Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: [1] 2 3 4 5 6 7 Next   Go Down
  Print  
Author Topic: Flex-5000A CW Impressions  (Read 33996 times)
W6UV
Member

Posts: 538




Ignore
« on: February 04, 2011, 08:45:03 PM »

I finally dug my key and keyer out of storage and tried operating CW with my 5000A.

Wow! What a can of worms!

I've heard that the built-in keyer is practically useless in the current beta version (2.0.16) of PowerSDR, so I tried using my AEA MM3 as an external keyer. The keyer itself works fine with the 5000A, but I ran into a number of issues. I first tried using the side tone generated by the 5000A itself. This was a disaster because I just could not send properly when listening to this side tone. There must be lag or something, because my sending became very uncoordinated and full of mistakes. I attributed this at first to my rusty CW skills, but even that couldn't account for the mess I was sending.

I then tried sending using the side tone from the MM3 and everything was fine. No more uncoordinated gibberish going out on the air. So in its current incarnation, the 5000A's built-in side tone is more that useless -- it leads to crappy, mistake-ridden sending.

Since I was now using the MM3's side tone, I tried to turn off the 5000A's side tone by checking the "Disable Monitor" option in the CW options since the manual says "check the Disable Monitor check box to avoid hearing the side tone from the internal keyer". This doesn't work--even with Disable Monitor checked, I still hear the 5000A's side tone in addition to the MM3's side tone. The only way I could operate was to mute the 5000A right before sending. There has to be a better way.

On the receive side of things, the 5000A is a rather pleasant CW rig. The filters work fine and can be set narrow enough to eliminate close-in interference. I could detect no ringing at all, even when set to the narrowest setting (25 Hz). The 0 Beat button quickly zero-beats a signal to my preferred 350 Hz pitch and centers it in the pass band. No issues here.

Operating split is straightforward, and the ability to "watch" a pileup on the panadapter display is a real plus. This makes it easy to locate the stations the DX is working and greatly helps determine his tuning pattern.

I'm disappointed that CW seems like a red-headed step child in the Flex world. I hope that these issues are addressed (as promised) in the next version, but I'm not holding my breath.

73,

Jerry

Logged
W4TQ
Member

Posts: 40




Ignore
« Reply #1 on: February 05, 2011, 07:32:09 AM »

I had the same problem when using an external keyer and the 5000A monitor.  My CW sending was terrible (it's not that good to begin with due to hand tremors).  I also concluded that the problem must have been due to the small delay, and my brain's inability to adapt.

I bought a Palstar keyer, which has a monitor volume control on the front panel (very nice keyer).  I now do CW with no problems at all.

73,
Dan - W4TQ
Logged
K0OD
Member

Posts: 2558




Ignore
« Reply #2 on: February 05, 2011, 09:39:15 AM »


 "I could detect no ringing at all, even when set to the narrowest setting (25 Hz)."


The last time I played with the 25 Hz and 50 Hz filters, maybe 9 months ago,  they worked no better than the 100 Hz one (which *is* quite good).  Their performance depends on buffer and sampling settings, computing power, and maybe a lot more.  You can probably get your filters to work narrower than 100Hz but other performance, including panadapter bandwidth and keying, will degrade severely.

See:
http://kc.flexradio.com/KnowledgebaseArticle50437.aspx

I don't detect any cry from users for improved CW filters. Ultra narrow CW filters introduce some undesirable fuzziness to high speed CW. Even on crowded bands, 250 Hz and rarely 100 Hz, are usually best.

An adjustable notch is needed more than "showy" ultra narrow CW filters. But we won't be seeing that soon.

Logged
HAMMYGUY
Member

Posts: 85




Ignore
« Reply #3 on: February 05, 2011, 11:36:43 AM »

This is the main reason I finally unloaded all of my Flex rigs.  CW transmit can be frustrating.  The CW monitor lag is just the cream of the crop.  What used to get me was the missed dits.  Hit the paddle and the software would try to switch to transmit, but the dit would never make it out the antenna.  I tried newer machines, old machines (which actually seemed to work better), and old machines with older software.  The older official versions worked the best since they don't have the new GUI.   

External keyers definitely worked OK MOST of the time. If the computer all of a sudden got busy with anything, the software would begin to miss characters.  Yes I had a dedicated machine with nothing running in the background. 

