Author Topic: [WorkInProgress] LM324/358 based, 0-20V, 0-1A, cc/cv PSU  (Read 9185 times)

0 Members and 1 Guest are viewing this topic.

Offline debininjaTopic starter

  • Contributor
  • Posts: 41
  • Country: us
Re: [WorkInProgress] LM324 based, single supply, 0-20V, 0-1A, cc/cv PSU
« Reply #25 on: June 25, 2017, 04:40:55 pm »
 :wtf: How is my answer childish? Of course there's a lot of theory behind it (which will take time and experience to learn), but making a basic PSU with some tradeoffs is fine to me. To you, it might not, and that's fine--to each their own. All I'm saying is that I'm fine with keeping expectations low when building something as a beginner, considering the cost of the build and the parts used, and my current knowledge.
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 3233
  • Country: gb
Re: [WorkInProgress] LM324 based, single supply, 0-20V, 0-1A, cc/cv PSU
« Reply #26 on: June 25, 2017, 04:52:39 pm »
People expect too much from people without an extensive engineering background. The massive income that comes from this sort of profession tends to spoil many,

 :-DD I hope you aren't considering a career in electronics based purely on the notion that you will have a massive income, as you are likely to be quite disappointed.

People are simply pointing out flaws with the circuit.  No one knows if you are already aware of these flaws or whether you are happy to accept the consequences of them.  Your responses make you look overly sensitive/defensive which is strange as it's not even your circuit design.
 

Offline debininjaTopic starter

  • Contributor
  • Posts: 41
  • Country: us
Re: [WorkInProgress] LM324/358 based, 0-20V, 0-1A, cc/cv PSU
« Reply #27 on: June 25, 2017, 05:06:04 pm »
Fair enough. Not being overly defensive, but I'm just saying, based on what I saw from the guy's data, it seems pretty good (to me). :-/O
I just swapped out some parts from his schematic to suit my needs.
 

Offline ZeTeX

  • Frequent Contributor
  • **
  • !
  • Posts: 610
  • Country: il
  • When in doubt, add more flux.
Re: [WorkInProgress] LM324 based, single supply, 0-20V, 0-1A, cc/cv PSU
« Reply #28 on: June 25, 2017, 05:11:57 pm »
:wtf: How is my answer childish? Of course there's a lot of theory behind it (which will take time and experience to learn), but making a basic PSU with some tradeoffs is fine to me. To you, it might not, and that's fine--to each their own. All I'm saying is that I'm fine with keeping expectations low when building something as a beginner, considering the cost of the build and the parts used, and my current knowledge.
Great,
Kleinstein attached a good simple PSU circuit, trust me, it's better than the most PSUs schematics on the web, and it's cheap, it uses simple part, and not that complicated to understand. he also made it stable which is the hardest, so probably if you just copy and paste it into a PCB with small modifications it would work. Why not use it? trust me he spent time on this even though it looks like it took 5 minutes for him, and you go and search for another schematic without first saying why not that is probably 4 times worse and complicated than his.

A top of the line PSU with ultra-low noise and ripple, UV precision, and ultra-fast performance don't cost much in terms of parts, they cost much in terms of time and knowledge.
the performance of most Bench PSU's compare to many things in life is not determined by money, because inside in general if you remove the transformer and the case they are really cheap, probably cost less than 10$ in terms of components.
So if you can improve the performance, just by adding a resistor there and there or just changing the designing without increasing cost, why not?


 
The following users thanked this post: debininja

Offline debininjaTopic starter

  • Contributor
  • Posts: 41
  • Country: us
Re: [WorkInProgress] LM324/358 based, 0-20V, 0-1A, cc/cv PSU
« Reply #29 on: June 25, 2017, 05:22:15 pm »
Thanks for getting it through my head. |O I will do that. Let's call for a peace treaty now.
I will update the post once I finish transferring it to KiCAD.  :-+
 
The following users thanked this post: ZeTeX

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14073
  • Country: de
Re: [WorkInProgress] LM324 based, single supply, 0-20V, 0-1A, cc/cv PSU
« Reply #30 on: June 25, 2017, 07:02:11 pm »
...
Here's another one I found, and it seems to fit the bill for me perfectly.
According to one of the commenters:

That circuit is not that bad. There are kits for a similar circuit without the 33 V regulator and different negative supply. The main trouble with these kits is they claim to much power and suggest to high a voltage so they violate the ratings of the OPs and transistors. But after scaling down to 20 V and 1 A they might be perfectly OK. The difference to many other circuits is that the current regulation turn down the set voltage instead of acting after the regulator. With rather slow OPs like the LM324 this might actually even be an advantage. for faster reaction.

