Car Fix DIY Videos

Proyectos LED says:

Hello Martin, I am making my own power supply to, but mine is switch mode,
based in the LM2596.
I think it is much easier and efficient.
the videos are on my channel, but not finished yet.

Cheers,
Gaston.

Léopold Lapaille says:

I was waiting for this one! This is one of the first more serious projects
I hope to do myself once I get decent enough. Thanks!

xanataph says:

I would look very carefully at that voltage reference stage. It does not
look right at all, especially on that multicoloured schematic. The
reference voltage should be taken from the junction ov the zener and the
resistor. In the circuits shown there it looks like you are trying to feed
out ov the gate ov the fet!!

Also a lot ov these older simpler power supply designs did not allow the
voltage to be lower than the zener voltage. Hence why they chose fairly low
ones. Perhaps your fet arrangement is an attempt to improve on this? But
either way I would give the original circuit a good scrutiny around this
area, as like I said that hook up around those components does not look
correct.

kalhana1 says:

Hi Martin, thanks for the video, great work! I would usually start with an
LTSPICE simulation when it comes to precision analog designs before going
to breadboard. Yes you will spend some time, but it can save you many hours
of head scratching later. And it also provides you with a point of
reference for test point voltages etc (in the absence of a working HW
setup). You may decide that you only need to model a certain section of the
circuit that’s most unknown/critical. Most manufacturers provide models for
their components, if not you can use an equivalent component model.
Sometimes if you’re confident with the simulation, you can skip breadboard
and jump straight to PCB CAD and it works 1st time. The time spent learning
something like LTSPICE is well worth it.

In addition, SPICE is a great way to learn and understand exactly how the
circuit works. I know a few professionals that don’t do simulation before
PCB design and end up spending days debugging and bodging and then having
to do a 2nd PCB rev.

Nimajamin says:

Could you perhaps use a ten-turn wire wound pot get the precision control
you mentioned at the beginning? :)

TubiCal says:

Ok this is all very basic, but you still have to go a long
road to your final PS….

First is: you need to redraw the whole thing as like it is it is very hard
to read…
the better way is to put the power transistor in the upper positive rail
with collector orientated to the left and emitter right. This is the normal
orientation for this kind of schematics.

Next is:
please do not mess with that zener diode (as it is acual a zener on the
schematics) As below about 6 volts they got a neg. Temperatur coefficent
and the series diode got that opposite TempCo so together they cancel it
out nearly and give you a stable voltage regulation. If you need another
pot. angle vs. voltage figure simply change the series resistor or the
OpAmp´s voltage reference.

Next is:
why OpAmps? Keep it simple and diskrete. You can handle much more voltage
and/or current by choosing an all diskrete design and you are not limited
to the OpAmps suppy rails…(= +/18V to +/ 22Vmax) or you need some real
expencive types…

Next is:
The current sense resistor sould be prior to the regulation transistor in
order to cancel out the voltage drop across it…
If you (like it is now) put the current sense resistor in the gound path,
you lift up the ground of your PS and you will run in various problems:

No chance to connect anything already grounded to your PS or it will go
“mad”
No chance to set up a dual rail PS as your ground path is not real “ground”
as it if floating around, depending on the current flow…
No chance for a steady ground if you hook up an audio circuit to power it
up, it´ll not be stable and do some nasty things while beeing fully ok,
it´s all to the floating ground. So better is to put all sense resistors
into the positive rail!

Next is:
I like the 2N3055, but be aware of its age and therfore its “flaws”
especially the second breakdown effect. if you plan to have something like
a 30V/5A PS you need to have more than one 2N3055 or it simply will blow up.

Here is a good example of a basic PS with all diskrete toppic and very
straight forward design and it is easily to build-up:
http://homemadecircuitsandschematics.blogspot.de/2012/01/how-to-make-versatile-variable-voltage.html

This is not my design and I have no realtionship to the creator….but i
like the design very much, as it could be found in a PS design book…;)

Anyway thanx for the video….
And do not worry much about your PS, as always it is better to build
something that does not function in first place…than to have no clue how
it works…;)
Good luck!! 

jix177 says:

Good vid. Glad to see it coming along so well.

logikgr says:

Glad to see that you’re doing good progress on your project, especially
after dealing with the flooding of the lab/basement. Cheers! 

KX36 says:

The circuit looks a lot better than the previous schematic. One thing I’d
say is I’d expect the transistors labelled 2N222 to be wired as OR, with
their collectors and emitters joined together in parallel, rather than as
AND as shown. With OR, either the current or voltage ppamp can pull the
output voltage down and the output will depend on the lower of the 2
voltages set by the opamps. AND should only work when both the current and
voltage are near enough to the set point to turn both transistors.

Grand Diego says:

Hello, can you give us the nice schematic of your power supply ?

C Marston says:

Where do you draw power for the op amps? It’s not on your schematic.

Robert Calk Jr. says:

Good job, thanks.