Due to reasons and curiosity, I want to know if I can recreate this circuit using NPN transistors instead of PNP transistors. I know functionally that they can be the same if we invert the polarities; however, I am not sure how to factor the L78 (voltage regulator) into the equation.
1\$\begingroup\$ It's not a very useful circuit given that there are adjustable voltage regulators out there that will outperform your circuit by some margin. \$\endgroup\$– Andy akaDec 12, 2022 at 12:44
4\$\begingroup\$ I think the NPN version would work with a negative regulator (ie LM79xx). \$\endgroup\$– ErikRDec 12, 2022 at 13:14
\$\begingroup\$ it is useful but the question guys is "how to factor the L78xx?" \$\endgroup\$– Tony Stewart EE75Dec 12, 2022 at 15:37
\$\begingroup\$ Wingmore, you've not responded to anyone. Probably should consider it. Also, why this particular circuit? What's the adjustable range you want for the output, what output current compliance, and what's the input source details? Or is this just an abstract question of "take this behavioral circuit, for whatever it does, and change it to use NPN instead to have the same behavior?" \$\endgroup\$– jonkDec 12, 2022 at 18:10
1\$\begingroup\$ Picked this circuit because I have the parts lying around and thought may as well try utilise them. Honestly it's more an abstract question but I'll provide answers to the other questions. Adjustable range isnt that important so 5-12V is good. Going to drive some pc fans so current doesnt need to be super high. Input's going to be 12V. \$\endgroup\$– WingmoreDec 12, 2022 at 22:22
Vxx is the fixed Voltage of the LDO and this makes it an ADJ LDO with a pot or fixed R ratio. ( but would you regulate power with an NPN voltage or current from a uC or thermistor?)
Also, you don't need a mirror. Just use one NPN transistor and fixed resistors to make a Boost LDO VCVS (voltage-controlled voltage-source).
The old LM317 is a 1.25V boost ADJ Regulator. This method of R ratios makes an LM78XX boost ADJ Regulator so the lower voltage LM317 types are more useful and have better regulation, but if all you have is a 5V LDO and need 12V, that'll work. Check out the link for more options.
What's all this other Stuff?
The question instead could be "How do you complement any design" for any application? (lol). Or how can you make any fixed LDO into a VCVS or use complementary LDOs and make an audio current source Amp or how do you make the LDO into a thermal Fan controller?
In this case, take a mirror of a current mirror.
The minor issue is that only the emitter currents are mirrored but will still work with a slightly different ratio due to hFE+1.
It is common to see linear ICs with both negative and positive current mirrors.
You can use a 3 terminal ADJ. REG like the LM317 or these with a pot and reply on the constant voltage between ADJ and OUT.
But how can you use an NPN to control an ADJ. REG. to drive a FAN with a negative thermal feedback using a 10k Thermistor and a few fixed resistors + NPN and maybe a gain pot? Maybe you can spend a day with a spreadsheet and compute the R values for a Thermally Controlled Speed Fan from Vcc to <3V for 60 to 50 'C and silent below 40'C using epoxy mounted to a SMPS core.
- This could use the mirror version with an NPN controlling a fixed LDO just as in your practical question. Bravo.
I did the same as above about 20 yrs go for a 19" rack for a 185K$ design contract to do for an 8 Port T1 phone supply for Lucent/Avaya Telcom using twin 1U high fans running off 48V for mass production.
1\$\begingroup\$ The LM317 is a linear regulator with a 2V typical drop out. It is not an LDO, which stands for "Low Drop Out". I know Tony Stewart that you disagree. This comment is for others who might wonder whether it is standard terminology to refer to an LM317 as an LDO. \$\endgroup\$ Dec 12, 2022 at 15:51
1\$\begingroup\$ @TonyStewartEE75 LDO isn't just "low drop out." It's a design topology, which is dissimilar to traditional non-LDO topologies. Because of the LDO topology, what you write about being more concerned about instabilities with respect to ESR is true. But even that varies with the specific details of the LDO design. \$\endgroup\$ Dec 12, 2022 at 18:05
1\$\begingroup\$ @TonyStewartEE75 There are adjustable LDO topology ICs out there with truly low-drop out and where they actually use an LDO topology. They exist. Dropouts definitely under a volt and adjustable and using the p-type approach to achieve the LDO topology. I need only one example to prove that. And I've done so. You'd need an exhaustive examination of all adjustables to make your point. Much harder to do and impossible now that I've demonstrated just one example. Some is the logical opposite of all. \$\endgroup\$ Dec 12, 2022 at 18:07
1\$\begingroup\$ Just read the pages. It's all there. I don't need to add anything to them and I'm not going to debate what they write. I think you are arguing at cross-purposes due to a misunderstanding of what an LDO actually is and what you think it is. So we won't go anywhere until you read the link I provided which defines an LDO as opposed to other topologies. You think the definition is about something else. Which it is isn't. No point debating that, though. Anyway, @MathKeepsMeBusy is correct so far as I understood their writing. And like Math, I'll just leave it there. \$\endgroup\$ Dec 12, 2022 at 18:59
1\$\begingroup\$ LDO's are defined by the pass-topology and the consequent stability. I've provided a page to read on that topic. The achieved drop-out doesn't define LDO vs not-LDO. The pass-topology and associated stability does. I'll leave it there as all we'll do is argue at cross-purposes. \$\endgroup\$ Dec 12, 2022 at 19:34