CW RF Drop Out

Mal G3PDH

I have downloaded and installed the latest version 3.4.21 which all went well on my 6500 with Windows 10 PC set up + Maestro. My main operational mode is CW.

Having suffered the RF transmission CW drop out problem for several years now I was hoping to find that this had been cured. However, after conducting several QSO’s it was disappointing to note that it still exists. The existence of this bug has been raised by several keen CW operators many times but there does not seem to be any acknowledgement or confirmation from Flex that they have managed to recreate it or that investigatory efforts have been made.

To summarise this annoying bug…. When manually keying CW using an external, or internal keyer, the RF output randomly drops to zero although the sidetone continues but no RF is going out. Upon allowing the rig to drop into RX it will then continue transmission again until the next time. Therefore the operator can be unaware that the rig is not transmitting unless continuously watching the power meter. This problem does not arise when the keying is being made from a computer macro as in say N1MM, only on manual keying. One would have thought that this fact might point to the basis of the problem for the software engineers at Flex.

Because of this problem I have taken to using my ELECRAFT K4 more and more for CW as the drop outs become embarrassing and classifies the Flex as ‘Unfit for Purpose’.

Having owned the 6500 for 7 years now I like the concept and operability but dismayed by the fact that this long term bug has still not been corrected.

Mal G3PDH

Mike-VA3MW

Hi Mal

I opened a support ticket for you you.

This is certainly very strange as it is not something that happens to everyone. We have been using 3.4 for months now and even on some pretty big contest stations. I know that had something like this shown up with them, we would have heard about it immediately.

Just as a thought, if you were to send some CW into a dummy load would the same thing happen?

73

Mal G3PDH

Hi Mike, I note your comments but all 3 CW operators in my local area here have the same problem and I know of others in Europe who also reported the same previously. It has nothing to do with RFI and does it on low power and into dummy loads etc. It has existed for several years now. It is noticed more when using a slightly longer TX/RX delay setting. Testing using a very small delay or QSK will not reveal the problem. It only happens when the rig is keyed manually and NEVER when keyed from a computer macro. The contest use that you mention is most likely to be macro based keying, not manual keying. The problem is very random and can be absent for hours. I have no other issues with the operation of my station other than this one problem.

I can appreciate that recreation of this random fault can be difficult but surely the fact that it never happens when keyed by macros and only under manual keying conditions must point to the problem. All of the operators involved in reporting this fault, including myself are all FOC members and obviously keen detailed CW. Operators. A casual CW operator may not spot this depending on the delay settings being used. The solution is not to just to reduce the delay setting as was once ventured by Flex. I do not know of any other rig with such a problem, certainly not my K4 which I am using more for CW now.

Thanks for opening a support ticket and I look forward to resolution of the problem.

73

Lasse SM5GLC

Guess it's time for me to join the discussion as I was mentioned by Mal :)

I have had this spurious RF drops ever since v 2.5 and upwards. I have tested all of them except for the latest as it requires a total reformat of the SD-card, and I assume it is a pain-in-the-rear to revert back (Sorry Mal).

Just to describe (once again) the problem: While transmitting CW (built-in Winkey), connected to the front of the radio, using audio from the radio, strictly "local", the radio (6500) will suddenly loose all RF output despite generating the correct side tone. So if not observant one will happily keep on sending CW and no RF output. The power will reoccure once the radio drop to RX and then back to TX again. Having full QSK means one only loose one char, but have the relays clicking like mad (no pin diodes in my amp), and I prefere to have 2-300 ms hang time, i.e. not have the RX to break-in whilst transmitting. The problem is detected by most of the Flex CW-ops I have been in contact with, but of course one have to be observant to notice and not always there.

If there is ANYTHING I can do to help solve this problem please tell me so... rather than once again open a help ticket. I have done this numerous times.

I really like my Flex, but I also like my thinning hair, now all grey. This issue have me pulling what's left of it.

Mal G3PDH

Many thanks for your support Lasse, It helps to prove that it is not just an isolated problem on our stations but in the Flex software/firmware. I agree with your description of the fault which is exactly as experienced by myself and the others reporting this problem. I hope that Flex will have their software engineers take a serious look at this and analyse our descriptions to hopefully discover the reason as to why it happens.

73

Mal G3PDH

K1ESE

I have also suffered from this for years with several HelpDesk tickets. The solution from the HelpDesk was to revert from Version 3 to 2.4.10.

