Updated, below. 27 Feb
The good:
- 23-27 watts out on all bands 10-80 (10 watts on 6m) with a LiFePO4 battery as DC source
- Harmonics down at least 55dB on most bands (-48 on 6m; -39 on 30m)
- Excellent receive selectivity, ie sharp skirts and no ringing, even at 100Hz
- Excellent fidelity from the built-in speaker
But then this:
The RIT will not stay activated during transmit - which completely defeats the purpose of having RIT. In other words, RIT is turned off by the act of transmitting - and stays off. Fixed by new firmware.
This is true in both CW and SSB.
On several occasions during this morning's POTA activation, I needed to adjust for a calling station's frequency being slightly off. I enabled the RIT, turned the dial slightly until the pitch was where I wanted it.
So far, so good.
But as soon as it was my turn to transmit, a single hit of the paddles immediately turned off the RIT and centered it back to zero offset.
My power output was set to 20 watts and the antenna was only 20 feet (6.5m) away so I thought that RF might be causing the problem. Sadly, no.
Even with only 1 watt into a dummy load, the problem still occurs, leading me to believe that the fix may be as simple as a firmware update.
I've emailed this info Yu BG2FX. I'll report here with any news of a fix for this issue.
For now, FX-4CR's and FX-4L's must be considered as not having RIT functionality, even though there is a button labeled "RIT" on the front panel.
This problem has been verified by another FX-4CR owner.
There are other non-trivial issues with this radio that I will eventually write about but this one is significant enough to deserve its own posting.
UPDATE 25 Feb:
The other (more) major issue:
Here are the RBN results of calling CQ at 22+ wpm as AE5X. Leading dits are omitted roughly 30% of the time.
Other FX-4CR owners (CW ops) can check this for themselves via a WebSDR by sending a character with a leading dot and listening to how it sounds on the distant receiver. An A becomes a T, a 5 becomes an H, etc.
Yu is aware of this and says that a FW update can fix it. I think this one will be a tough nut to crack since it involves the rx-to-tx transition.
Just to be clear: not all dots are affected - only those that are the initial characters causing the radio to go from rx to tx. And then, only 30% of the time, at higher CW speeds.
This, and other problems, are being addressed here.
 |
| Results of calling CQ with my callsign (AE5X) |
It was this issue that caused me to ask for a refund.
This radio may perform well on phone or digital modes - other reviewers can address that.
T/R keying and RF Out:
The T/R delay can be adjusted. It is not PIN diode QSK but it is quiet. The housing is so substantial that the relay would have to be very loud indeed to be heard by the operator. Fortunately, that is not the case.
As for keying artifacts - you will hear two varieties in the speaker/headphones: a slight "popping" sound after transmitting, when the rig switches back to receive, and a similar sound (infrequently) while transmitting (CW).
Finally, on a few occasions, the radio reverted to a 3-watt output even though I had dialed in 15 or 20 watts. This is not an issue of high SWR - it just seems to be random.
The issue with spurious emissions is dealt with here.
Is there anything good to say about this radio?
I've looked at this radio strictly from the perspective of being a CW op. For those who intend to use it for SSB or digital modes, it may be just fine as is. Many of the problems I've mentioned won't apply on those modes.
Ironically, CW is the simplest mode, and yet the most demanding in terms of the design required of a transceiver.
Also, the receiver is much more sensitive than the -120dBm MDS mentioned on Yu's website. And even better is the selectivity. I was amazed at how strong signals disappeared entirely when I selected a narrow (50 or 100 Hz) bandwidth and then slowly started to tune away from the signal. The skirts are sharp - razor sharp - and the actual bandwidth seems to match its stated value.
A big plus - maybe the biggest of all - is the speed and dedication with which Yu is addressing these problems.
All my criticism of the FX-4CR can be distilled down to this one statement: this is not (yet) a rig for a CW op. I believe that will soon change.
.
.