A time of 120 µs until the current limit sets in, sounds like long, but the charge flowing before this is not that much. It's 2 A times 120 µs, thus about 250 µAs or the charge of a 100 µF capacitor dropping 2.5 V.  Depending on the excess current (e.g. closer to a dead short) the delay can get shorter (down to about 25 µs for a dead short) or longer. It is just the limitation of the CC mode used OP. With an LM324 it would get even slower and might be still acceptable if the peak current is limited. This additional limit of the peak current is probably a good idea. The circuit shown is also just a simple close to minimal version. Except for the compensation in the CV mode (which is made to reduce / avoid voltage overshoot), this is more or less the basic lab supply circuit with an emitter follower output stage. Using a faster OP for the CC mode is likely a good idea, though there are limits too: too fast and it might oscillate, depending on parasitics.
 
The following users thanked this post: ZeTeX, debininja

Offline ZeTeX

  • Frequent Contributor
  • **
  • !
  • Posts: 610
  • Country: il
  • When in doubt, add more flux.
Re: [WorkInProgress] LM324 based, single supply, 0-20V, 0-1A, cc/cv PSU
« Reply #31 on: June 25, 2017, 08:54:01 pm »
...
Here's another one I found, and it seems to fit the bill for me perfectly.
According to one of the commenters:

That circuit is not that bad. There are kits for a similar circuit without the 33 V regulator and different negative supply. The main trouble with these kits is they claim to much power and suggest to high a voltage so they violate the ratings of the OPs and transistors. But after scaling down to 20 V and 1A they might be perfectly OK. The difference to many other circuits is that the current regulation turn down the set voltage instead of acting after the regulator. With rather slow OPs like the LM324, this might actually even be an advantage. for faster reaction.

A time of 120 µs until the current limit sets in, sounds like long, but the charge flowing before this is not that much. It's 2 A times 120 µs, thus about 250 µAs or the charge of a 100 µF capacitor dropping 2.5 V.  Depending on the excess current (e.g. closer to a dead short) the delay can get shorter (down to about 25 µs for a dead short) or longer. It is just the limitation of the CC mode used OP. With an LM324 it would get even slower and might be still acceptable if the peak current is limited. This additional limit of the peak current is probably a good idea. The circuit shown is also just a simple close to minimal version. Except for the compensation in the CV mode (which is made to reduce/avoid voltage overshoot), this is more or less the basic lab supply circuit with an emitter follower output stage. Using a faster OP for the CC mode is likely a good idea, though there are limits too: too fast and it might oscillate, depending on parasitics.
So the classical floating regulator and the emitter follower output stage are very similar, the difference is minimal in performance.
if you could get a separate transformer winding, is it recommended to go for the classical floating regulator or the emitter follower regulator? it seems like although you get slightly better performance for the floating regulator, it's more hassle and more complicated.
 

Offline debininjaTopic starter

  • Contributor
  • Posts: 41
  • Country: us
Re: [WorkInProgress] LM324/358 based, 0-20V, 0-1A, cc/cv PSU
« Reply #32 on: June 26, 2017, 05:29:40 pm »
Hi all. Changed up klein's schematic up a bit. End result is the same, works in the same way, same performance.

* Changed up the gain to be 8.05 now, so max vout is 8.05 * 2.5V ref from TL431, so 20.125V (set via potentiometer).
* Made some minor changes to some of the resistors. (22k resistors instead of 20k--that's what came with the resistors kit I have).
* Added in an output protection diode (reverse biased, probably throw in some random 1n400x diode).
* Changed up wire parasitic inductance to 100nH (10cm of wires, at worst)

Question:

* I don't get is the purpose of D1 (at U1). Is that supposed to be a LED for constant voltage mode indicator?
(I checked out the current limit mode by adjusting the input voltage at U2 non-inverting amp. D6 sinks about 3.6mA via the LED when CC limit is reached, so that's working fine).

