|
Let me start with I'm a noob at diagnostics.
Since my signal is really low I started to do some tracing.
One thing I noticed is odd if I inject a signal further away from the antenna source it gets softer listening to a 1khz tone. (Isn't that the opposite of what you would expect) (Injected 7.150mhz am signal with 1khz tone)
So for example injecting a signal at c200, 201, 202,203,204 gets progressively softer to? non existent at c204 with a 100uv signal injected.? If I touch the L4 to inject the signal I can hear the signal a little? but c204 I hear nothing injecting there.? If I touch the base of Q10 I hear nothing.? If I ramp up the signal generator significantly 30x higher I can then hear a signal injecting at the base of Q10.
What could cause this?
Usually I thought you would expect the signal to get stronger the further in the receive path you get but it actually got weaker and required 30x the signal at base of Q10 to even hear it.??
Any ideas what I should check?? Or is this expected?
Thanks
|
This is for a ubitx board v4.
|
If I am interperting what you are doing correctly, you have an AM signal at 7.15 Mhz with 1 k hz modulation.? You are injecting it starting at the antenna and progressing towards the audio section.? Is that correct?
I would suggest the you read the HF Signals circuit description:
The first thing to note is that the 7.15 mhz signal is mixed with the local oscillator of 45-75 mhz to get a 45 mhz signal that is then amplified ant further mixed back down to 12 mhz and then detected by beating an 11996500 BFO to get the audio back out.?
Note that injecting a signal further and further away from the first mixer will result in lower detection, if you are keeping it at 7.15 mhz.
Hope that helps explain what is happening, though not in solving your problem.
I will reread the thread ans see if I can help.? In the meantime someone with more experience than me might chime in with a suggestion.
73
Evan
AC9TU
|
My first thought is that it may be some kind of bandpass filter
that is not on-frequency. However, very often the obvious is
not the problem in such cases. Signal injection and tracing is
the way to go, though, in my opinion.
Bob — KK5R
--------------------------------------------
toggle quoted message
Show quoted text
On Sat, 12/22/18, Dave Space <davesspacebar@...> wrote: Subject: [BITX20] Tracing receive signal for low signal To: [email protected] Date: Saturday, December 22, 2018, 12:22 PM Let me start with I'm a noob at diagnostics. Since my signal is really low I started to do some tracing. One thing I noticed is odd if I inject a signal further away from the antenna source it gets softer listening to a 1khz tone. (Isn't that the opposite of what you would expect) (Injected 7.150mhz am signal with 1khz tone) So for example injecting a signal at c200, 201, 202,203,204 gets progressively softer to? non existent at c204 with a 100uv signal injected.? If I touch the L4 to inject the signal I can hear the signal a little? but c204 I hear nothing injecting there.? If I touch the base of Q10 I hear nothing.? If I ramp up the signal generator significantly 30x higher I can then hear a signal injecting at the base of Q10. What could cause this? Usually I thought you would expect the signal to get stronger the further in the receive path you get but it actually got weaker and required 30x the signal at base of Q10 to even hear it.?? Any ideas what I should check?? Or is this expected? Thanks
|
I also thought that injecting an AM signal is like listening to an
AM broadcasting signal where you have to zero-beat the carrier
and listen to the modulation. A 'scope would be better to see the
carrier. But "ear-balling" the 1-KHz signal can give an idea where
the signal is amplified and where it is falling off. This is good
theory, at least. I'm sure someone in the group has a better
suggestion.
