Go Down

### Topic: High Voltage Monitoring (Read 4309 times)previous topic - next topic

#### mclaughlinmax

##### Feb 15, 2013, 11:35 pm
So I need to regulate a capacitor's voltage and I was planning on doing it via an Arduino. The capacitor bank I'd be monitoring would max out at 500v, but I would like to be able to monitor it as it charges and cut the voltage off with the Arduino. I have no problem programming it, but how would I go about retrieving the voltage? Any help would be great.

Thanks

#### Krodal

#1
##### Feb 16, 2013, 12:05 amLast Edit: Feb 16, 2013, 12:11 am by Krodal Reason: 1
You could use a voltage divider, if it is okay to connect the grounds.
For example two resistors of 22M in series to reduce the voltage and a third resistor in series of 68k for the voltage to the Arduino.
Perhaps adding 1nF parallel over the 68k to reduce noise and perhaps some protection diodes.

The resulting voltage will be : 68k / (22M + 22M + 68k) * 500 = 0.77V
If you set the voltage reference of the Arduino to 1.1V, you can measure the voltage.

I have done something like that with my Geiger counter, and it is doing well.
I measured the voltage with my multimeter and adjusted the calculation in the Arduino to that.
This is my post with the schematic for my Geiger counter: http://arduino.cc/forum/index.php/topic,120390.msg907435.html#msg907435

#### mclaughlinmax

#2
##### Feb 16, 2013, 04:48 am
Wouldn't I still be stuck lower resolution results though? It would still be between 0 & 1024 right?

#### Krodal

#3
##### Feb 16, 2013, 08:49 am
Yes, the 10-bits (0...1023) have by default a range of 0 to 5V.
By selecting the internal reference, the 10-bits have a range of 0 to 1.1 V.

That is roughly 1mV resolution.
With the resistors I mentioned, the resolution to measure the 500V is 0.7V.
So you can measure the 500 with 0.7V precision.
However, due to the high impedance, and resistor inaccuracy, it is not that accurate.

http://arduino.cc/en/Reference/AnalogReference

#### Docedison

#4
##### Feb 16, 2013, 09:22 am
22 M ohms x2 isn't hard to do with precision resistors. The last recommended 68K resistor should be a 100k pot for final calibration. Be EXTREMELY CAREFUL, WHAT YOU ARE WORKING WITH CAN KILL YOU QUICKLY
It shouldn't be difficult to get an accurately calibrated voltage divider.

Bob
--> WA7EMS <--
"The solution of every problem is another problem." -Johann Wolfgang von Goethe
I do answer technical questions PM'd to me with whatever is in my clipboard

#### Krodal

#5
##### Feb 16, 2013, 10:44 am
A 100k pot could have less good contacts in a number of years.
That would cause the 500V to go to the Arduino.
In such situations I always use fixed resistors, and the correction can be done in the Arduino with a calculation.

It is very dangerous indeed. 500V is dangerous, and 500V in a capacitor is extremely dangerous because there is no current limit.

#### dc42

#6
##### Feb 16, 2013, 10:46 am
Resistors have a voltage rating as well as a power rating, so make sure your resistors are adequately rated. You will probably have to use several in series to get 500V rating. if you are powering the Arduino from an accurate 5V supply then you don't need to use a pot for adjustment, just use the standard 5V analog reference and use 1% resistors. Connect a 10nF capacitor across the lower resistor to alleviate the inaccuracy that will otherwise result if the lower resistor is greater than 10K.
Formal verification of safety-critical software, software development, and electronic design and prototyping. See http://www.eschertech.com. Please do not ask for unpaid help via PM, use the forum.

#### afremont

#7
##### Feb 16, 2013, 07:33 pm

Resistors have a voltage rating as well as a power rating, so make sure your resistors are adequately rated. You will probably have to use several in series to get 500V rating. if you are powering the Arduino from an accurate 5V supply then you don't need to use a pot for adjustment, just use the standard 5V analog reference and use 1% resistors. Connect a 10nF capacitor across the lower resistor to alleviate the inaccuracy that will otherwise result if the lower resistor is greater than 10K.

Everything he said plus another reminder that you will have enough energy stored in the caps to kill yourself.
Experience, it's what you get when you were expecting something else.

#### zoomkat

#8
##### Feb 16, 2013, 09:20 pm

So I need to regulate a capacitor's voltage and I was planning on doing it via an Arduino. The capacitor bank I'd be monitoring would max out at 500v, but I would like to be able to monitor it as it charges and cut the voltage off with the Arduino. I have no problem programming it, but how would I go about retrieving the voltage? Any help would be great. Thanks

You might consider the charging circuit in disposable flash cameras. The capacitor in them is charged to ~400v thru some type of automatic circuit. I can verify that a charged cap in one of these cameras will get your attention if you are in the discharge flow path.
Why I like my 2005 Rio Yellow Honda S2000  https://www.youtube.com/watch?v=pWjMvrkUqX0

#### Learning

#9
##### Feb 16, 2013, 11:07 pm
Make sure you don't sidestep any 'bleeder resistor' network that may be part of this kluge, you may be working with lethal voltages/currents. Be careful!

