Topics

QCX+ IQ Balance, AF Gain


Erik
 

I recently finished building the QCX+ and it powers up and receives. I haven't built a kit before so this has been quite interesting and fun!

Unfortunately I have noticed a few problems/quirks, probably due to some sort of error I've made but I wish to seek some assistance with thinking through them:

With a dummy-load attached, the BPF was able to be tuned well and within the specs outlined in the guide. However, the IQ Balance seems to be a challenge to adjust. The volume knob seems very sensitive and will reach peak volume (13) within a very small rotation, whereas the guide suggests half way rotation - this is seemingly too much on mine. I adjusted the volume down to get within the range specified but when adjusting the potentiometer, it barely moves one or two small lines over the course of many many rotations and seems to on one end, the same is true for both low and high phase adjustments - just barely a hair of movement on the screen with quite a bit of turns on the potentiometers, the lowest value for each being at one extreme end for each of those potentiometers. 

I have checked all the toroids and all are the correct amount of windings for 40m and I've checked the continuity for each and they are as expected - they are making good connection with the solder joints. The windings on T1 are all the same direction, as per the instructions. The solder joints appear ok and I've checked that the components are in their correct places and orientations. I wonder what I've missed, and where to begin.

Interestingly, if one uses another radio to produce a tone, at say 7.030, one will hear a mirror of that tone which is the same sound 1.4khz above the true signal. The same is true when listening to live signals using an antenna. 

Does anyone have advice? I don't have much test equipment - limited to other radios and a multi-meter and don't have an oscilloscope. 

Any help would be appreciated,

Erik
VA3JH


Gregg Myers
 

Hello Erik,

The IQ balance R17,R24 and R27 is part of the circuit which reduces your received signal from 'double sideband' to 'single sideband'. In your case, you seem to not be able to tune these properly... therefore you hear a mirrored 'double sideband' signal. See page 154 of the manual for the particular circuit section where this happens.

There is a troubleshooting guide on the QRP-labs site you can follow, which is always a good place to start. But since many problems are as simple as a bad solder joint or possible solder short somewhere, I would first go over your soldered connections very carefully in good light and with a magnifier. Reflow any suspect connections and then test again. A bad solder joint is often not as easy to spot as you might think. Hopefully, it is that simple.

73,
Gregg

On Wed, Jul 15, 2020 at 1:39 PM Erik <eriksomfay@...> wrote:
I recently finished building the QCX+ and it powers up and receives. I haven't built a kit before so this has been quite interesting and fun!

Unfortunately I have noticed a few problems/quirks, probably due to some sort of error I've made but I wish to seek some assistance with thinking through them:

With a dummy-load attached, the BPF was able to be tuned well and within the specs outlined in the guide. However, the IQ Balance seems to be a challenge to adjust. The volume knob seems very sensitive and will reach peak volume (13) within a very small rotation, whereas the guide suggests half way rotation - this is seemingly too much on mine. I adjusted the volume down to get within the range specified but when adjusting the potentiometer, it barely moves one or two small lines over the course of many many rotations and seems to on one end, the same is true for both low and high phase adjustments - just barely a hair of movement on the screen with quite a bit of turns on the potentiometers, the lowest value for each being at one extreme end for each of those potentiometers. 

I have checked all the toroids and all are the correct amount of windings for 40m and I've checked the continuity for each and they are as expected - they are making good connection with the solder joints. The windings on T1 are all the same direction, as per the instructions. The solder joints appear ok and I've checked that the components are in their correct places and orientations. I wonder what I've missed, and where to begin.

Interestingly, if one uses another radio to produce a tone, at say 7.030, one will hear a mirror of that tone which is the same sound 1.4khz above the true signal. The same is true when listening to live signals using an antenna. 

Does anyone have advice? I don't have much test equipment - limited to other radios and a multi-meter and don't have an oscilloscope. 

Any help would be appreciated,

Erik
VA3JH


James Daldry W4JED
 

Hi, Eric

What you are describing is a missing I or Q channel. The whole point of splitting the signal into I and Q is to bring them together and cancel out the upper or lower apparent signal.

So - where do we look? Check the DC voltage on IC7 pin 1 and IC6 pin 1. They should be very close to identical, around 2.5 volts. Whichever one is off is the channel that has the problem. The most likely suspect is that a wire on T1 didn't get stripped properly, but looks just fine. The active connections for this problem are wires 5, 6, 7, and 8. If you have the same voltages on R5, R6, R8, and R9 the transformer is OK, as well as IC4. Next most likely is on the feedback resistors 7, 10, 13, 16, 20, and 23. The easy way to find which one is that the pin 1 or pin 7 connected to it will be slam against ground or slam against 12 volts.

