The AM Forum

I was wondering if some of you other cats with these amazing radios have gotten any reports that there is an echo in your audio???
I turned off my CMPDR ( compander ) and COMP (compressor) and this effect went away.
 
I'll check my settings. I may have tweeked a setting for audio processing from none to a slightly faster setting to reduce latency and the audio card I presently have (M-DELTA) cannot keep up with the computer.

I was with a 40M SSB group working greyline and they heard this echo and thought an amazing phenomena was taking place with backscatter.

Thanks
Fred

Echo? How about roger beep?

http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/

Fred,

Maybe someone here knows off-hand. Otherwise:

After reading thru the archives, I have seen guys with echo complaints and suggestions to fix. Just do a search on their site for "echo" or pan thru the thread titles looking for similar stuff.


This HPSDR WIKI may help too.

http://openhpsdr.org/wiki/index.php?title=HPSDRwiki:Community_Portal

T

Echo is an RFI issue with the computer

Frank is right, Fred.  The typical RF induced audio anomolies from analog equipment is different from what the Flex does.  In the Flex, RF issues causes an echo in your audio on transmit.  You can usually snuff out much of this with snap on ferrite chokes on the audio lines at the Delta-44 breakout box and at the input of the card in the computer tower.  It's a trying process.  Also, make sure your Flex, breakout box, and computer tower are all grounded to your common shack ground.  This may take some patience.....been there and done that!  By the way, I heard you on 160 AM last night and you had it there, too.

OK thanks
I see a  lot of solutions from the links.
Jay, I was on the Flex and most of the echo went away when I turned off the processing in the Flex the echo went away. I'll have to look over what may have changed in the last couple of months.

Fred

By turning off the processing, you were probably just changing the amount of latency/delay in the system. The echo might be the difference between the rectified audio (possibly happening some place in the chain with litte latency) and the "regular" audio, which is processed with latency. This is somewhat of a WAG on my part. Either way, RF in the audio section is usually bad.  :D

Glad you found some help there, Fred.   I will be going through the same thang soon and glad to know it's RF getting into the system.  Gonna ground te computer well and add some beads to be sure.

Question: In general, could the problem be that when RF gets into the interface or computer, there becomes corruption errors in the data stream, so the data needs to be sent a second or third time to verify itself?   That seems logical to me, but may not be correct.

I should have some HPSDR boards by the end of the week to start building and assembling. Can't wait.

T

As others have mentioned, one of the more common causes of echo is RFI. Another cause that is not mentioned as often is when audio levels rise to the point where the ALC becomes very active (slamming against the pin).

I have not had any RFI issues at all with my Flex at any power level, on any band, however, it produced a fine echo effect (even when loaded into a dummy load) when I was (stupidly) underdriving the first audio stage and compensating by driving the next stage much harder in the software audio chain. This prevented the leveler from working correctly which allowed audio excursions to go way beyond the 0 dBm point on the ALC scale. In my case, when the ALC was pegged, it began to produce a ringing effect, which created a very impressive imitation of an echo sound. It was easy to fix by going one audio stage at a time in PowerSDR and setting the gain distribution at each stage so that it never exceeded 0 dBm. The TX metering setups for MIC>EQ>LEVELER>ALC>CPDR made it simple to adjust the audio drive at each stage so that the transmitter produces full output with no stage being over-driven. At this point, with each stage now set correctly, I never bother watching anything except the ALC level when transmitting in any voice mode to assure that it doesn't exceed 0 dBm on peaks and the ringing/echo is gone.

The single most common cause of echo is the soundcard itself and the soundcard settings.
However, that has nothing to do with the useless "compandor" setting....

Tom
It seems like the RF is detected and demodulated and sent back into the computer. The latency in the computer gives the echo affect.
Fred

I'm not it is in the sdr1000, but may be software dependant

Unplug the cable that goes to the LINE-IN of the soundcard, then transmit and see.

When I had the SDR-1000, you could unplug this cable and STILL see your transmitted signal on the display, and hear the audio via the MONITOR function.

