for abt. 45dB of Sideband Suppression, and because of simplicity of schematic you shouldn´t expect IC705 Performance. So instead of looking at the waterfall, try to work with the dB readings in the histogram. Looking at TX Drive 2 tests, this looks already within -45dB requirements. However: There is abnormal behavior in your case. 3. Your signal is clearly unclean and noisy. That´s your real issue. Like something is switching on and off at the wrong time. Try working with 3.2 Noise Gate, as your input threshold is clearly off. You can also try different external mics and see if it changes anything 4. You´re txing full bandwidth (no restrictions). If you apply SSB Filters on RX, this will also limit your TX Bandwidth. Try to compare apples to apples, if you do an unfair comparison as tx bandwidth of Icom is also limited. That being said: I find, your videos are not representing the (tr)uSDX Project, and we could have helped you with your issues (as we tried). A guy with your amount of followers should be more interested, and has the responsibility in "getting it right" before creating a public false impression, that (tr)uSDX is bad to the bone. This clearly mis-represents the (tr)uSDX Project, and is harming the project, that changed a lot of peoples lifes. I know you don´t mean it that way, but I get mails from people, that are now insecure because of your issues. You have a lot of subscibers and your range is far, so please be aware of your responsibility. I´m more worried about the people that did not write me, because they just drew the wrong conclusions right away, as they believe you..... But that´s your decisions. I´m just giving my opinion here, but I will no longer be "pushed around" by such Videos like yours. That´s freedom of speech, but me disagreeing is also. Good Luck ! 73 Manuel; DL2MAN
Manual,
Ok, Lot's to address, I'll take it a chunk at a time.
You say:
"The issue could have been resolved immediately if you just had accepted replacement. You insisted, you wanted to fix it yourself, and here we are. I´d have appreciated, if you just had worked with us, trying to resolve the issue, and then make a Video about the solution and the results."
At the point in time that replacement of the RF board was offered, I thought it was fixed. The problems I'd had with stability and rebooting were resolved and I saw no need for the replacement board. I was unaware of the modulation problem at that point.
However, I would welcome a replacement RF board at this point as I'm somewhat convinced the problem lies in that board.
Also, this series is an experiment, sharing each stage of my exploration of the radio. Not just a polished finale, but each phase that leads to the finale.
You say:
"From a ham perspective, I think you´re mis-representing the whole issue."
I'm attempting to be as transparent as possible, sharing exactly what I observe. Nothing is misrepresented or altered. They see what I see.
Now to your numbered points.
"1. There´s defiantely a different picture remote vs. local. It clearly shows mitigation remote, yet there´s an issue."
The entire signal was lower on the remote. Far enough above the noise to be discerned but much weaker than in the local test. The edges of the splatter were not as clear, but still present. It wasn't that it was better, there was just less signal to see.
As to Point 2. I have seen a much cleaner signal from another tr(u)SDX with no visible energy in the suppressed sideband. I know the radio can do MUCH better than what I'm seeing. Even at TX drive level 1. And you do concede that my case is abnormal. In line with my repeated statements that the problem is with my radio's build.
As to point 3. That is helpful, I'll look at the noise gate setting. I suspected something like that when it was chopping off the audio all together on 80 meters. However, the degree of splatter varies greatly on the different bands, so I doubt a setting that applies to all bands would be the problem. It doesn't chop off the audio on 20 meters and hardly at all on 40 meters. Being frequency dependent like that leads me to think it's something wrong in the RF section
Point 4. Bandwidth. The audio is rolling off just under 2 KHz, so there's nothing to modulate above, the filter shouldn't matter. However, I will look into that.
As to your last statement,
"I find, your videos are not representing the (tr)uSDX Project, and we could have helped you with your issues (as we tried). "
As I have stated more than once in the vids, I'm sure the problem is in my unit and not representational of the radio in general.
I don't know that I can be any more clear.
