NCO方波应该是什么样的？

你没有告诉我们哪张照片。如果连接NCO的引脚具有开路漏极能力，则可能需要上拉或切换到推挽。否则，它只是CMOS输出，所以高侧基本上是VDD假设没有太多的负载。可能有一些噪音在您的电源轨或碰撞（短）与其他一些信号。



      以下为原文

    You didn't tell us which PIC. If the pin the NCO is connected to has open-drain capability, perhaps you need a pull-up, or switch to push-pull. Otherwise it's just a cmos output, so the high side is essentially VDD assuming not much load.
 
There's probably either some noise on your power rail or a collision (short) with some other signal.

看起来像一个非常嘈杂的VDD轨道。你有旁路电容在每对电源引脚？（正如马克提到的，你没有告诉我们哪张照片，所以我们不知道它有多少个电源引脚。）



      以下为原文

    Looks like a very noisy Vdd rail.
Do you have bypass capacitors on every pair of power pins?
(As Mark mentioned, you didn't tell us which PIC, so we don't know how many power pins it has.)
 

啊，糟了，我用的是PIC16F18345。谢谢大家，嘈杂的栏杆听起来很有可能，只是用一个无焊的电路板把这些放在一起，我相信他们有缺点。对于电路的电源PIC，我只有一个帽子在PIC电源引脚（1 VDD和20 VSS）。虽然我忘记了我曾经用过的价值。（也许10uF？）不确定）你们有没有直接推荐PIC配置/值的建议？我没有在5V（NCP1117）电源的帽子，电源PIC，所以可能是另一个噪声源。



      以下为原文

    Ah my bad I'm using the PIC16F18345.  Thanks guys, noisy rails sounds likely, just working with a solderless breadboard to put this together at the moment and I'm sure they have drawbacks. 
 
For the side of the circuit powering the PIC I just have one cap across the PIC power pins (1 Vdd and 20 Vss).  Although I forget what value I even used.  (maybe 10uF?? not sure)  Do you guys have a recommendation for the cap configuration/value(s) for directly bypassing the PIC? 
 
I don't have caps across the 5V (NCP1117) power supply that powers the PIC so that could be another source of noise.    

10UF电容器可能是电解的，在这种情况下它具有相当大的电感，因此不能作为一个“旁路”电容器。你需要一个较低的值，最好是陶瓷或单片，大约100NF。（尽管你现在可以达到1或2个UF，但是100NF仍然是最常见的值。保持电容器的腿尽可能短，以便通过电容器从PIC的VSS到VDD管脚的总路径尽可能短。NCP1117数据表表示在其输出上必须至少有4U7电容，并且应该是低ESR（小于33毫欧姆）。具有100NF上限的Ralele将实现这一点。他们还建议在调节器的输入上加上10uF的上限。



      以下为原文

    A 10uF capacitor is probably an electrolytic one, in which case it has considerable inductance, so cannot act as a "bypass" capacitor.
You need a lower value, preferably ceramic or monolithic, around 100nF.
(although you can get up to 1 or 2 uF in that style now, 100nF is still the most common value.)
Keep the capacitor's legs as short as possible, so that the total path from the PIC's VSS to VDD pins through the capacitor is as short as possible.
The NCP1117 datasheet says you must have at least 4u7 capacitance on its output, and it should be low ESR (less than 33 milliohms).
Your 10uF cap in parallel with a 100nF cap would achieve this.
They also recommend a 10uF cap on the input to the regulator.
 
 
 

1或2 UF哈！我的照片上有47个UF 0603陶瓷（还有0201个0.1uF）。可能过度，但有助于ADC的性能。



      以下为原文

    1 or 2 uF Ha!  I've got 47uF 0603 ceramic on my PICs (along with a couple 0201 0.1uF). Probably overkill but helps with ADC performance.

来自NCP1117的数据表：“调节器的频率补偿是由电容器CUT提供的，它的使用是强制性的，以确保输出稳定性。4.7μF的最小电容值与等效串联电阻（ESR）在33 m（TYP）至2.2的范围内是必需的。



      以下为原文

   
From the NCP1117 datasheet:
"Frequency compensation for the regulator is provided by capacitor Cout and its use is mandatory to ensure output stability. A minimum capacitance value of 4.7 F with an equivalent series resistance (ESR) that is within the limits of 33 m (typ) to 2.2 is required."

我认为用4.7UF的帽子意味着更多的是33毫欧。请参阅数据表中的图12。



      以下为原文

   
I think with a 4.7uF cap they mean MORE that 33 milliohms. See fig 12 in the datasheet.

当电源进来时，你需要一个大的ELKO，然后从ELKO后面的所有点获取所有的源，特别是0V，你需要在这一点上达到0V的范围。如果你有一个调节器，确保使用330N和100N非常接近PIN +一些ELKO。在芯片上，你需要100N。非常接近。



      以下为原文

    Where the power comes in you need a big elko, then take all sources from that points behind the elko,
especially the 0v, you need to attatch the scope 0v at this point also.
 
If you have a regulator make sure to use the 330n & 100n very close to the pins + some elko.
At the chip you need 100n again very close.