Which meant you really are not monitoring your actual transmit signal.
Only the audio BEFORE it is sent to the transmitter.

May be the software/firmware allows you to monitor through the reciever... I haven't owned an sdr-1000 for 3 years or so.... so I can't tell which is the case now..

Thanks All,
I'll report back after I do a little snooping. This all started since the install of the Ameritron Amplifier. The "INS and OUTS" cables for the SDR1000 drape right over the amplifier cabinet.

I'll move the sound card interface and the black box under the operating table to separate the two devices.

Bruce, I find the CMPDR button does amazing things!!! I have the CMPDR and COMP selected and the settings of 4 for each. The mic level control on 23 and it pumps the audio up where the munkey is schwangin'!!! Huge pos peaks from the amp. I REALLY see 1500W P.E.P.!!! Typical Ham test equipment. Nothing is calibrated to NIST. Un-selected the modulation is average looking on my 'scope and Bird PEP meter.

BOB 1AEX, I'm using the next to the latest revision SDR software for a FLEX 1000 radio.
I do not see all the menus you mention for setting levels and ALC in this SDR stuff.
mic, leveler,eq,ALC........?????
How / where do I get into those settings? I'll take a second look

Another Question for all:::: where can I find the clamp-on ferrites that we usually see on computer/data products? Usually molded onto the power cord or data cables. EX: power cable for display monitor and the interface cable that connects monitor to the 'puter???
Fred

Fred:
Here's some:
http://parts.digikey.sg/1/3/snap-clamp-on-ferrites
http://www.mouser.com/Search/Refine.aspx?Keyword=clamp+on+ferrite+bead

Fred asked:

Rob: Post the e-mail you sent me yesterday. That should clear up Fred's question or at least started in the right direction.

Fred,

I've run my SDR-1000 with the TenTec Titan with good results. I have the FA66 strapped to the SDR box with a piece of braid, and took all the long audio cables and coiled them up into coils about 3 inches in diameter for RF choking and then strap both boxes to the station ground as well as the computer case. So far so good. I did have a problem a few weeks back when I was blinking out the computer during voice peaks, but grounding the CPU case fixed it.

Are you running 2.0.16? I'm using that on the F5K and will have to try the CMPDR and COMP settings. Haven't played with them much yet.

Clamp on ferrites are available from Mouser or DigiKey or probably places like DX engineering. I've been using type 43 but hear that 30 is good for HF also. Most of the ones you see at hamfests and RS are optimized for VHF and don't do much down in the HF spectrum.

hello
Kevin thanks for your direction on the FA66. I have one coming from the flooded country of Australia. I presently have the M-delta and the outboard interface, laying right next to the amplifier.  And thanks for the numbers for HF Ferrite

Thank you Pete for your links and request for the navigation aides in the SDR software.
The techie guy Gary?? said the SDR 1000 is happy with V1.8.???  for SDR 1000. There was a bug in V2.0..I think

Fred

Version 2.0xx is beta software and still being massaged. I'm running 2.0.16 and haven't found anything that doesn't seem to work to my liking.

Fred,

I have been reading some of your posts lately and I'm confused about your sound card. One of the SDR 1000 sound card offerings early on was the Delta 44 made by M-Audio. The Delta 44 uses a computer expansion slot sound card and an outboard audio breakout box with 1/4" tip and ring unbalanced line plugs.

I never heard of a M-Delta sound card and I suspect yours might be a Delta 44. In which case you might rethink your replies to this thread.

I had the same problem with echo early on and nothing would get rid of it until I found I was over driving the audio input from the mixer to the sound card as mentioned above. I grounded everything and snapped so many cores on the RF lines looked like something out of Star Wars. I even isolated every audio line with transformers. I was pumping in too much audio. Once the audio input was set to 0db the echo was gone and my SDR-1000 has been happy ever since.

If your using a mixer board and outboard processing or EQ you might try turning that down first than look at the Power SDR settings. Good Luck. I know how frustrating this can be. About Zero DB input to the sound card is more than enough.

