[moved] MCP7940N Real time clock issues

Status
Not open for further replies.
A

aliyesami

Full Member level 6
Joined
Jan 7, 2010
Messages
369
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,296
Location
USA
Activity points
4,190
I have a working clock on breadboard where I have soldered the crystal to the chip itself n it works perfect.
Now I'm building a circuit based on smd components but the clock is not advancing.
I did change the crystal that was recommended by Microchip support, with another crystal n the clock started to work but overnight it stopped n even after power reset it won't start.
I can set the time ok n it also retaining time but the time is not advancing.
 

Hi,

It seems you have a hardware problem.
Show us your schematic and a picture of your circuit.

Klaus
 

so last night I de-soldered the crystal and put it in upright position and the clock started to work but over night it lost 4 hrs.
There must be some stable hardware layout and component list for this chip.
loading the board layout.
please note that the crystal in the layout is not the one iam using , i am now using the crystal from Citizen CFS-20632768DZBB (digikey # 300-8301-ND)

https://obrazki.elektroda.pl/9883350100_1445522839.jpg
https://obrazki.elektroda.pl/6049329500_1445522906.jpg
https://obrazki.elektroda.pl/4230705000_1445522942.jpg
 

Hi,

with the given xtal you need about 20pF each load capacitance. I wonder why there are two different sizes of capacitors. Use ceramics NP0/COG

A shorter and low inductive connection to the red(top) ground plane is better, but should give no problems here.

What are the SCL and SDA line voltage for normal operating and batery operated?
Is the clock voltage (normal and battery) within it´s spec´d range?

You did not accidentally ultrasonic clean the xtal (or the whole PCB)?
What about flux residuals?

Or an accidental overheating of the xtal?

Klaus
 

KlausST said:
Hi,

with the given xtal you need about 20pF each load capacitance. I wonder why there are two different sizes of capacitors. Use ceramics NP0/COG
Click to expand...
I used 10pf and 8pf as per datasheet recommendations, their sizes are different as 8pf cap is not available in 1206 package.
What are the SCL and SDA line voltage for normal operating and batery operated?
Is the clock voltage (normal and battery) within it´s spec´d range?
Click to expand...
i will check these voltages later today and report

You did not accidentally ultrasonic clean the xtal (or the whole PCB)?
Click to expand...
no i did not clean the pcb , it was made by the board house professionally.

What about flux residuals?
Click to expand...

I dont know how to measure this .

Or an accidental overheating of the xtal?
Click to expand...
I tried few crystals and was very careful in soldering them with heat dial set on the middle on a 40 watt soldering iron.

Klaus[/QUOTE]

- - - Updated - - -

hi Klaus
do you think the crystal i mentioned itizen CFS-20632768DZBB (digikey # 300-8301-ND), can be good to use with mcp7940 ?
if yes what value of capacitance should i use with it ?
 

Hi,

The capacitors must fit to the xtal, not to the oscillator ic.

In the RTC datasheet you see xtal examples. Both xtal in the example use lower load capacitance than the used xtal.
Therefore I recommend to use higher value xtals.

I don't have the xtal datasheet by hand now, but if I remember right the recommended load capacitance us 12pF.
In your circuit there are two capacitors, they act as serially connected, therefore each needs twice the capacity like given in the datasheet, so 24pF. But the oscillator has I/O capacitance. This needs to be subtracted from the 24pF.

So 20 or 22pF may be the right choice.

Maybe worth a try.

***
The pcb is professionally soldered by the boardhouse, most probably they have used no clean solder paste.
So I don't expect flux residual problems.

Klaus
 

hi Klaus !

just to make sure i understand correctly, i can use this crystal CFS-20632768DZBB (digikey # 300-8301-ND) with 20 or 22pf capacitors ceramics NP0/COG ?

thanks
 

hi Klaus !

thanks for your advise i replaced the caps with 20pf and now the clock is working exact to the seconds
 

just as i thought all was i noticed that actually the clock is loosing about 5 secs a day
I am using 20pf caps for CX1 and CX2 . and the crystal is digi-key number 300-8341-1-ND
manufacturer number Citizen CMR200T32768DZBT .

the datasheet of the crystal is saying load capacitance of 12.5 pf ,should i be using 12.5 pf instead of 20pf ?

