On Thu, Jan 27, 2022 at 09:25 AM, Julian N4JO wrote:

Sorry, Paul, engineer-speak; what I mean, simply, is "how do I make that noise happen on my QCX?"
I'm interested in finding and understanding the root cause, and potentially mitigating it.

I can.... 

The tone seems to be coming from Clk2, as it is normally set to the RX frequency plus the TX offset.

Also, if you go to Menu "3.8  CW Offset" you can change the frequency of the tone heard on RX.


It is more noticeable on a QCX-mini than a QCX+, but, with antenna removed I can hear it on all my QCXs.


The problem on my QCX-mini is that it BEATS with the incoming signal, and since usually you cannot exactly zero beat to the received signal it creates this wobble or "growl" type sound...  For me, it does make it hard to copy the incoming signal, it's very distracting...


It really is noticeable in RIT and Split mode, where RX is offset from varying frequencies from TX, the bleed-thru from Clk2 is very load.

To demonstrate:

Hi, Julian.

Sorry for the delay in response.  The unit is as is configured from the shop, set up for 17m.  Nothing has been adjusted.

If I knew how to tag you in a post, I would, as I just posted about the "growl" or "idling engine sound" which is of far greater import at this time.  I really like this rig, but am puzzled as to why it passed muster yet it is plagued with some serious issues.  Where it works it is absolutely astonishing.

Thanks for writing and please see my recent post #17m #troubleshooting and I think "#growl". 

73,

Paul
V7/KA4WPX

Sorry, Paul, engineer-speak; what I mean, simply, is "how do I make that noise happen on my QCX?"
I'm interested in finding and understanding the root cause, and potentially mitigating it.
--
Julian, N4JO.

Hi, Rod.

Thanks for the helpful information.  That's a lot of fiddly you mention!  In that case I'll poke around for other less invasive cures, such as working a split.

As to the growl, I was able to temporarily dampen the level by holding and slightly moving the 3.5mm plug.  This change in level is elusive, however.  It would seem to be some type of grounding issue.  My next step is to remove the housing to investigate. (As an aside, the level does abate some if I change the ref. frequency up 2 MHz.)

73,

Paul V7/KA4WPX 

Hi,Julian.  If you could please provide a bit more detail in what you mean by configuration I'll gladly pass that info along.  Thank you.

73,

Paul V7/KA4WPX

Thanks for the info Paul.  You have the same issue as me with the 700Hz tone.  Hans I think still has it on his list - to investigate.  You'll probably have seen in the other messages on this topic that I've suggested:

"Two potential "fixes" came to mind - the "Move Clk2 frequency well away during Rx" that Mont describes or, potentially, remove the resistor that feeds CLK2 alignment signal to the antenna input.  There again, as Gregg found and Mont confirms, there's no signal actually at the CLK2 pin - so that may not work (but still might due to pick up along the track)."

The "move Clk frequency well away during Rx" would need Hans to modify the (already crammed) firmware.  The resistor removal is fiddly and may not work - it needs a brave someone who's experiencing the symptoms to remove that resistor on an already-aligned mini, to compare beofre and after. My mini is too beaten up for me to risk it, I'm sorry to say.  I made a ham fisted replacement of a couple of duff op amps (mine is a v1 mini) and I don;t want to chance my luck on that pcb unless I have to.

Good luck - I hope someone somewhere can help you fix that growl too.

73
Rod
G0VKX

Paul, can you give me a specific configuration I can set up on a QCX to experience it?
--
Julian, N4JO.

Hi there, Rod.

Thanks for writing.  Yes, when RX and TX frequencies are different no 700Hz tone is heard; in simplex operation the tone is audible. 

The 700Hz tone is audible with or without an antenna, as is the growl.  The growl, however, obscures all but the strongest signals within 500-1000Hz.  I do not know if they are related.

73,

Paul
V7/KA4WPX

Hi Paul
Yes - the 700Hz constant tone is caused by a leakage from the Si5351 CLK2, which is evidently set the the "next intended Tx frequency.  So, when the mini is set to "split", that tone can be shifted out of the Rx audio passband by shifting the tx frequency.  But with the tx and rx frequencies the same, I still hear that 700Hz tone.  Do you?
Do you hear the growling, and the 700Hz tone, when your mini is connected to a reasonably good antenna?  I can hear the tone (on a quiet band).
I can't help regarding the growl - sorry!
73
Rod
G0VKX

I've had the same issue with the 700Hz tone.  When I toggle to dual VFO mode the tone disappears.

Now, if I could determine the cause of the growling/idling engine sound at various intervals, which can be modulated by keying in practice mode or toggling the mini then this will be an amazing little machine, indeed.

73,
Paul V7/KA4WPX

Hi
I wonder whether someone is willing and able to to try removing the resistor that feeds CLK2 alignment signal to the antenna input?  That might attenuate the symptoms enough to be useful, albeit at the expense of losing the alignment signal(s).  I'm not going to mess further with my early-version QCX mini, which only just made it through my eventually successful efforts to replace a couple of op amps.

73
Rod G0VKX

Was there a clear resolution to this issue?
Thanks

James

On Fri, May 7, 2021 at 01:02 PM, KB9NBL wrote:

Was there ever a resolution to this problem?

As far as I know Hans has not had time yet to look at this...