But the cure was worse than the problem. I give up too much going to 2.4.10, SmartLink mostly. Going to 2.4.10 also caused me to need to reinstall Windows as the only cure for not being able to delete the newer COM ports. Again, at the direction of the HelpDesk.

So far with 3.4.21 I have not seen the dropout but it has been less than a day since I upgraded. I am disappointed to know the problem is still out there.

Mike, if you are contesting you are not likely using a paddle most of the time with an external or internal keyer. It's apples and oranges. And, comments like that are dismissive of a real issue.

K1ESE

Mike-VA3MW

Thanks

I passed on your comments to engineering.

If yo are still having an issue with 3.4, then I urge you to open a support ticket, all of you. This is critical.

thanks

Lasse SM5GLC

I agree, this is critical!

But given the problems to down-grade from newer versions, I am very hesitant to try it out. Right now 2.4.10 with warts and all, do allow me to operate CW with-out drop-outs.

This is a cut-n-paste answer from Tim on me submitting a ticket three (3) years ago:

""I have conferred with Engineering and the analysis of your HelpDesk support incident has concluded that the issue you reported is a probable software defect. I have transferred this issue to the Engineering/Software Development group and placed it in their bug tracker for further analysis.

This issue has been logged as defect number: G7759

At this time the only workaround is to stop sending and allow the radio to transition to RX and then start resending.""

And then two years ago, once again Tim replying to my questions:

""We are aware that the defect is still in SmartSDR for Windows. That defect has not been listed in Changelog so it has not been addressed yet.""

What more can I do to get this issue resolved????

/Lasse SM5GLC

Mike-VA3MW

I am trying to recreate this, so I need to make sure I have the radio set up the same way.

Can you answer the following:

• SmartSDR Version you observed this at (I see one person said 3.4.21)
• RF Power out level
• CW Speed
• Internal Keyer or External Keyer (or a Macro)
• Can you recreate this with CWX?
• CW Delay setting
• Iambic or Breakin

Based on what you are saying, that the radio has to drop into RX, you have the Delay set to a long number.

thanks in advance

Lasse SM5GLC

I'll bite... still on V2, and this has been an issue since 2.5.1 and onwards.

Both SSDR for Win7/10 and Maestro.

P out usually 10-50 watt.

CW speed: any 15-35 WPM

Internal keyer, with paddle at the front of radio (6500)

Iambic Enabled type B

Delay 350 ms

But remember this is what I use now... but this is my standard settings and should have been used when I had the problems. I did a quick search and see there are a number of people reporting this very issue, and I am sure we all have different settings :)

https://community.flexradio.com/discussion/8024499/

https://community.flexradio.com/discussion/comment/20535613#Comment_20535613

https://community.flexradio.com/discussion/comment/20595411#Comment_20595411

https://community.flexradio.com/discussion/comment/20451702#Comment_20451702

https://community.flexradio.com/discussion/comment/20586118#Comment_20586118 (comments by VA3MW!)

https://community.flexradio.com/discussion/comment/20583509#Comment_20583509

Mike-VA3MW

Thanks Lasse

When we try to recreate an issue, we can't assume anything.

Since all users run the same software at the same version, the radio will behave the same given the same parameters. While I am not totally eliminating software, I am leaning towards a hardware difference or a failing component.

But, at the moment it is too early to commit to anything.

I have a career in problem isolation in the software/hardware world (over 50 years) and I learned long ago, I have to deal with the facts I can collect.

Thanks

KD0RC

Hi Mike, I experienced the CW dropout for the first time (that I am aware of...) yesterday. Here are some data points that I hope are useful:

Flex 6400
V3.4.21.14056
RF Power 500 W (KPA 500)
40 Meters
External keyer on accessory pins 4 and 5
Delay 150 ms
WPM 25

Symptom: Sent my call and happened to see the RF drop as I sent the "R" in my call. W1AW/6 came back with KD0C 599. I resent my call and he got it corrected.

I will experiment more to see if I can get it to be a more reproduceable problem. Please let me know if the support team wants me to do anything to help identify the problem.

Mike-VA3MW

Len, that is exactly what I need. Thanks

Lasse SM5GLC

Len,

curious if this is the same issue, you say you lost the "R" but apparently the latter part of your c/s did come out correct? What we have been seeing is that once the RF vanish, one need to drop out of TX to RX and then back to TX again to get the RF back again. I did some testing a while back, seeing if the RF would magically come back whilst still in TX, but no, I kept trying for a "long" period staying in transmit and still no RF until I did the RX/TX cycle.

