Topics

QCX 40m troubleshooting #qcx #problem #qcx40


Leighton
 

G'day,

I have built and had my 40m QCX (firmware 1.01a) on air and have had some contacts with it with great success (apart from my lack of CW skills - hi hi). I have been really impressed by it and I want to thank Hans for such well thought out and great kit. 

After a some quite successful use, I seem to have ended up with a subsequent fault. I am hoping that someone might be able to point me in the right direction and fill in the gaps in my knowledge. 

I have gone through the troubleshooting traced back to a problem with the bias voltage at T1. My 5V supply rail seems ok (4.95V measured right side of D2). The voltage reading at T1 pins 5 & 8 is 0.26V.

I have carefully removed T1 pins 5 & 8 and measured the voltage from the divider at T1 pins 6 & 7 which measures 2.47V. This seems to be the correct bias voltage, but only correct with T1 (5 & 8) disconnected. Measuring the voltage on the disconnected T1 pads 5 & 8 show 0.27V. (I am not sure if it is significant, but I have confirmed that there is actually about 0.01V difference). Resistance between the right side of D2 and T1 pad 6 is 5.45k.

In summary, these are the measured voltages:
Right side of D2: 4.95V
T1 pins 5 & 8: 0.26V
The following with T1 pins 5 & 8 removed:
T1 pads 5 & 8: 0.27V
T1 pins 6 & 7: 2.47V

Any help or pointers would be greatly appreciated. I would love to have my QCX back on air.

Thanks,
Leighton VK2LI


Ted 2E0THH
 

HI Leighton

Greetings from a soggy Cambridge, UK.

Could you describe the actual fault that you encountered? Clearly the QCX was working fine beforehand.
Does the display still work? Can you still receive?
It will help the the wonderful learned people on here get you quickly on the air again.

73s Ted

2E0THH

 


Alan G4ZFQ
 

Measuring the voltage on the disconnected T1 pads 5 & 8 show 0.27V. (I
Leighton,

T1 pins 5 & 8: 0.26V
The following with T1 pins 5 & 8 removed:
T1 pads 5 & 8: 0.27V
T1 pins 6 & 7: 2.47V
Something around IC4, C43-46 seems to be shorting to ground.
Examine very carefully around pins of IC4 and everything connected to them.
Measure the resistance to ground on the four capacitors and T1 pads.

73 Alan G4ZFQ


Leighton
 

Hi Ted,

Greetings from sunny, warm NSW, Australia. :)

Sorry, I should have tried to explain the fault condition. Some of the symptoms are similar to those found in the troubleshooting guides and I have tried to go through those before hassling people here, but I may still have missed something.

Except for receiving signals, everything seemingly works (display, transmit, audio circuits, etc.). Encoder clicks can be heard and the audio circuits go into some sort of feedback and self oscillation if the audio is turned up beyond about 1/2 volume.

When going into the alignment menu to peak the BPF, peaking now occurs with C1 completely un-meshed. This is different to its original alignment behaviour, where C1 was about 1/3 to 1/2 meshed.

This behaviour is also affecting the further alignment of the radio and I was why I went looking further for the fault.

Let me know if you want any more info.

73,
Leighton VK2LI


Leighton
 

Hi Alan,

Thanks for the reply and assistance.

I had a similar thought, but there doesn't seem to be a direct short to ground.

These are the resistance measurements, to ground, that I have made on IC4:
Pin 1: 0
Pin 2: 6.68 M
Pin 3: 3.95 M
Pin 4: 18.5
Pin 5: 3.95 M
Pin 6: 11 M
Pin 7: o/c (T1 pin 5 disconnected)
Pin 8: 0
Pin 9: o/c (T1 pin 8 disconnected)
Pin 10: 3.94 M
Pin 11: 18.5
Pin 12: 3.94 M
Pin 13: 11 M
Pin 14: 6.8 M
Pin 15: 0
Pin 16: 1.95 k

Would I be correct in thinking that 18 ohms on IC4 pin 4 and 11 doesn't seem quite right? 
I also measured IC5 pin 5 - 118 ohms to ground.

I have examined carefully around the pads of C46, R8 and IC4 pins 4 and 11. I also carefully removed C46 to check that it wasn't faulty. (I know, the likelihood of such a component faulty is low). 

Measurements on between ground and T1 pins (T1 pins 5 & 8 connected):
Pin 1: 0
Pin 2: 0
Pin 3: 0
Pin 4: 0
Pin 5: 5.5 k
Pin 6: 5.5 k
Pin 7: 5.5 k
Pin 8: 5.5 k

Any further thoughts? 

Thanks once again.