I do know that on receive he turns off the Si5351a's Clk2 output.  And on transmit, he turns off Clk1, and reprograms Clk0 to 25MHz (or vise-versa).  This is needed to keep the caps charged that are attached to the FST3253.

I'm hoping that when he turns off Clk2, he can also change it's frequency setting by something (e.g. at least 1MHz up or down).  Then if there is any bleedthru internal to the Si5351a, at least it will be off frequency and shouldn't be detected or noticed.


Hopefully Hans will get to it soon and find an elegant solution I'm sure.   :)  :)  :)


After first noticing this in my QCX-mini, I went back to my 2 QCX+ rigs.  It is in all my QCX rigs, although not nearly as bad as in my QCX-mini.

It is really an issue when using RIT or split frequency.  In fact, split frequency has several spurs(?) on all 3 of my QCXs, as you tune the RX frequency away from the TX frequency...


73
km6wt

Was there ever a resolution to this problem?

James

Hi

A little more info from me that I'm cautiously optimistic will help diagnose and in due course cure this minor but niggling issue:

After futile efforts on my part to detect the source of the "next Tx" frequency that's causing that 700Hz tone - using another receiver on "narrow cw" bandwidth for example - it dawned on me that the mini itself was all I needed -  I could use the mini's own antenna Rx input terminal.

I probed very very carefully around the underside of the main pcb, with the mini's own antenna input via a capacitor (for belts-and-braces abundance of caution).  The 700Hz tone became clearly audible in the mini's connected earpiece when connecting to R55, or IC3 pins 4 or 10.  It's a weak signal but definitely above the noise; this wasn't the case when probing to the ground/ earth or to any other pin of IC3.

To my mind this demonstrates that the 700Hz tone is caused by a weak signal from CLK2 of the Si5351 on the "next intended Tx frequency".  Which is encouraging.

I imagine that something around perhaps -115dBm or less wasn't going to appear on a 'scope, by the way.

Nothing here is intended in any way whatsoever to pressurise or hasten Hans to do anything.

Cheers all

Rod
G0VKX

"When you understand the inside the outside will be just fine"
Thelonious Monk

On Tue, Feb 2, 2021 at 06:23 AM, Bob M. wrote:

Switching the oscillator on and off or changing its frequency by more than a few kHz takes some amount of time. This might interfere with the full break-in operating mode and may not be the ideal solution, even if it's a quick fix.

It's already turning Clk 2 On/Off for keying, and it changes Clk 0 Off/On frequency during keying.

I asked Hans previously about the affects on QSK with sending all this data through I2C to the Si5351a. 
Apparently with the I2C running at 400KHz, there's plenty of time between the dits & dahs without affecting QSK.


I'm sure Hans will find a solution once all the Fires are under control.

Fortunately, 99.99% of the time it doesn't affect normal operations.  Unfortunately, that still leaves the 0.01%


73
km6wt

Switching the oscillator on and off or changing its frequency by more than a few kHz takes some amount of time. This might interfere with the full break-in operating mode and may not be the ideal solution, even if it's a quick fix.

Thanks everyone for coming back with such helpful information.

As soon as I saw your post, Mont, the penny dropped.  Your description fits the symptoms I'm seeing.  Sure enough, in effect I have a low level "netting" tone that as you and Gregg describe must surely be caused by an internal signal leaking somehow from the Si5351 at the (intended next) transmit frequency.

If I lock my sign gen feeding -110dBm at the antenna socket to zero beat with the rogue "netting signal", then go to split mode, I hear a strong beat note at the difference between A and B frequencies. 

Two potential "fixes" came to mind - the "Move Clk2 frequency well away during Rx" that Mont describes or, potentially, remove the resistor that feeds CLK2 alignment signal to the antenna input.  There again, as Gregg found and Mont confirms, there's no signal actually at the CLK2 pin - so that may not work (but still might due to pick up along the track).

Daniel - good question regarding other ICs etc - my 74ACT00 voltages seem fine and it seems to be working OK.  Similarly the processor volts are OK and it seems to be doing its stuff re display, ADC, keying, sidetone, CW reader,and so on.  D1 and D2 are fine.  When I killed the voltage regulator I had only 8V at the supply socket, so ~7.3V? ended up on the 5V rail.  I think all the ICs bar the Si5351 are specified to  tolerate 7V max (another of several "if only" thoughts that haunt me!).  I'd expected the Si5351 to fail on account of, what, nearly 6V applied to it.  It's still possible I suppose that the (I'm assuming for now) internal Tx freq signal is the stronger on account of some damage done; but that seems a bit subtle for a failure.  Similarly maybe the 74ACT00 switching is a little less clean on the gate that feeds the alignment signal to the antenna (but again that would only be an issue if the assumed Tx signal is propogated along that route).  I'm still wating to come across some subtle failure - with fingers crossed that my mini's all now as good as it can be :o)

Finally a big thanks, Hans, for your comments - as you say, not a critical issue.  I hope that dust settles for you soon and that calm descends peacefully upon your shoulders.

Best wishes all

Rod
G0VKX

"When you understand the inside the outside will be just fine"
Thelonious Monk

There are actually several tones in Gregg's recording.  There is one tone there when the recording starts, a second tone starts at about 4 seconds in, a third tone starts at about 12 seconds in, and I think another starts at about 15 seconds in (but it's so weak I could be imagining it).  Not sure if that means anything, but that's what I observed.