Bob — KK5R
--------------------------------------------
toggle quoted message
Show quoted text
On Sat, 12/22/18, Evan Hand <elhandjr@...> wrote: Subject: Re: [BITX20] Tracing receive signal for low signal To: [email protected] Date: Saturday, December 22, 2018, 12:48 PM If I am interperting what you are doing correctly, you have an AM signal at 7.15 Mhz with 1 k hz modulation.? You are injecting it starting at the antenna and progressing towards the audio section.? Is that correct? I would suggest the you read the HF Signals circuit description: The first thing to note is that the 7.15 mhz signal is mixed with the local oscillator of 45-75 mhz to get a 45 mhz signal that is then amplified ant further mixed back down to 12 mhz and then detected by beating an 11996500 BFO to get the audio back out.? Note that injecting a signal further and further away from the first mixer will result in lower detection, if you are keeping it at 7.15 mhz. Hope that helps explain what is happening, though not in solving your problem. I will reread the thread ans see if I can help.? In the meantime someone with more experience than me might chime in with a suggestion. 73 Evan AC9TU
|
Dave
There is gain distributed rather evenly so not sure you are seeing anything unusual. Unlike a commercial rig with gobs of AGC, this rig requires using the gain control. Please evaluate into a good, full sized antenna with some headphones. Strong huge signals will require minimum gain setting, signals just above noise may require nearly all the gain.? also try this test using a good 40m antenna. Set ubitx for at least mid volume. The noise should go up when you plug in the antenna. If you can hear band noise, I suspect all is okay. If you want more audio, find a inexpensive amplified speaker.
Curt
|
Better thinking is that the carrier-balanced 1-KHz signal is heard
farther back toward the audio section, after the BFO has done
its work.
Bob — KK5R
--------------------------------------------
toggle quoted message
Show quoted text
On Sat, 12/22/18, Bob Lunsford via Groups.Io <nocrud222@...> wrote: Subject: Re: [BITX20] Tracing receive signal for low signal To: [email protected] Date: Saturday, December 22, 2018, 12:55 PM I also thought that injecting an AM signal is like listening to an AM broadcasting signal where you have to zero-beat the carrier and listen to the modulation. A 'scope would be better to see the carrier. But "ear-balling" the 1-KHz signal can give an idea where the signal is amplified and where it is falling off. This is good theory, at least. I'm sure someone in the group has a better suggestion. Bob — KK5R -------------------------------------------- On Sat, 12/22/18, Evan Hand <elhandjr@...> wrote: Subject: Re: [BITX20] Tracing receive signal for low signal To: [email protected] Date: Saturday, December 22, 2018, 12:48 PM If I am interperting what you are doing correctly, you have an AM signal at 7.15 Mhz with 1 k hz modulation.? You are injecting it starting at the antenna and progressing towards the audio section.? Is that correct? I would suggest the you read the HF Signals circuit description: The first thing to note is that the 7.15 mhz signal is mixed with the local oscillator of 45-75 mhz to get a 45 mhz signal that is then amplified ant further mixed back down to 12 mhz and then detected by beating an 11996500 BFO to get the audio back out.? Note that injecting a signal further and further away from the first mixer will result in lower detection, if you are keeping it at 7.15 mhz. Hope that helps explain what is happening, though not in solving your problem. I will reread the thread ans see if I can help.? In the meantime someone with more experience than me might chime in with a suggestion. 73 Evan AC9TU
|
OK, I reread your post.? ?I do see that you are before the first mixer testing on the low pass filter (c200-204).? I would assume that it would not be linear at all.? even though you are getting closer to the input to the first mixer, you are injecting into a totally different type of filter response: the prior components are still in the circuit.? To perform the test as you are attempting, you would need to disconnect the low pass filter and inject directly in the mixer.
Sorry for my confusion on the first post.
I will let it up to others to help you, before I point you in the wrong direction
73
Evan
AC9TU
|
Why I went down this route is due to low signal can barely hear anyone.? So I figured I'd try some tracing.. My understanding is 50uv (volts) should provide an s9 signal.? My signal generator lowest signal is 10mv which is way above that and sounds pretty good for a 1khz signal.? So I ran the 10mv signal through a -40db attenuator which if I calculate correctly should be then 100uv double the amount that should produce an s9 signal and injected at the antenna I can barely hear it.
Then injecting the signal further along the path c200, 201,202,203 gets software and softer until I can't hear anything at C204, a tiny bit if I inject at L4, nothing at Q10 base until I 30x the injected signal.
Since that should be just pass the low pass filter I would have thought injecting at the base of Q10 should work fine since that should be the first location the signal really makes it in,
but it requires a calculated 3mv signal to even hear it there which is way above the 50uv that I believe should result in an s9 signal if I'm understanding this all.
I also checked the voltages of Q10,11,12 and all of them match the diagnostic guide pretty closely within a couple tenths of a volt.
If I should remove the low pass filter when trying to inject at q10 should I just have a jumper for c200 to c204 to remove the low pass filter from circuit effectively?