I would be more than happy to do further testing, but the pain of down-grading is a bit too much, and I "depend" on having a working radio.

Mike-VA3MW

If this can be recreated with 3.4.21, it would be much better.

Roger G3LDI

This issue has been logged as defect number: G7759

Looks like all CW operators have G7759. As has been said it does NOT happen when using macros. However, when paddling a quick comment to somebody in a contest, that comment gets lost because the other party thinks you have stopped sending. Effectively you have because there is no RF going out. However, the sidetone remains, so if you monitor that instead of carefully watching the power out, then you will think you are still transmitting.

Please don't suggest lowering the change over time again. That is a fudge, not a fix. There are sufficient of us with this problem to know it is NOT RFI causing it. The reproducing of this on demand is the problem, like going to the dentist with toothache, which miraculously stops when in the chair!

I was using 2.4.10 but have now upgraded to 2.8.1. Regression is possible but 2.4.10 is not supported any more.

Malcolm G3PDH ( 3.4.21 ) and I spent this afternoon trying to replicate the fault and it didn't happen. Wonderful!! I was a TV/Video Tech all my life and I do know all about intermittent problems.

73 de Roger, G3LDI

Dan Trainor

I am sure Flex has the list of known (and unfixed) bugs related to CW mode operation - listed by bug number with descriptions. G7759 being one of them. And also I believe they must have a list of requested enhancements to CW mode operation. In the honor of transparency, I would like to receive these lists. 73 de Dan WA1QZX.

Mike-VA3MW

What we want to do is try to get to a way to recreate this reliably. If we can do that, then we can then work to narrow it down and actually isolate the root cause.

Intermittent issues are hard to isolate and resolve and right now, I don't have enough to go one as all we have is a single symptom. (Like having a head cold --- what is really causing you to feel like ****?)

Here are things to consider if you are helping to narrow this down.

• does it happen after a 'state' change like a band change or a mode change?
• is it after you send a command from N1MM
• does it happen all the time, or just once in a while (this is critical to know)
• if you reboot the radio, does it go away for some period of time -- or, it doesn't matter

If you are providing details, I can start to relate actions to the symptom.

In Len's case, is that a loss of power or it failed to send a CW element? Or, are they the same thing?

Thanks for your patience. We really do want to isolate this.

73

Val DM1TX

Yeap. This happened to me too since v2.5 I though myself that I am crazy or something till today. Low power, big power, dummy or not, local front jack or remote over Remote Rig, straight key, magnetic key, paddle etc. It simply drives me now and then total nuts. Yes over N1MM everything is fine. But I consider that keying over pC is not what is CW about. Sorry mean no offense. I go back now and then to my KX3 but I really love my 6500. I have not upgraded yet to 3.4. And yes not having the "luxury" of a K4 :-) 73

Mike-VA3MW

Thanks Val

The actual key won't make a difference, but the input to the radio is good to know.

Also good to know that Macro Keying isn't part of the issue, but could that be related to the fact that N1MM Macros are pretty short?

Do you also see it when hand sending a short message?

The K3 is a great CW radio. No argument there. :) The K4 is still a work in progress (I watch that group as well).

KD0RC

Hi Mike, "failed to send" might be a better description than loss of power. My sidetone worked all the way through, so it sounded like all was well. I just happened to see the RF drop during the one character. Then, sure enough, he came back with just that letter missing.

When I send, I tend to leave a little extra room between letters (CW enunciation?) to help the other station copy better. I wonder if that little extra time between characters let things drop for long enough that sending the next character worked normally? Perhaps this is a critical data point in all of this. Others have said that they need to stop sending long enough to return to receive to clear the issue.

Lasse SM5GLC

Len,

yes if the transmitter drops, you will get power back. If you (just for testing purposes) increase the delay (TX hang) and keep an eye on power output, you might be able to see if the power stays off for the whole transmission.

/Lasse SM5GLC

Lasse SM5GLC

I know we are now investigating 3.4.21, but as this has been with us since 2.5.1 I somehow recall that some of the releases appears more prone to this spurios problem than others. Also sometime a fresh upgrade did look as it had the problems fixed, and then out of the blue, back again. I sincerely hope we can hear from all other CW users with 3.4.1 installed to report their findings.

KD0RC

Thanks Lasse, I am now testing by transmitting 1 W into a dummy load with my external keyer set to send a typical CQ string followed by a 500 ms pause, then repeat indefinitely. I set the Delay to max (2000 ms) so that the rig does not return to receive (especially during the 500 ms pause). I have the sidetone turned off and am listening on another rig, so that I only hear the transmitted signal.

It has been running just under a half hour with no issues. I will let this go as long as I can stand it...

I normally turn my rig off overnight, but it runs all day (15 - 16 hours most days). Once in a while, I leave the whole thing powered up overnight, but that is a pretty rare occurrence unless I leave it up for remote operation. As far as I know, I have not had the issue remotely.

Lasse SM5GLC

Len,

as others have written. It seems this is only an issue when using the direct keying input on the radio. Not when using CWX nor keying with N1MM or other logging s/w.

My memory is a bit hazy, but if I recall correctly I did notice this gremlin both with the built-in keyer and an external keyer.

/Lasse SM5GLC

KD0RC

Hi Lasse, yes, I am using direct keying with an external keyer plugged into the accessory jack (pins 4 & 5). This is the same setup as when I experienced the lost character.

Mal G3PDH

Mike, Just to say that it is hard to recreate this on demand. I have tried today without success but will see how things go over the coming days and report back again.

There are some key facts being reported consistently which should allow the software engineers to start analysing the problem.

1. It only happens with ANY kind of manual keying input on any skt, but not from computer macro keying.
2. It is totally random and happens on short or long messages.
3. The sidetone continues as normal even though the RF output has stopped
4. Allowing the radio to drop into RX enables the return of RF output again EVERY time.
5. It now appears that many stations are experiencing the problem and have done for some time but may not have reported it.
6. It has existed for several years
7. Why does it not happen on v2.4.10. What is the difference in keying systems with that version compared to other versions.
8. I do not believe that any of us are operating with a strange setting somewhere that is causing the problem as we are all experienced users and the radio is operating normally apart from this glitch.

K3EL

I agree with Mal's list of "key facts", they summarize the problem quite well.

I installed v3.4.21 the day that it was released, and have spend several hours over the last couple of days either making hand-sent CW QSOs or just transmitting text into a dummy load to see whether the CW drop-out issue has improved with the new software. CW speed was typically around 30 wpm, and the delay was set at 350 ms. Unfortunately, I too still find the same drop-out problem occurring, although my early impression is that the dropouts may be less frequent than with other recent versions of the software.

In a previous thread on this topic I saw a suggestion that the CW drop-out issue may be related to paddle contact bounce. Over the last couple of days I've used two different paddles connected either to my 6600 or to a Maestro, or used an external keyer, and with all of these options the drop-outs still occurred. The external keyer uses an open collector switching output, not a relay, so there would be no issue related to contact bounce.

The other major issue I have with the 6600 in CW is that the radio randomly drops part of the first morse character in a new transmission - a real problem when calling in a DX pileup - whether hand sent or using CWX. Unfortunately this still seems to occur quite frequently with v3.4.21.

Dave, K3EL (or maybe A3EL or T3EL or just 3EL depending on my 6600's wishes)

KD0RC

OK, I am done abusing my transmitter for a while. I ran my test for a few hours with no occurrences of the defect. As I operate normally, I will try to watch more carefully to see if it does it again, and to try and understand what may have led up to the issue.

Lasse SM5GLC

Just go a report from Rune SM5COP running 3.4.21 and a 6600, and he says he has not noticed any dropouts for some time, not just the new release but also with the previous. Caveat, he has not been a heavy Flex user latelely.

Mal G3PDH

Further to this thread

I have carried out test this morning using about 15w into a dummy load and proven that the RF drop out still exists in v3.4.21 software. It is very infrequent and took some time but it eventually happened.

I managed to hold the TX on during the drop out and was able to establish the following:-

The TX icon remains lit

The MOX icon button remains ON and the Break-in button remains lit

The RF output remains at zero for as long as the keying is maintained. Used a series of dots.

I also noted that when the RF dropped the PTT line to the linear amp (not in use but on line) also drops. This may be of significance to the software engineers

The TX was restored as usual by dropping to RX then starting to key again.

I took a picture of the screen plus a video in which one can hear the dots being sent on the sidetone and have forwarded these to Ken Wells who is also handling this.

Roger G3LDI also carried out a similar exercise with video showing the radio in TX and sidetone but no power output after catching the drop out with his v2.8.1 software.