#### dc42

#10
##### Feb 17, 2013, 07:39 pm
Yes, a 500V capacitor bank is potentally lethal. The voltage divider will double up as a bleeder resistor.
Formal verification of safety-critical software, software development, and electronic design and prototyping. See http://www.eschertech.com. Please do not ask for unpaid help via PM, use the forum.

#### Docedison

#11
##### Feb 17, 2013, 09:58 pm
When I suggested a pot I didn't intend the wiper be a part of the series element rather the wiper is the output, failure of the wiper contact results in no/intermittent output, Not the high voltage at the Arduino.
There should also be a series resistor of about 10 - 20K between the pot and the Ardiono analog input pin.
Bleeders... devices to "Bleed" off the high voltage on a capacitor 'bank' are Not in the 50 Megohm range...
Many years ago I built amplifiers high power RF amplifiers, and I used power supplies that developed 900V or more... the Minimum value I was comfortable with was 20K ohms Not 44 Mohms, a value 2500 times larger.
Values between 10 and 100 K are more common as bleeder resistors.
500/44,000,000 = 11.4 uA, 500/20,000 = .025A or 25 mA.

Bob.
BE EXTREMELY CAREFUL THE VOLTAGES YOU ARE WORKING WITH CAN BE INSTANTLY LETHAL.
I WOULD, VERY, AND I HAVE BEEN SERVICING H V EQUIPMENT SINCE I WAS 16 OR 17... I'M STILL HERE.
REMEMBER SAFETY IS NOT AN ACCIDENT, IT'S A PLAN.
--> WA7EMS <--
"The solution of every problem is another problem." -Johann Wolfgang von Goethe
I do answer technical questions PM'd to me with whatever is in my clipboard

#### dc42

#12
##### Feb 17, 2013, 11:04 pm

Bleeders... devices to "Bleed" off the high voltage on a capacitor 'bank' are Not in the 50 Megohm range...
Many years ago I built amplifiers high power RF amplifiers, and I used power supplies that developed 900V or more... the Minimum value I was comfortable with was 20K ohms Not 44 Mohms, a value 2500 times larger.
Values between 10 and 100 K are more common as bleeder resistors.

The appropriate value of a bleeder resistor depends very much on the  capacitance of the capacitor bank. For example, if the capacitance is 10nF then a 44M bleeder resistor will bleed 500V down to a safe value in a little over a second. For an RF power amplifier, you would have been using a much greater capacitance than 10nF, so a much smaller bleeder resistor would have been used.
Formal verification of safety-critical software, software development, and electronic design and prototyping. See http://www.eschertech.com. Please do not ask for unpaid help via PM, use the forum.

#### Docedison

#13
##### Feb 18, 2013, 02:50 amLast Edit: Feb 18, 2013, 12:21 pm by Docedison Reason: 1
Quote
Quote
if the capacitance is 10nF then a 44M bleeder resistor will bleed 500V down to a safe value in a little over a second. For an RF power amplifier, you would have been using a much greater capacitance than 10nF, so a much smaller bleeder resistor would have been used.

Excuse me I must have missed this in electronics 101 but just how does a 100 nF capacitor bleed a capacitor 'bank' charged to 500 V in "a little over a second'? last time I checked capacitors didn't conduct DC.
Did anyone bother to calculate the time constant for 44 megohms and a .01uF' capacitor "Bank" @ 500 VDC.{Edit for a typo} RKJ
With the 44 meg resistor grounded and no other leakage?... try it.     [should read 10 nF but .01 uF is the same.]
@ dc42 I missed the 10nF comment in reply #6, Getting old I read it as .1uF and missed the decimal in the above comment.
--> WA7EMS <--
"The solution of every problem is another problem." -Johann Wolfgang von Goethe
I do answer technical questions PM'd to me with whatever is in my clipboard

#### dc42

#14
##### Feb 18, 2013, 09:22 am

Quote
Quote
if the capacitance is 10nF then a 44M bleeder resistor will bleed 500V down to a safe value in a little over a second. For an RF power amplifier, you would have been using a much greater capacitance than 10nF, so a much smaller bleeder resistor would have been used.

Excuse me I must have missed this in electronics 101 but just how does a 100 nF capacitor bleed a capacitor 'bank' charged to 500 V in "a little over a second'? last time I checked capacitors didn't conduct DC.
Did anyone bother to calculate the time constant for 44 megohms and a '1uF' capacitor "Bank" @ 500 VDC.
With the 44 meg resistor grounded and no other leakage?... try it.

Where did you get 1uF from? Unless I've missed something, the OP has not said what the capacitance of the capacitor bank is. I was pointing out that if it was only 10nF then a 44M bleeder resistor would be OK.
Formal verification of safety-critical software, software development, and electronic design and prototyping. See http://www.eschertech.com. Please do not ask for unpaid help via PM, use the forum.

Go Up