As for trying to use an oscilloscope since my scope resolution is only down to 20mv per division, I can't really see things at the 100uv level on the scope when dealing with low signal strength I believe.? So seems like signal injection is the best way to try?
A better way or places to inject???
|
Please understand that I am not an expert on this topic, or board.? I do have a background in Electrical Engineering, so have some training, though not in troubleshooting (other than for industrial controls).
With that, and if you are just looking to bounce ideas around then here goes:
Looking at the schematic (available from HF Signals), Q10 is past the first mixer, so injecting a 7.15 mhz signal there the signal will not be amplified, there is a 45mhz "roofing filter" limiting the signal to a 15mhz band around 45 mhz.? To verify amp here you would need to inject a 45mhz signal that would then be mixed down to 12mhz in the next section (starting with Q30).
Backing up to the original issue: I have found that the BFO setting is one of the real heavy hitters in apparent receiver sensitivity.? A simple test that worked for me was to use a free audio spectrum analyzer program on a PC with mic input to the audio card.? I adjust the radio to a frequency with white noise (static) only and check the spectrum on the PC.? It should show amplitude between 500 and 2,500 hertz.? When I had low volume, it was because the SSB filter was cutting off one end of the spectrum or the other.? I went back and forth adjusting the BFO until the spectrum was centered between 500 and 2500 hz.? After doing that simple alignment, the S/N ratio got MUCH better.? I then did the WWV zero beat to insure calibration.
Link to the free program:
http://www.techmind.org/audio/specanaly.html
In doing the above, 2 things;
1 - I am using KD8CEC software
2 - I used the corresponding Memory Manager software that allows for easy changes and ability to go back to the original settings.?
I highly recommend the CEC software AND upgrading to a Nextion display.? You still should be able to check the current BFO setting through the above process.
Link to KD8CEC HamsKey:
Let me know if this helps.? Also any feedback on my helping, or confusing you would be appropriated.
"Every criticism is a gift in disguise"? Learn to accept it as a gift, and use it to improve.
73
Evan
AC9TU
|
Evan, indeed bfo adjustment is a factor, my recent v4 was way off. Maybe this could he in play here.
Dave, good work using your diagnostic equipment. I would first he suspicious that the automated assembly of the ubitx may be okay. Between bfo setting and understanding what a good ubitx sounds like might be all that is going on here.
Curt
|
Thanks I see as folks suggested? Q10 is past the first mixer of clk2.? That still doesn't explain why injecting at L4 is heard significantly less than injecting at L1.? I would think the closer I got to the amplifier the stronger the signal should be not weaker right? (I can try dropping the low pass filter out of the circuit I guess)
I've tried zero beat the 1khz signal while adjusting the BFO and analyzing the spectrum and the best setting still seems to be the default of 11995 for BFO.
(I compared the sound with a 1khz tone off my pc and off the signal generated from my frequency generator to my Yaesu and both match the ubitx so frequency and BFO seem good from that perspective at least.)
|
Actually I redid the BFO adjustment to zerobeat and at 11996.6 sounds better for 1khz, and at least the conversations are intelligible better now.? So that is a bonus. (conversation is still faint)
Still a 100uv signal injection should give 2 times s9 power expected but it still sounds very faint.
The strength does seem impacted by which end of the spectrum I'm at near 25MHZ seems a bit harder to hear the 100uv tone than compared to 7.150MHZ.
|
I read (and had the issue myself) where the USB and LSB seem to get reversed.? This was related to the LO being off.? It was solved for me by going back to the default after setting the BFO and then do the zero beat with WWV.? Since you have the signal generator, I would try to zero beat it with the calibration process in the setup menu.? I had the CEC memory manager by that time, so did not need to get it close the first time (good enough to be able to receive WWV) and then calibrate with WWV.
You might want to try calibrating using your signal generator on one of the upper bands to see if that changes the rigs response.? Of course, you want a pure carrier without any modulation to do that.
I would also be curious if there is any effect on changing from USB to LSB or vise versa and if the characteristics of lower signal on the higher bands changes to on the lower bands.
