Topics

QCX+ Si5351A Error #qcx #si5351a #error


Mark R.
 

Hello,
I am receiving the 'Si5351A error' on my QCX+ 40m kit. On initial power up everything seemed to be working fine, it prompted for what band the radio was. 
I attempted to input the band by hitting the tune knob, realising that was not selected I then hit the actual select button. The radio power cycled and that is when it 
started giving the error message. 

I have:
Checked all solder connections - everything looks good

Checked the connection of the 28 pin IC - IC is seated with no bent pins

Checked polarity of all diodes - all of the black stripes are over the white stripe on the PCB
  - D1 and D2 are getting 3.5 volts

Checked for continuity between pin 28 and R4 - getting 1k ohm on one side of R4 and 0 ohms on the other
  - Checked for continuity between pin 27 and R3 - getting 1k ohm on one side of R3 and 0 ohms on the other


Al Holt
 

I suggest an almost 'microscopic' inspection of the traces and connections around the si5351a chip and its pathways to the processor. I haven't had that kind of problem.

I try and get the excess rosin off the board using a toothbrush dipped in isopropyl alcohol and moderately scrubbing just about all the soldered side of the board. I think it makes it easier to see any 'iffy' connections. It also helps to flush away any possible 'whiskers.'

Since the si5351a is surface mounted a judicious treatment with alcohol/toothbrush might help clear up the issue.

I hope it helps!

--Al 


@CurtisM
 

okay, I am not digging deep into the troubleshooting just now - but when power recycles it can be because the voltage on a regulator is drawn way too low.  if your power supply is robust enough (not marginal in current for receive - seems doubtful) then maybe there may be a local short.  I imagine (and hope) its not yet a faulty PLL IC causing this matter.  if you are at a loss, perhaps patiently troubleshoot with the designer.  do look around the PLL IC to see if you dropped any solder there - but I think more likely its somewhere else, maybe near the controller or the regulator feeding it.  be careful and patient, it may be a small matter causing this. 

73 curt


Mark R.
 

Thank you guys for your suggestions!
Al, I took your advise literally and looked at the si5351a chip under a microscope, all of the solder connections look good. When it comes to looking at traces I don't quite know what I am looking for, but I didn't see anything that looked damaged or out of place. I cleaned the board up with some isopropyl acholic and inspected all of my connections, everything still looks good. I don't see any bridging, cold joints, or solder flakes on the board.

Curtis, If I understand what you are saying the PLL IC would be the 28 pin IC? I have taken a close look around that and entire board and in doing so I found out I had installed C37 backwards. I have righted that issue to no avail. I have also taken a look at my power supply and realized it was only rated for 500mAh, I have switched that for a 2A supply. That has not resolved my problem.

Now here is where the problem gets even more confusing... somewhere between Al's suggestions and Curtis' the radio actually began to work. I don't have an antenna strung up (but had one connected) so I was not able to receive. I was able to tune around and go into menu's, then after maybe 3 minutes it power cycled again and I am back to the Si5351a error.


Al Holt
 

Mark, 
Thanks for the update! Sorry that your still haven't cracked it, so to speak. C37 is the electrolytic cap on the output of the +5V regulator. Have you verified that you're getting +5V to the system? (it's nice having those breakout points on the PC board!) 

Rereading the OP, I see you checked the data lines of the synthesizer to the processor and the diodes used to drop +5V down to +3.5V. Wouldn't hurt to check that voltage again making sure it's still there and feeding the chip.

Are you using the crystal or TCXO option?

Can you do a reset of the system, or are you stuck with the error message and can't get past it?

I've found the QCX draws about 100 mA in receive so you were probably safe with the original supply.

Since it did work briefly, maybe some gentle tapping and wiggling of board and components might make it spring into life again.

Thanks,


--Al


Mark R.
 

Al, 
I am just using the crystal, I do not have the TCXO option. I have checked and I am getting 5V from the breakout pins. 
I cannot reset anything, on powerup I get the error displayed and cannot navigate to any menus. 

I have also received an email from Hans, which I will include as well as my response so that this page is updated with everything I have tried.
Hello Mark
 
I saw your emails to the QRP Labs forum, I am not sure if that helped narrow it down? 
 
I saw that you had it working at some points then it stopped... it does seem like an intermittent connection somewhere. You are correct to concentrate on looking at the I2C bus, but also the power supply to the Si5351A; note that it has two Vdd pins, both must be at 3.3V (or 3.5 as the diodes give). The SCA and SCL connections must connect to the correct processor pins (27/8) and they must also both have 1K pull-ups to 3.3V. If any of that isn't right, or if there is a short from any of those signals to ground, or a short to any other nearby signals, then the Si5351A Error would be reported (because the processor cannot communicate with the Si5351A). 
 