The neat part of this exercise is that you have 2 identical circuits that should be doing identical things. When you find the point where the 2 sides are identical, and at a move to the right they're not, you've found the problem.

This should get you started on your trouble shooting voyage. It might even find you the problem. Once corrected, you'll have to re-do R17, R24, and R27. Have fun.

73

Jim W4JED

On 7/15/20 3:21 PM, Erik wrote:
I recently finished building the QCX+ and it powers up and receives. I haven't built a kit before so this has been quite interesting and fun!

Unfortunately I have noticed a few problems/quirks, probably due to some sort of error I've made but I wish to seek some assistance with thinking through them:

With a dummy-load attached, the BPF was able to be tuned well and within the specs outlined in the guide. However, the IQ Balance seems to be a challenge to adjust. The volume knob seems very sensitive and will reach peak volume (13) within a very small rotation, whereas the guide suggests half way rotation - this is seemingly too much on mine. I adjusted the volume down to get within the range specified but when adjusting the potentiometer, it barely moves one or two small lines over the course of many many rotations and seems to on one end, the same is true for both low and high phase adjustments - just barely a hair of movement on the screen with quite a bit of turns on the potentiometers, the lowest value for each being at one extreme end for each of those potentiometers. 

I have checked all the toroids and all are the correct amount of windings for 40m and I've checked the continuity for each and they are as expected - they are making good connection with the solder joints. The windings on T1 are all the same direction, as per the instructions. The solder joints appear ok and I've checked that the components are in their correct places and orientations. I wonder what I've missed, and where to begin.

Interestingly, if one uses another radio to produce a tone, at say 7.030, one will hear a mirror of that tone which is the same sound 1.4khz above the true signal. The same is true when listening to live signals using an antenna. 

Does anyone have advice? I don't have much test equipment - limited to other radios and a multi-meter and don't have an oscilloscope. 

Any help would be appreciated,

Erik
VA3JH


Shirley Dulcey KE1L
 

Is that mirror tone 1.4 KHz away just as loud as the primary signal, or is it significantly lower?

If it's the same volume, you're not getting any opposite sideband rejection. Image rejection in a phasing receiver like the QCX and QCX+ happens by the magic of mixing two signals that are at a 90 degree phase angle: the two mixer products add together for the desired sideband and cancel out for the other one. Because the phase and amplitude relationship of the two signals is not perfect you don't get perfect cancellation, but a properly aligned QCX should be able to get 40dB or more rejection of the other sideband at the center frequency of the audio filter. You do the phase alignment at that frequency so that's where you get the best rejection; opposite sideband rejection above and below that center frequency is lower but those signals are also attenuated by the audio bandpass filter so they're less of a problem.

There are two things that could cause that. Either your I/Q balance setting is way off, or one of the two paths (I or Q) isn't working at all and you're only receiving one of them (and therefore no cancellation). 

If you get your hands on an RF-capable oscilloscope at some point, things to look for:

1. The I/Q LO signals from the Si5351A are both reaching the mixer and have the correct phase relationship. You will need a two channel scope to check the phase relationship.

2. Both the I and Q signals are present at the outputs of the preamps IC5A and IC5B. The QCX+ has convenient jumpers where you can sample them, JP7 and JP8. On a QCX you'll have to find the appropriate places on the board to probe.

3, The QCX+ has convenient jumpers JP9 and JP10 that you can use to look at the signals just after they come through the phase shift filters but before they are combined. (Take off the jumper caps to take the summing resistor out of the circuit.) You should see the same tone on both I and Q, but if you're looking at the wrong sideband one of them should be inverted (180 degree phase shift).

If the mirror tone is much softer (hopefully at least 40 dB), your receiver is operating normally. You may still hear the opposite sideband of strong signals like the one coming from your local signal source.


On Wed, Jul 15, 2020 at 3:39 PM Erik <eriksomfay@...> wrote:
I recently finished building the QCX+ and it powers up and receives. I haven't built a kit before so this has been quite interesting and fun!

Unfortunately I have noticed a few problems/quirks, probably due to some sort of error I've made but I wish to seek some assistance with thinking through them:

With a dummy-load attached, the BPF was able to be tuned well and within the specs outlined in the guide. However, the IQ Balance seems to be a challenge to adjust. The volume knob seems very sensitive and will reach peak volume (13) within a very small rotation, whereas the guide suggests half way rotation - this is seemingly too much on mine. I adjusted the volume down to get within the range specified but when adjusting the potentiometer, it barely moves one or two small lines over the course of many many rotations and seems to on one end, the same is true for both low and high phase adjustments - just barely a hair of movement on the screen with quite a bit of turns on the potentiometers, the lowest value for each being at one extreme end for each of those potentiometers. 