I believe that there is a "factory reset" function built into the stock code that resets the calibration to the values used in HF Signals.? If my memory is correct, it is engaged by holding in the encoder switch when turning on the power.? Of course, this will erase the BFO setting as well as the LO.
BTW, I have been assuming all along that you are running the stock software and display.? Is that correct?
Evan
AC9TU
|
So ....What I did on my "seasonal vacation"
Quick and dirty trace thru V3 ubitx ...Test points correspond to V4 locations ... Set ubitx freq to 7.152 mhz dial indication ...Maybe off a little as WWV? 10 mhz was silent this AM Scope at VOL-H on all tests, not moved? !0x probe to 2215a Tec ...? 0.1 V P-P for reference ... ? SG at 7.152? Maybe off (just because) -50 dbm into antenna jack gave 0.1 V P-P ? at TP-17 -60 dbm (300 uV) 11.996200 gave 0.1 V P-P at TP-16 -90 dbm (10 uV)?????? ....?????? ....? ?? 0.1 V P-P at TP-14 -70 dbm (100 uV)? 44.98500 mHz gave 0.1 v P-P at TP-13 -60 dbm (300 uV???? .......... ....????? 0.1 v P-P at TP-1?? -80 dbm (30 uV)?? ...........???????????? 0.1 v P-P at TP-1? -50 dbm? (1000 uV)? 7.150 mHz gave 0.1 v P-P at TP-2 -60 dbm (300 uV)? 7.150? mHz? gave 0.1 v P-P . "Test on my bencb ..on my ubitx ..with my SG ....YMMV (your mileage may vary) All measurements made with NO effort to "match" anything with anything else SG output 50 ohm thru 0.1 DC blocking cap...To clipleads ...to whatever Z might be where-ever ...NO modulation ...I get confused easily when looking for AM modulation on SSB stuff A Serendipitous Solstice To ALL
Jim
On Saturday, December 22, 2018, 2:56:26 PM PST, Dave Space <davesspacebar@...> wrote:
Actually I redid the BFO adjustment to zerobeat and at 11996.6 sounds better for 1khz, and at least the conversations are intelligible better now.? So that is a bonus. (conversation is still faint)
Still a 100uv signal injection should give 2 times s9 power expected but it still sounds very faint.
The strength does seem impacted by which end of the spectrum I'm at near 25MHZ seems a bit harder to hear the 100uv tone than compared to 7.150MHZ.
|
I'm not sure what to make of Jim's measurements.
YMMV indeed!
Two measurements at TP1 with a difference of 30dB?
Somewhere between 10 and 20 dB of gain or loss from TP2 to TP1,
depending on which TP1 measurement you choose?
The nominally 16dB amp between TP16 and TP17 gives 30dB of gain?
Though the 10dB of gain between TP13 and TP14 could well be legit,
that nominally 16dB amp is having a very hard time doing the right thing at 45mhz
given the low Ft of the 2n3904 transistors used.
I would expect maybe 1dB of loss going from TP2 to TP1,?
certainly not a difference of 10 or 20 dB.
And would not expect a 50 ohm signal generator to significantly upset things there.
Sounds like the original problem is somewhere around L1,2,3,4 and T2.
Could remove L1, L4, C200, C205,? and then short TP2 to TP1.
No effect on the receiver performance except that we will also receive a strong image
from any FM broadcast station at 2*45mhz+(OperatingFreq).
For example, when tuned to 7mhz, CLK2 is at 52mhz so that the incoming 7mhz signal
gets mixed down to 52-7=45mhz by the mixer at D1,D2.
However, an incoming signal at 52+45=97mhz (normally 97mhz gets knocked out by L1,2,3,4)
will also produce a 45mhz result out of the mixer at D1, D2.
So choose an operating frequency where such an FM broadcast station is not present.
And remember not to transmit without the L1,2,3,4 filter operational.?
How Dave sees more response when injecting at TP2 than he does when
injecting at TP1 is a mystery.? My guess is there is a cold solder joint or
some toroid has a broken wire somewhere between L1,2,3,4 and/or T2.
And with cables lying about to inject 7mhz into TP2, some of that 7mhz
gets coupled into the stuff around the mixer at T2 somehow.
That L1,2,3,4 filter could misbehave a little when the 50 ohm signal source drives TP1,
but I would be surprised if this was anywhere near as severe as what Dave reports
on a properly working uBitx.