The RIT also does not work properly on my unit as well. Additionally I have been unable to get FT8 to function properly. I was able to get some FT8 receive to work at one point, but no one was able to hear my transmissions. Currently I cannot even get windows to recognize the radio so FT8 is a no-go. Finally, using earbud or headphones through the radio DOES NOT turn off the speaker. I have been successful with SSB contacts and CW contacts. The radio is solidly built but has some serious flaws that hopefully can be corrected in a firmware update.
ReplyDeleteAB4BA
Whether this issue gets corrected or not will speak volumes about any future problems and the willingness of the designer to address them.
DeleteI've always avoided Xiegu for exactly this reason - unfinished products, lack of support and customers-as-beta-testers. I think Xiegu has gotten better - they at least give the appearance of trying. Hopefully that trend toward better follow-up will hold true with products from BG2FX as well.
73,
John
You need to install the CM108 sound card driver, otherwise it will not work，Please rest assured. I will fix these problems as soon as possible
DeleteYu has emailed me and is attempting to rectify the issues.He is inviting reports on any other problems associated with the FX-4CR.
ReplyDeleteSorry I don't speak English, I use google translate and I saw the link to the video you posted. I understand what you said about RIT issues. This one is easy to fix. I hope you can ask more questions and I will fix them as soon as possible.
ReplyDeleteI welcome more people to ask questions so that the FX-4CR can be used better. I'm not going to shy away from problems. And you said that the output power in all bands is high, which is a serious threat to heating, you can set 20W power, after CW manual transmission or press PTT to maintain full power output, while rotating the VOL encoder counterclockwise to reduce the output power to around 20W, 6 meters output power should be limited to 5W
73 YU BG2FX
Thank you, Yu - I will limit power to 20W. Looking forward to a firmware update for the RIT issue and I appreciate your reply. The FX-4CR has no competitor: 10 bands, 20 watts and very small - I believe you have a fantastic radio once these issues are resolved.
DeleteThanks again,
John
I guess one man's fatal flaw is another man's irrelevancy. It's funny, my first thought was RIT? Radios still have RIT? Seriously, I can't remember the last time I used RIT. Back in the day, it was necessary due to drift and guys calling off frequency. But today, even cheap HF radios don't drift and at least on SSB, if a guy is off frequency it's probably because he left HIS RIT on! But as John has said, on CW people will sometimes call significantly off frequency. Why, I don't know. It's simple. First, make sure your RIT is off :), then tune the station you want to call until its audio tone matches your radios' sidetone. Then call, you will be on frequency. If you are sort of tone deaf and can't do that, make sure your CW filter is on and peak the station you want to work in the passband. You will be on frequency (or close enough).
ReplyDeleteNow I am not saying the RIT shouldn't be fixed, it should. But to me, it doesn't render the radio useless. At least to me, the question is does the Split Frequency function work properly? If it does, you can use this until the RIT is fixed to be able to tune off your TX freq for off freq callers if necessary. But more importantly, without a functional Split function you can't work rare and semi-rare DX stations that use that function easily.
73, Kevin K3OX
I intentionally call significantly off-frequency myself sometimes, even when a station isn't operating split. I do this if the station has other multiple callers who are zero-beating the station of interest. Calling 200-400Hz down from zero beat makes my pitch a bit higher in the other station's headphones, differentiating me from the horde.
DeleteI learned this by being on the receiving end of a pile-up (yes, I do generate them at times during POTA activations) and then being called by AD1C, an Honor Roll CW DXer who now chases parks, and knows how to do it. A cacophony of callers, all roughly the same pitch - because they zero beat me - and then AD1C at a high pitch, easily discernible, even though he may be weaker than many of the other callers.
He may be operating split with dual VFO's but it would be quicker and simpler for him to do it by moving his VFO down a bit, then tuning back up to me with RIT. Being an efficient op, I'll bet that's what he's doing. It's how I do it.
