# Voltage Level shifting

Discussion in 'The Projects Forum' started by sanketbarot, Mar 5, 2007.

1. ### sanketbarot Thread Starter Member

Dec 12, 2006
14
0
Hello friends,

Initially I have to convert the 24V signal to 5V which I did using voltage divider circuit but now the thing is my input signal level is not decided it may vary from 15V to 24V so the question is how can I maintain 5V.

My input is varying from 15V to 24V but I required 5V output.. how can I do this??????????/

2. ### Papabravo Expert

Feb 24, 2006
10,340
1,850
Comparator. LM393 dual or LM339 quad. This function may also be available in a single.

3. ### sanketbarot Thread Starter Member

Dec 12, 2006
14
0
I have studied all the three but they all are comparatoes..
If I am not wrong whatever my Vcc is I will get that output incase where output is logic high. So if I want 5V then I can give %v as Vcc right??
Bur once I connect 5V as my Vcc can give input between 24V to 15V??

My Basic circuit is I am deriving DC voltage from the AC voltage source. Ac voltage source may go down..So as calculated minimum DC voltage I can have is 15V and maximum volatge I can have 24V. I gave this DC voltage to one circuit and which will accordinlgly giving me the sigbnal what I want but the thing If DC voltage changes my Signal Level Changes

The second Idead I just thought is can I use Optocupler Like.. because My circuit is like
Suppose My Dc voltage is 24V then my signal strength will be of 24V only same way if Dc voltage is 15v then my signal strength will be 15V only..

So can I work like Whatever my DC voltage I can give that to optocoupler an opto will convert it into 5V..

4. ### hgmjr Moderator

Jan 28, 2005
9,030
214
I think the use of an opto-coupler/opto-isolator may provide you with the flexibility you need to monitor the dropout of your DC voltage. It can be designed to accommodate the variation of the DC voltage you are monitoring and still produce an output signal that will indicate the voltage has dropped.

If you happen to be using the voltage that you are monitoring to power your detection circuitry then you could run into a problem.

hgmjr

5. ### rwmoekoe Active Member

Mar 1, 2007
172
0
hi sanketbarot,

is this input and output are digital? 0 or 1 in nature?
if yes, we should know what the threshold level of the input will be? Should it be always at 50% of the max input level?
if no, just use the comparator, with a fix supply voltage of 5vdc, and tie the input (after a resistor say 10k) to a clamp diode to the vcc.

that'll do it

6. ### Papabravo Expert

Feb 24, 2006
10,340
1,850
You are incorrect. The Supply voltage on the LM393 can be 15 Volts or 24 Volts up to 36 Volts. The output is an open collector logic output which can be pulled up to +5V or +3.3V or whatever your logic supply happens to be. That's the whole point the comparator Vcc and the logic Vcc ARE NOT THE SAME!

In the case of the 26LS32 line receiver the logic Vcc is +5V but the inputs wil tolerate much higher voltages. Check the datasheets and read them a little more carefully before you dismiss the suggestions of experience.

7. ### mrmeval Distinguished Member

Jun 30, 2006
833
2
Do you have a graphic of the schematic?

8. ### Papabravo Expert

Feb 24, 2006
10,340
1,850
Not really, but it is the same circuit you find in any textbook on analog circuits or any datasheet. One input goes to the signal of interest. The other input goes to a voltage divider to select the threshold. The output is pulled up to the Logic Supply and goes to the processor input. By reversing the two inputs you can invert the logic state of the output.

Here is the important point. Comparators have analog inputs and open collector digital outputs.

9. ### lattesurf New Member

Mar 6, 2007
9
0
Why not try using a 5V regulator? An LM7805 would work well.

Inputs between 15-24V would be stepped down to a flat 5V. Even though there is a voltage drop of 2-3V for the regulator, that is still not a problem as long as your input voltage is at least 8-9V.

Though you might need a heat sink if the circuit runs for extended time.

10. ### Papabravo Expert

Feb 24, 2006
10,340
1,850
I think this is a really really dumb idea.

11. ### thingmaker3 Retired Moderator

May 16, 2005
5,072
6
I suspect Lattesurf simply misread the earlier post, mistaking "signal" for "voltage level." I'm sure it was an honest error, and not really "dumb."

