"PLL NOT LOCK PLEASE REBOOT"... Davis Vantage Pro2 Weatherstation Repair

[alexig]

I've probed pin 9 on the chip and it has a 2.99 voltage.  What does this mean? 
Locked or Unlocked?  The chip works with 3V.  Does this mean that the pin is high,
and therefore actually PLL Locked?

The datasheet says:
9 LOCK Digital output PLL Lock indicator, active low. Output is asserted (low) when PLL is in lock. The pin can also be used as a general digital output, or as receive data output in synchronous NRZ/Manchester mode

So if it's high 3V then PLL is not locked which is what you are seeing. But check with a scope as it could change very quickly.

With your scope, see if you can turn on low frequency reject in the trigger menu and see how you go with the 50Hz.

When the power is off, check the tantalum caps with your multimeter by switching to ohms (make sure they are discharged already before doing so).

[hgg]

Hello alexig,

So if it's high 3V then PLL is not locked which is what you are seeing. But check with a scope as it could change very quickly.

I missed that.   So, PLL is not locked indeed.
I probed pin 9 with the scope and it confirms the DMM measument.

Pin 9 with DC coupling:

Pin 9 DC

Pin 9 with AC coupling:

Pin 9 AC

Here a small 976mVpp oscillation is showing which looks similar with the signal of the crystal.

By the way the LF Reject helped a lot here!

Now for the capacitors.
I've measured all the tantalum capacitors with the DMM in Ohms range and all of them were
showing very high resistance in the MOhms range.  So no capacitor is completely dead.
When I measured them with the ESR meter though, I got the following results:

C85 = 0.036R
C76 = 0.036R
C14 = 0.032R
C21 = 0.029R  and
C11 = 0.124R...  This is the one sitting next to U11 (voltage regulator?).

All the capacitors apart C85 have a capacitance of 10uF.  C11 has 4 times greater ESR
than the rest of them!  Can this be an indication of a faulty cap?  Maybe if not enough
current can pass to the C1021 circuit it might not be able to start the crystal oscillation
as stated in the datasheet.

What do you think?

[dr_p]

I dont't think it's a problem, but to be sure you can add a good 10uF on top of it.

[hgg]

I've just tried a new 10uF on top of C11, but nothing happened.
Same PLL error... 

There is a tiny tiny capacitor though labeled C28, that has a 9 Ohms ESR!
It measures exactly 3nF.  Its right next to the crystal on the left side in the photo.
Isn't that too high?  Could it be that the 9 Ohms is a false reading?

I have contacted Davis support with all the details and they've just replied to me saying that they do not
repair the faulty consoles.  They just replace the boards.  (I hate that habbit of the 21st century... )
He also said that "sometimes it is the crystal Y51 (14.7) but it could be the radio chip or the CPU"

Now, it does not seem likely to me that it is either the CPU or the radio chip.  The weather
station after the PLL error, works as usual and responds to every command and function.
The PPL error is reported correctly as well.  Its not a false positive.

I have also asked him if they have taken into account Note 002 of the CC1021 Errata and
programmed the firmware correctly in order to avoid PLL lock errors that are inherent to
the CC1021 chip!  If he replies that they have, then the only thing that remains is to remove
the crystal and test it.

Don't you agree?

[hgg]

Pippy,

That is what I will do.    Maybe its just the crystal after all.
(Do you mean the tantalum ones or the tiny with the 9 Ohms ESR as well?)

Thank you guys.  You have been very helpful indeed!

[SeanB]

If you do remove the crystal try resoldering it back the other way round. this may help, it does work in some cases with sluggish crystals.

[hgg]

The two little brown ceramic smt caps by the xtal

Ok, I see.  Thanks again!

If you do remove the crystal try resoldering it back the other way round. this may help, it does work in some cases with sluggish crystals.

I will try that as well.  If nothing happens I will test it on a Pierce oscillator circuit.

[tsmith35]

I opened up my Davis VP2 wireless remote transmitter and found a similar crystal in there. Peeled back the conformal coating to reveal:

TXC G ("G" is very small)
14.7 Ur03 (or UrOC)

Might have a look at http://www.txccrystal.com/crystal.html or call them for details. Here's the overview.

Also, I read about another person who had the same problem. They removed the batteries and left the unit sitting for a long while, and the unit came back up fine when power was reapplied.

