LM338 regulated snubberized PSU for audio amplifiers

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Re: MKT vs. Ceramic SMD

agent.5 said:
I am eagle-ing my LM4780 SuperSymmetry with LM338 snubber. For the 300nf (connecting V+ and V-) and the 100nf (bypassing V+ and V-), should I use 1210 ceramic smd chip capacitors on the pins of the LM4780, or should I stay with the MKT type capacitors?

MKT.
100V rated for the 330nf (not 300).
The 100nf caps can be 63 or 100V MKT, which are small.
Put these as close as possible to the chip's PSU pins, as always.
 
Hi all,
some time ago i build the psu with TO3-regs, the only ones i could buy in germany (PSU with TO3 regs). Now i've got some TO220 types (one third of the TO3 price!) for my next project.
Thanks a lot, carlos, the psu runs really good, the LM4780 modules from peter couldn't run the psu to it's limit. ;)

Are there limitations i have to take care for? Thermal stability? Maximum current? In the datasheets i could find any diffs.

For the bridged version (not sure: one LM4780 or two LM3886), should i lower the voltage?

Michael
 
Hi!:)

I finally finished my new amp, wich is based on carlos`s ideas.
This is an dual mono integrated amp, with two lm3886 pr channel in paralell made as carlos suggested with low value resistors. The preamp section is made of opa627bp with a gain of 2,5.
The powersupply is this lm338 snubberized.

This amp is truly exellent!:att'n: It`s the best amp I have ever had in my system.
The sound is detailed, dynamic and powerful and sounds more like 200W than 50W. It actually has much more bass power than my old Electrocompaniet ECI-1 wich is an true 100w (and costed atleast 2500 $ some years ago)

The amp sounds very "clean" and "uncolored", and seems to reveal everything in the recordings.

I`m very happy!:D

Thanks

Tor Martin:)

BTW my speakers are the old, but exellent epos ES 14.
 
slackman said:


Are there limitations i have to take care for? Thermal stability? Maximum current? In the datasheets i could find any diffs.

For the bridged version (not sure: one LM4780 or two LM3886), should i lower the voltage?

Michael

The tradeoffs can be found using National's interactive web-tool, based upon Excel. When you download it you have to "enable macros".
http://www.national.com/appinfo/audio/files/Overture_Design_Guide15.xls
In the bridged version, PdMax is 4X that of one chip -- this is because each chip sees a complete swing from Vee to Vcc.

I use a whoppingly big heat sink on the Bridgeclone -- and can get distortion stats which are nothing short of sensational.

Because this is an opamp on steroids there is a tradeoff between running the chip hot (with very low distortion) and "mean time between failure".

Jack
 
Tor M said:
[BThis amp is truly exellent!:att'n: It`s the best amp I have ever had in my system.
The sound is detailed, dynamic and powerful and sounds more like 200W than 50W. It actually has much more bass power than my old Electrocompaniet ECI-1 wich is an true 100w (and costed atleast 2500 $ some years ago)

The amp sounds very "clean" and "uncolored", and seems to reveal everything in the recordings.

I`m very happy!:D

Thanks for reporting, Tor.
You description is quite accurate, that's it.

:cool:
 
This thread has wondered all over. The point needs to be made that the speaker will determine which GC version sounds the best. My experiance with unsnubberized andunregulated GC amps is they sound wonderful with easy to drive speakers. I used my simple GC amps with rear horn loaded Lowther PM2As. The sound was very good, amazed many friends with single ended amps. But when I used the GC amps with speaker that showed a more difficult load the sound wasn't nearly as good. I tried the GC amps on wide veriaty of speakers and the pattern was the same. Great sound on efficient simple loads and not so great on more difficult loads.
 
To continue, I hit the tab button and posted before I made my point.
The GC builder needs to aware of the the speaker load before deciding which version to build. If there is doubt then build the simple GC listen to the result. If the sound isn't what was expected then the snubbers and regulators can be added. As with everything there are no absoltes.
Just my opinoin.
 
Andy Bartha said:
My experiance with unsnubberized andunregulated GC amps is they sound wonderful with easy to drive speakers. I used my simple GC amps with rear horn loaded Lowther PM2As. The sound was very good, amazed many friends with single ended amps. But when I used the GC amps with speaker that showed a more difficult load the sound wasn't nearly as good. I tried the GC amps on wide veriaty of speakers and the pattern was the same. Great sound on efficient simple loads and not so great on more difficult loads.

That's correct.
The low capacitance GC is as picky with speakers as a SET amp.
Wich is obvious, no matter how big a trafo you use.
The 'juice' goes away fast with such small caps.
I test my amps with lots of speakers, I can only say they can drive most of them and they sound good practically everywhere.
The low capacitance GC is exactly the opposite: you struggle to find a speaker that matches it, or else you will have to make your own high sensitivity fullrange horn, which is not exactly the type of speakers I like.
Not that it will only work with this type of speakers, but the choice is very narrow.
It can sound good, yes, but only sometimes.:D
 
Hi!:)

I agree with that. The last chipamp I had in my system was a VBIGC. This had wonderful midrange, and impressive holgraphic soundstage. But with my speakers the dynamics were slow and bass were "mucky".
The amp sounded very good with slow vocal jazz and that kind of music, but when things needed to speed up.....

Probably a terrific amp for the right speakers! ;)


Tor Martin:)
 
Peter Daniel said:
I noticed the same boards being sold on two other European sites, both of them run by diyAudio Moderators. I was never approached before those boards appeared there. Some of my UK clients were actually under impression that I have UK agent, which is not the case.

Peter, I'm appalled to read this. I had assurances that you had no issue from me distributing some kits in the UK (and before I put them on madaboutsound.com). Granted, it didn't come from you directly but I took Brian's word as true. I'll remove them from the site.
 
derekyu said:
dear carlos
i would like to know if i change it to voltage variable. what values should i change to to r3 and r5?

What voltage do you need?
The formula is: 1.25 * (1 + R5/R3).

To make the PSU variable you could use 1K trimmers in series with 1.2k resistors in place of R5 and R6.
That would give you an adjustment between ~17.5V and 31V.
 

Attachments

  • cfm lm338 regulated snubberized psu.png
    cfm lm338 regulated snubberized psu.png
    23.4 KB · Views: 878
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.