It's too bad because they really are fun to play with.
Logged
W6UV
Member

Posts: 538




Ignore
« Reply #4 on: February 05, 2011, 02:07:55 PM »

This is the main reason I finally unloaded all of my Flex rigs.  CW transmit can be frustrating.  The CW monitor lag is just the cream of the crop.  What used to get me was the missed dits. 

Anyone know what architecture Flex uses when sending CW? I bet the key input gets sent up the Firewire port to the PC to be processed before being sent back down the link to the radio hardware as an RF waveform.

They would have been better off putting a $0.50 Pic MCU in the box and using that to directly key the transmitter, even if the radio's internal firmware has to generate the RF waveform directly. A simple on-off keyed CW signal can't be that hard to generate to require a fast PC's processing power.
Logged
W9OY
Member

Posts: 1318


WWW

Ignore
« Reply #5 on: February 05, 2011, 03:32:51 PM »

Jerry

I use the F5K almost exclusively on CW and very much enjoy the radio.  The way I have mine setup is in the driver I use "safe mode level 1" and I use 512 buffers.  To get to the driver go to the "start" button on windows menu bar and under "all programs" and find the "flex radio systems" folder.  Click on "flex radio" and it will start a GUI for the driver.  Just make your adjustments if needed and close.  some folks need to run 1024 buffers depending on your computer.  I use 1024 on a slow old laptop I have.  Just play with it using the lowest number of buffers that are stable. 

In PSDR setup screens go to: setup>audio tab I set the buffers to 512 and I use sample rate = 96000.  (if you had to use a larger buffer in the driver then set this buffer to match what you used in the driver)  The shorter the sample rate the greater the lag so 192 has the least lag (in terms of side tone) but it also has the worst filters.  96000 is a good compromise I think, but you can play with this also.  Under setup>DSP I use in the Buffer size CW field RX 4096 and TX 512.  You can shorten the latency by going down on the RX buffers but that also decreases the steepness of the filters 

I use an external keyer, a winkey USB and I feed the output of the winkey into a real serial port  (not a usb to serial converter).  As a buffer I use a PNP transistor between pin 4 and pin 6 on the serial port connector and drive the base from the keyer through a few hundred ohm resistor (see page 22 of the manual 1.14.0 on the flex website for pinout)  to turn on this means of keying in PSDR, go to setup>DSP>keyer and set "secondary" to the correct serial port and I use RTS as the key line. 

I have virtually no lag in terms of sending and no trouble using the internal monitor in the radio.  I generally don't go over 40wpm or under 15.  I set my delay on the front of the radio to 60ms and find that is a good compromise for between word QSK.  I use this setting in pileups all the time so I can stop transmitting when  the DX starts transmitting.  I am using 2.0.16 software and its associated driver and firmware.  I have never had good luck with the internal keyer.  I do not like the way the "memories" are set up.  The keyer is much better than it used to be, but I still don't like it much.  I do have a paddle hooked up to it in case something screws up but I much prefer the winkey. 

You can listen to KIUQ, the self appointed "anti-fan girl", flap her lips all day long about how horrible the radio is, but frankly I think she just doesn't know crap about radio and is a bit of a misanthrope to boot.  In the old days she spent all her time over on the ten tec reflector knocking the Orion and now she spends all her time knocking the Flex radio.  I wish she would just buy a Yaesu or a K3 so she could go spend her time knocking that radio, and leave us fan boys alone.  Can't wait to hear her response:  waa waa waa.

Since I went to this method of keying several years ago (SDR-1000 days) I have had no problems with any of the subsequent revisions of PSDR as far as CW goes.

Hope this is of some help to you

73  W9OY
Logged
W6RMK
Member

Posts: 656




Ignore
« Reply #6 on: February 05, 2011, 05:18:47 PM »

"Anyone know what architecture Flex uses when sending CW? "

The key closures wind up as MIDI note messages and wind up in the PC in that form. The latency on that part should be fast and very consistent.

Then, the actual keying envelope is processed through the usual transmit chain, essentially by creating an audio tone that is translated up.  Very similar to if you had an audio code practice oscillator feeding a SSB transmitter.

