High precision (sub 1ppm) low power RTCs?

[ajc]

Hi,

I'm trying to design an RTC that delivers a timing error of better than 1ppm.  So far the closest I've been able to get is the Maxim DS3232, but this is +/-3.5ppm (across full temp range) + around 1ppm / yr (ageing).

I've looked for 32.768kHz oscillators, but can't find any with sub 1ppm specs.

If power wasn't an issue, of course there are solutions out there... but in this case it is a battery powered device that needs to operate several years on a single battery...

Perhaps I've already found the state of the art in what's available, but thought I'd put the question out there in case someone can suggest a solution.

Thanks in advance!

[ivan747]

Get an accurate oscillatorin a different frequency and divide it down. Or use GPS.

[Paul Price]

High precision oscillators use temperature controlled "ovens" to set an exact temperature well above the expected ambient. Even then, over days to years, component aging cause some drift, but a very small drift. There are some xtal oscillator modules off the shelf, that have a built in temperature control for the xtal.

[amyk]

If power wasn't an issue, of course there are solutions out there... but in this case it is a battery powered device that needs to operate several years on a single battery...

What's the capacity of the battery? You've quantified the precision but not the power requirement, so give the latter and we'll have a better idea of the design contraints.

[ftransform]

I'm trying to design an RTC that delivers a timing error of better than 1ppm.  So far the closest I've been able to get is the Maxim DS3232, but this is +/-3.5ppm (across full temp range) + around 1ppm / yr (ageing).

Seems virtually impossible since an OCXO is out of the question. We need to know a bit more about your design, what conditions it operates under etc.

For what it's worth many years ago we bought a load of Dallas 2PPM RTCs but found it was very hit and miss if they met the spec.

It sounds like you need to periodically re-sync your clock, or find some way of mitigating drift as a problem. For example a product I have been working on needs extremely accurate timing (a few 10s of uS variation between two completely separate devices maximum) but it wasn't possible to do in a low power and low cost way. We eventually found a way of syncing the two devices even though there was no comms between them :-) I'd tell you how but then I'd have to kill you, but the point is there may be a way around your rather extreme requirement.

mmm

[mikeselectricstuff]

That is a very hard problem - you should maybe look again at why you need that level of accuracy and see if you can do it a different way.
Depending where you are geographically, a MSF/DCF radio-controlled clock may me an option to periodically adjust the internal timebase.
 

[ajc]

I'm trying to design an RTC that delivers a timing error of better than 1ppm.  So far the closest I've been able to get is the Maxim DS3232, but this is +/-3.5ppm (across full temp range) + around 1ppm / yr (ageing).

Seems virtually impossible since an OCXO is out of the question. We need to know a bit more about your design, what conditions it operates under etc.

For what it's worth many years ago we bought a load of Dallas 2PPM RTCs but found it was very hit and miss if they met the spec.

It sounds like you need to periodically re-sync your clock, or find some way of mitigating drift as a problem. For example a product I have been working on needs extremely accurate timing (a few 10s of uS variation between two completely separate devices maximum) but it wasn't possible to do in a low power and low cost way. We eventually found a way of syncing the two devices even though there was no comms between them :-) I'd tell you how but then I'd have to kill you, but the point is there may be a way around your rather extreme requirement.

Yep, I was kind of figuring a periodic sync may be the best option here.  We do have a wireless interface which could be used....intrigued by the no comms / sync technique though....

The power question is still a bit open in that there will be a single battery supplying power to the rest of the system as well (and this is still being worked on).  What is defined is that we need 5 year battery life and I was kind of hoping I could just throw a backup battery in there with the RTC so the time only has to be set once and is retained even if batteries are changed.   In short, I've kind of assumed I need uA consumption for time keeping rather than mA (which is of course a problem as soon as I look at any decent precision oscillators).

Separately I've also wondered about the distribution of timekeeping error across multiple devices and whether you could essentially correct for error by averaging the time from multiple devices.  I suspect though that much of the error will be temperature dependent and similar between devices.  This doesn't help with power consumption either...

That is a very hard problem - you should maybe look again at why you need that level of accuracy and see if you can do it a different way.
Depending where you are geographically, a MSF/DCF radio-controlled clock may me an option to periodically adjust the internal timebase.
 

The requirement is actually a regulatory one which kind of makes it hard to avoid.