Mike

Correct Mike
I have the M-Audio. I see Delta on the breakout box and I assumed it was M-Delta..

I'll start experimenting in the day or two. I have the mic gain set to 25, so maybe not over driving the audio card. I have never touched the "virtual sliders" on the sound card itself, coz when you go through calibrations for the P.A. and receive carrier meter cal they probably all change.
There was mention about setting levels to various stages and ALC and I have never seen any adjustments like that. YET!!
Thanks
Fred

Hi Fred,

I did what most guys do when a new box arrives. I plugged a mic in and ran the Mic Gain up until it sounded fine to the guy on the other end. What I found out from a discussion on the Flex net is that it's necessary to set the peak levels correctly from front to back in the software audio chain, otherwise the leveler and compander will not respond as they should. For the first few weeks of messing around, I had no clue how badly my poorly distributed gain settings were affecting the ALC, which of course had a major impact on things like sideband suppression, DSP artifacts, and ringing in the audio chain in all voice modes.

I am using PSDR version 2.0.8, but I believe the steps are the same for all 2.x.x versions, and probably all the 1.x.x versions as well.

As I understand it, the sequence to follow is this:

1. Set the TX meter to Mic Level and set the GUI slider so that the maximum reached by peaks is about -2 dBm

2. Set the TX meter to EQ, open up the PSDR Mixer interface, and set the Preamp slider so that the maximum reached by peaks is below 0 dBm.

3. Set the TX meter to Leveler. Open up the PSDR tab for DSP and go to the AGC/ALC settings page. Adjust the Leveler setting so that peaks on the TX meter do not reach 0 dBm. You can further adjust the attack/decay/hang times if the Leveler is not responding quickly enough, or if it seems slow to release. This is largely trial and error, but I found the decay and hang times to be too high, and reducing them smoothed things so that there was no audible pumping or excessive periods of gain reduction. (Page 156 - 157 of 2.x manual)

4.  Set the TX meter to ALC. Open up the PSDR tab for DSP and go to the AGC/ALC settings page. Apply audio and observe that the maximum level of peaks does not exceed 0 dBm. If everything in the preceding stages has been set correctly, you should see that your maximum peaks will end up between -5 dBm and -1 dBm. You can also adjust the attack/decay/hang times of the ALC if you see a peak sneaking through now and then. Again, I found that by reducing the decay and hang times in small steps I could find a setting that resulted in smooth audio without pumping or extended periods of gain reduction.

At this point, engaging reasonable levels of the Compander or the DX function will result in a very significant loudness difference in your signal, without invoking the ALC or any signs of ringing or distortion.

I haven't had anything but good reports in every voice mode since following the procedure above. What I thought was RF getting into the rig (which is what was repeatedly reported to me by people I spoke with on the air) was simply poor adjustment on my part. I would imagine that gain distribution with the SDR-1000 could be even more exciting with the external sound card being involved. I'm not even sure if all the steps in the process above would apply, but I know that you know your way around with that rig.

Good luck and I hope you can tame your hollow effect!

Rob W1AEX

OK Rob,
Thanks........ it's all in the meter and menus there geesh. I have seen these things but nothing was clicking or connecting.
Play time begins.
Thanks Rob
Fred

It's not you Fred. The manual hints at all this stuff, but it doesn't connect the dots. I happened to hear a discussion about this on the 20 meter Flex net and learned a lot from that. Also, in an explanation about factors involving the way ALC is incorporated at the link below, Tim Ellison's last paragraph got my attention.

http://www.mail-archive.com/flexedge%40flex-radio.biz/msg05481.html

When I checked my Mic Level on the TX meter, it was barely deflecting. However, the EQ Level preamp was at the maximum setting (apparently by default) and this really played havoc with the leveler, which was pegging the level meter to the max, and the ALC, which behaved in a very stange and peaky way. As a result, the compander was useless because it instantly made the ever-present ringing much more intense. Setting the gain distribution to appropriate levels corrected all the issues I was experiencing.