Since there is some potential variability in the latency from key to RF, it can be very, very confusing to listen to an RF monitor (or the equivalent of a loopback within PSDR).  Depending on how variable the latency variation is, it can be totally disruptive.  A really old version of PSDR, before they changed the buffering strategy, would only turn the Tx/Rx around on buffer boundaries, so the delay varied randomly by one buffer time.  Now they're using a more full duplex buffer processing strategy, but there still is some latency variation because the buffers to be processed all get queued up to be processed through one thread, and I don't know that there's any guarantee of ordering or latency.  Especially if there are multiple subreceivers or pan adapter things going on (since they also result in buffers going through the buffer processing chain).   (I don't recall whether the current version of PSDR uses a separate thread for Tx and Rx, but even then, threads have to compete for processor resources)

If you're using a keyer with an audio monitor, then your brain is hearing your keying as you do it, so it should work just fine, and only the other end of the QSO would hear any variations.  And, since the receiver processing has a lot more latency variation than the transmit, your signal might be quite good to the other end.

If you're using VAC, then there's another source of variable latency.

Yes, this *could* all be fixed (after all, pretty much every video/audio mutimedia app on the PC, especially games and video playback, manages to do it pretty consistently and seamlessly) but it's hard work, and the existing software works "good enough" if not excellent for many users.
Logged
W6UV
Member

Posts: 538




Ignore
« Reply #7 on: February 05, 2011, 09:22:41 PM »

W9OY:

I have the buffer size set fairly low in the Firewire driver (not at the rig right now so I can't look up the actual values) and I have the sample rate set to 96K and the Operation Mode set to Normal. I have a fast PC with four processor cores, 8GB of RAM, and don't run anything else on the box when using PowerSDR. I've gone through and turned off any unneeded services in Windows as well.

I have the MM3 keyer output plugged directly into the 5000A's Key jack and have DSP->Keyer->Connections->Primary set to "Radio".

With these settings, the internal side tone generated by the 5000A is enough out of sync that it messes up my sending. I'm just getting back into CW after a 15 year break, so my keying is bad enough without the rig's side tone lag contributing to it. Wink  I'll try tweaking the buffer size settings and see if this helps.

W6RMK:

I agree with you able latency issues not being a problem in other PC software. I've run some professional audio and video applications on PCs less capable than the box I run PowerSDR on and did not have any latency issues, so this can be fixed. I guess Flex just has to put their limited resources where it generates the most bang for the buck. I don't know how many Flex owners use CW, but I'm sure the numbers aren't nearly as high as for SSB and digital modes (where I have no problems at all). I just wish Flex would put more effort into fundamentals rather that fancy looking skins and other eye candy, but their marketing people probably told them that's what sells rigs.


BTW, I found a fix to disabling the internally generated side tone issue I mentioned in my initial post. It turns out that the "Disable Monitor" check box in the keyer setup dialog box is ignored if the "Mon" button on the front panel is pressed.

At this point, I'll stick with using the side tone generated by my keyer rather than the internal side tone. The only problem with this is that I often operate CW when others in the house are asleep and the side tone coming from the keyer is too loud when I have its volume set high enough to hear it when using headphones (which I always use when operating).

Anyone have any ideas how I can mix the audio output of the keyer into the headphones along with rig audio?
Logged
K0OD
Member

Posts: 2558




Ignore
« Reply #8 on: February 06, 2011, 07:51:51 AM »


W9OY says: "In the old days [Stan] spent all [his] time over on the ten tec reflector knocking the Orion"

K9IUQ says: " I have NEVER been on the TT Reflector, nor have I ever owned a TT radio"

And the rest of us wonder: "W9OY, got a link?"
Logged
WA8JNM
Member

Posts: 175




Ignore
« Reply #9 on: February 06, 2011, 08:13:24 AM »

W6UV:

I run CW on the Flex 5000A as well. I too am playing with the settings, especially now that I just got the radio back with the second receiver and the latest .16 beta.  One thing I noticed in your comment: you are using Normal for the Operational Setting. The manual and knowledge base does not recommend this. Apparently Safe Mode 1 is preferred. Normal works very poorly for me; I think it is misnamed. Also, I appreciate the comments here re using the external keyer  sidetone rather than the Flex sidetone, because of latency. I'm going to go try that right now, because my CW is still a bit rough too.

