Pure Signal / wiring/ ANAN 8000DLE

[Bill in Georgia]

download/file.php?mode=view&id=123

[w-u-2-o]

Thanks for posting that, Bill! That looks great, and should help a lot of people.

I'd recommend adding a 50 Ohm terminator to the unused PS FEEDBACK (output) port. Actually, it's a good idea to use these on all unused ports, just don't transmit into one at over a Watt!

SMA example
BNC example

73!

Scott

[ddecoons]

Note on above diagram -

I received a new Anan 8000dle on October 2, 2017. There have been changes where the external coupler now connects to the PS Feedback port and not the PS Input port referenced above. I am not sure where this changed.

Does anyone know the serial number range of radios that use the PS Input and the range where it now uses the PS Feedback? This would help reduce confusion for new users. Possibly the above diagram could be edited with the serial number range that it applies to?

Dave wo2x

[w-u-2-o]

Dave,

Based on another, similar report of the same problem in this forum, I'm guessing that there are a batch of these that got put together improperly, i.e. it was not an intended change.

I will see what I can find out. Sadly not a good example of quality control.

73,

Scott

[np2g]

Hello To all
It has came to my attention that some of the newer 8000DLE units seem to have had Some assembler ingesting too much white horse beer.

The result has been that PS Feedback ,, PS Input Connections have been placed “ass backwards” .
@ An endearing engineering term denoting WOOPS ., AHHH Crap . I cant believe I did that .

This perplexing problem seems to be confounding some new users .

So I have , after countless hours of consideration . massif math calculations , 4 trips to NASA . I have decided that this perplexing problem must be fixed . Yes 9 out of 10 ½ hams possibly could agree .

Please follow instructions as your result may vary.

1. Open Case Yes me hearty!!!
2. Do not throw away screws they will be used later .
3. Should I add for safety sake turn off the power . Your choice however as you may want to see the white light . Back to the Detailed instructions .
4. Remove SMA connector that is presently in round hole marked PS Feedback and put it in round hole where PS Input lives
5. Removed as per step 4 SMA connector in PS input Just in case I
6. I Know you have tried to the fit two of these suckers in that one tiny hole Yes just say no and Proceed to step 7
7. Take case and put it back on. Instructions :::: They can l be provided for this operation

8. Use screws from step 2 . You should have none left unless you decided to chrome plate these out .

9. Now (yes you guessed It) it matches the PDF document . No Serial number needed .
10. Your efforts have been rewarded now PS feedback with your Feedback back signal works GEE !!Just like documentation .
11. Pat yourself on the back for a successful job . If you are still waiting for Chrome job please go back to step 1 .

Photos to follow



For those real thinkers Yes there is an alternative Everyone that has them correct should reverse them to match the mistake . Change the PDF . Do not pass go And yes you deserve to pay that $ 200 a year that brand X demands .


I do hope this does not confuses you .


NP2G

[Jim]

If you have one of the new Amplifier's that has a Pure Signal input the only change to the cabling diagram would be to disconnect the connection to PS input from the back of the Rig and simply attach it to the PS input on the back of the Amp, all else stays the same, correct??????

[w-u-2-o]

Amplifiers do not have "Pure Signal inputs". Not trying to be pedantic, but it's important to get this correct:

Certain new amplifiers now include internal directional couplers that provide a coupled output signal with typically -50 to -60dB coupling factor, depending on the amplifier (check your amplifier manual for exact specifications).

This coupler output signal may be used in lieu of the output of a separate, external coupler as the source of a feedback signal for PureSignal processing. In such a case, the external coupler is not needed at all.

Just as with the external coupler, the coupled output signal should be applied to the correct connector on the radio. In the case of the 8000 that is the PS Input connector. On the newer 100D, 200D and 7000, that would be the RF Bypass connector. On older 100's, 100D's and 200D's it would be the EXT1 connector.

One thing to be mindful of with regard to amplifiers offering coupled output signals: the amp manufacturers have been very conservative in their coupling factors. In some cases, too conservative, such that you may have trouble obtaining enough signal level when operating at lower power levels (e.g. 100W or less).

73,

Scott

[Jim]

Scott, thanks much.