Hopefully, the Flex tech group will add a segment covering this in future manuals. It would decrease the number of times I hear people on the air complaining that the leveler and compander don't work. They actually work beautifully when the preceding stages are set correctly.

Good luck!

OK guys here's the scoop,
I went through the adjustments and nothing was out of spec at all. None of the audio levels were beyond what the calibration called out for from Rob's steps. The echo was still there no matter what. Turning adjustments way down extremely low only reduced the echo.

This is typical computer magic and software voodoo. Out of curiosity I went to the sound card tab in the setup and noticed the sampling rate was 48000. I switched to 96000 and the panadapter  display changed a little. And the echo is gone, no matter how badly I misadjust the audio settings.
I went through all of the settings again. The levels are just where they are supposed to be and no echo. It was not RFI coz I was testing into a dummy with just the FLEX radio barefoot set for 20W carrier.

I noticed less distortion or pumping by reducing the leveler gain to 0. Less background noise, room noise...fans, my son running around upstairs. Like a radio studio. The ALC is happily around a -5 dBm max. And not a steady -5 like I'm banged up against it but a moving needle with my voice and never goes beyond -5dBm. Basically just tickling the COMP and CMPDR goodies. They are both set to 3 and the audio jumps out of the scope when they are on. Huge POS peaks.
I wonder how that sampling rate changed? OR what caused this echo, because I'm thinking it has been 48000 all along.
Very good thoughts here and a mind twister for sure.
The merry-go-round starts all over when the Edirol FA 66 card arrives. I'm anxious to hear any difference with the condenser mic plugged into the FA 66. It can provide the phantom power I need and there would be less electronics in the audio chain. Right now there is a Symmetrix 528 and a DBX EQ with balanced INs and OUTs to feed the M-Audio interface.
Fred

That's great news that you tracked it down Fred. It does seem very odd that the echo appears to have been caused by a change in the sampling rate, but I suppose that it's possible that 48000 didn't agree with a setting somewhere in the mix. Like you said, a little voodoo and magic was probably involved. Of course it begs the question as to whether the echo returns if you drop the sampling rate back to 48000...

I'm sure you'll tame the Edirol FA 66 when it arrives. Should be fun!

Rob W1AEX

Yes Rob,
The echo returns when the sampling rate is 48000
The Fa 66 can go to 192? 196? You can see almost the entire 40M band on the Pandapter
Fred

The FA66 does 192 Khz. You can set it for 48, 96 or 192. One thing, when you change it, you have to have it turned off and then turn it back on. It apparently takes the setting on power up. I ran it with an MXL 990 on the phantom power and just used the software equalizer and got good reports.

That's one of the fun things about QS1R.
You can see the ENTIRE HF spectrum on the display if you want..

Too bad it doesn't transmit.

 >:(

I would like to resurrect this back to the top to AXE a question about the cabling of the SDR 1000 to the sound card interface.
The M-Delta break-out box has four TIP/RING/SLEEVE jacks. I assume EACH jack has a left and right channel. The SDR 1000 came to me with HOSA cables. ONE1/8 inch stereo plug for the INs and ONE 1/8 inch stereo for the OUTs for the black SDR 1000 box.
The sound card end of these HOSA cables has two 1/4 inch plugs tip/sleeve for INs and 2 1/4 inch tip/sleeve plugs for the OUTs.
How is the distinction made for the electronics for left and right audio when both 1/4 inch plugs are the same? Does the SDR software "figure out' or configure the left and right audio info?
I know that if I reverse the "Receive" cables, the USB and LSB detected audio is also reversed. I can hear lower Sideband signal on Upper sideband and vice versa. So something somewhere knows left and right audio.

The Flex knowledge base is not showing anything.

The reason I am asking is I want to revamp the system using another sound card option the FA 66 Edirol, and that has balanced INs and OUTs and I want to run that audio through audio isolation transformers to isolate computer noise which is usually in the common ground connection.

