[Open_electroporator] kvboard volt divider (max voltage the ADC can read)

[Nathan McCorkle]

On Thu, Dec 28, 2017 at 1:11 PM, John Griessen <john at cibolo.com> wrote:

> On 12/28/2017 01:36 PM, Nathan McCorkle wrote:
>
>> Can we change that to around 1/2600 to get better range? We're easily
>> blowing >2400V way out of range, no chance we're going to do any failure
>> analysis with that, and I don't think there's a need for sub-volt
>> resolution.
>>
>>
>>     Besides running at lower volts, the only way to get more bit
>> resolution for current measuring
>>
>> Sorry, wasn't concerned with current measurement at all in my last
>> comments, I think your idea of keeping it within power availability
>> range/envelope makes sense.
>>
>
> We could go up some...The top volt level of much interest as is now is
> 3000V since the capacitors are rated that.
> Going to higher capacitors has some real design and purchasing time
> required -- not super easy.  Diodes would need a little work also.   I had
> hoped to spend some time on those parts going for all surface mount
> assembly and lower costs, small bulk for PCBA shipping, and high
> reliability.
>

Not asking or suggesting we change the voltage rating at all, let's be
clear on that.

I was merely thinking that since I experimented with pushing it's limits,
and that's when it broke, we might want that range to detect if damaging
conditions have occurred. This assumes we have no interest in single-volt
resolution.

However, if we consider probably the smallest practical microfluidic for
electroporating E.coli:
18000 V/cm * 1 microns
that equals 1.8V needed to pulse across the cell... probably much lower
than we can even produce, I'm guessing.
http://www.wolframalpha.com/input/?i=18000+v%2Fcm+*+1+microns

2400/4096 levels == 0.585 High-Volts per level

the other common field strength for E.coli is 15 kV/cm, substituting that
in the previous math yields 1.5V. So that's a diff of 0.3V... meaning even
if we could produce single-digit volts on the HV side, we'd be just unable
to discriminate if it was 15kV/cm or 18kV/cm (because they're only
different by 0.3V... but the ADC range is only capturing ~0.585).

So if we flip the equation, and ask instead *how small of an
electroporation cuvette or interelectrode distance could we have? To easily
discriminate 18kV/cm vs 18.1kV/cm* (assuming no noise, or averaging
compensates):
https://www.wolframalpha.com/input/?i=solve+((18100%2F10%
2F1000)*x)+-+((18000%2F10%2F1000)*x)+%3D+0.585

we'd need a *58.5 micron* distance. Easily within range of DIY and @home
operations, as well as amenable to production if that ever came to be.

On the other hand if we constrain the *electrode distance to the 0.1 cm (1
mm, 1000 micron) cuvette*, we see the voltage will be 1800V vs 1810V, a
difference of 10V... 10/0.585 == 17 meaning *we'd have ADC reads spanning
17 values* instead of 1 for the microfluidic.
https://www.wolframalpha.com/input/?i=solve(+((18100%2F10%2F1000)*1000)+-+((18000%2F10%2F1000)*1000))%2F+0.585


Sounds pretty good to me (with the 800 reduction factor, since 2400/800 ==
3, and the ADC input max is right around 3.1V targetted with battery).

Glad to have that basic math done when I need it.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://cibolo.us/pipermail/open_electroporator/attachments/20171228/772dc9f7/attachment.html>