Attached the modified asc.
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14073
  • Country: de
Re: [WorkInProgress] LM324/358 based, 0-20V, 0-1A, cc/cv PSU
« Reply #33 on: June 26, 2017, 07:22:37 pm »
The Diode D1 is there for the analog minimum function. So the lower voltage from U1 and U2 sets the actual output voltage. In some circuits D1 could be replaced with a resistor and R1 removed.

There are a few circuits that use LEDs at this position. However the reverse voltage for most LEDs is limited and the higher forward drop can prevent the voltage to go all the way to zero. So the better way is to use a kind of comparator and compare the outputs of the 2 OPs. This could be the 4 th OP in the LM324.

Another point to think about might be replacing R1 with a kind of current mirror, that can also be used as an output enable / disable and prevent voltage spikes during turn on. A limited, constant current would also act as a secondary fast current limit to prevent current spikes.

Having Q1 an 2N3904 might be just acceptable for an 1 A output current and 25 V. Normally this transistor might need to be a little higher power, maybe like an BD139. So the 2N2222 was already an the small side.

Remember that the circuit is simulated for fast OPs than the LM324. At least for U2 the speed of the OP is relevant. So using a LM324 would make current limiting even slower. For a 1 A current range the shunt should be a little larger - this also helps a little with a faster current limit.

R7 = 1 Ohms can be removed. It is a kind of left over, just in case a significant resistor is needed for stability.
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 6924
  • Country: ca
Re: [WorkInProgress] LM324/358 based, 0-20V, 0-1A, cc/cv PSU
« Reply #34 on: June 26, 2017, 07:34:07 pm »
OP you are forgetting there are situations where you'll see a positive output voltage, that is above your V setpoint (or also in CC mode). It can be due to power outage, charging batteries, loads with huge capacitance. It can happen for msec to minutes.

Think you are charging a 12V battery. The output is around 12V and if the supply is in CC mode, U2 has it's output pulled down to back off current.
Unfortunately, current can backflow through D5, LED D6, U2 and somebody gets damaged.

Same scenario in CV mode, your battery is 12V and you either dial it down a bit, switch off power etc .
Voltage-correction op-amp U1 will be trying to keep the output voltage lower than it is, and backfeed occurs. D1, D3 or the reverse-biased E-B junctions of Q1, Q2 conduct causing over-current for U1 and it blows.

Number one for me is reliability in a design and performance is second.

For your LTSpice circuit, some of my opinion:
I see no use for D3 or D5.
I think C8 is a bad idea, why slow down the voltage correction so much there? 1uF/220R
I recommend a protection diode across +Vout and V1 (C-E on Q2)
Q1 is too light duty, unless you add current-limiting resistor for it. Calculate Pdiss at say 10V 1A output.
D4 limits differential overload on U4 to 0.7V, I've seen this diode with older LM741 designs. Again, where your commanded voltage setpoint is above your actual output voltage, U1 can saturate and take a very long time to recover and you get big overshoot.

 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14073
  • Country: de
Re: [WorkInProgress] LM324/358 based, 0-20V, 0-1A, cc/cv PSU
« Reply #35 on: June 26, 2017, 07:59:41 pm »
D3 and D5 have the purpose to act as a kind of simple down programmer. So exactly in the situations that there is a higher voltage at the output, like a charge capacitor. Nearly all OPs have an internal current limit. So the current through these diodes is limited by the OP and nothing bad happens. There is no absolute need for both D3 and D5, one of them is enough. However without them one might exceed the reverse BE rating of Q1. The diodes can also help to avoid some modes of oscillation with a highly capacitive load and low current. D3 can slightly speed up the response in a few cases, when fast turning off Q2 is needed.

C8 is only filtering the set voltage. So it only limits the speed how fast one can change the set voltage. It is there to take out the frequency response of the differential amplifier around U3.