Bottom line though, I need RIT to work in a rig that has it ;-)
73,
John
As a died-in-the-wool search 'n pounce CW operator, the button I use the most on my Elecrafts is the 'SPOT' button, which faithfully zero-beats me to the station of interest. So, for any CW rig to be of any use to me, I need a functional 'SPOT' button that doesn't force me to try to match my sidetone to the received tone. Funny, after a dozen years of piano lessons, I think I'm still tone-deaf. N8SBE
ReplyDeleteWhat about stations who don't zero-beat you - do you use RIT?
DeleteIn addition to the above issues, Windows does not recognize the CAT/USB port or the onboard sound card (so no FT8). A search for the CM108 sound card driver gets you a noncompatible Windows Vista driver. I would hope Windows 10/11 would have a compatible driver for the onboard sound card. The bigger problem is if Windows cannot recognize the USB port, then firmware updates would be impossible on this radio. I attempted the STM32Cube programming software but again it does not recognize the USB port on the radio. I'm not sure if this is only an issue with my radio or it is more widespread. I too have emailed BG2FX for suggestions. To be fair, I have made multiple SSB and CW contacts with this radio so it mostly works for me. This is a 1.0 version and there is no current manual. Welcome to the beta test program.
ReplyDeleteSame here, Brad - STM32Cube unable to recognize the radio. And, as you mentioned, the driver download Yu mentions is for Vista & XP. I'm using W11 here. The firmware fix for the RIT will be impossible to implement without USB recognition...
DeleteI did an activation yesterday with the FX-4CR (2022 MacBook) and some testing today (Win11 PC, Win10 PC and Ubuntu laptop) - all worked without issue with the USB sound card and didn't have to install any drivers (for the sound card). The only driver I had to install was for the Prolific serial port. Not sure what I'm doing different from you guys :shrug: -Michael, N1XRR
DeleteGood question - hopefully the upcoming FW update will go as smoothly.
Delete73,
John
Just a FYI. The FX-4CR manual is up on the BG2FX website.
ReplyDeletehttps://img1.wsimg.com/blobby/go/15752049-75c0-4c53-9b88-c4e3901e4f90/FX-4CR-User-Manual.pdf
DeleteI contacted Yu, BG2FX, and he provided testing procedures for my USB port. We confirmed the port is dead. Yu is arranging a replacement radio. Additionally he has a solution for the RIT bug and the speaker/earphone bug which will be available soon. In summary, the USB port problem was likely unique to my radio, and Yu is providing excellent service to make it right. What more can you ask.
ReplyDeleteFantastic! I do wish beta testing would take place behind the scenes, before a new radio is offered for sale. These things would have been found if only the tiniest amount of a look-see would have been given by a 3rd party.
DeleteThere's a dollar amount, below which I'll accept some "issues" but it is much less than $550.
73,
John
I've been a beta tester for several rigs - from the US, the UK and Bulgaria. They were loaners and the conditions were all agreed upon first. I was confidential with my findings and, by the time the rigs were available for sale, they were pretty much finished products. I don't expect to be a beta tester for a rig I paid full price for, regardless of which country it's from. For me, the RIT-goes-away-when-transmitting issue wasn't so much about that, per se, but rather the lack of QC that failed to find that (very obvious) fault. It made me wonder what else might be lurking - and I found it.
ReplyDeleteThis is not meant as a criticism against Yu - I have the utmost respect for his response. But either I suppress what I've found and let others buy the rig when they might not want to if they knew a few details, or I treat the rig with a blind eye, let others buy the rig and then they've spent $550 they'd have rather kept.
Here's the issue I'll show in an upcoming YT video: leading dots are omitted (30% of the time) from transmission. You'll hear them in the sidetone, but they don't get transmitted because they're lost in the Rx-to-TX transition. This only occurs at 22wpm and faster. For me, it's a show-stopper. My callsign is not TE5X and my name is not Oohn.