I have checked all the toroids and all are the correct amount of windings for 40m and I've checked the continuity for each and they are as expected - they are making good connection with the solder joints. The windings on T1 are all the same direction, as per the instructions. The solder joints appear ok and I've checked that the components are in their correct places and orientations. I wonder what I've missed, and where to begin.

Interestingly, if one uses another radio to produce a tone, at say 7.030, one will hear a mirror of that tone which is the same sound 1.4khz above the true signal. The same is true when listening to live signals using an antenna. 

Does anyone have advice? I don't have much test equipment - limited to other radios and a multi-meter and don't have an oscilloscope. 

Any help would be appreciated,

Erik
VA3JH


Tony McUmber
 

Don't know about anyone else, but I get inconsistent results when I slam a component or circuit on the ground.

I will check that suggestion, however, as I have trouble with sideband rejection on my QCX 40.

Thanks!

73 Tony N0BPA


N3MNT
 

Check the volume control with an ohm meter, looking specifically for the resistance between the two end terminals.  The solder terminals


N3MNT
 

Check the volume control with an ohm meter, looking specifically for the resistance between the two end terminals.  The solder terminals make contact with a printed resistive element in the volume control.  It is very easy to loosen this connection when you are bending the terminals to solder them.  You can sometimes make the connection reliable by squeezing the terminal directly at the round hold in it.  If you do not have a connection on either of the end terminals it will do exactaly what you are seeing. It will also have a very loud click on the first character sent. 


Erik
 

Hi Jim,
Both IC7 and IC6 pin 1 are at about 2.6 volts each. 
Voltage for R5, 6, 8, and 9 are all about 2.54 volts. 
The voltages between IC6, IC7 (pin 1 or 7) and  resistors 7, 10, 13, 16, 20, and 23 is near 0 for each, but not quite... 

From pin 1 on IC6:
R7 0.050v
R10 0.051v
R13 0.016v
R16 0v
R20 0.015v
R23 0v

from pin 1 on IC7:
R7 0.035v
R10 0.035v
R13 0.016v
R20 0v
R23 -0.015v
Definitely nothing at 12v though. 

Erik


Erik
 

Shirley,

I wish I had a scope and understood how to use one! Indeed the mirrored tone 1.4KHz up is identical in tone and volume to the original tone. Therefore it would appear the unwanted sideband is not being rejected at all. 

and to N3MNT - I checked the volume control, resistance between the two end terminals is 5.12 kilohms, including on the other side of the board, and there is variable resistance between each outer pin and the middle pin, again on each side of the PCB.

Erik


Gregg Myers
 

Erik,

Try checking for continuity of R27 wiper contact back to pins 1 of IC6A and IC7A. If either of those paths are open you will not get sideband' rejection.

-Gregg

On Wed, Jul 15, 2020 at 3:38 PM Erik <eriksomfay@...> wrote:
Shirley,

I wish I had a scope and understood how to use one! Indeed the mirrored tone 1.4KHz up is identical in tone and volume to the original tone. Therefore it would appear the unwanted sideband is not being rejected at all. 

and to N3MNT - I checked the volume control, resistance between the two end terminals is 5.12 kilohms, including on the other side of the board, and there is variable resistance between each outer pin and the middle pin, again on each side of the PCB.

Erik


Erik
 

Hi Gregg, 

It appears that there is no continuity between R27 and pin 1 on IC6 or IC7. 

Erik


Gregg Myers
 

Hmmm... one of the phase shift paths must be getting through to R27 wiper or you wouldn’t receive any signals. At least one path must have some continuity through to the next stage. I don’t have the plus model QCX but it appears there are some jumpers (3 of them) right around R27. They are all in place and good continuity through them, correct?

On Wed, Jul 15, 2020 at 4:20 PM Erik <eriksomfay@...> wrote:
Hi Gregg, 

It appears that there is no continuity between R27 and pin 1 on IC6 or IC7. 

Erik


Erik
 

There are indeed three jumpers there. JP11 has continuity through to JP10 (q out) and 0.5kilohms out to jp9 (i out). There also appears to be 0.5kilohms between the q and I jumpers


Shirley Dulcey KE1L
 

Excellent point, Gregg! If one of the two receiver phases isn't working but the other is, you should be able to identify the dead one by removing JP9, and then putting back JP9 and removing JP10. If removing one of those makes the receiver go quiet, the fault is in the circuit on the other side. If JP9 out and JP10 in works, look for the fault around IC5A, IC7, and the X0, X3, Y0, and Y3 outputs of the QSD (pins 3, 6, 10, and 13). If JP9 in and JP10 out works, look for the fault around IC5B, IC6, and the X1, X2, Y1, and Y2 outputs of the QSD (pins 4, 5, 11, and 12). Also check that both the CLK1 and CLK2 signals are reaching the QSD; they come in on pins 2 and 14, and can also be tested at pins 1 and 2 of JP2.


