Voltage Regulators / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Voltage Regulators

Move Thread

LAN_403

ZakWolf

Thu Jun 26 2014, 12:54AM

Registered Member #3114 Joined: Sat Aug 14 2010, 08:33AM
Location:
Posts: 608

Can you put LMXXX regulators (adjustable) in parallel to increase total output current?

ex. LM317 datasheet

Output Current in Excess of 1.5 A

Output Adjustable between 1.2 V and 37 V

I want to make a small power supply, I want to use it to power random projects and other things like ZVS drivers. Meaning I will need around 10A minimum output current.

If Im not mistaken the only real problem is current hogging and uneven load etc.

Thanks

EDIT: I have a heavyduty transformer capable of about 45v rectified at 15A Im not to experienced with designing circuits so I figured this would be a good learning experience, I would like the output to be adjustable. Is there a better ay to go about this then paralleling regulators?

Edit:These look good but whats the difference ?

I think i like the first one better but it uses a regulator IC and the second one doesn't, which one would be the most reliable?

Sulaiman

Thu Jun 26 2014, 04:58AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

There are many considerations when designing a power supply,
in this case you have a 45V at up to 15A raw dc supply,
my first consideration would be heat,
e.g. for a linear regulator, when the output is 5V there will be 40V accross the regulator,
at 10A output that's 400W dissipated in the regulator, that would require a LARGE heatsink
or 'forced air' (a fan/blower) heatsink, neither are cheap.
An alternative is 'switch mode' but design and fault finding are more difficult.
Both linear and switchmode need careful (or lucky) design to be stable when the load is reactive
(inductive like a 'zvs' or capacitive).

Of the two circuits that you pointed to, I prefer the second
because it has current limit and good voltage regulation.

I have a couple of simple alternatives;

1) Variac - current trip (fuse or mini circuit breaker) - Transformer - bridge rectifier - capacitor.
A variac is not cheap, and the output voltage regulation is not very good BUT
the power supply would be very rugged and easy to design, build and fault find.

2) A simple p.w.m. regulator..
I have two if these

on order
I'm expecting them within days, so I'll test them and if useful for your application I'll report back very soon.

dexter

Thu Jun 26 2014, 11:35PM

Registered Member #42796 Joined: Mon Jan 13 2014, 06:34PM
Location:
Posts: 195

paralleling regulators is doable but as the difference between input and output voltage get higher so the heat dissipated by the regulators

better use that transformer with a variac or build one of the circuits from the links provided

ZakWolf

Fri Jun 27 2014, 06:22AM

Registered Member #3114 Joined: Sat Aug 14 2010, 08:33AM
Location:
Posts: 608

I think im going to build the second circuit

It seems pretty straight forward any recommend alterations before i build it? I haven't heard of anyone making this so i want to be sure it works alright

Or even this, it looks the same but uses the TIP version of the 3055.

Sulaiman

Fri Jun 27 2014, 12:28PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

I'm so sorry, I confused myself when opening the two links,
it's the first one using the 723 regulator that I meant to reccommend
because it has current limit and good voltage regulation.
There is a problem though,
most 723 regulators are only (continuously) rated for 40V supply,
your 45V supply may be a problem.

The second circuit has no current limit (if you short circuit the output, something bad will happen)
and the voltage regulation will be poor because the output transistors are in 'emitter follower' mode after the regulated voltage.
The 317 regulator is only rated for 40v Input-to-output so it also may have problems with your 45V dc supply
when the output voltage is set near zero.

So, given that both regulators have a 40v rating what should you do?
Go with the second circuit, the LM317 type, and ensure that you do not have more than 40v accross the LM317.
An easy/safe way to do this is to put a resistor between
the bottom end of the 10k potentiometer R1, and 0v.
somewhere around 560, 620, 680 ohms should be ok,
or put a 1k Ohm pot and adjust.......
with the power off, set R1 for minimum and the 1k for maximum resistance,
with a load on the output, power on,
adjust the new 1k pot so that the voltage between the IN and Out terminals of the LM317 is less than 40v.
Now you can adjust R1 to control the output voltage from this new minimum, up to maximum.

ZakWolf

Sat Jun 28 2014, 07:28AM

Registered Member #3114 Joined: Sat Aug 14 2010, 08:33AM
Location:
Posts: 608

Sulaiman wrote ...

