Yes, “nanofarads” is the upstart, Johnny-come-lately, unit of capacitance. Even after a couple decades of common usage, I have to think for a moment when I encounter it, and mentally translate to either uF or pF. And in my superannuated, chronologically gifted, state . . . . “thinking” makes my brain hurt. Fortunately, I believe values with the “nanofarad” designator have always been whole numbers greater than “one”. I can’t recall an occasion when I saw a value expressed as, e.g., “0.022 nF”. (Think of the little capacitors around the crystal in your microcontroller’s oscillator section.)
I guess you young-uns never had an opportunity to make the acquaintance of “micromicrofarad” (“mmf” for short). Now, there’s a unit that could eat up more than his allocated space on a schematic or BOM listing! I guess he needed the long name to compensate for his small size.
FYI . . . somewhat off-topic . . . . There are several common ways to reduce the consequences of reverse polarity and/or wrong voltage applied to a PCB assembly. Your series diode may be the most obvious and venerable. It’s not popular among those who want to squeeze the last femtocoulomb of charge from a battery, or who must handle more than a few amps.
Another method is to place a diode, zener, MOV, or varistor anti-parallel across the supply rails. A fuse (either standard metallic fuse, or self-resetting PTC “polyfuse”) may be inserted in series.
And sometimes you will find a P-channel MOSFET connected in series (and “backwards”) with the supply.
Few years ago I was reading datasheets of TPS54060 serie. I found the unit I have never ever seen - µMhos.
I read it as micro-mega-hoses. As mikro-mega reduces to 1 so hoses left. It took me some time to find that they use the unit non existing since about 130 years when was replaced by S. I understand that Mho is 1/Ohm but where from is the s at the end - may be it is to say that there are more than one of it there.
As I found many counting errors in 9 datasheets I have send them the list of corrections. The man who was mailing with me said: “Ok. You are right. I am sending all these datasheets to be updated.” After some time I found new versions. No bug was corrected. They only added one paragraph saying that from that place are the examples they are not responsible for being correct.
But µMhos they replaced with µS
I did a quick look online to see if something like was used for reverse polarity protection, but I did not find anything. Interesting, I had used this configuration for OverVoltage protection in a project that I am currently working on:
I use that method but with diode after it, and because I have that diode then I don’t use Zener but bidirectional transil (fuse not stressed when wrong polarity power connected).
People working in EMC lab told me that many DC powered devices have a problem with negative surge at power supply and not have a problem with positive surge. I thought a moment and decided to use the diode to protect my capacitors against being discharged during negative surge.
I have never done any experiments about it, but I suppose that P-MOS method is perfectly good to protect against wrong polarity battery connected but if you consider another DC supply where surges can be expected that P-MOS will allow to discharge the capacitors may be not to 0V but too much I think.
This is important. Waldo32, can you tell us more about the whole system? I don’t find documentation for the module, only for the chip ESP8266. That doesn’t take 9V. My best guess is that the module takes ~5V from either the USB connector or from the VIN pin. In that case you should use a regulator with 5V output. Maybe L7805 if it exists. You also haven’t told what kind of power source you can have. I have assumed that a barrel jack means any generic DC power source between, say, 9 and 24 V.
Thanks all - happy to give more detail. Also, after this conversation I think I was definitely wrong with the 7809. I think I want a 7805.
Here’s the background.
Per image below, I have a NodeMCU connected to a transmitter. The ModeMCU can be powered by microUSB (I thought it was 6-12V but someone recently told me 5 was ok). When it is (as in the picture), then power and ground on the transmitter go through the nodeMCU.
I want optionality to power both the nodeMCU and transmitter from a barrel jack (whatever is sufficient - I think 9V; 5V might be but seems like going a bit bigger is better). So if I power from a 9V barrel, then I need to step it down to 5V. One additional item I noticed was that the transmitter doesn’t have wires also going to the power from the regulator. I think it should; I’m not planning to ever power the whole system with micro usb and barrel. So having wires to the transmitter from both sources should be fine.
Lastly - the reason I didn’t connect every ground pin from the NodeMCU to the ground plane was because I was only using one of the GND pins (there are multiple) - but I guess it doesn’t matter and I should connect them all.
Does that help? Do folks agree with my analysis / have other thoughts on the design now?
We still don’t know much about the module or the transmitter. You should give us direct links to the datasheets. For example, it’s possible that the module already has a regulator which can handle the voltage coming from your power source. Without datasheets this all is guesswork, at best leading to design which has unnecessary parts, at worst leading to destroyed parts and starting again with new design.
The esp8266 is a 3.3v device. Your nodemcu carrier board has a 3.3v LDO regulator. The 1117 is a bog standard LDO, used on arduinos and the like. Its rated for up to 13v input, so you should be able to power the board straight from a 9v wall wart without an extra regulator. Just make sure the 1117 doesn’t overheat.
Check the polarity of your barrel jack. Walwarts are usually center pin positive.
I see from the picture, that the transmitter has a coil antenna, and your pcb has copper on the area that the antenna will lay across if mounted flat. This will destroy the radiation characteristics of the antenna. Even mounting the antenna close to a free-of-copper area of the pcb ( top and bottom) will affect the antenna tuning.
Thank you Andrew_Law! So I guess for my ground pour I should keep a non poured area under where the transmitter will lay? Except I have traces going to the pins under that area. Are traces themselves a problem or just the pour? (not sure what you mean by a “free-of-copper” area). Will also check the barrel jack.
Definitely a non-poured area where the transmitter aerial is located. To operate most efficiently any unipole aerial should be mounted normal to the ground plane. The traces to the pins are probably not going to have too much effect, as they are a tiny fraction of the transmitted wavelength, but they will have some effect, as will the pcb fibreglass. Free of copper is bare fibreglass.
Great example!
