内部振荡器的准确性怎么样

数据表的第一页：这将给出+/-1.8US，你的1981年测量有多精确？



      以下为原文

    First page of the datasheet:

This would give +/- 1.8us, how accurate is your 198us measurement?

我找不到准确的信息为我的TDS2014，但我刚才有它校准，它是优于几个纳米索检查我的信号发生器。我没有看到你引用的0.9%，所以谢谢你。我有点惊讶，两个设备是相同的，但我想我们可以假设这只是一个问题的振荡器，而不是我的配置。



      以下为原文

    I cannot find accuracy information for my TDS2014 but I just had it calibrated and it is better than a few nanosecs checking against my signal generator. I did not see the 0.9% that you quoted so thanks for that. I was a little surprised that two devices were off identically but I guess we can assume it is just an issue with the oscillator and not my configuration.
 
thanks for the reply
fritz

它是RC振荡器吗？你没有提到电源电压，在同一个夹具中尝试一个新芯片不会改变芯片的电源电压和温度。



      以下为原文

    Is it an RC oscillator?
You didn’t mention supply voltage, and trying a new chip in the same jig doesn’t change that.
The supply voltage of the chip, and it’s temperature.
 

谁来阅读数据表？你一定是在开玩笑吧，兄弟！



      以下为原文

    Who - me reading datasheets? You must be kidding, bro!

谢谢布莱克。几个星期后，我仔细观察了这一点，当我得到一份礼物时，我会用水晶做的。



      以下为原文

    Thanks Brek. I looked into this more a few weeks back and will set up with a crystal when I get a chance.
 
fritz

你真的希望从这样的设备中获得精度不到2nSEC吗？或者甚至需要得到准确的？苏珊



      以下为原文

    Do you seriously expect to get accuracy to less than 2nSec from a device such as this?
Or even need to get it that accurate?
Susan

2纳秒？当然不是。但是在2个周期的周期内，有超过一半的用户使用了。我正在处理一个电化学过程，它有一个足够大的松弛时间来过滤抖动。我不在乎系统是否会像化学过程那样随着温度和压力在几秒钟内漂移——但是在短短的几十个使用周期内，我想要合理的抖动，或者我们进入化学中的各种不利问题。实际上，我甚至不需要水晶，因为缓慢漂移并不重要，但我会在业余时间玩一玩，只是为了看看这能得到多好（我在实验室里还有很多其他物理问题要解决，有些问题需要更精确的处理）。正如各种响应者指出的，绝对时间精确到2uSecs，但我希望抖动比这更好，根据我使用PIC24s的经验，这是合理的。我原来的帖子的主要驱动力是，我看到一致的2usec偏差与一些不同的IC样本，所以我认为可能是我编码在PWM/OCx规则错误。但不是。谢谢你的评论。



      以下为原文

    2 nSec?   Certainly not. But 2 uSecs over a period of a few cycles--yes. I am dealing with an electrochemical process that has a relaxation time large enough to filter the jitter so to speak. I don't care if the system drifts with temperature and pressure over seconds as the chemical processes will as well-- but over a short period of tens of usecs I would like reasonable jitter or we get into various adverse problems in the chemistry. In actuality I don't even need the crystal as slow drift is unimportant, but I will play with one in my spare time just to see  "how good" this can get (I have a lot of other physics problems to solve in the lab here and some will require more accuracy).
 
I certainly don't expect the internal RC to be accurate to 2uSecs in absolute time as the various responders point out but I expect the jitter to be better than that and in my experience with PIC24s that is reasonable.  The main driving force for my original post was that I was seeing a consistent 2usec deviation with a few different IC samples so I was thinking perhaps I was coding something wrong in the PWM/OCx regs. But not.
 
Thanks for the comment
Fritz