73,
Leighton VK2LI


Steven Dick
 

18 ohms on IC4 pins 4 and 11 are not right as they are different from the other IC4 outputs. But you also say that with T1 pins 5 and 8 connected you measure 5.5K on T1 pins 5 and 8 (with IC4 unpowered.) So everything looks normal with T1 connections into IC4 as well as IC4 inputs.  I would carefully check all  IC4 pins for shorts to adjacent pins or an unsoldered pin such as a power or ground pin which could cause IC4 to exhibit strange effects.

-Steve K1RF

------ Original Message ------
From: "Leighton Judd" <leighton.judd@...>
Sent: 9/23/2020 5:49:15 AM
Subject: Re: [QRPLabs] QCX 40m troubleshooting #qcx #qcx40 #problem

Hi Alan,

Thanks for the reply and assistance.

I had a similar thought, but there doesn't seem to be a direct short to ground.

These are the resistance measurements, to ground, that I have made on IC4:
Pin 1: 0
Pin 2: 6.68 M
Pin 3: 3.95 M
Pin 4: 18.5
Pin 5: 3.95 M
Pin 6: 11 M
Pin 7: o/c (T1 pin 5 disconnected)
Pin 8: 0
Pin 9: o/c (T1 pin 8 disconnected)
Pin 10: 3.94 M
Pin 11: 18.5
Pin 12: 3.94 M
Pin 13: 11 M
Pin 14: 6.8 M
Pin 15: 0
Pin 16: 1.95 k

Would I be correct in thinking that 18 ohms on IC4 pin 4 and 11 doesn't seem quite right? 
I also measured IC5 pin 5 - 118 ohms to ground.

I have examined carefully around the pads of C46, R8 and IC4 pins 4 and 11. I also carefully removed C46 to check that it wasn't faulty. (I know, the likelihood of such a component faulty is low). 

Measurements on between ground and T1 pins (T1 pins 5 & 8 connected):
Pin 1: 0
Pin 2: 0
Pin 3: 0
Pin 4: 0
Pin 5: 5.5 k
Pin 6: 5.5 k
Pin 7: 5.5 k
Pin 8: 5.5 k

Any further thoughts? 

Thanks once again.

73,
Leighton VK2LI

Virus-free. www.avast.com


Leighton
 

Thanks Steve.
Checked for shorts on IC4 between pins and adjacent pins. The only ones at 0 ohms are pin 5 and 12 as per the intended trace on the board. No other pins show any signs of a short circuit or significantly low resistance.
IC4 pin 16 confirmed 0 ohms to right side (anode) of D2.
IC4 pins 1, 8 and 15 confirmed 0 ohms to ground. 
Also checked IC2, 3, 5, 6, 7, 8, 9 and 10 ground pins - all 0 ohms to ground.

Is there a possibility that IC4 pins 7 and 9 have internally gone short circuit (18 ohms)? I realise that it is not the most obvious possibility (soldering, bridging pads/tracks, dry joints, etc. is certainly most obvious).

I have attached a pictures of the pins on both sides of IC4. As far as I can see, they visually look clean on the outside.

73,
Leighton VK2LI


Evan Hand
 

Leighton,
I would agree with the internal issue with IC4.  The circuit equivalent is enhancement-mode MOSFETs between input and output.  One or two are most likely faulted to ground.  The only way to tell would be to remove and replace it.

The more difficult question is what caused it.  For that, I do not have a clue.

You might want to order 2 just in case the replacement fails.

Good luck with your hunt for the issue.
73
Evan
AC9TU


Leighton
 

Hi Evan,
Thanks for the input. 

Looks like my next step is to set about replacing IC4. 

I am not sure what caused it either. I found out when I wasn't hearing stations and went through the alignment process again. Most likely will be something that I have done, but I don't know what. Maybe I cooked it somehow?

I have now removed IC4 and have ordered a replacement (or two). I have also checked the FST3253 IC. I measure 18.5 ohms between pin 8 and pin 11 where nothing can be measured between pin 8 and the other outputs.

Hopefully this is the only fault that I have and replacing the IC will resolve the issue.

73,
Leighton VK2LI


Alan G4ZFQ
 

I would agree with the internal issue with IC4.  The circuit equivalent

Leighton, Evan,

A few have had internal shorts. I did not want to say this at first in case it influenced troubleshooting.
As to the reason, who knows? There have been a few here that have failed in a similar way.
I have been in the Softrock group for years, two FST3253s are hand-soldered in and I do not recall this problem. Maybe not reported?
Leighton, at least we do not have to persuade you to try removal:-)

73 Alan G4ZFQ


Leighton
 

Hi Alan,

I appreciate your consideration and not suggesting I go through the pain of removal first up. :-)

Replacing IC4 was not high on my list of components that I would like to replace. I didn't want to risk further damage and considered how I might remove it. With my modest equipment and a bit of care, it wasn't all that bad to remove.