12. ### Papabravo Expert

Feb 24, 2006
10,340
1,850
How can you be so sure? As a moderator you're almost required to give people the benefit of the doubt. Lattesurf can certainly speak for himself on this subject and probably should. It's also possible that I misunderstood what he was suggesting.

13. ### Dave Retired Moderator

Nov 17, 2003
6,960
145
It is irrelevant if a Moderator must give the benefit of the doubt to anyone. Perhaps if you think this is a "dumb idea" (and you are entitled to express your opinion) then you could indicate to Lattesurf as to why this is the case. Lambasting something as "dumb" without a reason why doesn't help anyone.

Dave

14. ### n9352527 AAC Fanatic!

Oct 14, 2005
1,198
4
Actually I used that 'dumb idea' in a manufactured system once. If it does fulfil the requirements, then I don't see any reason not to use it.

15. ### Papabravo Expert

Feb 24, 2006
10,340
1,850
You're right I should have explained why I think the idea is dumb. The first idea is that devices are characterized for use in their intended application. Many voltage regulators require specific capacitence on the inputs and outputs in order to achieve stable operation. The next issue with using a voltage regulator in this way is response time. You seldom see things such as rise time and fall time or propagation delay specified. A voltage regulator by its very nature has an undesireable characteristic of its input. The more current the output consumes the more current is drawn from the input. So the input to the voltage regulator starts to load down the driving source. A comparator has none of these problems along with the considerable benefit of an adjustable threshold.

You may have gotten away with this hack in a manufactured product, but not if I had been on the design review team. I am curious as the the reason why you thought this was a reasonable approach.

16. ### n9352527 AAC Fanatic!

Oct 14, 2005
1,198
4
Cost. Pure and simple. There were already several identical regulators in circuit. Adding another one means that inventory is simplified, purchase price is lowered, no need to qualify another device, get a supplier and another reel on the pick and place machine.

There are regulators with fairly flat quiescent current against different load current. The signal was capable of sourcing the current needed, and the timing specifications weren't that strict. It was not a typical approach, I agree, but my view of engineering isn't exactly ideal either. Designing a circuit, in my opinion, is not a form of art contest, as long as it worked alright and within the required specs, then whether it was an elegant solution or not have little or no bearing when the total cost difference is considered.

I agree it is probably not a good thing to tell others, especially when they are still early on their learning phase to use devices not as intended. But also bear in mind that different designers have their own views on how to design circuits. Some like to produce elegant and probably slightly more expensive solutions, others put emphasis on other considerations such as cost, testability or manufacturability and maybe countless other factors. There is no one single correct approach and the only dumb approach is when the circuit doesn't work to the specs.

17. ### lattesurf New Member

Mar 6, 2007
9
0

2. First post by sanketbarot states he/she uses "voltage divider circuit" to "convert the 24V signal to 5V".

3. You have your reasons for using 393, whilst i have my reasons for 7805.

4. Like what you said in your reasons for saying the 7805 is dumb, i could also say using the 393 is dumb. You have 2 comparators within an 8-pin package and you end up using only 1, wasting money & resources, dumb.

5. You need an additional 5V power supply to connect the comparator's Vcc, so how do you get the 5V? An LM7805? Voltage Divider using Resistors? More and more components if what sanketbarot needs is just a voltage step down ?

I don't give a damn if i'm banned from this forum or not, since this would be my 3rd and final post. My very first post and i got flamed? Wow, very nice welcome from this forum huh?

The moderator(s) & administrator(s) do not need to give anyone the benefit of the doubt, neither do I need them to give me any.

Even Thomas Edison makes errors with his work, it is human nature. You think you are cleverer than Edison? So Only your idea is good and others are dumb? In electronics, there is never a "Perfect" circuit design or component.

How do you know that using 7805 in sanketbarot's circuit would fail or not work as good a comparator, and be a dumb suggestion?

You should know pretty well about design trade-offs, the 7805 may not be very stable, but still gives a relatively close to 5V output. 393 MAY be very stable, but needs more components. When your circuit board is pressed for space, and output regulation is not very important, which would you chose?

So before you say someone's suggestion is "dumb", think about all these points. Unless your suggestion/idea is 100% foolproof and 100% workable, then say others are dumb.

You think you are so great? You are so clever? Go fly a kite!

Dec 10, 2005
67
0