Small issues, hopefully not thinking something to death
Fred

Hmmm, it's been a year since I fussed with that on the FA66 Fred. I do recall one issue though, and that was that the 1/4 inch plugs that were going into FA66 were not mono tip-sleeve connections, but rather used tip-ring-sleeve plugs (what I had been conditioned to know as 1/4" stereo plugs) wired as TRS where the balanced connection is between the tip and the ring, and the sleeve is ground. Maybe that is old hat to everyone else, but it confused the heck out of me. I had never heard of TRS balanced connections before that.

I think to sort the left right thing out, I did what you did and tried them one way and then the other till the sideband came out right, and then setup the transmit audio cables to match it.

You're right, the knowledge base seems to have gaps in it. Some of the stuff that may be obvious to some users, is not so obvious to the rest of us. I'm trying to sort out a low transverter output issue now and not getting very far there. I find the flexradio and flexedge lists are pretty good to follow and learn things as they go by.

I'm getting my ducks in a row here. I'll figure it out. I took the M-Delta break-out box apart and they are Stereo jacks for 4 channels in and out of the sound card.

I'm preparing for the FA 66 and now I discover that I should have a Firewire interface card capable of 400mbps. No problemo. I'm also taking the suggested route, by Frank GFZ to use quality audio isolation transformers to keep CPU noise out of the signal paths of the SDR. That hash noise that you pickup around a computer case with your short wave radio.

Fred

There is a PDF artical that explains in detail how to set up the FA66 with the flex.
A quick google will find it...
I may even have it saved here.

You do not use the ring and tip on the edirol, unless you are using balanced input/output.
If you use transformers, you could set it up that way.
If not, you use tip, and sleeve mono plugs, OR connect the sleeve and ring together on a stereo plug.


Fred,
You don't need to worry which is left, and which is right...
It makes no difference.

If when you fire it up, the signal is on the wrong side of the LO, just click the box to reverse it...under the "audio" tab.. ( swap I/Q)   or, change the plugs around.
No big deal there.

That was quickkkk

Here ya go Fred

http://support.flexradio.com/Downloads.aspx?id=46

That s  the SDR-1000 / FA-66 "Quick-Start Guide"

Thanks Bruce,
I was able to find the quick start and the nice pictures.
And thanks for the 'easy fix' if USB and LSB are reversed.

The FA 66 box is still en route. Fella in Australia had it on the Bay and the flooding has delayed the shipment. I hope it arrives soon.
Fred

UPDATE::: 2/4/11

I got impatient waiting for an EDIROL FA 66 sound card and ordered a new CAKEWALK by ROLAND FA66.....................
Seems to be possible 30 day shipping from Australia...............won from an Aussie from eBay...oh well

I received the CAKEWALK by Roland...........FA 66 audio interface. Loaded drivers and whatever else it needed and set the switches according to FLEX.Firewire card is alive and well.
BUT starting the FLEX SDR software 1.8 whatever,,,,,,.... it says missing  FA 66 ASIO driver.
The supported sound card lists an EDIROL FA 66. So the games beign.

Edirol FA 66 and Roland/Cakewalk  FA 66 the same thing??? Apparently not for drivers. Roland website very confusing for downloading driver that works for Cakewalk and won't work for Edirol..........I Email Greg @ Flex for a solution.
I'm also checking my computer system for possible misplacement of the ASIO driver.
Any other ideas???
I know this is not a Flex radio site, but the knowledge base is UN-knowledgeable about this too.
Fred


Side note....................
I experimented by cliking (in sound cards) UNSUPPORTED SOUND CARD and it almost lets me start the hardware....SDR 1000 but comes back with "wrong sampling rate" and quits. Sampling rate is set for 192000, which was my original goal.

The echo effect was there as induced by RF getting into either the flex or the computer.  The compander just made it worse.

..... or do you have an echoey room and the compandor made it worse?

Al

HI AL
My particular echo went away by selecting 96000 sampling rate. It had somehow gotten to 48000 and caused the echo. It was not RFI or audio from over-processing.

Fred