您可能会惊讶于这些小的高值MLC的电容随施加的电压而变化。典型地，它们可能在接近额定电压时失去其额定电容的一半，有些甚至会下降到额定值的10%。HTTP//www. VisHay.COM/DOCS/40144/CAPPANE。PDFLY是我在S.C.Engig.Mead新闻组的帖子的副本，在这里我展示了巨大负电压系数的例子。T：HTTP://CARS FUMUSEX.COM/A/M/S/P1211197-0258469-HUGE负电压系数，用于高容量MLC电容器。



      以下为原文

   
You might be surprised how much the capacitance of those small high value MLCs changes with applied voltage. Typically they may lose half their rated capacitance at close to rated voltage, and some will even drop to 10% of rated value.
 
http://www.vishay.com/docs/40144/capchange.pdf
 
Here is a copy of my post on the sci.electronics.design newsgroup where I show examples of the huge negative voltage coefficient:
 
http://cars.forumsee.com/a/m/s/p12-11197-02584469--huge-negative-voltage-coefficient-for-high-value-mlc-capacitors.html
 

你绝对正确-电压降额必须占任何陶瓷不是C0G/NP0。村田的SimulfE-SKEAM或来自KEMET或TDK的S- PARAM工具非常有用。在3.3V处，一个47μF的63V X5R 0603是12uF，但是它具有与任何其他通用0603 MLC相同的封装电感，因此比其他0603的SRF低得多的阻抗，并且在其上具有相同的阻抗。



      以下为原文

    You are absolutely right - voltage derating must be accounted for on any ceramic that's not C0G/NP0. Murata's SimSurfing or S-param tools from Kemet or TDK and the like are very useful for that. A 47uF 6.3V X5R 0603 is 12uF at 3.3V - but it has the same package inductance as any other general purpose 0603 MLC and thus much lower impedance than other 0603's below their SRF, and the same impedance above it.
   Attached Image(s)

超级电容建议大家感谢所有的输入！我唯一不懂的是埃尔科？那代表什么？我有很多不同的陶瓷和钽低ESR帽在我的收藏，所以我应该能够解决这些问题在我的PIC和电压调节器。



      以下为原文

    Fantastic capacitor advise guys thanks for all the input!  The only thing I don't follow is elko??   What does that stand for?
 
I do have many various ceramic and tantalum low ESR caps in my collection so I should be able to fix these issues on my PIC and voltage regulator.  

一些工程师使用“ELKO”作为氧化铝电容器的速写。我认为它来自德国“EelkTurLythChier-Knnistor”。



      以下为原文

    Some engineers use "elko" as shorthand for Aluminium Oxide capacitors.
I think it comes from the German "Elektrolytischer Kondensator".
 

啊，感谢QHB，也许我花了一段时间才明白这一点！对所有帮助过我的人。只要把100NF和10UF上限平行于调节器输出和一个10uF上限的投入。全多层陶瓷。我的NCO方波现在看起来很锋利！



      以下为原文

    Ah thanks qhb, might have taken me a while to figure that out myself!
 
And to everyone else that has helped me.  Just put 100nF and 10uF caps paralleled on the regulator output and a 10uF cap on the input.  All multilayer ceramic.   My NCO square wave is now looking sharp! 
   Attached Image(s)

很不错的。我也从未听过“埃尔科”这个词，我设计电子产品已经有30年了。然而，我对铝电解没有多大用处。在高电压下不能摆脱它们的高电容，但是对于数字和混合信号应用来说，ESR太多，在温度下寿命差，而且通常太大。



      以下为原文

    Very nice. I also had never heard the term "elko" and I've been designing electronics for 30 years. I don't have much use for al-electrolytics, however. Can't get away from them for high-capacitance at high-voltage, but for digital and mixed signal applications way too much ESR and poor lifetimes at temperature, and generally too big.

关于偏置下的电容变化的说明：尽管X5R介质的电压降得非常大，但X5R电容器在相同的情况下，通过商业温度范围的电容，优于X7R电容器。电容漂移较大，但最差电容仍高于X7R。



      以下为原文

    A note about capacitance change under the biasing: despite very large drop under voltage for X5R dielectric, X5R capacitor outperforms the X7R one by capacitance across commercial temperature range in the same case size. The capacitance drift is huge but worst capacitance is still higher than X7R.

哦，对不起：ELKO或ELCO在荷兰ELKROTROCIC冷凝器中。美国人也应该使用这个词，这可以节省大量的打字。



      以下为原文

    Oh sorry : elko or elco is in dutch elektrolithic condensator.
Americans should adobt that word also, saves me alot of typing.
Glad it works.
 

我采用X7R、X5R、C0G、NP0这样的单词，并节省了更多的输入：



      以下为原文

    I adopt by using words like X7R, X5R, C0G, NP0 and save even more typing :)

那么埃尔科是什么词呢？



      以下为原文

    What word is a elko then ?

我很难过，表情/表情符号/象形文字/埃尔科可能很快取代我们这个行业的大多数人已经理解的名字。



      以下为原文

    I am sad to think that emoji / emoticons / hieroglyphics / elko may soon replace names that most of us in this industry already understand.