EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: paul18fr on September 12, 2014, 02:17:34 pm
-
Dear All,
I'm trying to understand if there 1 or 2 power supply in the figure 1 herebellow ?
From my understanding, the goal is to provide +/- 15 V (from DC adapters), as we can see in the figure 2 with 9V batteries: am I right ?
Thanks
Paul
-
you have two there, one is negative and one positive
-
that's what I've been thinking to ... thanks
NB: I've been wondering why in Fig 1, the 2 capacitors are differents ... hummm ??? it sounds like "unabalanced" signal
.... many things to learn (not just "mimic"), just the beginning of a (long but interesting) walk ;)
Paul
-
I've been wondering why in Fig 1, the 2 capacitors are differents ... hummm ??? it sounds like "unabalanced" signal
Cost effective integrated circuit processes generally have NPN power transistors and not PNP power transistors. Because of this, the negative regulator (7912) uses an NPN transistor in as a common emitter configuration which is more unstable and requires larger input and output capacitors for stable operation. This is also why the 7912 pinout differs from the 7812 pinout.
-
yes it used to be the thing that regulators needed lots of input and output capacitance but the positive regulators which are the most used by far have come a long way and are now able to work with as little as 100-330nF rather than the proverbial 100nF ceramic + 10-100uF electrolytic.
-
yes it used to be the thing that regulators needed lots of input and output capacitance but the positive regulators which are the most used by far have come a long way and are now able to work with as little as 100-330nF rather than the proverbial 100nF ceramic + 10-100uF electrolytic.
That is still the case but higher performance designs pay more attention to it.
Modern low dropout positive regulators which use PNP or P-MOS pass elements have the same input and output capacitor requirements as old negative regulators which use NPN pass elements and for the same reason. The common emitter/source pass transistor has voltage gain which depends on the output load making frequency compensation more difficult. The common collector designs used in high dropout regulators are a lot more stable and can often get by with no bulk output capacitance.
-
well,
The schematics come from the Girino project (A big challenge for the newbie I am, but nothing more than a project to learn).
I think I understand almost all the philosophy (not in detail, the big lines):
- I think the design have been made so that either LM324 or TL084 can be used ... with the help of the jumpers
- as the author says, LM324 needs a single power supply whereas TL084 needs two,
- ok for the offset role (practical example of the non-inverting summation role of the op-amp),
Nevertheless I'm a bit disturbed by the two power supplies:
- LM7812 and LM7912 involve a drop in the voltage, that's why 15V is necessary
- But why 15V for the Arduino that accepts 12V max from the specification?
- I would have used 2 specific 15V power supplies and another 9V for the Arduino separately
- The author speaks about power supplies in series (??????)
Ambitious project (personal) challenge calling for additional tools (LTspice for example)
Paul
-
the author document (zipped due to its size)
-
- I think the design have been made so that either LM324 or TL084 can be used ... with the help of the jumpers
- as the author says, LM324 needs a single power supply whereas TL084 needs two,
- ok for the offset role (practical example of the non-inverting summation role of the op-amp),
There is an interesting symmetry between the LM324 and TL084 which they might have considered; the input common mode range of the LM324 extends to the negative supply while the input common mode range of the TL084 extends to the positive supply.
These days lots of amplifiers have those properties but in the past an input common mode range which included the positive supply was relatively rare or maybe unappreciated. The even older LM301/LM301A has this useful property also which is why it is sometimes found in high side current sense circuits.
-
well,
The schematics come from the Girino project (A big challenge for the newbie I am, but nothing more than a project to learn).
Good learning project, IMO. Just don't expect to get an oscilloscope out of it :)
Nevertheless I'm a bit disturbed by the two power supplies:
- LM7812 and LM7912 involve a drop in the voltage, that's why 15V is necessary
- But why 15V for the Arduino that accepts 12V max from the specification?
Arduino uses a 7805 regulator at the input. According to the spec (https://www.fairchildsemi.com/datasheets/LM/LM7805.pdf (https://www.fairchildsemi.com/datasheets/LM/LM7805.pdf)), these can take up to 35V input, so 15V should be safe. The Arduino doesn't likely take much current alone (the "shield" is not powered from it), so it should be OK.
The 9-12V recommendation for Arduino is due to the fact that the "extra" input voltage gets turned into heat and has to be dissipated. If you have 15V input, your Arduino + some shields take 500mA, you are going to have to dissipate (15-5) * 0.5 = 5W of heat. That will make the regulator extremely hot very fast, it doesn't have much of a heatsink. It will most likely turn off or even fail. Dave did a good video on this type of problem (https://www.youtube.com/watch?v=y-KkPLWZJko (https://www.youtube.com/watch?v=y-KkPLWZJko))
If you use only 9V as input, then you are dissipating (9-5)*0.5 = 2W only, with the same load, requiring less cooling. That's why you need to keep the input voltage as low as possible or use a different type of regulator - an LDO or (better) a switching regulator.