I didn't feel I was getting help initially. When I emailed pictures of the bad soldering, you initially replied that you didn't see bad soldering. After I detailed it in the video, it was taken seriously and you worked with your builder to get that problem resolved. That was a good result.
When I observed the bad splattering, you told me it was artifacts from the digital processing only visible in the near field. That didn't make sense given the magnitude of the observed issue of clipping and splatter, but I gave you the benefit of the doubt and tested it anyway.
I will get to the bottom of the issue on my unit and share the solution eventually. I'm sure it will help others who build the kit to avoid whatever the issue turns out to be.
This also ties to your final statement:
"A guy with your amount of followers should be more interested in "geeting it right" before creating a public false impression, that (tr)uSDX is bad to the bone, which clearly mis-represents the (tr)uSDX Project. I know you don´t mean it that way, but I get mails from people, that are now insecure because of your issues. I´m more worried about the people that did not write me, and just drew the wrong conclusions....."
Again, I have stated publicly and in the vids, more than once, that the problem with my unit shouldn't be a reflection on the radio in general. I don't know how I can be more clear.
There will be people that jump to conclusions, I can't help that. I'm not hiding anything, I'm not misrepresenting anything. I'll pin your comment and my reply to the top of the comment list so your input can be considered as well. I'm just trying to be fully transparent in my exploration of this radio. Warts and all.
I'm sorry you're getting some flack, that is not my intention. When I come back to this project, I'll open the video with a strongly worded statement, again, that my issues are with my one unit and should not be considered a reflection on the radio as a whole. I've said, publicly, in the videos, I've heard from many people that have perfectly fine units without issues. And I've pointed it out yet again here.
On the subject of the radio's settings, What should the working defaults be for TX drive, Noise gate, Bias min and bias max? I assumed the out-of-box initial boot up firmware defaults should be close to correct, but that assumption could be wrong.
When I come back to the project, I'm planning on going through a complete alignment as one would if building the kit. I still think it's a hardware fault somewhere on the RF board.
EDIT: John, AE5X also did a test, looking at the spectrum and using two tone injection to measure IMD.
czcams.com/video/iPmv1g9rv7s/video.html
Luckily DL2MAN found a defective trusdx in his own stack and now recognizes the problem... Sorry that there is so much hate with the haters, ignoring them isolates them...
Love that you and Manuel are able to work together on this.
Manuel you are going to make yourself sick if you keep worrying about every little doubt about your fine work. The value of Kevin's video is showing some troubleshooting techniques, but all you see is "harm" to your project. Stop it (please).
@@KC3DJV they look at odds w/each other to me.....
Kevin, I applaud you for doing such good analysis.. I am well aware of how much effort is required to pull all that footage together. First class.
Thanks Kevin for your very well done analysis.
Simply let's all admit that the tr(u)SDX is an overly simplified DSP receiver and transmitter.
Don't forget that the origin of it is a technical challenge : to modify the QCX to make some real time signal processing, on a ATMEGA328P processor (a very low end processor from the Arduino). A bunch of trade-off to make this, using a 8 bits 20 MIPS processor without a DAC to generate the TX signal from I/Q vector.
This is genious to make it work, but it pushed the concept to the limits. Thus the RX and TX signal quality is somewhat limited.
73, F1SOC
Limited! Now you are over simplifying things, sir. It's terrible.
Hi there, this is in the RF stage, i know you now all this. Two Suspects in all this probably accounts for the hit-and-miss working correctly from one device to another, T1 and T2, pay attention to how they have been wound, even more particularly to start and end windings, make sure the start and end of the windings have been fitted correctly on the board, this will definitely affect the transmit. I would check the rest of the windings is on all the inductors, maybe even take a reading of what they are to what they should be. I would love to get my hands on one to build, maybe video and upload any findings. Thanks for all your time videoing and uploading. The DM (UK)
T1 and T2, got it. Thanks for the suggestion.