I have the external coupler and assumed.......that's how we get in trouble that the 7000dle uses it's EXT1 like the older 100's, 200D's
so it's the RX bypass. I see in in the manual........got to read the manual.

We were aware the 8000 DLE uses the PS Input connector.

So if I don't want to use the Xtronic Directional coupler I simply run the cable from the coupled output on my Amplifier to the RX bypass on my 7000dle and I am all set.

[w-u-2-o]

Correct, IF the coupler output level the amp manufacturer has provided is suitable. Most are for full power operation. Some aren't for low power operation, e.g. if you are running barefoot and relying on the coupler in the amp to do the honors.

[cLicari]

Scott..
I am using a Xtronic XDC-4 coupler for PS feedback. With the loop connector disconnected on my 8000 and the "PS input" being fed, does the "PS feedback" SMA need to be terminated? I presently just have the loop connector attached and dangling. With the XDC-4 I am getting an S-ATT reading of 18dB at 1500w with a 10dB attenuator inline from the "forward" SMA output feeding the input on the 8000. Occasionally I get a strange momentary display (see image) on the pan adapter if feedback is lost. What do you think is going on here? I'm a new Anan user. (spare me the rod)
Thx
Carl
[img]

IMG_4538.jpg

[/img]

[w-u-2-o]

Hi Carl,

Those numbers sound spot on.

"They" say that you don't need to terminate the unused port, but I do, and I recommend it. It's cheap insurance (50 Ohm SMA terminators are cheap )

You shouldn't see these "blow outs" associated with the feedback indicator, but you can see them if the correcting indicator goes dark.

Question: are you reliably driving the ALC indicator to 0dB? In fact, it is usually better if you drive things hard enough to see +1dB on the ALC COMP meter. (The ALC and ALC COMP meter together are really the entire ALC meter, but for some strange reason reading below 0dB can only be seen on the ALC meter and positive indications on the ALC COMP meter.

Anyhow, PureSignal requires that 0dB ALC be reliably reached to make a PureSignal measurement.

What amplifier are you using?

73,

Scott

[cLicari]

Last check I was right at 0dB on the ALC but I'll recheck it. I'll order a 50 ohm terminator for the feedback SMA. The amp is an OM Power OM2500HF. I don't like seeing the blowouts and it usually is when feedback goes dark. I will bring ALC up to +1dB to be sure I'm hitting it hard enough. Can't believe they wany $12 to ship that terminator, lol.
Thanks Scott,
Carl
NX5T

[K9RX]

Carl,

I've reported this action previously.... it is there every once in a while. Just BLOWS UP the screen - is 10's of Khz wide - and I can't imagine what others on the band hear - has to be horrendous. It lasts for a split second but then it takes another couple of seconds before PS comes back working properly. My suspicion is this is a bug. It is not something that you could 'command/cause' - i.e. not a user error. Why others don't see it is of course beyond me - I suspect more do and just don't pay much attention to it as it comes back around. Also if DUP isn't turned on you wouldn't see it - I have DUP on all the time. Also it seems more and more to me that the PC is the determining element. I went from an i5 to a 'twice as fast' i7 and haven't really seen any improvements in the anomalies I've experienced. My CWX won't work - using N1MM+ to do CW won't work (had to get an external K1EL keyer) ... no explanation ...

the rado truly is a love/hate relationship.

Gary
K9RX

[w-u-2-o]

Carl,

Last check I was right at 0dB on the ALC but I'll recheck it.

Allow me to reinforce the idea that it's better to have just a little in the bank in this case, i.e. do be sure to be hitting +1 on the ALC COMP at least. Warren Pratt's ALC implementation is an advanced, look-ahead algorithm and I no one will hear the difference. It is not like the horrible hard clipping type of ALC that was used in the old, legacy version of PowerSDR, whereby if you went over 0 you sounded terrible. Exceed 0dB with confidence!

I'll order a 50 ohm terminator for the feedback SMA.

As I said, cheap insurance, except for shipping

The amp is an OM Power OM2500HF.