- I would have used 2 specific 15V power supplies and another 9V for the Arduino separately
- The author speaks about power supplies in series (??????)
You could use a separate supply for the Arduino, but it would be an overkill in this case. It would be justified if you wanted to make sure that the digital noise from the micro isn't getting into your sensitive input circuitry or something similar, but that is not really going to matter within usable limits of this design. Good bypassing of the power supplies is a good idea, though.
Two 15V power supplies are in series, so that he gets +-15V against the common pin for powering the opamps. It is an overkill, IMO - he could have used +-9V or even a single supply and rail-to-rail opamps and saved the extra circuitry. He mentions that possibility, though. The Arduino is connected only to one of the power supplies, so it gets only 15V, not 30.
Ambitious project (personal) challenge calling for additional tools (LTspice for example)
Paul
Naw, not really. There is no point in simulating the power supply, that is just standard datasheet circuit. Build it up on a breadboard and measure it to make sure you understand it, though!
The opamp part is also simple - all he has there are two buffers (he calls them "emitter followers" for whatever reason :-// ) to buffer the input signal (he wants 1M input impedance as on a regular scope) and to buffer the output of the summing amplifier. The second buffer is likely not even needed, the opamp should have more than enough power to drive the relatively high impedance analog comparator and ADC inputs of the ATMega. The summing circuit is there to add 2.5V offset so that he can measure negative signals too (the ATMega can take only signals between 0-5V).
To understand the circuit, I suggest looking at some opamp tutorials, e.g.:
http://www.eevblog.com/2014/04/06/eevblog-600-opamps-explained/ (http://www.eevblog.com/2014/04/06/eevblog-600-opamps-explained/)
or
https://www.youtube.com/watch?v=OMJ9WGrRf6A (https://www.youtube.com/watch?v=OMJ9WGrRf6A) (Sparkfun - probably simpler for start)
Or look at any EE textbook, opamp circuits are for sure going to be covered.
Good luck!
-
The opamp part is also simple - all he has there are two buffers (he calls them "emitter followers" for whatever reason :-// ) to buffer the input signal (he wants 1M input impedance as on a regular scope) and to buffer the output of the summing amplifier. The second buffer is likely not even needed, the opamp should have more than enough power to drive the relatively high impedance analog comparator and ADC inputs of the ATMega. The summing circuit is there to add 2.5V offset so that he can measure negative signals too (the ATMega can take only signals between 0-5V).
To understand the circuit, I suggest looking at some opamp tutorials, e.g.:
http://www.eevblog.com/2014/04/06/eevblog-600-opamps-explained/ (http://www.eevblog.com/2014/04/06/eevblog-600-opamps-explained/)
or
https://www.youtube.com/watch?v=OMJ9WGrRf6A (https://www.youtube.com/watch?v=OMJ9WGrRf6A) (Sparkfun - probably simpler for start)
Or look at any EE textbook, opamp circuits are for sure going to be covered.
Good luck!
Not that I'm an expert, or even close, but I found the "Op Amps for Everyone" book to be very helpful in beginning to understand op amps and what you can do with them. These little chips are amazing in their versatility.
-
Good learning project, IMO. Just don't expect to get an oscilloscope out of it :)
My children and I have different projects, and in parallel to the purposes previously described, we just want to be able to see some specific signals at low frequencies (I do not expect signals lower than few KHz) - the signal generator is one of the next steps; I've planned to invest to an x-scope, in a near future (the ones I found in internet - 2nd hand - are not really interesting so I wait to purchase one such as Siglent or Rigol one)
NB: there are very happy to learn how to apply schematics on the breadboard, how to braze components on the veroboard, how , to build electronic devices and so on … I don’t know who plays the much (father or children :-DD)
If you use only 9V as input, then you are dissipating (9-5)*0.5 = 2W only, with the same load, requiring less cooling. That's why you need to keep the input voltage as low as possible or use a different type of regulator - an LDO or (better) a switching regulator.
One more thing to see ...
You could use a separate supply for the Arduino, but it would be an overkill in this case. It would be justified if you wanted to make sure that the digital noise from the micro isn't getting into your sensitive input circuitry or something similar, but that is not really going to matter within usable limits of this design. Good bypassing of the power supplies is a good idea, though.
Two 15V power supplies are in series, so that he gets +-15V against the common pin for powering the opamps. It is an overkill, IMO - he could have used +-9V or even a single supply and rail-to-rail opamps and saved the extra circuitry. He mentions that possibility, though. The Arduino is connected only to one of the power supplies, so it gets only 15V, not 30.
I need to assimilate that …
Naw, not really. There is no point in simulating the power supply, that is just standard datasheet circuit. Build it up on a breadboard and measure it to make sure you understand it, though!
I’m interested in simulating the whole circuit in order to learn how to use this kind of solver (it is a professional distortion :P) and maybe to verify I’m right in my analyses
Thanks for your feedbacks and advices
Paul