- - - Updated - - -

correction: the load capacitance for this crytal is listed as 6pf , should I be using 6pf for CX1 and CX2 instead of 20pf ?

- - - Updated - - -

this is the formula in the datasheet to calculate the CX1 an CX2 values , can you tell me how to interpret it ?
 
Last edited:

Hi,

Now you changed the crystal...
correction: the load capacitance for this crytal is listed as 6pf , should I be using 6pf for CX1 and CX2 instead of 20pf ?
Click to expand...
..I went to digikey, used your order number, downloaded CMR200T datasheet.
I can't find anything about 6pF. It clearly says load capacitance is 12.5pF.

****
this is the formula in the datasheet to calculate the CX1 an CX2 values , can you tell me how to interpret it ?
Click to expand...
Did you even try to use it?
The only value that is somehow unknown is C_stray. It depends on layout. Usually from sub pF to some pF. It may be 0.5 ...2pF.
Let's calculate with 1pF.
C_osc is given in the datasheet to be 3pF (parameter D8)
And let's calculate wiht two 20pF external capacitors.

Then CX1 = CX2 = 20pF + 3pF = 23pF.

Then the formula is: 23pF x 23pF / ( 23pF + 23 pF) + 1pF = 529 pF / 46 pF + 1pF = 11.5 pF + 1pF = 12.5pF.
You are free to calculate with your own capacitor values...

*****
You uploaded the picture with the informations of the datasheet.
Did you read the last two sentences of chapter 5.2.1 in your picture?

Klaus
 

sorry iam not getting it . . i have to derive the values of CX1 and CX2 and i know the CL =6pf and C_Osc =3pf.
so I have two unknowns in this equation how do i get their values?
yes i read the note it says the crystal load capacitance should be between 6-9 pf , the crystal iam using its load capacitance is 6pf .

thanks
 

Hi,

--> Datasheet says you should use a load capacitance of 12.5pF. Don't you agree?
***
The value of your external capacitor (20pF) is NOT the load capacitance. To calculate the load capacitance you must use the given formula.
For sure you can decide to use the same capacitor for CX1 and CX2, then there is only one unknown.
You can reorganise the formula and calculate the value. ( Or you can read my example to see that 2 x 20pF give the perfect load capacitance of 12.5pF).
For raw estimation:
If you need 6pF of load capacitance, C_oscillator = 3pF, C_stray = 1pF.
6pF - 1pF = 5 pF
5 pF × 2 = 10 pF (two same capacitors in series)
10 pF - 3pF = 7 pF (each if the two external capacitors).

Klaus
 
hi Klaus !
i was talking to the tech support guy from digi key and he told me that datasheet does not belong to the part i bought, the crystal i bought has a load capacitance of 6pf he confirmed.
so i should be using 7pf for CX1 and CX2 ?
 

Hi,

only you know what you have bought....
(My hint: Rely more on the manufcaturer (informations, datasheet) than on the seller)

Take the values as calculated above...or calculate it on your own.

+/- 10% of the calculated values should not result in a big frequency differ.

Klaus
 

I was not interpreting the datasheet correctly , its a 6pf crystal . see the note below.
I will take the values you calculated , so 7pf for CX1 and CX2 correct?

Mike_ext_2213: I see what you mean in the data sheet. Below that link on our site we have another that calls out the part number breakdown, click here for this document. If we look at the manufacturer part number, CMR200T32768DZBT the letter B in the breakdown does indicate 6pF load cap. I think the first link is more of a "general" device data sheet.
 

hi Klaus
the 7pf did the trick , I observed the clock over days and it didn't even loose or gained a second .

thanks for your help
 

Status
Not open for further replies.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…