A quartz crystal oscillator has power consumption in the uA range for kHz outputs.
This MEMS product is as stated above "a power HOG!"
Current Consumption Idd – 5.9mA No load condition, f = 400 KHz,
Piyush mentions they have another option that cuts the power by 1/2; that still puts it at 3 mA which is orders of magnitude above uA power consumption and unusable for most applications!
Notice how the power consumption is rated for: No load condition.
So it is a great power HOG that you can hook up to nothing!
Notice how they carefully worded the fact that they compared in to “offering from another player”
The other player must have been manufacturing vacuum tube oscillators!
Please send me email at firstname.lastname@example.org to discuss further.
I may have the wrong part link - the part that you listed is not programmable (a key value for the 8503) and also has worse specs (+/- 100 PPM stability over comm. temp. and 0.7 MHz min freq.).
I agree with you that low power is a must in less than 1 MHz applications.
For an example of better Crystal-Osc-Icc's look at
- and also the better Microcontrollers, that have configurable power oscillators, also cover the 200KHz-1MHz range.
If someone is going to choose a sub 1MHz ref, they are going to expect low Icc.
I'm Piyush Sevalia, VP of Marketing at SiTime.
We compared ourselves to the other programmable oscillators in this 200 kHz – 1000 kHz product segment and there are not many providers in this segment.
a) Comparing the SiT8503 to an industry-standard 7050 package offering from another player, our active power is 75% lower.
b) Comparing the SiT8503 to a non-standard footprint device from another player, our active power is about the same, but our standby power is 50% lower.
We have a lower power option available, which cuts the power in almost half. In the variety of applications that we’ve seen, power is not an issue at these frequencies.
Obviously, this 8503 device should not be compared to a 32 kHz RTC oscillator, which is for a completely different market segment.
The comment about non-MEMS silicon oscillators is definitely interesting. Many of our customers require total frequency stability of less than +/- 50 PPM (and in some cases, as low as +/- 10 PPM), which is not offered by non-MEMS Silicon oscillators and is probably difficult to achieve. Our customers say that this requirement is for improved timing margins and system reliability. We'd love to hear some feedback on whether this total frequency stability spec is important or not.