Jerry, KE7ER
toggle quoted message
Show quoted text
On Sat, Dec 22, 2018 at 04:09 PM, jim wrote:
So ....What I did on my "seasonal vacation"
Quick and dirty trace thru V3 ubitx ...Test points correspond to V4 locations ...
Set ubitx freq to 7.152 mhz dial indication ...Maybe off a little as WWV? 10 mhz was silent this AM
Scope at VOL-H on all tests, not moved? !0x probe to 2215a Tec ...? 0.1 V P-P for reference ...
?
SG at 7.152? Maybe off (just because) -50 dbm into antenna jack gave 0.1 V P-P
?
at TP-17 -60 dbm (300 uV) 11.996200 gave 0.1 V P-P
at TP-16 -90 dbm (10 uV)?????? ....?????? ....? ?? 0.1 V P-P
at TP-14 -70 dbm (100 uV)? 44.98500 mHz gave 0.1 v P-P
at TP-13 -60 dbm (300 uV???? .......... ....????? 0.1 v P-P
at TP-1?? -80 dbm (30 uV)?? ...........???????????? 0.1 v P-P
at TP-1? -50 dbm? (1000 uV)? 7.150 mHz gave 0.1 v P-P
at TP-2 -60 dbm (300 uV)? 7.150? mHz? gave 0.1 v P-P
.
"Test on my bencb ..on my ubitx ..with my SG ....YMMV (your mileage may vary)
All measurements made with NO effort to "match" anything with anything else
SG output 50 ohm thru 0.1 DC blocking cap...To clipleads ...to whatever Z might be where-ever ...NO modulation ...I get confused easily when looking for AM modulation on SSB stuff
A Serendipitous Solstice To ALL
Jim
|
Hi,
I have not looked at the schematic. This problem reminds me of the failure I found in the front end of a different radio. There was indeed a broken winding on one of the inductors, one of those donut looking toroid thingies. I rewound the inductor and that restored receive operation. The signal just about vanished when the probe was in the circuit with the broken inductor, much like the description of the findings here. When I jiggled that inductor just right the signals came pouring in. Due the nature of the broken wire that was surprisng but the broken ends would make temporary contact. I was able to confirm the open with my ohm meter while the inductor was in-circuit (power off of course). I could *see* the fracture with a magnifying glass.
This may be helpful or may send the OP on a wild goose chase. I hope it can help.
73,
Bill KU8H
toggle quoted message
Show quoted text
On 12/22/18 7:57 PM, Jerry Gaffke via Groups.Io wrote: ---snip---- How Dave sees more response when injecting at TP2 than he does when
injecting at TP1 is a mystery.? My guess is there is a cold solder joint or
some toroid has a broken wire somewhere between L1,2,3,4 and/or T2.
And with cables lying about to inject 7mhz into TP2, some of that 7mhz
gets coupled into the stuff around the mixer at T2 somehow.
That L1,2,3,4 filter could misbehave a little when the 50 ohm signal source drives TP1,
but I would be surprised if this was anywhere near as severe as what Dave reports
on a properly working uBitx.
Jerry, KE7ER -- bark less - wag more
|
different frequencies ....
Jim
On Saturday, December 22, 2018, 4:58:05 PM PST, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:
I'm not sure what to make of Jim's measurements.
YMMV indeed!
Two measurements at TP1 with a difference of 30dB?
Somewhere between 10 and 20 dB of gain or loss from TP2 to TP1,
depending on which TP1 measurement you choose?
The nominally 16dB amp between TP16 and TP17 gives 30dB of gain?
Though the 10dB of gain between TP13 and TP14 could well be legit,
that nominally 16dB amp is having a very hard time doing the right thing at 45mhz
given the low Ft of the 2n3904 transistors used.
I would expect maybe 1dB of loss going from TP2 to TP1,?
certainly not a difference of 10 or 20 dB.
And would not expect a 50 ohm signal generator to significantly upset things there.
Sounds like the original problem is somewhere around L1,2,3,4 and T2.
Could remove L1, L4, C200, C205,? and then short TP2 to TP1.
