Regulated PS w/ LM317 & LM337's

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
oldfatbubba

oldfatbubba

Disabled Account
Joined 2006
Regulated PS w/ LM317 & LM337's

Please check my math & logic.

I'm making LM3875-based Gainclone monoblocks with regulted power supplies. Can I safely use LM317 & LM337 regulators using this schematic? (ignore the voltages, filter caps will 1,500uf)

An externally hosted image should be here but it was not working when we last tested it.


I think I can because:

LM3875 disapation at full power (Vs=25v+/-, Pout=30watts, & Rout = 4 ohms) = 33 Watts per National's documentation...

Therefore... Itotal = about 1.3 amps...

and, since I+ = I- = 1/2 of Itotal, each regulator will pass about 0.66 amps

Since Imax for each LM317/337=1.5 amps, each regulator should be running at 44% of its capacity when the is amp running at full power. Is my logic correct?

Thanks for looking, oldfatbubba...
 
E

ErikdeBest

Member
Joined 2003
Hi

I like it you thought about it before posting the question. Most people would just drop the question: can I use ML317/337 to regulate the power rails for a gainclone?

For low listening levels...and for maybe 90% of the time these regulator would be fine...but, whenever there is some Peak in the music current draw will be larger! Think about it this way...why are the recommended trafo's rated for at least 4 amp on the secondary?

Erik
 
A

AndrewT

R.I.P.
Joined 2004
Hi,
30W into 4ohms will need an output voltage of about 15Vpk.
Your 4ohm speaker could have a minimum resistance of about 3ohms.
The peak current into the load is about 5Apk.

If you want 30W into 4ohms and it needs 15Vpk at the output why are you planning to run +-25Vrails? Is the chipamp that bad?

Can I recommend you build your chip amp without regulated supplies?

Just use a decent sized transformer, rectifier and smoothing capacitors much bigger than 1m5F.
For Ipk=5A, I would recommend +-10mF to +-15mF per channel.
It should perform well on an unregulated PSU.
 
darkfenriz

darkfenriz

Member
Joined 2004
You've calculated mean current, but at peaks output current of regulator will be nearly the same as speaker output current, because regulators use feedback, capacitance does not matter very much here.
But this shouldn't harm if regulator switches off due to short circuit protection at some peaks if you have high capacitance at regulators' output.
For that time it just won't regulate.

regards
 
sek

sek

Member
Joined 2003
Hi,

I've been using the 317/337 in a regulated supply for an LM3886. ;)
But only because I wanted to test my schematic and board layout before spending money on suitable regulators!

It's interesting to what loudness levels such an arrangement can drive a speaker with average efficiency. But as the 317/337 have internal current/temperature limiting functionality (and way too low current capability), they are in no way suitable to power 4Ohm speakers, let alone deliver 30 Watts of output power.

As mentioned by others, there are better (yet more expensive) regulators available, of which the LM338 is the cheapest one, while something like the LT108x series is recognized as among the best.

Basically, what you want to set up is one of the proven designs of regulated gainclone power supplies, about which there are many threads in these forums. Just do a search... ;)

Cheers,
Sebastian.
 
oldfatbubba

oldfatbubba

Disabled Account
Joined 2006
ErikdeBest said:

For low listening levels...and for maybe 90% of the time these regulator would be fine...but, whenever there is some Peak in the music current draw will be larger! Think about it this way...why are the recommended trafo's rated for at least 4 amp on the secondary?
Click to expand...

Point taken. Perhaps my mind was thinking "Class A tube" where the fluctuations in current are minimal.
 
oldfatbubba

oldfatbubba

Disabled Account
Joined 2006
jaudio said:

Have you looked into adding input bypass transistors to increase current?
Click to expand...

I could see how that would work well, but it would make for a more complex build.

FWIW, there's a filament supply circuits in Ranier zur Linde's book, Audio Valve Amplifiers, which bypasses a 7806 regulator w/ a
BD242. I was oft tempted to use that design in high gain guitar amps where noise is usually a problem.
 
G

gmphadte

Member
Joined 2003
Use LM338 and use the schematic to get negative supply.

Gajanan Phadte
 

Attachments

  • neg.jpg
    neg.jpg
    16.7 KB · Views: 565
A

AndrewT

R.I.P.
Joined 2004
Hi,
just to pluck a couple of numbers, from the posts' worst case currents, could range from 0.66A to 5Apk.

Which end of these extremes do you design for?

I know which end I will aim for.

BTW. the current variations in a ClassA stage can vary from almost zero to double quiescent current. Again that is an enormous ratio that has to be accomodated by the PSU.
 
raromachine

raromachine

Member
Joined 2006
I note that the minimum current from the 317T is 1.5A, but the typical maximum (for the Nat Semi LM317T) is 3.4A!

If you could parallel two of those suckers you'd probably be on the money - so the question is, will 2 operate in parallel? Or would you need two sets of rectifiers for each unit? Mind you, it's probably easier to use the pass transistor style hookup...
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom