Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Resistor Divider help needed

Status
Not open for further replies.

graygem

Member level 1
Joined
Sep 21, 2010
Messages
33
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,286
Activity points
1,508
I am trying to make the resistor divider below.
I would like to figure out how to calculate the resistors for the different taps.
The taps are going into op amps to turn on LEDs.


3tap Resistor Divider.jpg





Don't hesitate to tell me if I'm an idiot.
 

from resistor divider we have these equations

Code:
7.08 = 15 * (R4 / (R4 + R3 + R2 + R1))
7.1 = 15 *  ((R4 + R3) / (R4 + R3 + R2 + R1))
7.12 = 15 * ((R4 + R3 + R2) / (R4 + R3 + R2 + R1))

we have 3 equations with 4 unknown so assume a value for one of them and solve equation to find rest of them
for example assume R4 = 7.08 thne you will have R3 = R2 = 0.02
R1 = 7.88
 

Hi,

Quick way is making total resistance equal V+. I do it because I'm not good with numbers...

Code:
Rtotal = 150K

R1 = 78K8
R2 = 200R
R3 = 200R
R4 = 70K8

(R2 + R3 + R4)/(R1 + R2 + R3 + R4) * Vin =  0.4746r* 15V            = 7.12V
(R3 + R4)/(R1 + R2 + R3 + R4) * Vin =         0.473r * 15V             = 7.1V
(R4/ R1 + R2 + R3 + R4) * Vin =                  0.472 * 15V              = 7.08V


Sometimes worth doubling/multiplying or dividing resistor values to find ones closer to standard resistor values.

No interest in doing that with op amps to get a more precise and reliable voltage for each tap?

- - - Updated - - -

Hi again,

Another, longer way is:

a) Divider voltages/Vtotal
b) Select Rtotal
c) Multiply results from a) by b)
d) Subtract R2 from Rtotal to get R1 value.
e) Subtract intermediate resistor vaules from each other progressively and last from bottom resistor.

e.g.

a) 7.12/15 = 0.47466666666666666666666666666667
7.10/15 = 0.47333333333333333333333333333333
7.08/15 = 0.472

b) Rtotal = 100K
100000 * 0.47466666666666666666666666666667 = 47466.666666666666666666666666667
100000 * 0.473r = 47333.333333333333333333333333333
100000 * 0.472 = 47200

100000 - 47466.666666666666666666666666667 = 52533.333333333333333333333333333

47466.666666666666666666666666667 - 47333.333333333333333333333333333 = 133.333333333333333333333333334

47333.333333333333333333333333333 - 47200 = 133.333333333333333333333333333

Sorry for the lengthy decimals there, just inculded so numbers tally.

So rounding up or down,
R1 is 52K5
R2 is 47K5
R3 is 130R
R4 is 130R

I prefer the other method, it's quicker, but this one has it's uses.
 

Also note, that the range of values you are trying to create with this divider is only about +/ 2%, so if you use this approach, you must use precision resistors, or simply the normal variations in resistor values will cause you to miss these specifc voltage values.
 

Also note, that the range of values you are trying to create with this divider is only about +/ 2%, so if you use this approach, you must use precision resistors, or simply the normal variations in resistor values will cause you to miss these specifc voltage values.
Maybe not so much, because there is only 0.04V span between lowest and highest out value.

If I where to use cheapest possible components, I would first decide for a couple of low value resistors for voltage drop over R2 and R3 - 100Ω for example.
Thereafter I would find the corresponding value for proper voltage drop over R1 and R4.
Also I would allow myself to substitute R1 and R4 with two resistors in parallell to get closer - or better, just use a potmeter.
 

Your not an idiot but your 15V supply error will quickly dominate your results.

20mV thresholds differences with No Tolerance makes your spec INCOMPLETE.

20mV range even with 10% accuracy is 1mV/15V or 66 parts per million.

TRY AGAIN. write new specs. with tolerances !!

or choose 10 ppm Vref at 7V and pullup to 120mV offset easily with taps..
 
Last edited:

Hi,

0.47466666666666666666666666666667
I wonder what's the idea behind writing 30+ digits? Nonsense.

For a 1% tolerated resistor 3 digits are sufficient, maybe write 4 digits (integer + fractional). Everything above 4 digits makes no sense.