Dave
Logged
WA8JNM
Member

Posts: 175




Ignore
« Reply #10 on: February 06, 2011, 09:12:59 AM »

W6UV,

Later. I am able to send CW much better using only the external keyer sidetone. Thanks for the tip. I too had trouble shutting off the Flex 5k sidetone, even after I chose that setting in Setup. But now it is off, and I don't know what I did to kill it. Wish I did. I also listened to the transmitted CW on a separate receiver and it sounds fine to me. I only send about 20 wpm, though, so I can't say if any problems develop at higher speeds. By the way, my settings are as described by the other op in this thread, except I use a sample rate of 48000. My external keyer output, though, goes directly in the CW jack of the Flex, not to a USB port.

Just ignore the rants by Stan. I do. He must be short on things to do.
Logged
W6UV
Member

Posts: 538




Ignore
« Reply #11 on: February 06, 2011, 10:09:10 AM »

One thing I noticed in your comment: you are using Normal for the Operational Setting. The manual and knowledge base does not recommend this. Apparently Safe Mode 1 is preferred. Normal works very poorly for me; I think it is misnamed.

Dave,

I used to use Safe Mode 1, but switched to Normal after a long conversation with Dudley in Flex Tech Support. Normal mode works better than the "Safe" modes on my setup -- they probably recommend the safe modes (and name them that way) because of differences in PC hardware people use. My PC is very fast and has very low DPC latency, so Normal mode is appropriate. YMMV.

73,
Jerry
Logged
W6UV
Member

Posts: 538




Ignore
« Reply #12 on: February 06, 2011, 03:56:57 PM »

So - blame me, my computer, or my 5K if you wish, but something is wrong with the CW mode on MY 5K. And I have talked to many many others who feel the same way but are afraid to go public because of the nastiness that this subject brings...

Stan,

Based on what I've experienced so far, I think you've hit the nail on the head when you say the 5000A is not a good CW rig. I've only used mine on CW for a week, and I shudder to think what other issues I might encounter. I only have one receiver at the moment, so I can't listen to my signal on the air.

I like the ergonomics of PowerSDR a lot, but if CW is not up to snuff (or can't be brought up to snuff--some things are just not fixable in software), then I might buy a K3 for use on CW and relegate the Flex to SSB/data modes. There's probably a reason many of the big DXpeditions use K3s and not Flex.

73,
Jerry
Logged
W9OY
Member

Posts: 1318


WWW

Ignore
« Reply #13 on: February 06, 2011, 05:13:23 PM »

I love the way Stan spends his time being the "messenger of bad news" yet hundreds of us use the radio every day to make CW contacts.  You will notice if anyone says anything constructive about the Flex radio he is labeled a "fan boy" or is some part of some super secretive underground Masonic testing society.  If his nasty rantings are then responded to in a likewise manner he is a victim, just a poor humble messenger out to save ham-manity from the evil Flex Radio... What a hoot   

I worked VP8ORK a couple of days ago on 80 and 40 CW.  Because of the Flex system combined with DDUTIL and Skimmer, on 80 I was able to determine exactly where the VP8 was listening and worked him on the first call even though there were about 100 stations in the pileup.  I don't have a super station, just a Flex a vertical and about 1300W.  Because of the Flex system I was able to determine within in a minute that the VP8 was alternately listening to the top and bottom of the pileup and worked him in 3 calls.  On 40 I also use a simple vertical.  The pileup on 40 was a solid wall of RF.  You can listen of Mr Bad News Stan if you like.  You can buy a noisy old K3 if you like, I will just continue to rack up the DX in my log with my Flex 

As to Dxpeditions   They use K3's because Elecraft sponsors Dxpeditions.  They use Acom amps for the same reason and DX Engineering antennas for the same reason.  This is marketing 101.  You will note the first entry on the list of Orkney equipment sponsors

http://www.vp8o.com/vp8equipment.html

I'm not sure why Flex doesn't do this (even though I'm a fan boy and in the super secret Masonic society of testers)  My guess is they are more interested in using their resources for developing radios for the government and military. 

73  W9OY

Logged
K0OD
Member

Posts: 2558




Ignore
« Reply #14 on: February 06, 2011, 07:50:37 PM »

W9OY
Quote
"I don't have a super station, just a Flex a vertical and about 1300W. "

And a Florida QTH. LOL
Logged
Pages: [1] 2 3 4 5 6 7 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!