I corresponded with some other owners of this amp. They all report that the RF devices in it heat up and cool down quite quickly. This can cause a momentary burst of IMD at the beginning of transmit because the PureSignal algorithm cannot keep up with the rate of change in the linearity curve during heating. This type of problem is known as an "amplifier memory effect". Once the critical components achieve thermal stability PureSignal will work fine. I see this on my KPA500, too, where the FETs heat and cool quite quickly, literally within seconds. So if you are seeing these issues at the start of a TX, that is probably why.

I don't like seeing the blowouts and it usually is when feedback goes dark.

In your screen shot I can see quite clearly that the correcting indicator is dark. That is the really the indication you care about.

73,

Scott

[w-u-2-o]

I've reported this action previously.... it is there every once in a while. Just BLOWS UP the screen - is 10's of Khz wide - and I can't imagine what others on the band hear - has to be horrendous. It lasts for a split second but then it takes another couple of seconds before PS comes back working properly.

Gary,

You have to carefully distinguish between the PureSignal algorithm failing to correct vs. drop-outs in the data stream. A wide burst of energy is generally indicative of the latter problem and not related to PureSignal. When PureSignal fails to correct the result is simply the normal looking IMD you get on any non-linearized radio.

My suspicion is this is a bug. It is not something that you could 'command/cause' - i.e. not a user error.

When PureSignal fails to correct it is either user error or due to a memory effect in the transmit chain that PureSignal is incapable of correcting. With respect to the data drop-outs you could argue that it is due to inadequate software/firmware design. IMHO there are certainly things that could have been done better, e.g. TCP instead of UDP, multi-threading instead of single threading, etc. However, I can't complain too loudly because I can't write a stitch of C code myself.

Why others don't see it is of course beyond me - I suspect more do and just don't pay much attention to it as it comes back around.

Others certainly do see it. I see it here all of the time. But it is not so often that I choose not to use the radio. I just recognize it as a limitation of the implementation and, while I strive to limit its occurrence, recognize that there is only so much you can do with a non-real-time operating system. I get a glitch maybe once every 6 or 10 minutes that I notice.

Also if DUP isn't turned on you wouldn't see it - I have DUP on all the time.

Very true. Everyone should always run with DUP ON.

Also it seems more and more to me that the PC is the determining element. I went from an i5 to a 'twice as fast' i7 and haven't really seen any improvements in the anomalies I've experienced. My CWX won't work - using N1MM+ to do CW won't work (had to get an external K1EL keyer) ... no explanation ...

I and others built fast PCs with great success, or at least better success than you have had. Why the solutions that work for us don't work for you continues to remain a very elusive mystery

73,

Scott

[cLicari]

Scott/Gary... Again, PowerSDR newbie question. I can't seem to get ALC above -0dB using either Line In Gain or MIC gain. What adjustment do you recommend to use? Max Gain in DSP > AGC/ALC tab?

I know of one other using the OM2500HF and he is not seeing these blow-outs, but he is only one.
I have changed the cable out from XDC-4 to PS Input, (waiting for weather to clear to try) but doubt that's the issue.

For reference, the computer is i7 8700K, 32GB ram, 500GB M.2 SSD, EVGA GTX 1060 3GB video card I just built two months ago.

The photo I posted is probably not the best example I shud have used. MOST of the time the blow-outs occur with the "correcting" and "feedback" in green but the "feedback" going dark. This photo shows a straight linear profile but many times the shape has symmetrical semi-circles out each side. I will try to capture a better image. Are videos possible to post here?

Fyi.....No one on RX end reports anything visually (panadapter) or audio anomalies when I see the blow-outs.

Thanks
Carl

[w-u-2-o]

Scott/Gary... Again, PowerSDR newbie question. I can't seem to get ALC above -0dB using either Line In Gain or MIC gain. What adjustment do you recommend to use? Max Gain in DSP > AGC/ALC tab?

