I made this wonky hodgepodge of kits. I took a buck converter kit and a linear regulator kit that came with a switching vac to dc supply instead of a transformer.

I took the voltmeter from the one kit and wanted to read whichever regulator output was active , and didn't want a switch, and didn't have a multipole switch either, which could have probably worked fine. With a more-pole switch of course I would maybe just have a pole handle switching the voltmeter monitoring wire. But, i thought it would be cooler to just have some diodes, except the voltmeter isn't really easy for me to calibrate so now it's just always .6v low. No worries, i know that, but I wanted to know if there were some easy fixes for this. I can't just have the outputs connected because then the voltage bleeds all over the other circuit, even though it's off, and I don't want a digital circuit becuase while that would be cool, it would be active and technically not efficient, were I to really need this kit I would just order a multipole switch. Instead, I'm wondering if there is any way to overcome this voltage drop without an electronic or physical switch. Some kind of more passive solution not dissimilar to using some diodes.

Surely there's not,but I thought I would ask!

 

How do you plan to use these two supplies? You can't just connect them to your load because then you have the same problem as with your meter. If you use the same approach of using two diodes to select whichever one is at the higher voltage, then your meter reading will correspond pretty closely to what the load is actually seeing.

You can also use lower-voltage drop diodes such as Schottky diodes.

 

How do you plan to use these two supplies? You can't just connect them to your load because then you have the same problem as with your meter. If you use the same approach of using two diodes to select whichever one is at the higher voltage, then your meter reading will correspond pretty closely to what the load is actually seeing.

You can also use lower-voltage drop diodes such as Schottky diodes.

each has it's own output, I just move the alligator clip leads from one to the other. That's not a terrible idea to have them each fed via diodes, but the pcbs have nice terminals and actually I store it with the leads not plugged in, so i just install them on which side I want to use,,, of course the only thing I've used it for is to use the oscilloscope to see the difference in noise on each output. Interestingly I do seem to lose the main flyback switching noise at the buck converter, I only really see noise that I am sure is the buck converter because it's different than the one coming out of the buck. I also do a pretty good job of losing the noise out of the linear regulator, so I feel it is working to be less noisy using it. Not that i have a use case, or plan to use it like that, it was just an experiment in the first place.

I do like that plan though, just have the voltage drop on the output. If I decide to mount lugs where everything runs to a single output set , I might do that.

 

Thecircuit shown with two power supplies and diode isolation will work, BUT it will not be quite right as to the voltage is is showing. The reason is that it has series diodes in the meter connection, and DIODES ARE NOT AT ALL LINEAR. That is intrinsic to being diodes. As a diode shifts from being an open circuit to being conductive, the voltage drop varies greatly.

 

Question I have is: since they both are adjustable, what voltages are you wanting? The buck and the linear regulated circuits do the same thing. Another question: why do you want or need two supplies into a single load?

 

Question I have is: since they both are adjustable, what voltages are you wanting? The buck and the linear regulated circuits do the same thing. Another question: why do you want or need two supplies into a single load?

you only use one or the other. there's a switch to choose which one is connected to the 12v supply. there is no equipment it's for, it's just to put it together like a hobby. it's just a little bench power supply for kits. also, it's for the purpose of using the oscilloscope to see the difference in the outputs. For example, i could see the switching noise on the 12v supply, but I could not see the 12v flyback switching noise on either output. on the buck output I saw noise that came from the buck but it was not mixed with the flyback noise, which did appear to be gone, and on the linear, there was noise but more like just ambient emf noise picked up from the power lines and wifi and such, which incidentally did appear to be stronger on the linear supply than on the buck supply, where the buck switching noise appeared quite pronounced, but as a cleaner signal.

 

For example, i could see the switching noise on the 12v supply, but I could not see the 12v flyback switching noise on either output. on the buck output I saw noise that came from the buck but it was not mixed with the flyback noise, which did appear to be gone, and on the linear, there was noise but more like just ambient emf noise picked up from the power lines and wifi and such, which incidentally did appear to be stronger on the linear supply than on the buck supply, where the buck switching noise appeared quite pronounced, but as a cleaner signal.

So why the need for two different supplies? Seems the Buck is where you want to stop. Instead of directing two supplies into one just add some filtration. Caps and inductors would serve better at cleaning up the last bit of noise. I would think.

 

Certainly it is unclear just what the goal is! Post one shows a "basic 12volt" DC supply feeding two different regulators. And certainly calling the one a converter is a bit confusing. That device called a "converter" could be a switch-mode regulator able to supply a higher output voltage than the input voltage.
One thing to be aware of when looking for noise with an oscilloscope is that BOTH SIDES of the connection must be tiedto the circuit of interest. That means that the common side as well as te "signal" side must be switched. That "perfect ground" used for a reference is a myth!
The common side of a circuit is seldom perfectly quiet. And the scope will always display the sum of all the voltages present between the "signal" input and the "signal common" input. And the circuit attached to post #1 does not even show the "common return" sideof the supply output. That leaves a lot open to guesses and assumptions.

 

To all:

So, I got some more similar voltmeters from amazon, and these have a calibration adjustment, which will solve the problem, since I can just calibrate the meter 0.6v higher than the reading.


i was under the impression both the buck and boost were called "converters" , while the linear regulators never were called converters, so I thought using the word converter would be an easy way to distinguish between the buck converter and the linear regulator.

The goal is just to make a device that can be switched between linear and buck, so that if linear regulation is desired, it's available, but if not then buck is more efficient and therefore will do things like use less energy, be less warm, or if there are other differences that might be beneficial then these too would be available. But honestly, I didn't have this need, it's just a toy kit experiment for the sake of putting toy kit experiments together. By now I have a small army of some kind of power supply or another, I just kind of like them at the moment and think they're interesting and fun.

Another reason I wanted to build it was to see just how much the noise from the primary 12v supply would be seen on the output of the linear regulator and the buck converter. As I mentioned above, I did try to measure this. I put a 100ohm resistor as a load hoping it would pull enough current to increase the ripple on the capacitors, and then I used the oscilloscope to try to find the switching noise at 3 points. I tested after the 12v supply, and then after each adjustable supply. I found that the 12v supply switching was only visible (in my experiment) just after the 12v supply, and perhaps slightly in the linear output although my confidence in this measurement is low especially as there was a fair amount of high frequency noise on that signal (or at least on the scope) resulting in a fuzzy line. I also found, however, that on the output of the buck converter there was a very clear and distinct signal that was different from the one after the 12v supply, which I supposed was from the buck converter itself. In fact, this signal was so clean and clear that I was absolutely confident that the 12v supply noise was completely removed from the signal. So, there was still a strong switching pulse to be seen, but only the one set.

Anyway, I'm quite pleased with these other little voltmeters from amazon, as they have a small potentiometer which allows me to either calibrate to another voltmeter, or in my case will allow me to calibrate to overcome the voltage drop of the diode separating the two read lines.

 

So, I got some more similar voltmeters from amazon, and these have a calibration adjustment, which will solve the problem, since I can just calibrate the meter 0.6v higher than the reading.

Maybe.

Note that the "0.6 V" often quoted as the forward voltage of a silicon diode is nowhere near a standard or a constant. Depending on things like construction, temperature, and current, the actual value can vary from around 0.4 V to over 1.5 V. You can "calibrate out" the voltage drop for one set of conditions (primarily, part number and operating current), but that calibration will not hold when the current changes.

ak

 

Maybe.

Note that the "0.6 V" often quoted as the forward voltage of a silicon diode is nowhere near a standard or a constant. Depending on things like construction, temperature, and current, the actual value can vary from around 0.4 V to over 1.5 V. You can "calibrate out" the voltage drop for one set of conditions (primarily, part number and operating current), but that calibration will not hold when the current changes.

ak

I think it will not be a terribly big issue here, as this whole rig will not be suitable for much current in the first place. 1A is probably more than I would think is a good idea, it's more aimed at projects that fall under .5A

 

WHY NOT USE AN ACTUAL SWITCH??? And, one more consideration is about common side isolation! MANY of the less expensive, DC powered digital meters simply do not have adequate isolation between thier power supply connections and their voltage input connections, especially on the input supply voltage negative side. An input switch will provide at least a few megohms of isolation.
So there is a scheme that is better than diodes, both for isolation and accuracy. No diode drop at all.