Great videos. I think I’ve seen all the ones related to this radio. I do own one. I purchased a pre-built unit from an authorized dealer. I’ve had a lot of POTA hunting success with it from my front porch with a wire in a tree. With that said, I have not tested like you have. I hope I’m not splattering on the bands.
I think what your doing is very helpful for the community and potential builders of the kit version. It helps them with troubleshooting steps when theirs don’t work quit well. Also, I printed the power pole holder, works well, ty for that.
I’ll keep watching and learning. 73.
It's worth the money for a great little CW rig, for sure. That's primarily what I bought it for and I have no complaints on using it for CW.
I love it when you tinker & troubleshoot with kits like this, Kevin. But, I would have already sent this one back to see if the engineers could do something with it! 🇺🇸 😎👍☕
Engineers! Why? Try fixing it yourself, you'll learn loads!
@@awaismushtaq5719 I learned loads when I began actively seeking anything else other than Chinese crap, which has to be 'fixed' right after it is purchased.
@@lilblackduc7312 LOL 😂😂😆! I love it. Too good! Natural!
@@awaismushtaq5719 I bought a vacuum cleaner years ago and made sure to get an extra drive belt while I was at it. I have parts for my coffee maker & window fan. 😎👍☕I know nothing in the last 20yrs is the quality we used to purchase from Sears. Being an "Electronics Junkie" since after I was born in 1958 just makes my outlook that much worse, Awais. So, there's a "thumbnail sketch" of my 'Model of Reality'...I've become a Daffy, lil' Black duck...🤪😜😁 🤣
@@lilblackduc7312 LOL! You are too good! Yes, and please forget about the quality your and my generation were used to back then, in good old Golden Days of Electronics. I'm pretty much a nerd in electronics, although I work in textiles but because of my undying love for electronics, I have spent last 18 years of my useless life in reading, doing and learning by other means about electronics as much as I could. I own a National two band radio receiver with Cassette Player. It was a masterpiece back in its time yet still serves its master with all its parts in tact although its service life expired about a decade and a half ago.
Wow! Great demonstration, and I was not aware you could perform such a remote test (WebSDR). The problem was immediately clear from the waterfall display, and it was nice to link _what I saw_ with _what I could hear._ Of course with the _sharp edges_ of digital you are going to have sidebands that need to be filtered - sadly power is also wasted in those sidebands.
I've used the Web SDR method myself when I've been told my audio is distorted like I have RF on it. Unfortunately in my case my signal came through perfectly on the Web SDR with no distortion so I'm none the wiser!
Hi Kevin, while this is beyond my understanding (technically), I have watched this series closely with great interest. As a result, I now understand things a little better! I am fine with through hole construction and the bigger surface mount, I think you may be onto something with the ability, or dare I say it, skill of the person who built the boards. Without knowing exactly what the approved builders actually do, it is hard to pin down where your problem lies.I do not think that shield will resolve the issue as this problem is not present on every build, and while shielding may sort it out, it will not cure the problem, but just mask it. At the end of the day, and bearing in mind my limited technical ability here, I think that this will only be solved by going to the component level. I await the next installment!!! 73 Jim M7BXT
The inconsistency of the results across users may suggest quality control issues related to manufacturing or assembly. Statistical process control has been used to effectively mitigate discrepancies in production rather than trying to correct later. It seems there is already rigor in the expected design performance, so focusing on specific production tests and associated test period may offer a potential solution.
Kevin,
I built the low band (original) version about a month ago and although I have only made 2 contacts they were at 933mi. (weak) and 714mi. (better), the latter complemented the audio. Both were using a DX Commander from the home QTH.
Herb W5TEK
"Some people see splatter, some don't."
But what percentage of owners are looking for splatter on a remote SDR or asking for honest signal quality reports?
It's possible a large percentage of these radios splatter their signal and the owners either don't know it or ignore it if they get poor signal reports.