Now to wait for a replacement to turn up. :-)

73,
Leighton VK2LI


Evan Hand
 

Leighton,
Good job on the IC4 removal.  I have found that a desoldering iron, either the plunger or bulb type, works well.  If you are uncomfortable with removing again it again if it fails again, then you might want to put an IC socket in to allow for easy replacement.  I would not recommend that, as there are other potential problems as Hans has pointed out in the construction manual.  The one exception is the microcontroller, that has soo many pins, and has more potential to be replaced for updates as well as failure.  It also has little impact on the RF side with slightly longer leads.

Your tests confirmed that the device did fail.  Alan's comment is interesting, as this is the first that I have seen on the QCX group about IC4 failure.  With over 11,000 sold and assumed built, I would guess that you were just the unlucky one to get a bad part.  I doubt that it was something that you did unless you went poking around in the circuit and zapped it with some static.  That is one of the disadvantages of CMOS, they are more susceptible to static discharge.

Good luck with the repair.
73
Evan
AC9TU


Steven Dick
 

This discussion about the FST3253 got me curious. A good reference for
the bus switching chips like the FST3253s is On Semiconductor's app note
AN-5007. See:
https://www.onsemi.com/support/design-resources/appnotes?rpn=FST3253
originally written by Fairchild Semiconductor.

The good news is that these are robust devices with internal protection
diodes on inputs and outputs, so they are fairly robust with regard to
possible damage from static electricity. However, it is still a CMOS
device and could be potentially damaged with any static electricity
anywhere along the manufacturing from non-static proof packaging to
improper handling once out of the package to static discharge exposure
during assembly to static discharge to the PC board. The damage may or
may not show up immediately. Or it could have been a defective device,
which would be highly unusual unless the chips are not directly from the
manufacturer. When assembling the QCX+ PC boards, it's a good idea to
follow good anti-static practices. In my working days, I had an
expression: "CMOS loves the summer. CMOS hates the winter." Another
common problem (which is not the problem here) is power sequencing,
where a driving device and a receiving device are powered by two
different power supplies. If the source device is powered first, it
forward biases the protection diodes exceeding their maximum current
rating in the receiving device and can eventually damage it. A third
occasional problem (which is not a problem here) is biasing CMOS
complementary PNP-NPN inverter circuits in the wrong place so both
transistors turn on simultaneously and draw a lot of current from the
power supply, eventually damaging it. I have seen all of these, many
times.

-Steve K1RF

------ Original Message ------
From: "Alan G4ZFQ" <alan4alan@...>
To: QRPLabs@groups.io
Sent: 9/24/2020 2:50:08 AM
Subject: Re: [QRPLabs] QCX 40m troubleshooting #qcx #qcx40 #problem


I would agree with the internal issue with IC4. The circuit equivalent

Leighton, Evan,

A few have had internal shorts. I did not want to say this at first in case it influenced troubleshooting.
As to the reason, who knows? There have been a few here that have failed in a similar way.
I have been in the Softrock group for years, two FST3253s are hand-soldered in and I do not recall this problem. Maybe not reported?
Leighton, at least we do not have to persuade you to try removal:-)

73 Alan G4ZFQ





--
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus


Alan G4ZFQ
 

Alan's comment is interesting, as this is the first that I have seen on the QCX group about IC4 failure.
Evan,

I have looked at most posts in this group since before the QCX was introduced. I forget most things but I do remember a few FST3253 failures. A very small number compared with the number of kits..
Maybe ESD? But they do not seem particularly susceptible, I've had about 6 Softrocks connected permanently to antennas for more than 10 years and no failure. But we do not get too many storms or static electric here.

73 Alan G4ZFQ


Alan G4ZFQ
 

I had an
expression: "CMOS loves the summer. CMOS hates the winter."
Steve,

I was surprised when someone in the USA said that.
Here, if we have really dry conditions it is far more likely to be summer.

73 Alan G4ZFQ


Steven Dick
 

OK - here summers (hotter) are humid and winters (colder) are dry. Guess
the saying needs to be modified for international use.

-Steve K1RF

------ Original Message ------
From: "Alan G4ZFQ" <alan4alan@...>
To: QRPLabs@groups.io
Sent: 9/24/2020 9:32:03 AM
Subject: Re: [QRPLabs] QCX 40m troubleshooting #qcx #qcx40 #problem

I had an
expression: "CMOS loves the summer. CMOS hates the winter."
Steve,

I was surprised when someone in the USA said that.
Here, if we have really dry conditions it is far more likely to be summer.

73 Alan G4ZFQ





--
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus