rdpzycho:
there will be times that ‘yes’ the regulators will conflict…
but doing precautions and using caps around them usually work, all the time…
a lot of IC will be doing an internal regulation too (to somewhere around 2.++ volts)…
even industrial and commercial PSU’s do use regulators in series with pre-regulators and other regulators…
it should just be noted that the load of the regulator is also a regulator in which the control should be able to compensate…
I think everybody is missing the point of what I was saying: there is no advantage to him on this board to put the two regulators inline with each other. There are, however, potential disadvantages. We are not talking about every single possible situation. We are talking about this one dude’s PCB.
You still haven’t provided any evidence for the potential disadvantages, or specified exactly what they are, apart from mentioning ‘weird oscillations’. Are both regulators affected, or only one of them, and what causes them?
NleahciM:
I think everybody is missing the point of what I was saying: there is no advantage to him on this board to put the two regulators inline with each other. There are, however, potential disadvantages. We are not talking about every single possible situation. We are talking about this one dude’s PCB.
I am just saying that it’s not wrong to do so…
I, myself, would have tied the 3.3V regulator to the 12V input…
one potential problem that is more critical (than the series loaded regulators) is that the 3.3V regulator is directly connected to a supply for an RF module…their loops must be well localized in the PCB so the RF load doesn’t affect the 3.3V regulator…
leon_heller:
You still haven’t provided any evidence for the potential disadvantages, or specified exactly what they are, apart from mentioning ‘weird oscillations’. Are both regulators affected, or only one of them, and what causes them?
So I know I’ve read in various app-notes or datasheets to not inline vregs - but so now you’re forcing me to actually think about why this advice was given. Damn you for making me think! :evil:
Anyways, the way I picture it - I see multiple control loops driving each other. That’s a recipe for badness. It’d be a much bigger deal with DC/DC converters. I agree that it would almost never be a problem - and have done so in the past on some boards when I didn’t have a choice.
My point here is that the OP does have a choice. There are no advantages to inlining them for him while there are potential disadvantages. Probably some that I’m leaving out that a better analog guy could fill in.
The AME1117’s have an absolute maximum Vin of 12V whereas the 7805’s Vin can go to 35V. I have a bunch of 12V wall-warts laying around that I had planned to use to power the board(s) with which is why I thought it was best to run the LDO off of the 7805. Perhaps I would be better off picking up a smaller supply so that I’m not pushing the limits of the 1117 if I decided to supply it directly.
To continue my original line of thinking, I thought that using the second layer of the board as a ground plane and running all of my traces on the top layer (which I ought to be able to do since there are so few components) would eliminate or reduce to an acceptable level any negative effects of the RF module… but then I saw so many differing opinions on grounding that I begain to have a lot of doubts. That mixed signal document that rdpzycho linked (thanks btw) is the most concise that I’ve seen yet and reinforces the “single plane is best in nearly all cases” therory.
Using the bottom layer as a ground plane is a good approach. I often make RF boards with through-hole parts, with the top layer left unetched with clearance round holes that are not connected to ground.
houta69:
The AME1117’s have an absolute maximum Vin of 12V whereas the 7805’s Vin can go to 35V. I have a bunch of 12V wall-warts laying around that I had planned to use to power the board(s) with which is why I thought it was best to run the LDO off of the 7805. Perhaps I would be better off picking up a smaller supply so that I’m not pushing the limits of the 1117 if I decided to supply it directly.
To continue my original line of thinking, I thought that using the second layer of the board as a ground plane and running all of my traces on the top layer (which I ought to be able to do since there are so few components) would eliminate or reduce to an acceptable level any negative effects of the RF module… but then I saw so many differing opinions on grounding that I begain to have a lot of doubts. That mixed signal document that rdpzycho linked (thanks btw) is the most concise that I’ve seen yet and reinforces the “single plane is best in nearly all cases” therory.
Yep, that’s my bad for not fixing it on the schematic that I uploaded, I used a predefined component and failed to update it, but I did specify the exact component in my 2nd post.
Yep, that’s my bad for not fixing it on the schematic that I uploaded, I used a predefined component and failed to update it, but I did specify the exact component in my 2nd post.
Hmm - don't do that on your schematics - that's just a recipe for problems.
Don’t trust an unregulated 12V wall wart to give 12V. I’ve got one that I’m using that claims 12V and 1000mA. At low current draws it is usually giving me over 15V. So if the 3.3V regulator has a max of 15V, you probably don’t want to hook it up directly to a “12V” wall wart.
Hmm - don’t do that on your schematics - that’s just a recipe for problems.
Hmmm I didn’t expect you to notice that level of detail. Even though I provide my name (which you don’t), you’ve referred to me as “dude” (my fav) and “op”. Find another thread to play in kid.
Hmm - don’t do that on your schematics - that’s just a recipe for problems.
Hmmm I didn’t expect you to notice that level of detail. Even though I provide my name (which you don’t), you’ve referred to me as “dude” (my fav) and “op”. Find another thread to play in kid.
Take a careful look at my user name bro. I'd be worried about you if you couldn't figure out my name. Though I don't really see why that would matter to you... By the way - you do realize what "OP" means, yes? OP = Original Poster. More commonly used in newsgroups, but us forum goers use it as well.
I’m actually trying to help you. Yes I’m a bit harsh - but I find that if you’re nice about it people don’t get it and have to be told the same thing many times. I don’t like repeating myself.