You have to use the ALC COMP meter to see how far above zero you are. Switch to that meter and report back

[cLicari]

Ok.. With ALC Comp meter I'm actually seeing 1dB w/peaks to 2.5dB

Carl

[WA0VY]

Carl, I assume you mean 1 to 2 on peaks in ALC Comp. ALC comp is showing levels in db above 0 db and those levels look fine. Upon further review of your screenshot, it looks like for some reason PS just stops functioning. I do not think it has anything to do with your cable, the terminator or either of your samplers. You might try reconnecting the jumper and use the internal coupler. I expect you will see the same thing, i.e. the correcting and feedback lights go dark, but its worth checking. I'm still at work awaiting your video but need to leave here within the hour.

[cLicari]

Scott...
Here is a YouTube link to a short video showing what I am experiencing with Pure Signal.
Carl
NX5T

[K9RX]

Carl,

This is not what I'm seeing, at least re longevity. Mine is only there for a very brief moment - maybe 1/4 second or so... possibly 1/2 second... I don't remember what the lights are doing but I do believe the right one remains green.

Scott,

I'm not talking about the FT8 issue here. this is on SSB. If that is what you thought then no, its not the VAC/computer latency issues that I experience on FT8. I DO live with it of course. But knowing that the whole raison d'etre for PS is a clean signal this flys in the face of that. When it does it it is BROAD ... on SSB I most often have the full 96Khz span showing in the PAN and it will go a good 1/4 of that at least. It looks very similar to what Carl is showing - only it is very brief. Its like for lack of a better phrase, it loses lock. Like its a PI controller where the integrator windup overflows ... Could it be non-linearity? I couldn't refute that... but I'm running the amp with gobs of headroom (2.2K out capable, typically run 1.2K) - it shouldn't go non-linear ... who's to say it is or it isn't.

Gary

[cLicari]

After reloading firmware, another PS test video with same basic results....

[w-u-2-o]

Great job posting those videos!

The PS Test 3 video was most interesting in that it showed the PS control panel and amp view windows.

It would appear that you are either a) occasionally get bad data or b) your amp is doing something really weird.

If you would do me a favor, run the PS Test 3 over again, but disconnect the external coupler, reconnect the jumper on the back of the 8000, and only linearize the radio. Let the amp do it's own thing. Let's eliminate the external coupler and amp as the source of the problem (or confirm it).

[cLicari]

Scott/Brent....

Here is the video testing Pure Signal with the stock coupler reconnected. Barefoot power levels at 50w, 100w and 120w. Sorry about the audio and pauses between power levels. No amp used.

[w-u-2-o]

Carl,

What I'm seeing in the latest video looks totally normal. There appears to be a little heating going on at the start of a transmission, but PS catches up as soon as that memory effect stabilizes.

I'd suggest another couple of experiments to narrow down the issue:

First experiment: do the barefoot test again, at 100W, but use the external coupler instead of the internal coupler. This will tell us if the external coupler and connections are any factor.

Second experiment: do the opposite--run QRO at 1500W, but only have the internal coupler in place. This will only linearize the radio, not the amp, but that's OK, let's just see what happens.

In all cases, please make sure the advanced PS properties and amp view windows are visible, as well as the step attenuator indicator.

Thanks,

Scott

[cLicari]

Scott..
Pure Signal test with external coupler attached running barefoot 100w
on 17m

[w-u-2-o]

OK, not exactly the same (20M vs. 17M), but it looked perfect. Step atten. value was spot on. So nothing wrong with PS or the external coupler and cabling in this example.

Now do the test where you linearize the radio but not the amp. In fact, just leave external coupler in place, between the radio and amp.

[cLicari]

Scott...
Pure Signal test with amplifier and external coupler at 1300w on 17m. I know it's not with the internal coupler but thought I'd try this config as it was already intact. I can do it with internal coupler and QRO if you would still like. FYI... I took the top case off earlier today because the the PS case bulkhead nuts were loose. I tightened the nuts and checked for cold solder joints on the PS board while it was open.

[w-u-2-o]

Where was the external coupler on this last test? On the output of the amplifier? Based on step attenuation values it would appear so. Run the test with the internal coupler and only linearize the radio, but run it QRO, i.e. let the amp do its own thing.

[cLicari]

Scott..
The external coupler was on the output of the amp with the 10dB attenuator inline on PS test 6.

PS test 7 on 17m, 1350w with internal coupler.
PS test 8 on 20m, 1430w with internal coupler.