I mentally filter that. After my vid looking at the audio quality, I heard from several that duplicated the test method. That was the dataset I drew from. Ignoring the purely anecdotal input.
My determination was made on the real data.
Hi Kevin, I have yet to see or hear clear SSB from a uSDX based radio. RX demos outnumber TX demos 10:1. Yes, of course, it is a fantastic effort to create a class E SSB transceiver with an atmega32, but it is extremely difficult to get right with plenty of processing power let alone an 8 bit micro.
same result if an external mic is used?
I really like your reviews and troubleshooting on this. Have you tried an external mic on it? And if so do you have any reccomendation?
Thx
Is there a decent sounding recording of the usdx TX signal? I haven't heard a single one.
I have another suggestion - can you test a digital signal (PSK31, RTTY, FT8?) That may give you some clues about the nature of the splatter cause. (DE WB0POH )
Tick, Tick, Tick.
Trivia for the members: What is that sound?
Kevin's just checking to see if that (tr)uSDX is throwing out Gamma rays...
Yep! I left it in to see how long it would take for someone to guess. :-)
🤦♂️it’s really hard to deny the rampant splattering problem on trusdx - the capability is still very impressive for the physical size but buyers really should just accept this problem as a hardware limitation.
Conscientious and responsible hams don't knowingly transmit malformed signals that can cause *radio interference* to other stations or services, that's why not.
Can the plastic case be the culprit? But then that wouldn't account for some radios with no distortion and others with it.
Yep.
Way to think it through. That's what ya gotta do.
Maybe that would be a good T-shirt. :-)
It's also putting the kabash on my shield idea.
Back to electronic flaw hypothesis.
(Tr)uSDX’s ancestor uses Hans’ QCX circuit, it has little intermodulation due to filtered PWM audio modulating a cascaded transistor in linear mode to supply modulated 12V to the class E BS170 mosfet drains for the restoration of the SSB envelope.
The current (Tr)uSDX has its audio filtered PWM and SI5351 clock output (square wave) tied together to the BS170 gates. which affects the BS170 transfer characteristics of linear and saturated regions in the same time, a simple explanation of your unpleasant experience with splattering noise.
It is better to address this issue to PE1NNZ, DL2MAN has nothing to do with it, he is just a designer of the Class E amplifier based on WA0ITP spreadsheet
Check the bias on the finals 🤔
Input audio filtration?? I haven't followed closely.
A couple of vids back I looked at the audio and signal in AM and FM. There was no distortion, just low modulation. I got right down on the mic and spoke loudly and the audio was still clear, no clipping. So it's not a problem with the mic or any analog circuitry behind it.
@@loughkb I'm referring to frequency roll-off both above and below audio spectrum limits. Not audio modulation amplitude (including high clipping). I need to eyeball a schematic, I may be talking out my butt (pun intended). Great videos by the way !! Been meaning to write for some time now.
I just sent you a FB MSG.. Hope I got the right guy... Or someone is going to think I'm as weird as I really am. ... Let me know what you think as I may be looking in the wrong place, it's getting late, I'm half blind and my brain don't work this side of darkness outside...
The audio in am and fm was rolling off below 2KHz. Nothing high to modulate. No, I'm pretty sure it's something in the RF section.
"Sausage being made" in a typical small Chinese PCB assembly sweat shop. The (tr)uSDX was probably assembled in a similar place. Notice how the workers handle the boards starting at 22:30. Even if the assembly job was given to a big-name Chinese assembly house, the little jobs get farmed out to the dozens of supporting sweatshops.
czcams.com/video/HF0E8WeEUVM/video.html
What about on FM or AM ? G7VFY
Two or three videos back in the series.
@@loughkb Have you tried using an audio signal generator with adjustable attenuation ? I don't suppose another microphone. Maybe putting some ferrites on the mic lead? just a thought.