D4 is limiting the input difference for OP one might consider a second one for the other direction too. It is not only for protecting the OP, it is also about limiting the charge in C4 and discharging C7 when the current limit sets in. This helps to prevent voltage overshoot one a fast CC to CV transition. It kind of slightly slows down the recovery of the voltage, but it usually more acceptable for the voltage comes up slow instead of an overshoot.
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 6924
  • Country: ca
Re: [WorkInProgress] LM324/358 based, 0-20V, 0-1A, cc/cv PSU
« Reply #36 on: June 26, 2017, 08:13:08 pm »
Hmmm I'm not sure why then I change exploded LED's so much, lol. They pop across the lab.
Example in this B&K 1670/1671 lab PSU, kinda similar to OP's circuit but has floating aux. supply for the op-amps. (sch missing a resistor for base drive to TIP31 I think).
 

Offline ZeTeX

  • Frequent Contributor
  • **
  • !
  • Posts: 610
  • Country: il
  • When in doubt, add more flux.
Re: [WorkInProgress] LM324/358 based, 0-20V, 0-1A, cc/cv PSU
« Reply #37 on: June 26, 2017, 09:56:55 pm »
I have redrawn @Kleinstein schematic so it will be more readable.  :D
I have added an overvoltage protection crowbar circuit, but I think this is not the best solution because you will have to have a fuse to blow up and replace, so I rather find something else.
I have changed the Darlington transistors and this caused some oscillation, so I had to increase C5, the capacitor across the CV op amp.
I also added a constant-current led indicator.
You might want to replace R2 with a current source, I have seen people do that although I don't see any difference between just a resistor and a current source.
 
The following users thanked this post: debininja, Neomys Sapiens

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: [WorkInProgress] LM324/358 based, 0-20V, 0-1A, cc/cv PSU
« Reply #38 on: June 27, 2017, 09:29:47 am »
R3 is much smaller than necessary, D2 and D3, and probably R1, are unnecessary, D8-R5-M1 are a crude shunt regulator not a crowbar (there's no latching, and certainly not an unlimited current draw -- use an SCR), U3 and associated components look superfluous (why not do the same thing with U2?), and C7 looks suspiciously large, or not needed.  C1, C2 and C3 are superfluous in a simulation of ideal caps, and the structured values are paranoia anyway; a single part is almost always better.

C4 and C5 aren't helping with integrator windup and it's not clear how well compensated this will be, but they can be adjusted.  C6+R7 is well placed, and the same for U1 might be good.

Just small changes, doing better. :)

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: ZeTeX

Offline not1xor1

  • Frequent Contributor
  • **
  • Posts: 716
  • Country: it
Re: [WorkInProgress] LM324/358 based, 0-20V, 0-1A, cc/cv PSU
« Reply #39 on: June 27, 2017, 11:40:36 am »
U3 and associated components look superfluous (why not do the same thing with U2?), and C7 looks suspiciously large, or not needed.

there is an error in ZeTeX's schematic
C7 should be connected to the negative output not to ground
the purpose of U3 in Kleinstein's schematic is to tie the negative pole of the voltage reference to the negative output rather than to ground

without U3, the reference voltage depends on the load current too
 
The following users thanked this post: ZeTeX

Offline ZeTeX

  • Frequent Contributor
  • **
  • !
  • Posts: 610
  • Country: il
  • When in doubt, add more flux.
Re: [WorkInProgress] LM324/358 based, 0-20V, 0-1A, cc/cv PSU
« Reply #40 on: June 27, 2017, 05:07:08 pm »
U3 and associated components look superfluous (why not do the same thing with U2?), and C7 looks suspiciously large, or not needed.

there is an error in ZeTeX's schematic
C7 should be connected to the negative output not to ground
the purpose of U3 in Kleinstein's schematic is to tie the negative pole of the voltage reference to the negative output rather than to ground

without U3, the reference voltage depends on the load current too.
Thanks, fixed it.  :)
R3 is much smaller than necessary, D2 and D3, and probably R1, are unnecessary, D8-R5-M1 are a crude shunt regulator not a crowbar (there's no latching, and certainly not an unlimited current draw -- use an SCR), U3 and associated components look superfluous (why not do the same thing with U2?), and C7 looks suspiciously large, or not needed.  C1, C2 and C3 are superfluous in a simulation of ideal caps, and the structured values are paranoia anyway; a single part is almost always better.

C4 and C5 aren't helping with integrator windup and it's not clear how well compensated this will be, but they can be adjusted.  C6+R7 is well placed, and the same for U1 might be good.