No effect on the receiver performance except that we will also receive a strong image
from any FM broadcast station at 2*45mhz+(OperatingFreq).
For example, when tuned to 7mhz, CLK2 is at 52mhz so that the incoming 7mhz signal
gets mixed down to 52-7=45mhz by the mixer at D1,D2.
However, an incoming signal at 52+45=97mhz (normally 97mhz gets knocked out by L1,2,3,4)
will also produce a 45mhz result out of the mixer at D1, D2.
So choose an operating frequency where such an FM broadcast station is not present.
And remember not to transmit without the L1,2,3,4 filter operational.?
How Dave sees more response when injecting at TP2 than he does when
injecting at TP1 is a mystery.? My guess is there is a cold solder joint or
some toroid has a broken wire somewhere between L1,2,3,4 and/or T2.
And with cables lying about to inject 7mhz into TP2, some of that 7mhz
gets coupled into the stuff around the mixer at T2 somehow.
That L1,2,3,4 filter could misbehave a little when the 50 ohm signal source drives TP1,
but I would be surprised if this was anywhere near as severe as what Dave reports
on a properly working uBitx.
Jerry, KE7ER
On Sat, Dec 22, 2018 at 04:09 PM, jim wrote:
So ....What I did on my "seasonal vacation"
Quick and dirty trace thru V3 ubitx ...Test points correspond to V4 locations ...
Set ubitx freq to 7.152 mhz dial indication ...Maybe off a little as WWV? 10 mhz was silent this AM
Scope at VOL-H on all tests, not moved? !0x probe to 2215a Tec ...? 0.1 V P-P for reference ...
?
SG at 7.152? Maybe off (just because) -50 dbm into antenna jack gave 0.1 V P-P
?
at TP-17 -60 dbm (300 uV) 11.996200 gave 0.1 V P-P
at TP-16 -90 dbm (10 uV)?????? ....?????? ....? ?? 0.1 V P-P
at TP-14 -70 dbm (100 uV)? 44.98500 mHz gave 0.1 v P-P
at TP-13 -60 dbm (300 uV???? .......... ....????? 0.1 v P-P
at TP-1?? -80 dbm (30 uV)?? ...........???????????? 0.1 v P-P
at TP-1? -50 dbm? (1000 uV)? 7.150 mHz gave 0.1 v P-P
at TP-2 -60 dbm (300 uV)? 7.150? mHz? gave 0.1 v P-P
.
"Test on my bencb ..on my ubitx ..with my SG ....YMMV (your mileage may vary)
All measurements made with NO effort to "match" anything with anything else
SG output 50 ohm thru 0.1 DC blocking cap...To clipleads ...to whatever Z might be where-ever ...NO modulation ...I get confused easily when looking for AM modulation on SSB stuff
A Serendipitous Solstice To ALL
Jim
|
On second glance, those measurements show a 30 dB loss from TP16 to TP17, not a gain.
Also a 20 dB gain from TP14 to TP16, going through the 45mhz filter and D3,4 mixer.
That one should be on the order of a 10dB loss.
The only one that makes sense to me is the 10dB gain from TP13 to TP14
Very weird.
Jerry, KE7ER
toggle quoted message
Show quoted text
On Sat, Dec 22, 2018 at 04:57 PM, Jerry Gaffke wrote:
Though the 10dB of gain between TP13 and TP14 could well be legit,
that nominally 16dB amp is having a very hard time doing the right thing at 45mhz
given the low Ft of the 2n3904 transistors used.
|
Assuming your signal generator can correctly inject the levels shown in the chart,
you should get rational results here regardless of the different frequencies.
Namely,?
? Perhaps 1 or 2 dB loss through the LPF at L1,2,3,4
? 6dB through the mixers
? Maybe 3 or 4 dB through the two crystal filters
? And around 10 or 12 dB through the 45mhz amp, 16dB through the 12mhz amp.
I do like the way you went about the measurements though,
looking only at the audio with the scope means just about any scope could be used there.
Or perhaps a diode peak detector and a dc voltmeter.
The signal generator could be an si5351 driving a step attenuator.
Jerry
toggle quoted message
Show quoted text
On Sat, Dec 22, 2018 at 05:51 PM, jim wrote:
different frequencies ....
?
|
Bob Lunsford