I'm so sorry, I confused myself when opening the two links,
it's the first one using the 723 regulator that I meant to reccommend
because it has current limit and good voltage regulation.
There is a problem though,
most 723 regulators are only (continuously) rated for 40V supply,
your 45V supply may be a problem.

The second circuit has no current limit (if you short circuit the output, something bad will happen)
and the voltage regulation will be poor because the output transistors are in 'emitter follower' mode after the regulated voltage.
The 317 regulator is only rated for 40v Input-to-output so it also may have problems with your 45V dc supply
when the output voltage is set near zero.

So, given that both regulators have a 40v rating what should you do?
Go with the second circuit, the LM317 type, and ensure that you do not have more than 40v accross the LM317.
An easy/safe way to do this is to put a resistor between
the bottom end of the 10k potentiometer R1, and 0v.
somewhere around 560, 620, 680 ohms should be ok,
or put a 1k Ohm pot and adjust.......
with the power off, set R1 for minimum and the 1k for maximum resistance,
with a load on the output, power on,
adjust the new 1k pot so that the voltage between the IN and Out terminals of the LM317 is less than 40v.
Now you can adjust R1 to control the output voltage from this new minimum, up to maximum.

Thanks man im gonna start on it :)

Dragon64

Sat Jun 28 2014, 08:07PM

Registered Member #1438 Joined: Sat Apr 12 2008, 12:57AM
Location: Canada
Posts: 218

Why not use a pass transistor? It will easily allow the LMXXX regulators to be able pass 10A assuming you have the suitable transistor.

Here's an example of a LM7812C for a 30A PSU

And some guide about general PSU construction and pass-trasistors

ZakWolf

Sat Jun 28 2014, 09:26PM

Registered Member #3114 Joined: Sat Aug 14 2010, 08:33AM
Location:
Posts: 608

Dragon64 wrote ...

Why not use a pass transistor? It will easily allow the LMXXX regulators to be able pass 10A assuming you have the suitable transistor.

Here's an example of a LM7812C for a 30A PSU

And some guide about general PSU construction and pass-trasistors

thanks mang

Sulaiman

Sun Jul 13 2014, 03:16AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

Even the 78xx series will have a problem with a 45 Vdc supply
(if 45 V @ 15 A then at 0 A the supply will be even higher than 45 V )

I'm not saying they can't handle 45V, just that the specifications all are <= 40V

P.S. The eBay PWM regulators seem to have been lost in the post,
the seller has said they've posted two more ...........

UPDATE: Two pwm controllers arrived this morning, quick look ... guessing a simple 555 circuit,
needs an output inductor + capacitor for use as a d.c. supply (meant for a motor)
. Voltage rating of onboard 7812 regulator limits input to 40V max !!!!!!
. 10A switch but tiny heatsinks ... optimistic ;)
So, not even powered up I can say 'not suitable for your needs'.
Luckily I want them as dc motor controllers :)

P.P.S. I just had a look at datasheets for smps ic that I'm familiar with, none >40V
I'm sure there are, just that I don't know any.
A simple discrete linear power supply is not too complex, with a few compromises.
I'd start by considering the worst case; 45V @ 15A input, 0V @ 15A output = 675 Watts
The bridge rectifier will also provide a surprising ammount of heat.
That's a lot of heatsink - to be practical a forced-air cooling system minimum.
Then you want many (e.g. 10) cheap output transistors in parallel, sharing the current,
(thermal resistance, junction(s)-to-heatsink)
then follow any discrete linear power supply design.
Not counting the output transistors, probably 6 or less transistors.
Even a zener reference (39V ?) with potentiometer and compound emitter-follower is good enough for most purposes IF you have current-limiting. Also VERY stable.
(some loads can confuse 'cleverer' power supplies)
Later you may want to build a 'cleverer' psu, but I think a simple supply is really useful.
(and a lot easier to understand and upgrade)
OR, because it's how my first 'good' power supply was made,
A switch, variac, fuse, transformer, bridge rectifier, capacitor, Fuse combination is super rugged,
it behaves like the raw dc supply of typical projects, BUT variacs aren't cheap.
(or stockpiles of 'replacement' fuses!)

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob