6600M transmitter stops intermittently on CW

James Charlton

My 6600M running the latest version of V 3.x software started exhibiting intermittent transmit failures during the SKCC WES CW contest yesterday.  Specifically, I would be sending and during a dash (key down, transmitter transmitting) the RF output from the 6600M would drop to zero.  Letting the key up and then pressing it down again to sent the next character caused the transmitter to come back on.  However, as long as I held the key down, the RF would stay off.

My setup is a 6600M driving an Ameritron ALS 1306 amp driving a dipole.  I have a tuner between the amp and the antenna but it is usually in bypass.  

There is an EA4TX remote antenna switch and a logging PC connected to the 6600M via  USB cables.  Neither the switch nor the PC seem to be doing anything when the problem occurs. 

The 6600M is limited to 50 watts output and I have measured that with a separate meter. It's built-in ATU is in bypass, no ALC connection or other feedback to the 6600M is present, the SWR between the 6600M and the amp is about 1.1:1.  SWR to the antenna is 1.3:1.  No measurable common mode current in the shack.  The ambient temp in the shack is about 68F and the 6600M air vents are completely clear and power is steady.  No alarms, lights or error messages appear.

The problem is hard to trap because it occurs only every several hundred characters and the 6600M keeps the amp keyed on so its "transmit light" stays on, just no RF comes out of the 6600M.  It does not matter whether or not the amp is on.  I observed the problem on 20m and 40m bands.  Others weren't tested.

Although it is not very new, the last thing added to the shack is the EA4TX switch and it connects to the 6600M via a USB cable.

I would like some help fixing this because people say I'm missing the occasional letter.
Regards,
Jaimie "Jim" Charlton   AD0AB

Michael Pitre

I am having the exact same problem with my Flex6500. Has this been resolved?

Bill -VA3WTB

Tom W3FRG

James Charlton

Michael Pitre

I am experiencing the exact same problem with or without logger running or other software, the only difference being my radio is a Flex6500 running same version of SmartSDR. It is much less noticeable or non-existent with cw QSK delay set to zero. This happens with or without amplifier running, or whether using internal or external keyer.  I have tried going back to SmartSDR 2.0 with no improvement. 

Tom W3FRG

I neglected to comment during this anomaly, that the Receive signal also dropped down in level significantly during this RF Power drop out.
However, both Tx & Rx returned when either the VOX or MOX buttons were clicked.

Lasse Moell

Lasse Moell

Tim - W4TME

This is defect G7759 and it is currently being investigated by the software team.  At this time the workaround is to stop sending and allow the radio to transition to RX and then start resending.

Mal G3PDH

Tom -WA2TP

Joe N3HEE

Doug Hall

I've had my Flex-6300 for a number of years. Running SSDR 2.6.2 on Windows 10. I often use a different rig on CW, but tonight I spent some time on 80M CW with the Flex. I noticed the issue that people have reported on this thread, namely that the RF output would occasionally drop to zero while sending CW. Allowing the radio to revert to RX seems to fix the issue, at least momentarily. During a 30 minute CW QSO it happened to me more than half a dozen times.

If this is the same defect (G7759) as mentioned previously then Flex has known about it for quite some time. I consider this a fairly serious issue for CW operation. It sounds like it's related to contact bounce.

73,

Doug K4DSP

Erik Carling EI4KF

It is so disappointing that critical faults like this exist. It is what puts people off buying a Flex. It was logged as a defect in January and we are now in November with no resolution.

Roger G3LDI

Like Lasse, I have had this problem for months. So, the only way to get over it at the moment is to revert to version 2.4.10. I have been running that for a few months now.

73 de Roger, G3LDI

g7soz

Ah, so I'm not the only person to suffer from this. I am having exactly the same problem on all modes - one second RF out, the next none. Highly embarassing.

I was loving my Flex 6400 until this started, but I'm afraid that now I'm getting to the stage where I might send it back to the supplier and ask for a refund.

73

Simon G7SOZ

Doug Hall

Simon,

If getting a refund is an option for you I'd seriously consider it. I don't understand Flex's reluctance to address this issue. By now they have had more than sufficient time to investigate the problem, so they should have 1) figured out the problem and devised a fix, or 2) determined that they can't fix the problem. But like the Flex CW, they have gone "radio silent."

Incidentally I have also experienced this problem on SSB. Once in a great while keying the mic results in no RF out, and releasing the PTT and keying again restores TX. Others have reported this as well.

Admittedly It may be that there are only a very few of us experiencing the issue. But in that case Flex could at least say so. They have assigned a defect number, but that's about all we know.

73,

Doug K4DSP

M0UNN

Got same issue here as well on my flex-6300. TX stops intermittently , when sending cw with the paddle. It happens once in a while and very annoying, when in qso you notice, that you have no power out and the other station is already replying to your half sent message . Happens also , when operating remote using Maestro, but Maestro is also down at the moment and Flex is working on the issue....grrrrr......

Doug Hall

OK, I spent some time looking at this issue in depth, and now I am reasonably sure that it is due to paddle or key contact bounce. To recap, some Flex 6000 users have seen the TX stop intermittently on CW, and when it happens the rig won't transmit until you stop sending and let the rig return to receive mode. The CW Delay setting will have some impact on how likely you are to see it happen. With low delay values the problem might happen and you'd never notice, but with longer delay values if it happens and the TX doesn't drop out between characters you are more likely to observe the issue. N3HEE observed that it was worse with a bug or side swiper, and I can confirm that. I hooked up my old 1954 vintage Vibroplex and it was nearly useless. A Nye straight key was better, and it took a while before I observed the problem with that key, but it still exhibited itself. I cleaned the contacts on my Jones paddle and the problem got better but was not eliminated.

Then I switched to an external keyer and the problem was never observed over a number of QSOs.

Using my oscilloscope I can see contact bounce on all of my keys and paddles. Some are worse than others. The Vibroplex bug is terrible. But all mechanical switches (and hence ALL straight keys and paddles) have some contact bounce. It appears that the Flex firmware is not handling contact bounce properly. My external keyer does handle the debounce from my paddle correctly and I think that's why the problem goes away with an external keyer.

This same issue exists on SSB, and if you have a PTT switch that exhibits a lot of contact bounce you may occasionally key the radio only to find that you aren't transmitting any RF. Releasing the PTT button and pressing it again generally solves the problem. I experimented with a number of switches, looking at their contact bounce on the 'scope before selecting one for my PTT footswitch and now I rarely see the issue on SSB. But it does occur.

As stated in a previous post, this is not a new issue. It was reported some time ago. Given that it apparently showed up in later releases of firmware one must assume that it's a firmware issue and not a hardware issue. And maybe it's not a contact bounce issue at all, but it sure seems like it. In any event I wish Flex would fix it. But rather than holding your breath you might consider dusting off that external keyer and seeing if it solves the problem.

73,

Doug K4DSP

Mal G3PDH

Doug, All very interesting but I don't think its is contact bounce. I use an external logikeyer feeding into pin 4 on the accessory socket and I see the problem. I never use a key straight into the key socket. It never does it when the CW is keyed by a macro on such as N1MM+ only when using the key sockets. Admittedly mine seems to be very infrequent at the moment but very annoying when it happens as I prefer to keep the delay higher. I am on version 3.1.12 software.

Malcolm G3PDH

Mauro HB9FBG

I also had the same problem, in a very severe and persistent way. It didn't happen with the internal keyer and in the digital ways. With Flex USA I also tried old software but the problem remained. The attempts of technical interventions have been useless.

I am not a technician but I remain of the opinion that FlexRadio USA should retire ALL THESE RADIO and understand where is the problem which is not software. It is a hidden problem in the car, most likely a nonsense but embarrassing for the unfortunate who cannot use the radio as I did.

I hope you understand quickly why FlexRadio is, as I expected, a great radio.

Mauro HB9FBG

Doug Hall

Thanks, Malcolm. I have only used the front panel key jack for CW, but I assumed it's the same connection as the pin on the accessory socket. So far I have not seen the problem when using an external keyer, but I have only had it hooked up for a few days now. In any event using the external keyer seems more reliable than when I was using the paddle with the Flex internal keyer, so I'll stick with that for now.

73,

Doug K4DSP

Doug Hall

Update to my previous post. Malcolm G3PDH, after using my external keyer for a few days the problem showed up once during a QSO on 20M this afternoon. Up to this point I had declared the problem "fixed" or at least "worked around" but apparently not so. I was sending with the CW delay set to around 300 mS when I realized that there was no RF coming out of the radio. Pretty disappointing. So I don't know if contact bounce plays into it or not, but like you, I have seen the problem with an external keyer as well as the internal one.

I'm out of theories.

73,

Doug K4DSP

Lasse SM5GLC

Doug,

I reported this in summer 2019 and was issued a defect number G7759. The only remedie right now is to fall back to SSDR 2.4.10

I did revert back to 2.4.10 a few months ago and so far, I have yet to see this issue. Comparing with 2.6.2 my gut tells me there is slighlty different keying timing, i.e. I find it easier to send fast CW with the 2.4.10 vs s.6.2. Not sure if this part of the issue though. A shame that we CW operators are forced to use old software and not beeing able to enjoy all the new features. But hey... guess FT8 is where the action is, right? :( My bet is that noone at Flex is using CW.

Lasse SM5GLC

Just one more comment, I use the built-in winkeyer and have my paddel attached to the front of the radio. Have tried to use an external keyer and yes, it seems there are less occurance but not all good. Also (as I have posted in a separate thread) it seems 2.4.10 do behave a bit different regarding speed setting, as I am unable to change speed when transmitting CW from my logging s/w (tried a few other s/w too). This was no problem earlier on.

Lasse SM5GLC

Just a quick comment, as some might be interested in trying the new SSDR V2.7.3 (or even 2.7.5 and assume this goes for V3 s/w too), well don't :( The defect is still there! I was told that G7759 is not listed as fixed in the release notes and thus it's still with us. I ended up rolling back to 2.4.10. If you, like me, tried but want to do a roll back, contact helpdesk, as one need to do it in a specific order to avoid problems. Following the instructions, I had no problems to re-install 2.4.10.

Mike-VA3MW

I was reading a number of threads on this from the past.

@Lasse SM5GLC If you test this with an external keyer, does the problem still exist on a current version of the software?

Lasse SM5GLC

Mike,

first thanks for taking interest in this old and odd issue. I did some tests earlier on 2.6.2 and had this issue showing using external keying. Others seems to verify that using a vibroplex or straight key even fares worse.

But I have yet to see the power loss happen when using N1MM or other s/w based keying. So for contest operations it's ok. But you may understand my frustation when I was calling a DX, he comes back to me and when I reply with my 5NN, no power :(

As I wrote I gave up on 2.7.3 and have reverted back to 2.4.10. If there is some specific issue I can look into and bring some light to the bug, I could go back once again to 2.7.3, but as you know is a bit "messy" to go back to 2.4.10 :)

If you really want to dig into this, try key the radio, using paddel or an external keyer (I use a paddle to the front of my 6500), and set the delay so it will not drop during your transmission. It can take days to show, and then wham... you get it pretty much every transmission. And don't forget to keep an eye on the power output :)

/Lasse

James Charlton

Hi,

I ran into a related problem this morning during CWops and while troubleshooting it I experienced the old power loss problem again. First the new problem.

I am using a 6600M with the latest v 3. ? load driving a PGXL amp. The 6600m and amp are connected exactly per the PGXL book, i.e. the amp is keyed via the CAT cable connection through my local LAN. The PGXL book specifically says to set transmit delay to 0. I suspected that I was dropping first dots because people thought my call sign began with a T (it is AD0AB). Finally, one contact tole me that I as dropping dots.

The on-line manual says to set a transmit delay (with the PGXL) to 20ms. I did and that seems to have restored the first dots. I also set the "hang time" in the 6600m to 125 ms just so I don't have to listen the the relays switching. It has no effect on the lost first dots. But I also experienced the second power loss problem.

I was using a paddle with a Winkeyer running around 30-40 wpm. Since I was looking for leading dot failure, I was sending hundreds of dots very fast. Then I did occasionally experience power loss. The 6600m simply stopped outputting RF. The side tone was there and the on-off button was red, but as long as I kept up the string of dots, the RF stayed off. I don't know if the Amp was keyed on or not. Stopping the string of dots for even a very brief instant restored normal operation.

I have not bothered to open a new trouble ticket as the problem is well documented in the existing tickets.

Jim Charlton AD0AB

Roger G3LDI

Lasse SM5GLC and I have had this problem for ages now and like Lasse, I had hoped that the new version would solve this proble, but hey ho, it's still there. To quote Lasse:

I reported this in summer 2019 and was issued a defect number G7759. The only remedie right now is to fall back to SSDR 2.4.10

I am about to revert to 2.4.10 yet again to regain confidence in my sending. If you are not a CW op, a casual look for the problem may not even show it. However, during our weekly net this week, it happened five times. I have posted on another thread about the problem too, so hopefully somebody may now investigate and solve it!

Using the Flex with full breakin, i.e. delay set to zero, and the fault will not show. However, I run mine with a delay of around 200. If it was even longer, the fault would appear and you could carry on sending with no power output.

73 de Roger, G3LDI