Just small changes, doing better. :)

Tim
- I increased R3 to 150ohms, usually, I see people using about 100-150ohms, saving a few mA :P
- About D2, D3, R1 - for D2, and D3, Kleinstein explained their reason, for R1 you might be right, but as kind of protection, I think I rather leave it there: "D3 and D5 have the purpose to act as a kind of simple down programmer. So exactly in the situations that there is a higher voltage at the output, like a charge capacitor. Nearly all OPs have an internal current limit. So the current through these diodes is limited by the OP and nothing bad happens. There is no absolute need for both D3 and D5, one of them is enough. However without them one might exceed the reverse BE rating of Q1. The diodes can also help to avoid some modes of oscillation with a highly capacitive load and low current. D3 can slightly speed up the response in a few cases, when fast turning off Q2 is needed."
- I'd figure a proper overvoltage protection later
- C1 C2 C3 have different ESR, it's important for simulation and real life.
- C4 & C5 are for stability, I don't want to play with them too much, I just trust Kleinstein that he didn't just throw them in there with a random value.
Thanks for your help! :D
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14073
  • Country: de
Re: [WorkInProgress] LM324/358 based, 0-20V, 0-1A, cc/cv PSU
« Reply #41 on: June 27, 2017, 05:38:03 pm »
C3 is just a leftover from simulating a capacitive load. So no need for it in the circuit. Having two caps at the output kind of helps as an electrolytic cap has some ESR than helps with stability.

C5 is there just in case the OP is too fast. It might not be needed, but it would be a good idea to have the option, just in case U2 oscillates or as another point to optimize.  So one might leave it out (unpopulated) initially.

I tried a compensation with a series R + C instead of C4 too, but it turned out to be best with R = 0. Also having the compensation from the other side of the diode did not make a big difference, as much of the speed is set by teh GBW of U1 anyway. It is only with a fast OP that the compensation here is really needed.
 

Offline not1xor1

  • Frequent Contributor
  • **
  • Posts: 716
  • Country: it
Re: [WorkInProgress] LM324/358 based, 0-20V, 0-1A, cc/cv PSU
« Reply #42 on: June 29, 2017, 04:12:20 am »
- I increased R3 to 150ohms, usually, I see people using about 100-150ohms, saving a few mA :P
- About D2, D3, R1 - for D2, and D3, Kleinstein explained their reason, for R1 you might be right, but as kind of protection, I think I rather leave it there: "D3 and D5 have the purpose to act as a kind of simple down programmer. So exactly in the situations that there is a higher voltage at the output, like a charge capacitor. Nearly all OPs have an internal current limit. So the current through these diodes is limited by the OP and nothing bad happens. There is no absolute need for both D3 and D5, one of them is enough. However without them one might exceed the reverse BE rating of Q1. The diodes can also help to avoid some modes of oscillation with a highly capacitive load and low current. D3 can slightly speed up the response in a few cases, when fast turning off Q2 is needed."
- I'd figure a proper overvoltage protection later
- C1 C2 C3 have different ESR, it's important for simulation and real life.
- C4 & C5 are for stability, I don't want to play with them too much, I just trust Kleinstein that he didn't just throw them in there with a random value.
Thanks for your help! :D
I modified all capacitors by choosing real capacitor models of same value, 50V rated, and checked for 1s transient 10mA/.9A pulse load.

The circuit seems stable, but as soon as I change either transistors with more common ones, like BD139 and TIP41C, both or just one of them, the circuit oscillates wildly (LTspice XVII 2017-02-06 build).

I'll later check the phase margin with stable loads of 10mA and .9A and or when in CC mode.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: [WorkInProgress] LM324/358 based, 0-20V, 0-1A, cc/cv PSU
« Reply #43 on: June 29, 2017, 12:42:30 pm »
That's one reason for the larger value B-E resistor: less bias current gives less bandwidth in the output transistor(s), and therefore less prone to oscillation.  (We're talking ~10MHz of bandwidth, well in excess of what the op-amp can deliver -- the transistors don't need to be that fast, and it can hurt to run them that fast, for this reason.)

