diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Chip Amps (http://www.diyaudio.com/forums/chip-amps/)
-   -   high voltage lm3876 (http://www.diyaudio.com/forums/chip-amps/15004-high-voltage-lm3876.html)

zx3chris 11th May 2003 02:26 AM

high voltage lm3876
 
i want to try my hand at making a chip amp and i have a lm3876 and also a surplus tranny i ripped out of an old amp--the tranny is center tapped and 33vac on each side.. after rectify and smoothing this will be around 40v--is this too much for this chip?? data sheet shows 84v max but just wondering if anyone has used this high of a voltage

-chris

leadbelly 11th May 2003 05:42 AM

Well, I can't speak from experience, but I think you won't have problems if you (1) oversize the heatsink from what the datasheet says, and (2) only run well-behaved 8R speakers.

For the heat sink, I would avoid isolating them, and instead come up with a chassis design that would allow them to sit at V- for better heat xfer, making them internal and isolated from the chassis.

joensd 11th May 2003 05:55 AM

Personally I wouldn´t exceed the maximum ratings.
If I understand you correctly you´ve got 66VAC centertapped so
you´d get around +-46VDC which is well above the +-42V maximum rating National Semiconductor recommends.
You can of course give it a try but be prepared it won´t sound good (or even blow???).
You might wanna have a look here for probable consequences of too high a voltage transformer : http://www.diyaudio.com/forums/showt...threadid=13593

Regards
Jens

UrSv 11th May 2003 09:59 AM

Re: high voltage lm3876
 
Quote:

Originally posted by zx3chris
i want to try my hand at making a chip amp and i have a lm3876 and also a surplus tranny i ripped out of an old amp--the tranny is center tapped and 33vac on each side.. after rectify and smoothing this will be around 40v--is this too much for this chip?? data sheet shows 84v max but just wondering if anyone has used this high of a voltage

-chris

Much to high voltage. If you absolutely need to use that transformer I would recommend regulating the voltage to something more useful which is anything some distance away from 42 V.

And yes somebody has already used that high voltage and he publicly apologize for it.

ted 11th May 2003 11:52 AM

I was thinking that a possible solution is to "slave" the +/- rails. That is, you use a set of transistors to drive the lm3876 power rails with a voltage that is proportional to the output audio signal. It's something similar to what older amplifiers did to solve the SOA problem of high rail voltages. I think this was called a totem-pole output stage, but in this case, the whole amplifier is being driven this way, even the input stage.

Maybe somebody could comment on if the input stage of a lm3876 would like a dynamic power supply input?

I am not sure if such a trick would work with a lm3876.

joensd 11th May 2003 12:42 PM

Has anybody read the review of Bob Carver´s Sunfire amplifier?
It keeps the supply voltage 6V above what is needed for the momentary output power to minimise dissipation.
The 20 output transistors are just mounted to the case bottom.
Delivers 600W into 8Ohms with an "endless" current capacity.
Should work below 0.5Ohms.:bigeyes:
And just took 25 years of development time.:bigeyes: :bigeyes:

Jens - doesn´t need 600W but likes the idea

colwaiting 11th May 2003 03:25 PM

The 33-33 would be just ducky for the R. Elliot 3a amp. Keep it around.
the 3876 is set for 30-30dc which is around 22.5,22.5vac.
There is not much point in exceeding the ratings as the per-chip dissipation is really on the low side with these beasties! althoug the 3886
sez it is ok for 38-38 dc it will reach its thermal limits easily, and is for 8-ohms only.

zx3chris 11th May 2003 05:47 PM

the speakers i plan on driving are a set of peerless speakers i made a whle back and unfortunately they are 4ohm.. they do sound excellent, and the reason for 4 ohm is that they were going to go in my car, but i had trouble with the clearance of the woofer and my door/window--so i made them for my house. right now i am powering them off an old car head unit and a computer AT supply--they are only getting ~15 watts but sound nice--however they can go up to 150watts so id like to push them if i could--so i wanted to make an amp based on the 3876 (since i have it already) and drive 4ohm safely. i will keep on the lookout for some 22-22 transformers. thanks for the information

-chris

leadbelly 11th May 2003 05:51 PM

zx3chris,

Another thing you can try is straying from the IGC philosophy, and instead building a half-wave rectifying stage for that xformer with big caps to smooth the DC. That'll give something like +/- 22 VDC, so you'll be struggling with lower power rather than thermal protection :)

zx3chris 11th May 2003 07:27 PM

if i make half wave rectifying instead of full wave, how can i estimate the dc out?? would [ (VAC/.7) - .7 ] /2 be good?? in that case it would be +/-23VDC... if this works, in most gainclose i always seen people using simply 1000uf caps, what value caps should i use??

-chris


All times are GMT. The time now is 05:08 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2