There's one little problem with Class A circuits... you need a comparatively beefy power supply and heatsink, which just happen to be areas that are a bit $$$. AB matching the "blameless" criteria may be a better bet.
It's a good question though. It would almost certainly involve at least partly discrete circuitry, as chipamps tend to be optimized for moderate current draw instead.
It's a good question though. It would almost certainly involve at least partly discrete circuitry, as chipamps tend to be optimized for moderate current draw instead.
Thanks. I have the LM3886 currently. It runs on walwarts - dc dc boost converters - LM350 regulator (this I added just yesterday after someone on the forum suggested). It seems to make a difference in quality. I know it may not be the best way to run the LM3886 but cost wise that was low. Now I have an old receiver with 58V + and - supply was wondering how I can use that PSU and build a higher quality amp. Again I need about 20W from both channels.
Hi danielwritesbac,
I was looking at the LM1875 turbo circuit, any links to why these modifications to the circuit in the datasheet?
thanks
Since you ask, consider that these little black chips don't exist in a vacuum. The factory has to provide not only a chip, but the circuit it works in.
The data sheet schematic was created by the same team of engineers who created the chip in the firs place. Keyword "team," these are teams of several senior and junior engineers checking each other's work. They use testing equipment that costs tens of thousands of dollars. Take into account the fact that Texas International, for instance, doesn't hire the bottom ten percent of the class, those guys are designing toasters.
Consider prototypes and testing. Humans can't build these little chips, it takes a machine. So after everybody's best ideas have been worked out on paper, special machines create a few prototype chips for testing. These machines cost more tens of thousands of dollars.
After long human/machine testing, prototyping, and testing again, at last the micro-accurate production machinery is ordered for the factory. By now the dollar investment is staggering.
But with all this expense, how can these chips be sold for a couple of bucks--or for 50 cents? Because costs are spread out over production runs in the millions. Customers all over the world are served. Customers who have labs that test incoming deliveries, and won't hesitate to sue if the chip, and the circuit it's advertised to work in, are not up to par. Things had better be right, and they'd better be right every time.
So after all this investment of engineering expertise and money, some expert jumps up and says, "Hey, those factory guys don't know what they're doing, look at my circuit, it's a lot better."
I don't know how the chip ever got built to begin with if the factory guys don't know what they're doing. And I can't figure out why the factory couldn't have hired this expert instead of the clueless engineers they got stuck with. And I can't prove that the expert's circuit is not better, and there are even some more experts who declare that yes, the first expert's circuit really is better, and what's more I ran some highly scientific tests in my garage.
In my experience it all ends up with everybody disagreeing with everybody. I guess it boils down to like the man says, you pays yo' money and you takes yo' choice, it all comes out of the same barrel anyway.
.
Last edited:
With your reciever with the 58 v + and - supply something comes to mind. Sometimes the transformer has a prevision to run it on 220v instead of 110. If this is the case you could rewire the transformer for 220v but use 110 and your power supply voltage will now be about + and - 29v just about perfict for your Lm3886 .
VSSA by Shaan or LME49830 with lateral FETs would be my choice and at that voltage Id go for 4 output pairs of 2SK1058/2SJ162 for 8 ohms if using LME49830 if its AC then maybe 5 but thats going to be a expensive one as they cost quite alot. Here in Estonia LME49811 chip costs 37€ and pair of those fets 20€ So this idea is way too expensive.
Last edited:
Thanks, an interesting thought.
When I look at the back of old reciever it says 110V, 290W. There is no mention of 220V on the back. Thoughts?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.