You can test this for yourself with a 22wpm+ code speed and your transmissions as received on a websdr receiver. I send an "A" - a "T" gets transmitted, etc.
Now the real question for the ethicists: Do I suppress this info or reveal it? Either way, someone gets rewarded & someone else gets the shaft. Is my loyalty to a single designer or to potential buyers?
I have asked for, and received, a refund for my FX-4CR.
ReplyDeleteYu issued the refund within a couple hours - I won't even be able to get the rig in outgoing mail till Monday. This speaks volumes of Yu's character, just as his emails to me always did. If he ever develops a *beta-tested* FX-5, I'll be all over it.
I really wanted to love the FX-4CR (a 20W POTA rig - awesome!) but it just has issues, significant to me as a CW op, that I'm not willing to pay $550 with the hope that they get sorted.
BTW, the FX-4CR's receiver is nothing short of phenomenal.
While I'm glad to see that you got a resolution to your situation, I'm disappointed to see it had to go this way. I had not even seen/heard of this particular model when I saw your post of having received it and immediately wanted one and was holding off to see your review.
DeleteI personally feel that 20-50 watts is a sweet spot for a POTA rig and it's small size and features made me excited, especially at that price point. I do wish it had was a built-in antenna tuner for my non-resonant antennas, to be a VERY complete all-in-one package. I also would have preferred to see a more standard (PowerPole or 5.5x2.5) power plug.
Based on previous comments here, it sounds like the RIT issue is getting fixed by a firmware update so that wouldn't be so bad. I'm a bit curious how the break-in was (full, smooth, and quiet like the KX-2 or more noisy semi-break-in like my Yaesu rigs?) but that leading dot issue is definitely a deal-breaker.
I appreciate your transparency regarding the issues you found and Yu seems to be responding appropriately. You're not bashing the product or designer, simply stating the facts of what you found and how they affect you and why they make you feel the way they do. Yu has said that he 'is not going to shy away from problems' so it's encouraging that he seems to welcome your open and honest review as well.
With the skills, abilities, test equipment, and knowledge that you have demonstrated here on your blog, I REALLY wish a resolution could have been found that would have you keeping the radio (partial refund perhaps?) instead of sending it back, if only to make it easier for me to follow the development and improvements made via firmware updates. Once you reported the biggest of the issues (the leading dot and RIT) were fixed I would likely purchase one with confidence that it was going to be a good rig. For now, I will likely sit back and wait for the FX-5 and hopefully that one will have a built-in tuner and maybe you'll get to be an official beta-tester for it.
Thanks for your blog and the efforts you put in to it and your willingness to show your findings about this as well as the other rigs that you have run across your bench.
Nick
KC0MYW
Hello Nick,
DeleteI ordered the FX-4CR in September, knowing there would be a long wait. My plan was as you mentioned - to use it as a POTA rig. 6dB better than QRP; 6dB less than 100W - a 20/25W rig is halfway between QRP and 100W, not 1/4 of the way as watts alone would suggest.
In other words, the sweet spot you mention.
When I saw the RIT behavior, then the speaker-stays-on while headphones are plugged in, I tried to imagine how a radio is released with flaws that the briefest of use by the designer would reveal. Just what is the scenario that a person can design features into a radio, then release it for sale with those very basic features being untested? These issues don't even rise to the level of needing beta testers - they're right there on the surface for the designer to see!
It was at this point that I wrote this posting - and it wasn't so much about the specific flaws, but rather what they represent - apparent apathy.
Later, while comparing the receiver to that of my IC-705, I tinkered further with both radios and found that I was getting more hits on the RBN with 10W CQ's of the IC-705 than with 20W CQ's from the FX-4CR. It was then that I decided to "listen to myself, distantly", ie to each rig as it sounded on various WebSDR receivers.