73 Hans G0UPL
http://qrp-labs.com

Response 
Thank you for your reply Hans,
If I understood correctly what you were asking, I should see 3.3V on pin 1 and pin 7 of the Si5351a chip, I tested those and it was reading 3.4V. I am not certain what a SCA connection is, but from what I googled it should be the connector I soldered onto the PCB that the actual 28 pin IC slots into? As for the SCL, is that the pathway illustrated in red running between IC4 and IC1 on page 14 of the instructions? I measured both R3 and R4, they both read 2K ohm on the left pin(closest to LCD side of PCB) and 1K ohm on the right pin(closest to antenna). Is this what I should be seeing, or should it be 0 ohm then 1K ohm? I re-measured the voltages of D1 and D2, D1 was reading 3.4V and D2 was reading 4.2V, from what I understand both of these should be reading 3.5V + or - a small amount? I am continuing to look over all of my solder connections, and other than having C37 installed in reverse I have not found any problematic connections. 
 
Thank you,
Mark


Al Holt
 

Mark,
IIRC, D1 and D2 are in series and each one drops ~0.7V, so you'll see that voltage difference between the two diodes.

I think you're getting close to the point of needing an oscilloscope. Any chance of borrowing one? The idea here is to see if you have data communication on the I2C bus, but I'm not sure if you can tell much with a voltmeter. 

I wonder, if the crystal is bad would it prevent the si5354a from signaling? 

Another idea is to reflash the microcontroller, or order another one from Hans. There are ways to do this with the chip you have, but you'll need to research it on this forum to see what the best approach for you to take. 

If you have time to do some probing around to see if you have an intermittent connection, as you probably know, use an insulated tool.

Good luck!

--Al


Mont Pierce KM6WT
 

Hi Mark,

Do you think it might be heat related?   If you leave it sit off for awhile, will it work for a short period before resetting and failing?

Do you have a can of compressed air, like the ones used to blow the dust out of a keyboard?

Some of the hardware engineers I used to work with in past lives, used different tools to heat or cool components on a failing PCB to see if anything was heat sensitive..
Sometimes micro-fractures in a PCB, or trace, or solder joint, or a cracked component can be very difficult to see...  

If they did find something heat sensitive, then they would get the microscope out for closer inspection.


Just a thought...


73,
km6wt


James Daldry W4JED
 

Hi, Guys

There is a "canned cold" available as component cooler. It sprays liquid freon or similar low boiling point solvent onto the component in question to test for thermal sensitivity. If you needed to dust off something you could turn the can upside down and get just the vapor. In the same way, "canned duster" which blows out solvent vapor can be used as component cooler by turning the can upside down and carefully wetting the part in question.

73

Jim W4JED

On 10/12/20 5:45 PM, Mont Pierce KM6WT wrote:
Hi Mark,

Do you think it might be heat related?   If you leave it sit off for awhile, will it work for a short period before resetting and failing?

Do you have a can of compressed air, like the ones used to blow the dust out of a keyboard?

Some of the hardware engineers I used to work with in past lives, used different tools to heat or cool components on a failing PCB to see if anything was heat sensitive..
Sometimes micro-fractures in a PCB, or trace, or solder joint, or a cracked component can be very difficult to see...  

If they did find something heat sensitive, then they would get the microscope out for closer inspection.


Just a thought...


73,
km6wt


Mark R.
 

Thank you guys again for your suggestions, 
I've tested the radio under a few conditions. The radio does not work after being powered on from cold (room temp without having power connected), all I get is the error. I also left the radio powered on for 15 minutes, it still did not work. I carefully felt around for any hot components, but didn't feel anything. I should have canned air somewhere. Am I spraying parts one at a time to see if they might get the radio to work?

As for oscilloscopes, what exactly would you testing for with it in this case? I don't know a whole lot about them, but my local ham store does repairs so I would imagine they might have one. Its worth giving them a call but I don't know what I need them to test. 

You said something about the possibility of the crystal, would that be XTAL2 the 27.00MHz one? I look around for a replacement, worth a shot replacing. 

Thanks,
Mark


Alan G4ZFQ
 

You said something about the possibility of the crystal, would that be XTAL2 the 27.00MHz one? I look around for a replacement, worth a shot replacing.
Mark,

The message simply means that there is no communication between the Si5351 and IC2. IC2 is working. IC1 is not.
Yes, listen for the approximate 27MHz, often around 27.004 with a RX.

I'm not sure if the 27MHz is required for IC1 to respond to IC2, does anyone know?

Otherwise it is a bad connection on one of the two control lines SCL, SDA. Bad joint, track broken or shorted to ground.

73 Alan G4ZFQ


Mark R.
 

I am going to replace the 27MHz crystal. It should be here in a fews days. I will also call my local store and see if they would be able to assist with an oscilloscope. I'll post some updates when those things happen 


Al Holt
 

Mark,
Were you able to try picking up the 27 MHz signal on another receiver? I checked mine and had to hold the tip of the PL-259 connector about 1/2" away from the synthesizer chip.

If not, you might be able to 'kick start' it by tapping (not too hard) on the crystal's housing with something firm like the plastic end of a small screwdriver.

Reheat (reflow) solder connection to the crystal?

Just checking, thanks.

--Al