Multiple Voltage Dividers

Thread Starter

Ayden1717

Joined May 17, 2022
7
Hi, I am currently working on designing a circuit that will take measurements of a high voltage tractive system (583V) from an accumulator. On startup, there are two events that need to occur before the system reaches its nominal at 583V.

1. At 60V a comparator must switch a light on
2. At 525V a contactor must switch closed

I understand how to do everything with the exception of the voltage division. Ideally, it would be able to run two voltage divider circuits for each event but this yields a parallel resistance. I am wondering if there is any way I could run two voltage dividers on the same bus together or would I have to just be selective with my resistance values and run it off of one. I have been looking into voltage followers and how they can offer high input impedance in voltage divider circuits to be able to use the divided voltage without fluctuations due to low load impedance.

Also while on the topic of this parallel resistance issue, I am not sure what the actual tractive system resistance will be, and because I am taking a measurement of the battery voltage, I will technically be running this system in parallel with my voltage divider which then again will mess up the timing of the events. I am thinking I will just need to have one voltage divider with extremely high impedance to ensure that the tractive system is significantly smaller than the voltage divider. Some feedback on my thought process would be great!
 

dl324

Joined Mar 30, 2015
14,704
It would be helpful if you posted a schematic for the voltage divider configuration you're trying to describe.

If both dividers are on the same voltage, they'll be independent as long as the voltage source isn't being loaded.

If you're using a divider on a divider, having a current in the first divider that's at least 10 times the current drawn by the second divider would work. The stiffer the divider, the better.
 

crutschow

Joined Mar 14, 2008
30,127
Use just one voltage divider to generate a low voltage for two comparators (e.g. LM339/LM393).
For example, using a near 200:1 divider (e.g. 1.07megΩ and 5.11kΩ) will give 0.285V from the divider for 60V, and 2.48V from the divider for 525V.
The two voltage comparators are then set to detect 0.285V for the light and 2.48V for the contactor.
A voltage reference, such as the TL431, can be used give stable trigger voltages for the comparators.

What voltage(s) do you have to power the light and contactor?
 
Last edited:

BobTPH

Joined Jun 5, 2013
5,259
You can also use 3 resistors in series to provide two different divisions from the same input voltage. For example, 3 resistors of the same value would give you 2/3 and 1/3 of the input.

Bob
 

MisterBill2

Joined Jan 23, 2018
12,440
It is entirely possible to have a voltage divider with more than one tap point. What I believe is the sequence is that when the voltage impressed on "the accumulator" reaches 60 volts, one action happens, and then as the voltage rises to 525 volts a second action happens.
So really, what you need is one voltage divider and two voltage sensors.
So the first question is "how accurate does this need to be?? Is +/- 1% acceptable? Or does it have to switch at exactly 60.00 volts? and 525.00 volts?
5% accuracy is not so very hard, 1% accuracy takes a fair amount of effort, and getting to +/- 0.1% error, or less, gets both tedious and expensive.
The second question is similar: How repeatable does it need to be?
The third question is how fast is this voltage changing?
And the final question is how big is the project budget?

The really simple scheme is to use a high accuracy digital voltmeter that has at least four digits of resolution, and monitor the digital value with a fast computer to initiate the desired action as the voltage reaches those values. (This is expensive but it will work.A less expensive scheme will use one divider 100.00 :1.00, giving an output of zero to 5.75 volts. Then a pair of good comparators, such as LM111 types, with a very well filtered DC reference voltage and supply voltage, will work. Probably a good buffer amplifier for the measured value to avoid any error producing loading of the voltage to be monitored.
 

Thread Starter

Ayden1717

Joined May 17, 2022
7
It is entirely possible to have a voltage divider with more than one tap point. What I believe is the sequence is that when the voltage impressed on "the accumulator" reaches 60 volts, one action happens, and then as the voltage rises to 525 volts a second action happens.
So really, what you need is one voltage divider and two voltage sensors.
So the first question is "how accurate does this need to be?? Is +/- 1% acceptable? Or does it have to switch at exactly 60.00 volts? and 525.00 volts?
5% accuracy is not so very hard, 1% accuracy takes a fair amount of effort, and getting to +/- 0.1% error, or less, gets both tedious and expensive.
The second question is similar: How repeatable does it need to be?
The third question is how fast is this voltage changing?
And the final question is how big is the project budget?

The really simple scheme is to use a high accuracy digital voltmeter that has at least four digits of resolution, and monitor the digital value with a fast computer to initiate the desired action as the voltage reaches those values. (This is expensive but it will work.A less expensive scheme will use one divider 100.00 :1.00, giving an output of zero to 5.75 volts. Then a pair of good comparators, such as LM111 types, with a very well filtered DC reference voltage and supply voltage, will work. Probably a good buffer amplifier for the measured value to avoid any error producing loading of the voltage to be monitored.
The division accuracy does not need to be super accurate, within a few volts is good enough for our application, it should be fairly reputable because the 60V is a high voltage indicator light so we would like to be within reason for that. Both of the events will be triggered in under 5 seconds (our tractive system precharge to 583V is near 2.5 seconds).
 

Thread Starter

Ayden1717

Joined May 17, 2022
7
It would be helpful if you posted a schematic for the voltage divider configuration you're trying to describe.

If both dividers are on the same voltage, they'll be independent as long as the voltage source isn't being loaded.

If you're using a divider on a divider, having a current in the first divider that's at least 10 times the current drawn by the second divider would work. The stiffer the divider, the better.
Okay i apologize for no schematic, i like that rule of thumb so essentially if my resistance of one divider is significantly higher than the other (factor of 10) I should be okay to get decent readings on the one I need? I think I am going to go for one divider and tap into it twice, i am now just trying to figure out what me ESR for the divider should be relative to the tractive system by a factor of 10 sounds like it should be enough.

Thanks!
 

Thread Starter

Ayden1717

Joined May 17, 2022
7
Use just one voltage divider to generate a low voltage for two comparators (e.g. LM339/LM393).
For example, using a near 200:1 divider (e.g. 1.07megΩ and 5.11kΩ) will give 0.285V from the divider for 60V, and 2.48V from the divider for 525V.
The two voltage comparators are then set to detect 0.285V for the light and 2.48V for the contactor.
A voltage reference, such as the TL431, can be used give stable trigger voltages for the comparators.

What voltage(s) do you have to power the light and contactor?
Both of these devices operate at 12V. I was looking into the TL431 and thats a really great solution to my problem as well thanks for including that
 

MisterBill2

Joined Jan 23, 2018
12,440
Consider also that as the voltage gets divided the resolution is reduced. If the divider is 100:1, then a 100 volt change provides a one volt change in the output. This is something that you must consider relative to your accuracy requirements.
If you are feeding comparator circuits that have a small amount of hysteresis added, the rising and falling trip levels will be a bit different. So you certainly need to do some analysis.
Fortunately there is a great deal of information available about comparators. And i would not trust one word of what is found on you toob. Manufacturersapplication notes will be the most valuable source of explanations and information.
 
Top