I suspect a problem solder joint might be acting as diode and creating some weird mixing effect. G7VFY
Try doing something really old school--look at the signal on an O-scope. If the finals/driver is flat-topping, it will be obvious. If it's not flat-topping, then the finals/driver are not running in an AB or B bias region (effectively crossover distortion). de kg5ar
Hello Bill,
this is NOT a 'Class-A' or '-AB', nor a '-B' -radio!
Please RTFM before telling something like that, ... !
It's not a 'linear-amplifier' in this "PA" !
Only 3 small BS170-FET's, generating close to 5W, with an 'eta' of 0,8 and better, ... ! (Eta> 80%)
It is a little bit 'more complicated', not as 'rocket since', but nearly on, ... !
Modulation of phase and amplitude inside the frequency domain! Not a 'simple ring-' or 'diode-mixer', or something similar simple device. It's all a kind of DSP-formula inside the 'Arduino-IC' that generates the SSB-Signal.
And, ... don't forget, it's a simple experimental SDR-TRX, for less then 100€ . No one is advertising a 'Yae-Com-Wood' for +10k$ with 'absolute' preformance, ... !
Please hold this in mind !
73 de Markus ; db9pz
[ Owner and 'proud' user, of a 'homebrew' (TR)_U_SDX ]
I wanted to let you know that the YUMA HAMFEST has been canceled.
FECES!
I would look at the balanced mixer. That’s what it sounds like to me. I used to adjust PCM equipment in the military.
That's done in software on this radio, there is no physical balanced mixer.
@@loughkb IMHO, That would make me suspect it even more, since CW is fine. Balance in the mixing stage is extremely critical for carrier suppression and would affect bandwidth directly.
@@paulblock1367 You can't really apply classic analog theory here. They are creating SSB through software in an entirely different way. Nothing at all like a balanced mixer.
I will stick with name brand products like Yaesu FT-818 or Yaesu FT-891 and ICOM IC-705. Even Xiegu G90 has its quirks which make it unacceptable. I say don't waste time and money on second rate products.Thank You for exposing problems and saving others time, frustration and money.
Sounds like a bad cartridge on your Mac phone
Earlier in the series, I tested audio on AM and FM and the audio was clear and under modulated, even with me speaking much louder and closer to the mic. So the mic and any analog circuitry between it and the digital side is fine.
studio.czcams.com/users/video8PDcy7vYduY/edit?o=U
An update on the problem: czcams.com/video/iPmv1g9rv7s/video.html
Thanks for bringing that to my attention Louis.
For others reading the comments. This vid is John, AE5X, looking at the IMD of the output with two tone injection. Go watch, it's informative.
Sounds like the "designers" are in psychological denial. Doesn't help that they're getting smart with you about your testing.
Will be tough to help them while they hold that mindset, but I'm sure you'll figure out the root cause of the splatter in some of the assembled radios.
If I were DL2MAN, I would buy the defective radio from you at twice the price you paid and have it shipped to me via DHL so I could determine how the China assemblers are screwing some of them up.
This is a parade example of what I mean in my comment: Kevin just "pushes" and you immediately assume, it´s the designers fault. There´s a lot of inconsistencies in his testings and conclusions. That is denial.... I am DL2MAN, so I can tell you what I think, he should do: Accept he has a bad device, accept the replacement, that was offered by sunny right away, and be happy. If he wants to tinker: Good: But don´t misuse your high subscriber count to make it look, like what he´s showing is normal, and Kevin needs to save the world from us. Shame ! 73 Manuel; DL2MAN
No need to attack me Manuel. I think your anger at the comment by Louis above was misdirected my way there.
I have stated several times in the videos that the issues are with my radio and not representational of the project in general.
And, the replacement was offered and declined *before* the issue with modulation was discovered.
At this point, with the continuing issues I would welcome a replacement RF board.
@@dl2man You yourself are creating bad publicity for your product, VK3HJW
I'd toss it and find a different qrp rig. This one I have no plans of buying.