I'll have to look into whether there are clock sync signals available we could use (in Australia). 

[mikes]

Rakon makes a tcxo with a claimed 5 year 1.5 ppm stability (RFPT400), which draws 3 mA. You could clock a low power microcontroller to get RTC functionality.

[amyk]

The power question is still a bit open in that there will be a single battery supplying power to the rest of the system as well (and this is still being worked on).  What is defined is that we need 5 year battery life and I was kind of hoping I could just throw a backup battery in there with the RTC so the time only has to be set once and is retained even if batteries are changed.   In short, I've kind of assumed I need uA consumption for time keeping rather than mA (which is of course a problem as soon as I look at any decent precision oscillators).

Depends how big the battery is... 150Ah is enough to power that 3mA part above for over 5 years.

[baoshi]

I'm not sure if it is related but those token issued by banks for 2 factor authentication contains very accurate clock that produces pass codes derived from time. Maybe I can tear down one to see what's inside.

[mikes]

I'm not sure if it is related but those token issued by banks for 2 factor authentication contains very accurate clock that produces pass codes derived from time.

No, they don't. The rotating code gets resync'd at the server side each time the dongle is used.

[rr100]

That's correct, they don't have better precision than a normal watch xtal (we're talking probably tens of seconds per month). The server compares with a very few neighboring (time-wise) codes and then if it matches, you're in. It can't compare with too many because it would increase too much the chances of guessing the code so if you drift too much it will ask for multiple codes (so you can still prove with a high probability that you own the token) and then sync from that.

[SeanB]

They use a 30 second window, so calculate both the current code and the next 3, and remember the last 3 so they can compare with the one you present. They probably reckon the battery will die before it gets out of the read window, and as the battery is not a changeable part ( you need to buy a new one then) a drift of 60 seconds per year on a cheap low cost watch crystal is acepptable.

[ajc]

Rakon makes a tcxo with a claimed 5 year 1.5 ppm stability (RFPT400), which draws 3 mA. You could clock a low power microcontroller to get RTC functionality.

Good suggestion... I'll look into the part.  3mA is probably still too much for the system, but I'll run the calculations and probably put a couple of design approaches forward.  One with a much bigger battery!

[mc]

Yep, I was kind of figuring a periodic sync may be the best option here.  We do have a wireless interface which could be used....intrigued by the no comms / sync technique though....

It is truly ingenious, and no I didn't think of it :-)

I'm intrigued by this.
One method I can think of is if both are somehow connected to a common trigger, you could use that common trigger and an algorythm to ensure each clock 'syncs' to that trigger. That common trigger could be a common AC mains supply, or whatever is being monitored itself.

[ivan747]

Why don't you use the most accurate crystal you can power and at the end of the day sync it with GPS or a radio broadcasted clock like Mike Harrison says?

[ajc]

Why don't you use the most accurate crystal you can power and at the end of the day sync it with GPS or a radio broadcasted clock like Mike Harrison says?

You are right and this is the approach we are going to take for now.  Use the 3.5ppm Maxim RTC (ignoring ageing) and try to sync it at least every 12 months to maintain sync to within 1ppm.  The trade off in this case is that we now need GPS included in the design, but this seems to be the best way to go about it.

Thanks everyone for your suggestions and help!   

Adrian

[rr100]

So in fact you can live with 4+s drift? What about one of the radio clocks from your area? They need much less power than GPS (also you can get much easily the signal indoors) and you could sync every day.

[ajc]

So in fact you can live with 4+s drift? What about one of the radio clocks from your area? They need much less power than GPS (also you can get much easily the signal indoors) and you could sync every day.

Yep, we had a bit of a look into this too, but GPS just seems like the simplest option in this case (and since the frequency of update is pretty low, it wont have a big impact on battery life).  This said, I'm sure either method can be made to work.

[LaurenceW]

Depending on your real need, you might be pleasantly surprised by the Maxim "Extremely Accurate Clock" series. I've used them for 3-4 years, and my own experience is that they are well within spec over a period of many months. I have not exercised them over particulalry wide temp ranges, though.

[veryevil]

I use this method http://ww1.microchip.com/downloads/en/AppNotes/01155a.pdf

[Jebnor]

What about syncing with a GSM signal from cellular telephone systems?  It is my understanding that they are very accurate to allow the tower to hand the phone from tower to tower.