[hgg]

tsmith35 Hi,

Thank you for the link.
I think I have found the exact crystal on RS Electronics.
http://uk.rs-online.com/web/p/crystal-units/6719264/?searchTerm=CAA-14.7456-18-3050-X&relevancy-data=636F3D3226696E3D4931384E4B6E6F776E41734D504E266C753D656E266D6D3D6D61746368616C6C7061727469616C26706D3D5E5B5C772D5C2E2F252C5D2B2426706F3D313326736E3D592673743D4D414E5F504152545F4E554D424552267573743D4341412D31342E373435362D31382D333035302D582677633D424F544826

I am having the problem in the console, not the transmitter.  The transmitter works fine.

If I leave it for a long time with no batteries or power, then when it starts again it will receive some
data packets but then, it will display again the PLL NOT LOCK ERROR...

So the faulty component whichever it is, has not died yet completely.

I am waiting for a response from Davis just in case they have addressed the issue described
in the CC1021 errata, in a recent firmware.  If they have, then I will remove the crystal to test it.

My SMD skills are not the best, so I will wait... 

[tsmith35]

I am having the problem in the console, not the transmitter.  The transmitter works fine.

Yeah, I know the issue you're having is in the console, but I figured Davis would use the same frequency for locking both...

FWIW, any difference running on batteries vs wall wart?

For that matter, are you running the latest (v3.12) version of the console firmware?

[hgg]

FWIW, any difference running on batteries vs wall wart?

No difference whatsoever.

I am not running the latest firmware, that is why I've asked Davis is they have addressed
a similar problem in their firmware updates.  I am waiting for an answer.

I cannot update the firmware now and check for myself, because I don't have the USB adapter
with me at the moment....

[hgg]

From what I see, to check if a crystal is working is not so straight forward.

I will use the following Pierce oscillator (maybe the inverter oscillator is more simple) which
outputs the main frequency of the crystal.



Pierce Oscillator

It uses an N-Channel JFET but I don't have any and they are no so easy to find as well.
My question is, can I use a MOSFET instead of the JFET?  They both have a high input impedance
and the MOSFET even higher.  But with higher input impedance you get a lower gain.
Do you think that this substitution will work?

I have some FQP30N06 lying around...

[hgg]

Hi,

I have just replaced the crystal and it does exactly the same thing!   
I have run out of ideas...

Any suggestions?   (apart throwing it in the trash...)

[TrickyNekro]

The temperature and the humidity is measured by a small part at the left side of the microcontroller. I am not exactly sure about the code, the part is from Sensirior,
probably a SHT11 or something similar. The "blue" chips seems to be an atmospheric pressure sensor. I´ve seen something like this from a Chinese company, can´t
remember more though.
The PLL not locking is probably happening because you have more jitter than allowable in your osc circuit. First thing I would check would be the power rails.
Check all the possible power rails please, with your oscilloscope. Get the oscilloscope in AC mode and measure the maximum ripple. That would be a good start.

Forgive me if something like that was said before, I´ve read most of the posts but I might have missed something.

If you insist on throwing it away, I can accept the grand as well :-P

Best Regards, Lefteris
Greece

[tsmith35]

hgg, when did you buy the unit, and from what company/location? Perhaps I can pester Davis a bit...

[hgg]

I was wondering what the little SHT chip was.  Its on a tiny board place on the main board.

I don't have any schematic diagrams, so I checked the ripple on all of the pins of the smd transistors/voltage regulators. 

The middle pin of the three of U11 :  (178mVpp...)  Supply voltage is 5V and the multimeter
reads 4.3V now.


U11

Below the 3rd (right) pin of U11 :  (Multimeter shows 2.999V steady.  That goes to the CC1021)


U11 3rd pin

Both the above were measured with AC coupling.

I also get a very strange signal on the base pin of Q11.  This time DC coupling was used.


Q11 Base

Do you think there is a problem with the 5V rail?  Can I check anything else?

tsmith35 very kind of you!    I have it many years now, I cannot remember how many.
It should be from 1996 because that is the date of the first weather data I can find in the PC. 
I remember that I bought it from the provantage.com company.  I can search for the receipt. 
I might have it.  What can you do though?  They don't seem very cooperative...

[TrickyNekro]

If I am correct U11 is the voltage regulator, or at least it looks like this, which also seems to be lineal since I can´t spot any inductors around
(plus only 3 pins means no feedback pin so yeah).

