[TriEmbed] POTM

[Pete Soper]

Hi Paul,

These resistors are designed to survive overloads for short periods. The 
only thing that's going to overload a 10M resistor is a direct lighting 
strike. With that scenario, it's very difficult to predict what will 
happen. For instance if the fence wire makes a sharp turn on its way to 
the 10M resistor the lightning might treat it as an open anyway by 
virtue of the inductance.

But with amateur radio 160/80/40m receive antennas, up to thousands of 
feet of wire in a "Beverage" antenna (named after a guy named Beverage) 
is typically terminated with a resistor of just a few hundred ohms 
connected to ground. Using the most sturdy resistors possible is a hedge 
against replacing them frequently vs occasionally, as the induced 
potentials of close lightning strikes regularly drives them to high 
current levels. But these are very much lower value resistors than with 
your application. But for this amateurs go a step further, using carbon 
composition resistors vs metal film, to avoid the tuning side effects of 
the inductance of metal film parts. You don't care at all about that, so 
I think the ones you found will be excellent. The very high resistance 
values at the beginning of the string should make the power dissipation 
irrelevant (for receive antennas we'd go with the biggest available and 
that's two watts for a single part that can be put in series/parallel 
combinations for higher power situations).

So a string of 10M resistors followed by the lower and lower values, 
with the lowest closest to ground should be safe for everything but a 
close strike. The shunt resistors used to measure current from laser 
cutter power supplies use this technique, just putting the resistor 
between ground and the tube's ground return path so the voltage drop 
across the resistor is limited. And the datasheet of the resistor you 
found specifically says it can handle a 10KV pulse, so I think you're 
golden.

However, in my opinion if you want the MPU to survive out there you 
should take steps to avoid possible current paths in and out of it. For 
example, a long dangly piece of wire such as for an antenna would be 
terrible. A long cable for the "data" would be a recipe for destruction. 
Just imagine everything there is like the winding of a transformer, with 
a lightning surge providing the current through the other winding. 
Everything  but the  last resistor of the string inside a metal box tied 
to a ground rod with just a little plastic window for a WIFI antenna 
might withstand a close strike. With an external antenna you MUST have 
an antenna with an unbalanced feed where the coax can go through a 
lightning surge suppressor specifically designed to survive it. But of 
course if the system being blown up doesn't cost much, you could just 
rebuild it and ignore the whole subject of survival with a suppressor  
(here's a page of them, but designed for very much higher transmit power 
than you need: 
https://www.dxengineering.com/search/part-type/coaxial-lightning-protectors 
There are no doubt smaller/cheaper ones designed for the type of radio 
connection you would be using.)

-Pete

On 3/21/20 11:47 AM, The MacDougals via TriEmbed wrote:
>
> This month’s problem of the month was about an electric fence monitor.
>
> http://everycircuit.com/circuit/5798493590650880/electric-fence-voltage-monitor
>
> During the meeting, I was pointed to Panasonic high surge resistors at 
> Digikey.
>
> After a bit of searching, this is what I think would work.
>
> https://www.digikey.com/product-detail/en/HVR2500001005JA100/BC4429CT-ND/7351859/?itemSeq=321321346
>
> 10000V
>
> R1 – 10Mohm
>
> 7514.9V
>
> R2 – 10Mohm
>
> 5029.8V
>
> R3 – 10Mohm
>
> 2544.7V
>
> R4 – 10Mohm
>
> 59.6V
>
> R5 – 220Kohm
>
> 4.97V ßArduino monitors here
>
> R6 – 20Kohm
>
> GND
>
> Any comments about these resistors?
>
> I was thinking of soldering them in series and then potting them with 
> this:
>
> https://smile.amazon.com/Clear-Electronic-Grade-Silicone-Squeeze/dp/B0063U2RPW/ref=sr_1_19?crid=2RXV4MC2DX1Y2&dchild=1&keywords=potting%2Bcompound%2Bfor%2Belectronics&qid=1584805104&sprefix=potting%2Bcom%2Caps%2C165&sr=8-19&th=1
>
> ---> Paul
>
>
> _______________________________________________
> Triangle, NC Embedded Computing mailing list
>
> To post message: TriEmbed at triembed.org
> List info: http://mail.triembed.org/mailman/listinfo/triembed_triembed.org
> TriEmbed web site: http://TriEmbed.org
> To unsubscribe, click link and send a blank message: mailto:unsubscribe-TriEmbed at bitser.net?subject=unsubscribe
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.triembed.org/pipermail/triembed_triembed.org/attachments/20200322/76d08507/attachment.htm>