Since you seem to have a problem with me helping you I’ll spare you any further distress and avoid your threads from now on.
khearn:
Don’t trust an unregulated 12V wall wart to give 12V. I’ve got one that I’m using that claims 12V and 1000mA. At low current draws it is usually giving me over 15V. So if the 3.3V regulator has a max of 15V, you probably don’t want to hook it up directly to a “12V” wall wart. Keith
That certainly is true. I would suggest obtaining a lower voltage wall wart (6v or 7.5v), if possible. With a 12v nominal input, that 7805 (and AXE1117 if it is also connected to the wall wart) is going to run warm (some may even say hot). The cooler you keep the regulators, the happier they’ll be.
Regarding the chaining of the voltage regulators, there are many commercial designs that take that approach, and I have done the same on my personal projects. Providing that the recommended bypass capacitors are used, I don’t see what the issue is. However, I would be interested in any references to the contrary.
if the wall plug would give a voltage near 15V (filtered supply only) then it is best to tie the 1117 on the 7805 output, or you would have to resort to a preliminary limiter which would cost more components…
riden, the mechanism of the possible problem within series connected regulators are oscillations on the regulator acting as a load which was not considered on the regulator acting as the supply (which is also sometimes the same problem on digital loads due to its switching nature)…since most oscillations are already well taken care of internally in the regulator loop, adding good bypass and reservoir caps would make it much more highly improbable for the two regulators to oscillate…this problem is magnified on Switching Regulators, I noticed especially on Power Inverters where you load the preregulator with a 50/60Hz sine wave (was my previous job)…
one more thing to consider in series connected regulators (which is often overlooked) is the highest current available within the series…if the regulator acting as the load suddenly required a current at the limit of the first regulator, the first regulator would be giving off a slightly smaller voltage (or sometimes shut down temporarily)…as in the case here, 7805 has more than enough than what is 1117’s peak is…
My apologies to the forum community for my earlier negative post directed at Nleachim. I felt that this persons posts where increasingly becoming more intended to be condescending rather than productive. Regardless of whether this was actual or perceived, I regret allowing myself to become irked and taking a pot shot.
I really do appreciate all of the feedback/insight that I’ve received to my questions here. Thank you
On the topic of multiple regulators, one reason for connecting a lower voltage regulator to a high voltage regulator’s output is thermal management. Remember that a regulator’s biggest drawback is that it dumps more power the greater the difference between input and output voltages. That power is dumped as heat.
As for the groundplane issue, it also depends on whether you need impedance matched traces. It is almost impossible to get controlled impedance lines without some sore of plane.
first - don’t sweat the rude posters. I suspect they have their own set of issues to deal with. Usually, if you ignore them, they will go away. Thus the saying “don’t feed the trolls”.
On the wall wart voltage. I believe the intent is that at the rated current, the WW will deliver the specified voltage. At the specified input voltage, as well. So that 12V WW should be close to 12V with 100mA load (120 ohm). Intent being what it is, you should definitely verify that on a case by case basis. I would also not design so close to a limit. I prefer to have a 20% safety margin or better. You could put a couple of rectifiers in series with the input to drop the voltage a bit. A rectifier isn’t a bad idea as it will prevent polarity reversal - defensive design. or better yet, a bridge would allow either + or - tip WWs to be used.
As for cascaded regulators, I’ve never seen or heard of any cautions against it an certainly have seen it in multiple designs.
rdpzycho:
riden, the mechanism of the possible problem within series connected regulators are oscillations on the regulator acting as a load which was not considered on the regulator acting as the supply (which is also sometimes the same problem on digital loads due to its switching nature)…since most oscillations are already well taken care of internally in the regulator loop, adding good bypass and reservoir caps would make it much more highly improbable for the two regulators to oscillate…
Thanks for your explanation. I am coming from the perspective that the proper capacitors are being used per the data sheets, as they are required to avoid the problems load switching that you mentioned. If these capacitors are omitted, you could get oscillations even without adding the second regulator - in fact, I still have a burn mark on my index finger that I got years ago testing to see if a 7805 was getting too warm. As the regulators referenced are linear regulators, I can’t see that chaining the regulators will present a problem in the application the OP cited.
The wall warts I have appear to be regulated as I measure ~12.1V without any load. So is this a triple whammy to have three regs in series?
I ran a couple tests by hooking up these regulators in series and in parallel on a bread board to see if they’d run hot at all. When in parallel they “felt” slightly warmer than ambient temperature but certainly not hot. When connected in series the 7805 felt about the same while the 1117 felt cold or ambient. In both cases I had minimal loads attached. Also, I always use caps as specified in the components data sheet.
Now while the document does not explicitly show a 3.3V regulator running directly off of a 5V regulator, the fact that they are talking about running the 3.3V regulator from a 5V rail implied to me that one regulator was running off of the other somewhere along the way. Of course, it’s entirely possible that I may have misinterpreted the intended meaning.
Interestingly I just found on page 385 of my copy of “The art of electronics” under the heading of circuit ideas, figure C, a tracking regulator that uses two LM317’s in series. That fact that these are adjustable regulators would not make any difference in terms of possible drawbacks would it? Thanks
As the regulators referenced are linear regulators, I can’t see that chaining the regulators will present a problem in the application the OP cited.