Note that SPICE models often fail to model transistor phase shifts correctly:



I built the circuit on the left, which oscillates.  The simulation shows it is stable.  I had to modify it, to the circuit on the right, to get the simulation to oscillate.  Unfortunately, the parasitic components are about 10x larger than what is realistic for the circuit.  I am left to conclude there is some additional phase shift or delay in the real transistor(s), which is not present in the provided models (which I obtained from ON Semi, IIRC), or which SPICE cannot model at all*.

*SPICE contains two infamous hacks to approximate real phenomena:
1. A PN junction, in forward bias, stores charge.  This charge must be built up before the junction "turns on" (forward recovery), and depleted before it "turns off" (reverse recovery).  SPICE models this as a dependent capacitor with a dependent current sink, so that the V(I) curve is logarithmic (following the Shockley equation -- the static diode curve), and the stored charge behaves like a battery (charge is ~exponential with voltage).  A dependent capacitor cannot model forward recovery, because it's always capacitive, and the capacitance increases as it's being charged, whereas forward recovery is more like a resistor that decreases as charge is stored.  Reverse recovery is modeled reasonably well at least, but diode losses are usually underestimated (recovery loss is significant in real converters).  But even more tricky is where charge is stored: it's not stored uniformly in the junction: it diffuses through the semiconductor.  You can apply a short forward-bias pulse, then hold reverse bias, on a diode; if it's the right kind of diode (the doping profile is involved, as well), the reverse recovery transient will not move softly, but explode open in a fraction of a nanosecond: this is called drift step recovery (aka Grekhov diode).  The short forward-bias pulse dumped a wad of charge into the junction, which happened to diffuse out all at the same time, then suddenly, poof, the junction goes open-circuit!  This sort of thing is, in principle at least, impossible to model accurately in SPICE, because it's a bulk semiconductor thing.

2. Transmission lines.  This is a much less lengthy description, okay...  Transmission lines: you put waves in one end, wait a delay, and they come out the other end (and reflect back and forth, depending on impedances).  This is a hack because SPICE uses variable timesteps, which means what waves entered the TL are propagating at a nonuniform rate (that is, not at a constant distance per timestep).  It's not as simple as spinning around a circular buffer.  Most simulators simply restrict the maximum timestep, so the steps become uniform inside the TL -- which means no shortcuts can be taken during slowly changing times, and the simulation crawls.  (Or it may be more complex than that, I haven't looked.  It 'feels' like this is the case.  It may simply be that it's doing extra math to resolve the variable timesteps, and that increases the overhead significantly.)

In any case, these are components which cannot fundamentally be realized in SPICE: SPICE is the domain of matrices, where there is no speed of light (or, if there is, it's infinite), where transfer functions have no real delay, only poles and zeroes.  You can make approximations, but only that; an RLC transmission line approximation needs dozens or hundreds of stages, or the SPICE transmission line hack requires lots of internal memory** and runs slow.  One must keep these limitations in mind, when one wishes to construct some more unusual simulations (you know, like things in reality). ;)

**Okay, to be fair, this was back when you were lucky to run SPICE on a mainframe with 64k+ of core.  Now who here remembers what "mainframe", "64k", and "core" were :-DD

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14073
  • Country: de
Re: [WorkInProgress] LM324/358 based, 0-20V, 0-1A, cc/cv PSU
« Reply #44 on: June 30, 2017, 09:10:29 am »
The circuit shown needs some (usually just the parasitic) inductance of the shunt to get a stable current regulation. There is a reason this inductance was included in the circuit. Besides some positive effect on current regulation, parasitic inductance can make the voltage regulation more difficult.

Besides limitations in the models, the parasitic inductance of low impedance parts is a real problem with such a circuit. So even if it works in the simulation, it might oscillate real life, especially if the circuit is fast. The low impedance intended for a voltage regulator makes parasitic inductance important - it is a little complementary to the difficulties with parasitic capacitance in high impedance circuits.

Some of the transistor models are supposed to be very good, but some are also simplified. At least the reverse recovery is modeled rather good (though usually not much differentiation between hard and soft recovery diodes). With transistors there is not much of a forward recovery. This is more like an effect important in PIN diodes (especially high voltage high frequency rectifiers) - normal PN diodes are rather well described with the capacitor model, voltage overshoot is due to lead inductance and not due to the semiconductor part.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf