Hi all. So in looking at stans tech brief "Birth of a New Technology" I came across several sections that talked about the voltages that he was getting ranging anywhere between 20,000 volts and 90,000 volts depending on if your talking bout the fuel cell or the resonant cavity.

Now obviously theres not a hand held multimeter around that can measure that kind of voltage that im aware of, and an oscilliscope needs to have a reference to ground in order to get an accurate measurement if i understand correctly. Thats why a differential probe is needed to measure the difference between one pin or another.

If you look at differential probes they range in the thousands of dollars for voltage ranges of this caliber so for alot of us that would be cost prohibitive. So what im wondering is if it is possible to come up with a schematic or a low cost method of measuring across the cells in order to get a accurate reading.

It would seem to me if you have ten equally built cells and you measure 500 volts across one of them, you should be pretty close to 5000 volts total.  If that's not good enough, measure each one and just add the ten readings up.

I've always wondered how Mr. Tesla pronounced some of the voltages he worked with so matter-of-factly when he had rather primitive measuring devices.  I would have to guess he used spark gap width as his primary tool.

To me, at the end of the day, I would like to know how much voltage I really need to do the job and engineer into the device about a 10 percent margin.  Unless someone can prove that 5000 volts is good, but 10,000 volts is much better.  I haven't seen that proof yet.

Quote from cory991 on March 1st, 2014, 11:55 PM

Hi all. So in looking at stans tech brief "Birth of a New Technology" I came across several sections that talked about the voltages that he was getting ranging anywhere between 20,000 volts and 90,000 volts depending on if your talking bout the fuel cell or the resonant cavity.

Now obviously theres not a hand held multimeter around that can measure that kind of voltage that im aware of, and an oscilliscope needs to have a reference to ground in order to get an accurate measurement if i understand correctly. Thats why a differential probe is needed to measure the difference between one pin or another.

If you look at differential probes they range in the thousands of dollars for voltage ranges of this caliber so for alot of us that would be cost prohibitive. So what im wondering is if it is possible to come up with a schematic or a low cost method of measuring across the cells in order to get a accurate reading.

the reason why to have to use a differential probe differs a bit from your description.
normally both channel grounds and the trigger in port of a 2 channel scope are connected internally to same ground wire.

if you take measurements you have to know that otherwise you might create a shortcuts in your circuit using ch1 GND at one measurement point and ch2 GND at another measurement point.

and there is another reason:

in a Stan Meyer type application the water bath may have no ground wire reference. otherwise the specific splitting process won´t take place. that might be a hint to the point that there is some "cold electricity" effect taking place as referenced in tesla applications where ground wire reference always plays an important role.
 
there are high voltage probes < $ 100 which are not differential. they would fit for a 1 channel scope usage if the scope AC connection is operated thru an isolation transformer.
then the 2nd channel could only be used with the same GND reference as ch1. a useful application could be to measure amp flow thru a measurement resistor with same GND reference like for the hv measurement.

in this specific case you can use 1 non-differential high voltage probe for << $100 and a standard probe for the amps. but then you must be well aware that your scope is operated at a hv potential and avoid to touch any metal parts.

but if you want to be flexible you must use at least 1 differential probe because the preamplifier of the differential probe has an extremely high resistance towards scope internal GND reference.

for more details of a self made differential probe
-> http://www.eevblog.com/forum/projects/oshw-diy-1kv-100mhz-differential-probe-(dilemma-vs-hope)/

but

handling voltages above 1000 V range is a real expert issue! there are aspects of ionization, special passive components etc.

you can easily zap yourself into heaven doing it the wrong way.

so to make a "cheap" calculation

think about taking a 10 days part of your earned income and buy a hv differential probe and have fun instead of building a probe for 3 weeks, blowing it up at first application and zapping yourelf into heaven :-/
that exclusively pays off with an adequate life insurance ...

Quote from Matt Watts on March 2nd, 2014, 12:52 AM

It would seem to me if you have ten equally built cells and you measure 500 volts across one of them, you should be pretty close to 5000 volts total.  If that's not good enough, measure each one and just add the ten readings up.

I've always wondered how Mr. Tesla pronounced some of the voltages he worked with so matter-of-factly when he had rather primitive measuring devices.  I would have to guess he used spark gap width as his primary tool.

To me, at the end of the day, I would like to know how much voltage I really need to do the job and engineer into the device about a 10 percent margin.  Unless someone can prove that 5000 volts is good, but 10,000 volts is much better.  I haven't seen that proof yet.

amazing how Tesla made use of his primitive equipment ...

at the end of the day we all will know but now it´s just noon.

@Matt Watts You would be correct on measuring across a single cell, i kinda forgot about that detail but it only would work below 1000 volts if i remember right. I too would like to have a sure number on what voltage it required to make it work. Im just going off the numbers called out in the tech brief.

@Bussi04 Thanks for the clarification and link. Ive seen where one can use the math function and invert function to accomplish the same task i was wondering if that is indeed possible and how accurate you think it is.