On Wed, Jul 15, 2020 at 6:26 PM Gregg Myers <gregg.w7grm@...> wrote:
Hmmm... one of the phase shift paths must be getting through to R27 wiper or you wouldn’t receive any signals. At least one path must have some continuity through to the next stage. I don’t have the plus model QCX but it appears there are some jumpers (3 of them) right around R27. They are all in place and good continuity through them, correct?

On Wed, Jul 15, 2020 at 4:20 PM Erik <eriksomfay@...> wrote:
Hi Gregg, 

It appears that there is no continuity between R27 and pin 1 on IC6 or IC7. 

Erik


Theo
 

Did you ever figure out what was wrong? I have almost the exact same problem as you, only difference being JP11 has continuity to JP9 and 500 ohms to JP10. 


Dean Smith
 

JP7 has track layout error!
this affects the I channel.
Found by Paul M0bmn..
Hans is aware..


m0bmn
 

The only problem with JP7 is that it cant easily be cut for isolation if your trying to fault find,  this track problem will not in itself cause any problem other than that. for those that haven't built the QCX+ the word 'Jumper ' is a little misleading, there are no jumpers on the QCX+ , each JP'x' point is a break point to help with fault finding, they are a solid PCB track with a cut point if needed, once cut then a link can be fitted between the the two points on the circuit board  given a reference JP'x'.
it is most unlikely that there will be a break on the PCB between JP points if the builder hasn't cut the track themselves.
the biggest help for me was to use the trouble shooting guide on Hans site, with a low value resistor across R43 i could see the signal on my little scope and follow it through the the receiver till i found i had a small short on two tracks. it will be a harder task without a scope, the other way i tested it at each stage was with a old crystal eaphone, one end to gnd and the other to a probe which i could move around, it has about 20 Mohm impedance so didn't upset anything, a big help if you haven't a scope.
good luck, its a great radio when you get it finished so stick with it.
73 Paul M0BMN
 


Hans Summers
 

Hi all

Yes, as Paul says... the "JP7 problem" is a very minor thing. It does not affect anyone using a QCX in the most normal kind of way. Just building it and using it. The "JP7 problem" is just a deviation from my declared intention of how the JP (pairs of pads at 0.1-inch spacing) are to be used - in so far as I intended that you could break the signal path by carefully cutting the trace between the two pads, and then use a jumper wire or connector for later repair, or for connecting to the Dev board for alternative experimental circuit blocks etc. So the "JP7 problem" is very minor indeed and does not affect ordinary operations. 

I have already corrected the trace in my PCB layout and so when I next manufacture a batch of boards, that change will be included. 

73 Hans G0UPL
http://qrp-labs.com

On Sat, Jul 25, 2020 at 5:02 PM m0bmn via groups.io <m0bmn=yahoo.co.uk@groups.io> wrote:
The only problem with JP7 is that it cant easily be cut for isolation if your trying to fault find,  this track problem will not in itself cause any problem other than that. for those that haven't built the QCX+ the word 'Jumper ' is a little misleading, there are no jumpers on the QCX+ , each JP'x' point is a break point to help with fault finding, they are a solid PCB track with a cut point if needed, once cut then a link can be fitted between the the two points on the circuit board  given a reference JP'x'.
it is most unlikely that there will be a break on the PCB between JP points if the builder hasn't cut the track themselves.
the biggest help for me was to use the trouble shooting guide on Hans site, with a low value resistor across R43 i could see the signal on my little scope and follow it through the the receiver till i found i had a small short on two tracks. it will be a harder task without a scope, the other way i tested it at each stage was with a old crystal eaphone, one end to gnd and the other to a probe which i could move around, it has about 20 Mohm impedance so didn't upset anything, a big help if you haven't a scope.
good luck, its a great radio when you get it finished so stick with it.
73 Paul M0BMN
 


Theo
 

I'm at a complete loss for where my fault could be. I've followed the debugging guide here with both an oscilloscope and DVM and everything seems to be as expected. I just can't make any adjustments to IQ Balance, Phase Lo, or Phase Hi, by turning the trimmers. Does anyone have any ideas where to look?


Arv Evans
 

It seems that with a dual-trace scope one would be able to see relative levels and relative phase
offset between I and Q, or at the summation of the two.  This looks like an opportunity for someone
to write an article on phase shift balancing using various readily available test equipment.

Arv
_._


On Sat, Jul 25, 2020 at 12:49 PM Theo <theotheo36@...> wrote:
I'm at a complete loss for where my fault could be. I've followed the debugging guide here with both an oscilloscope and DVM and everything seems to be as expected. I just can't make any adjustments to IQ Balance, Phase Lo, or Phase Hi, by turning the trimmers. Does anyone have any ideas where to look?