Replace every tantalum capacitor you see. If you can´t find tantalum caps around, use electrolytic caps but at least 10 times bigger in value.
I see that most of the tantalums are 10uF, replace them with either the same, and if not possible in the immediate future use 100uF electrolytic.
In parallel to the electrolytic caps use some 100nF ceramic also, that´s gonna help by a bit.

But I think my friend that your problem is with the voltage rails. Replace the caps test the device and if it´s not ok then also replace the regulator.

That´s all for the moment, I wait for your reply!!!

Best Regards, Lefteris
Greece

[hgg]

A lot of work!   

U11 is almost certainly the voltage regulator.  It the only component that outputs 3V & 5V
at the same time.

A small detail.  When you switch on the unit it goes into a data entry mode first in order to input
some info about your coordinates, units and from which transmitter it will receive its weather
data.  If you measure the 5V & 3V rails during that time, they are perfectly stable with 24mV
ripple max on the 5V supply rail.  Crystals Y21 & Y22 are working normally and crystal Y51 is off.
After that when you press ok it goes into its RF data acquisition mode and crystal Y51 starts and
stops every two seconds approximately.  That's when the 5V power rail has the large ripple and
pulsing. 

5V are going into D12.  (A large barrel diode) Then 4.3V are coming out of D12 and go into the
C11 capacitor.  All the other power rail points have either 3V or 4.3V after that.

[TrickyNekro]

Can you power it from something like 7+ V let´s say? Please try that and report back, though I don´t feel like that´s the problem, still please do it,
maybe it´s a part of it though.

But yeah what you describe above is failed caps. The caps are bad so they can not handle the RF burst that is emitted?
I have the feeling by now that the communication between the devices is bidirectional.

If yes, then it´s really bad caps... Try to find low ESR caps for the replacements, that´s important ;-)


Best Regards, Lefteris
Greece

[hgg]

Lefteri if you help me fix it, the frapedes are on me!   

I just tried the 7V test and it looks like that the regulator might be sick...
The 3V output went from 2.999V to 3V but the 5V output (middle pin) and base went to 7V !
Do you think that this is the problem

If its the caps here as well, then its the most common problem in electronic repairs.
Can't they just make a capacitor that works!?

[AndyC_772]

Depends who you mean by "they". If you mean any of the reputable manufacturers, then they already do make reliable caps. If you mean the manufacturers that make cheap crap, then they don't. It's all down to your definition of "they".

[hgg]

Andy if you have followed the repair of my SONY FH-909 stereo, it was a faulty relay but three of its
capacitors were bulging and they were Nichicons...

[hgg]

I have also checked all the tantalum capacitors with the new DE REE DE-5000 LCR and they all have an
ESR of 0.14 @ 100Khz and a Tan D of 0.11   @ 120Hz except C11 next to the voltage regulator that has an
ESR of 0.30 @ 100Khz and a Tan D of 0.066 @ 120Hz.  (although Tan D looks way better for some reason)

Looks like that there is a problem with the regulator and its capacitor.
Lefteri what do you think?

(Ok, Update.  U11 is definitely a voltage regulator with an output of 3V.   Since the power supply is 5V
it cannot have a 5V rail as well.  It only has a 3V rail.  The only other place I can measure 4.9V is the
little Q23 transistor below the buzzer's electrolytic capacitor. When I lower the supply voltage below
3V a low battery error is displayed on the screen.  I will replace all the tantalum capacitors and see what
happens.)

[TrickyNekro]

What I see in your power rails seems to be something like an RF burst.
An RF burst takes a lot of power, in a very short time. My best guess is this, I really don´t know mate.

You are using the unit more than 10 years, or at least it is that old. Your power rails have that noise,
which indicates that something is drawing a lot of power in short pulses and the pulses seems like
data. RF burst consume a lot of power, 0.5A, 1A depends on your output power really.
Everything points to the caps, but you can never be 100% sure. Change them and let us know :-)

Best Regards, Lefteris
Greece

[hgg]

Hello Lefteri,

I have ordered today new tantalum capacitors.

The unit can be powered from mains and batteries as well.  My main power supply has very little
noise.  When I power it up from the battery terminals, the ripple is much less but the problem
remains the same.  The battery power comes to the board from the big central ribbon connection.

I will replace the caps and see what happens.
Thanks!