Resistive voltage divider

[glatocha]

Stupid, or maybe not, question:

Are there any max limitation for the resistors in the voltage divider?
Through the divider we will always take current, so for the low power application would be advantage to use as big resistors as we can.
Same if it is used as a divider for the voltage measurement, it is always better to have bigger input resistance.

Do I need to concern about something or I can go to megaohms?

[kt315]

You will be limited in the current you can draw out of your divider. It all depends on the impedance of whatever you are feeding with the divider.

So there is no limitation you have to worry about besides the Ohm's law.

[c4757p]

A higher resistance divider will also be more prone to picking up nearby EM interference, as well as generating its own thermal noise. Not a problem at low frequency/DC (just add a capacitor to LP filter it), but you want lower resistance for AC or low noise applications.

[edpalmer42]

Stupid, or maybe not, question:

Are there any max limitation for the resistors in the voltage divider?
Through the divider we will always take current, so for the low power application would be advantage to use as big resistors as we can.
Same if it is used as a divider for the voltage measurement, it is always better to have bigger input resistance.

Do I need to concern about something or I can go to megaohms?

Remember that you want to measure the output of the voltage divider.  Make sure that your meter won't load the voltage divider and change the reading.  For example, if you used two 10 MegaOhm resistors as your voltage divider and then tried to measure the output with a meter that had a 10 MegaOhm input impedance, your error would be about 33% due to the loading effect of your meter.

Ed

[SeanB]

Resistors have a voltage rating, independent of the power they are dissipating. For small 0.25W resistors it is 250VDC, for a 0.3W metal film generally 300VDC, for a 0.5W one generally 500V. You get high voltage resistors that have ratings of 1000V, 2000V and higher. Most SM resistors have much lower ratings, typically around 150V or less. Just take a note of this in designing the divider, as you need to not exceed this voltage or they probably will arc over internally, and eventually go open circuit.

[M0BSW]

Just have to say I'm really impressed with all the answers given, you boys really know your stuff, which makes me realise how little I know

[glatocha]

Yeah, looks like the question was no so stupid after all.

Even If I take my voltage directly to the OpAmp input I need to involve the Input current into the consideration.
After all I think I will use an OpAmp with 0..1 gain to get as high impedance input as I could. Which will bring other problems.

[IanB]

Resistors have a voltage rating, independent of the power they are dissipating. For small 0.25W resistors it is 250VDC

Of course, if you apply 250 V across a 100 k resistor the power dissipated will be 0.63 W, so that 0.25 W resistor is going to get rather warm...

[SeanB]

True, but for the 1M resistor it will be 0.063W, and the resistor will be cool, but will tend to arc over at higher voltages. I often see 470k 0.25W resistors used in power supplies as input capacitor discharge resistors, and they almost always go open circuit in a few years with power applied, often quietly, sometimes arcing over and burning up leaving a soot char on the board.

[Neilm]

Resistors have a voltage rating, independent of the power they are dissipating. For small 0.25W resistors it is 250VDC

Of course, if you apply 250 V across a 100 k resistor the power dissipated will be 0.63 W, so that 0.25 W resistor is going to get rather warm...

Which brings us nicely onto the effects of temperature on the resistor (Temperature Coefficient of Resistance or TCR). As a resistor warms, the actual resistance will change causing your bridge divider ratio to change. There are ready made dividers out there that claim that the ratio will stay the same as the divider is built into one package. I've never tested one so can't comment.

In the same way there is a Voltage Coefficient of Resistance (VCR). At higher voltages the apparent resistance will change. These are generally only specified on high voltage resistors though. For anything less than 500V it can be pretty much ignored.

Neil

[SeanB]

I have some NOS carbon composition resistors, they do this at quite low voltages, some are more like thermistors or varistors than resistors.