Nah, as I said in the vid, it's a great little CW radio.
I really don't understand why any of the tr(u)SDX tests were done without a mic or even the same mic. Really not crazy to me that the included on board microphone is potentially really bad.
I'm going to have to start copy pasting this reply.
As seen earlier in the series, when we tested AM and FM, the audio was clean and free of distortion. Even when I got close to the mic and spoke loudly.
So the mic and any analog circuitry behind it is not the problem.
🖐🏻73's🎙🎧KD9OAM📻
If you suspect build quality is a factor, surely it's now time return the unit to the designers, so they can check this aspect themselves and compare performance with their prototypes? You have have already shown this is not "finger trouble", and that you DO know what you're doing. Yes, I know diagnosis can be fun, and you'll get a few videos out of it, but life's too short to ruin your eyesight staring at SMT boards, And because you have a high profile YT channel, those guys must have a REAL vested interest in sorting the problem out!
The designers don't sell these radios directly. They authorize, endorse and promote specific overseas Chinese sellers to fabricate the PCBs, source the components, then assemble, QC and sell the units to the consumer.
If the assemblers in China are free to specify components manufactured for the domestic Chinese market, they may be of widely varying quality. It's my suspicion this may be the problem.
If there is a kickback to the designers based on the quantity of radios sold is not clear, at least not to me.
Kind of an odd arrangement they have. Puts them out of direct control of the manufacturing, but also shields them from direct responsibility for product defects, returns, and refunds.
Whatever the business arrangement, the designers contend that the signal spattering and intelligibility issue is a rare occurrence. And that most reports of it are either due to user error or improper test methods to observe the signal quality.
@@R50_J0 Yes I know a lot of the history, but surely they must have retained some sort of control over the licencing of their intellectual property? If not why are they now just giving money to Chinese manufacturers? Unless of course they simply sold the whole IP to a Chinese middleman "franchising company" for cash and washed their hands of all further involvement? As you say an odd arrangement, if you value your reputation.
Go back and watch the earlier vids in the series. Info from my first vid about the poor quality soldering is now being used to train the builders.
After my experience, Manuel requested samples from stock to further inspect their quality.
We put things in motion. I'm sure they're doing a better job now.
@@hectorpascal You're thinking along the lines of a traditional businessman. I suspect the designers don't have a strong profit motive, if any, nor are they trying to start a full time business with the radios.
They have come up with this oddball manufacturing and sales plan leveraging the cheap costs in China. Probably to avoid the time, effort, and responsibility involved running a business. The Chinese companies probably finance the materials and labor and are rewarded by sales driven by the designers marketing to a niche crowd here in the West. It's a win-win for everybody as long as there aren't quality problems.
Basically they seem like inexperienced, naive hobbyists trying to make a ham toy available to us at a cheap price without actually starting a business and all the hassle associated with one. The down side is a lack of direct control over the manufacturing and an unclear responsibility when there are product defects or complaints.
OK i look in my crystal ball and found that IT WAS DUN DELIBERATELY AS JUST ABOUT ALL STUFF (cheep low grade parts ect) FROM CHICOM HAVE YOU FORGOTTEN WE ARE AT WAR..
Thanks Kevin! I appreciate the issue you have raised as well as the brilliance and low-cost of the tr(u)SDX. It is truly a ground-breaking design. This issue illustrates my reticence to buy one and instead wait for the long delayed QSX from QRP-Labs. Hans puts an impressive finish on his products. The cases of the QCX-mini is beautiful and the feature set impressive. The QCX-mini isn't just a single-band CW transceiver. It does WSPR and CW beacon mode. It decodes CW and offers Iambic modes. It is rock stable and has built-in test and calibration features. Bought separately, the _added features_ would cost more than a QCX-mini sells for. Since I have a desktop SSB rig, I can afford to wait (and wait and wait) for the QSX. I know it will be worth the wait. qrp-labs.com/qsx.html