And that, as they say, was the final straw. Initial dits don't make it through to the antenna with the FX-4CR. To be clear, I'm talking about dits that cause the rig to go from rx to tx. If the rig's T/R switching has already placed it in tx and hasn't timed-out to go back to rx (a settable parameter, in milliseconds) those dits get transmitted fine. But for brief pauses that surpass the T/R time, allowing the radio to go back to rx - the next character - if it's a dit - will not be transmitted 30% of the time, if CW speed is 22wpm or faster.
I tested this further by simply sending the letter A, than pausing a bit & sending another A...or U.
On the WebSDR rx, I'd get a T for an A; an A for a U.
A check of TE5X on the RBN confirmed it - my CQ's as 'AE5X' were being distorted by the radio.
My FX-4CR went out via UPS this morning. There are other reviewers with FX-4CR's on the bench - they will be posting their findings soon. If they don't operate CW, this will not be an issue for them - they may not even find this problem. I'm looking forward to QST's review and also to Yu's ability to address this problem. Is this an issue that firmware can fix? I have my doubts. but hope to be wrong.
Thank you for realizing that I'm not bashing the radio or its designer. I've tried to make that clear but some people are pre-determined to believe what they want to believe, despite evidence to the contrary.
73,
John
PS
DeleteI ordered this adapter for the radio that some new FX-4CR owners may be interested in having:
https://www.ebay.com/itm/324596863299
(Anderson PP adapter)
@ John, AE5X
ReplyDeleteHere are MY RBN Facts (10W, Vertical)
[img]https://up.picr.de/45268079uz.jpg[/img]
Well, maybe it's also because of your touch work
Hi Daniel,
DeleteI tried this with my mcHF and IC-705 as I listened on a WebSDR. Leading dots get transmitted.
On the FX-4CR, they get transmitted only intermittently.
Same keyer, same operator, same paddles...only the rig differs - what am I to make of that?
73,
John AE5X
Hi John. Just curious if you tried with a winkey or winkey mini. Still have the problem?
DeleteNo, I was using the rig's internal keyer. I no longer have an FX-4CR but am awaiting one (so far) with the new main board that will be in it.
Delete72,
John
For what it's worth, the RIT issue and the CW issue above 22wpm have both supposed been fixed in a firmware update with the later just being released today (03/01).
ReplyDelete73,
Nikko
That's excellent news, Nikko. Hopefully someone can verify that via a webSDR - send a letter I and see what is received by the webSDR: and I or an E.
DeleteOnce the new boards are available for the spur issue, it'll be time to re-order. Any word on when that might be?
73,
John
Hi John,
ReplyDeleteThere may not be new boards now. Yu posted this in the FB group today:
"Notice: All users who are using FX-4CR do not need to mail the PCB main board first, and should adjust the settings to reduce the spurious to below -40DB to -50DB.
I tested on two newly assembled FX-4CRs, most of the spurs can be reduced to about -45DB to -50DB,
Only very few at -40DB"
73,
Nikko
I received my FX-4CR and have done some minor testing using SSB. Yu is currently working on multiple fixes for a near-term release. I have noticed that the power surge that was supposed to be fixed with the 3/8 version shows up on mine sometimes, have not figured out the pattern. I can key the mike into a dummy load while making no sound and sometimes the power surges up and down, without showing on the radios power meter. When I transition from TX to RX there is a enough of a delay that I sometimes miss a word or two of a response. I am hopeful things will look much better with the upcoming software release.
ReplyDeleteBy setting the AGC to Fast (2), or medium (1) I no longer miss any conversation going from TX to RX. Also the surge is much better with the mic gain set to 20. There is still some noise in transmit when I am quiet and that is supposed to be resolved soon. I have made quite a few contacts using SSB, up to 5,000 miles already. I am using it a lot for POTA hunting now.
DeleteI activated K-0034 (Great Smokey Mountain National Park) with the FX-4CR using SSB and the noise floor was great and I could appreciate it there unlike at my home QTH. I heard and made contact with people I could never do at home.
DeleteUsing the factory mic, what have you found is the best mic gain setting?
ReplyDelete