RTC Circuit

I’m Adding an RTC to my G120 Project and thought I’d start by looking at the cobra extension board schematic. See the attached image. Can anyone recommend what capacitor size that should be? Also. I’m getting some mean clock creep. A bit over a minute a day. Any suggestion how to combat that. My crystal circuit is also attached.

Thanks in advance.

clock creep: buy a better crystal ! Likely to be the PPM variability that you’re falling victim to.

What is the intent of the capacitor ? It says “DNP”, ie don’t place it - if you were using this from a GHI schematic, it’s probably unnecessary.

Is your battery rechargeable?

I’d agree with Brett on the crystal being the most likely issue. What’s the tolerance on the one you are using?

Also, what is the frequency of the crystal? I made a goof in the past where I ordered 32Khz and not 32.768Khz ones before and couldn’t figure out why the clock was so far out.

1 Like

I’m using a CR2032. I don’t think they’re rechargeable? Here is the crystal I used:


I’m not sure what PPM is. Or how I came to using 18pf capacitors. Any advice is much appreciated.


You need to change the design if using a CR2032 and not rechargable. You are applying voltage to the cell via the VCC and 1.5K resistor.

Change the resistor for another diode and then you will be fine. This way VCC applies power to the RTC circuit and when this fails, the battery supplies the voltage.

The only thing I can think on the crystal is the 9pF might be an issue. I tend to use 12.5pF as this is what the RTC normally calls for load wise. I have seen timing drift with an 18pF cap before so is worth checking this.

Thanks Again. If I put a diode in place of the 1.5k resistor, I’ll just need to make sure the forward voltage isn’t to high. The LPC1788 sheet says it needs a minimum of 2.1v.

Also, Maybe this crystal would be better?


Also how do you choose the size of C106 and C107?

Both the parts you pointed to are 20ppm tolerance. Same Same. :slight_smile: PPM in this context is parts per million, the deviation of the crystal to spec. A lower number is better.

And forgot to mention your question on crystal load caps. You basically need a load capacitance to have the crystal work at the stated spec. Your load capacitance required from the part dictates how accurate it will be as well, so if your part says 12.5pF required, then you need a total capacitance of 12.5pF, including residual capacitance in your PCB/circuit and capacitance in oscillator in the processor. To get 12.5pF (ignoring everything else) you would simply use two 25pF caps (as the caps are in series). If you assume you might have a few pF in stray capacitance you only might need 20pF caps, or 21pF (which can be how you further tune your result as @ Dave was mentioning).

THANKS HEAPS! That totally makes more sense now. Thanks Everyone for the help.

Sorry to bother you again but I want to be sure.

Will this diode work with this circuit do you think?



I’d say it’d be ok. Personally, before I committed it to a board I’d be testing it on the bench :slight_smile: If you’re in Sydney duck over to the trade counter and pick them up :wink: .

:wall: :wall: I’m in Melbourne. Didn’t even know they had a trade counter up there. Would be wicked if we got a trade counter down here? Looks like I’ll be placing another order tonight then. just got to get up to free shipping with a bunch of other parts.

Thanks for the tip.

Great to have another Aussie on the forum! Would love to hear what you’re doing if it’s not a commercial secret, always love to hear what the local industry is going on with. And if you need anything “mainstream” fro GHI in a hurry, let me know as I possibly have it sitting around (often in as new unopened condition :slight_smile: )

It’s a bit secret ATM, But it will soon be all revealed. I’ll Send you a PM though.


1 Like