Klaus
 

On obvious way to get any arbitrary resistor values is to put two 1% resistors in series or parallel of the proper value.
That way the worst-case arbitrary value will be within 1% of the desired.
This website will calculate the optimum values for any desired resistance.

Thus, for example, to get the 70.8k resistor value calculated for R4 in the first part of post #3, you can put 140k in parallel with 143k to give 70.8k within 0.08% (±1% of course).
 

On obvious way to get any arbitrary resistor values is to put two 1% resistors in series or parallel of the proper value.
That way the worst-case arbitrary value will be within 1% of the desired.
This website will calculate the optimum values for any desired resistance.
Here is one another such calculator. This one supports configurable stock, which could be useful.

Thus, for example, to get the 70.8k resistor value calculated for R4 in the first part of post #3, you can put 140k in parallel with 143k to give 70.8k within 0.08% (±1% of course).
Using this same example, my tool recommends putting 5.9k and 64.9k in series to obtain exactly 70.8k. Of course subject to the tolerance of the resistors used. I think 143k is an E96 value, but both 5.9k and 64.k are from E48 series, so I prefer this combination.
 

Calculation methods and implementation with standard series fixed resistors have been explained.

I wonder if it makes sense to derive exact voltage levels from a 15 V "reference". Precision references are sourcing 10.0 or 10.24 V, industry standard 15 V fixed voltage regulators have +/- 2 % tolerance.
 

Hi,


I wonder what's the idea behind writing 30+ digits? Nonsense.

For a 1% tolerated resistor 3 digits are sufficient, maybe write 4 digits (integer + fractional). Everything above 4 digits makes no sense.

Klaus

Obviously so, "nonsense", the idea - there was none - was just copy/pasting the results from the calculator, and I'm sure the thread poster understands such nuances, thanks.
 

Hi

the best TL431 B grade is rated 0.5% = 5000ppm accuracy.

Would a TL431 accomplish this?
No.
****

No Tolerance makes your spec INCOMPLETE.
I vote for tolerance....;-)

Klaus
 

your 15V supply error will quickly dominate your results.

In fact, the difference of only 0,02V ( =7,12-7,10 ) on a 15V supply means a voltage gap of only 0,13%, practically within the range of the noise seen in most power supplies. No analog parts could reach such an resolution like that with passive components.
 

Hi,

Is it for a LiIon battery charge status detector?

(if so, then I doubt it will give meaningful results)

Klaus
 

Hi,

Is it for a LiIon battery charge status detector?

(if so, then I doubt it will give meaningful results)

Klaus

Besides possible overheating, fire/explosions and hospital stays in the near future for the OP? If this is the method for displaying charge status what does the charger use (or is this what the charger will use)? :shock:
 

Is it for a LiIon battery charge status detector?

It's just a simple audio peak detector, doesn't have to be super accurate. The tap voltages are just what I measured at where I wanted the LEDs to come on at.
 

It's just a simple audio peak detector, doesn't have to be super accurate. The tap voltages are just what I measured at where I wanted the LEDs to come on at.
I use an LM3915 to drive the LEDs in my sound level indicator that also has a peak detector. The LM3915 has the resistor ladder and 10 comparators that drive the LEDs with an adjustable constant current. Its detection points are 3dB (0.707 times or 1.414 times) apart which is a small but noticeable change in level. Nobody will be able to hear the tiny change in level (less than 1%) with your resistors.
 

I use an LM3915 to drive the LEDs in my sound level indicator that also has a peak detector. The LM3915 has the resistor ladder and 10 comparators that drive the LEDs with an adjustable constant current. Its detection points are 3dB (0.707 times or 1.414 times) apart which is a small but noticeable change in level. Nobody will be able to hear the tiny change in level (less than 1%) with your resistors.

I didn't even notice that the project I linked to was yours! :laugh:
I went back to take a look at it as I thought the recommendation for what part to use sounded similar, then I noticed the name of the user that submitted the post.

I'll vouch for you if another moderator thinks you are promoting your own blog ;)
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top