Any one know any good easy solid state 10-20W amp circuits?

[sir_modulus]

I have seen a few circuits like the LM383 one (craig anderton and Richardo X) and the little gem and Ruby amps, but i'm looking for a bigger one, but one that is not too hard to build.

[Tim Escobedo]

There are TONS of ICs out there that are fairly easy to build in this power range. Many are designed for automotive 12V power, too, which may simplify the most complex part of the circuit: building the power supply.

These things turn up in surplus channels all the time at reasonable prices. Particularly the more obscure ones.

Some off the top of my head: LM1875, TDA2003... You might be surprise how often these type of chips turn up in commercial amp designs.

[smoguzbenjamin]

Check out the LM3886 for a 50W power amp IC :mrgreen: Go to //www.national.com and search for 'LM3886'

[Johan]

http://home.eunet.cz/rysanek/pdf/tda2050.pdf

complete with layouts and schematics...The LM1875 ( my favorite ) is a direct replacement if TDA2050 is hard to find..

Johan

[Fret Wire]

http://www.redcircuits.com/Page52.htm

[javacody]

Anyone know of any schems of practice amps specifically for TDA2003?

[Ben N]

The Frontman schematics are available at mrgearhead.com.  I wouldn't recommend one as a DIY project, but you can probably learn something useful just by looking at them.

[Brian Marshall]

i've been seriously thinking about making an amp lately.

Anyone know anything about power mosfets.  i hear they can sound really good, but i immagine you still have to have a bulky output transformer.

[Brian Marshall]

Check out the LM3886 for a 50W power amp IC :mrgreen: Go to //www.national.com and search for 'LM3886'

how do you think one of those would do wiht an 18 oir 24 volt supply?

ever actually use one?

[smoguzbenjamin]

It says that it can output 68W of power on +/- 28v and 50W on +/-35v, or something like that. Take a look at the datasheet. I haven't used one yet for lack of a suitable PSU, but thanks to Marcos I realised that you can use charge pumps... So I'm going to try in the near future

[Fret Wire]

http://www.redcircuits.com/Page52.htm

Couldn't the Thunderchief or the "18" be subbed for this amps' pre-amp stage?

[Peter Snowberg]

Sorry, but chrge pumps output very tiny amounts of current in most cases. I've never heard of a charge pump that could create anything near 50W unless you start talking about nuclear research equipment.

You can make little charge pumps for the milliamps required by the preamp. The Max1044 is a perfect example.

To make a high voltage high wattage supply, start with transformers from broken transistor based home stereos. The preamp supply should in most cases be +/-18 Volts absolute MAX with +/-15 Volts being better and/or easier in many cases.

You can use two 24 VAC transformers connected together in series and in phase to produce a +/-33VDC supply if you can't find a suitable single transformer. 50W??? That's about 750mA @ 66V so better to use two 1.5 or 2A 24VAC transformers, or if you could find a transformer with two isolated 24VAC secondaries each supplying 2A, you would be all set. I don't like running transformers at the edge.

You could then use two zener diodes and two regulators like 7815/7915 to get the preamp supply, or something that is more forgiving about the 33V+ supply voltage would be nice instead of having to rely on series zeners.  

Sounds like a nice project if you also have a heavy duty high voltage (100PIV or better) bridge rectifier and a couple of giant caps.

Surprisingly, old office equipment can yield fantastic amp supplies. Old high-end copiers in particular are amazing for parts and power supplies. Old printers are good too. I have a supply from a Diablo630 that weighs a ton and has three heavy duty regulated outputs. I've been thinking about turning it into a heavy duty amp using one of these chips for a while now.

Take care,
-Peter

[Peter Snowberg]

http://www.redcircuits.com/Page52.htm

Couldn't the Thunderchief or the "18" be subbed for this amps' pre-amp stage?

Sure, just remove the volume pot and connect the signal to the connection to C9 that the wiper used to go to.

The V+ on that amp is 33.6V so you'll want to probably look at increasing R7 which is where the TC or "18" would get power. A series zener might be a good choice in this case. You might also want a regulated input supply for the preamp there. In this case an LM317 is perfect.

Take care,
-Peter

[Fret Wire]

Thanks Pete

[Mark Hammer]

Note that a great many of the power-amp chips are intended for the automotive market.  This means they are designed around 13.2v (12v, more or less) supplies.  In many instances, though, they are also rated under the assumption that they are being sent to fairly low-impedance loads - a great many car speakers are 2 to 3.2ohms.  

What that means for the home practice-amp builder is that you will either have to score yourself a 4ohm speaker to get the sort of power you are looking for, or else use a multi-speaker setup to keep the impedance low. (e.g., a trio of 8ohm speakers in parallel, yielding a 2.7ohm load).  Of course, just because you are sucking that much current through the power amp chip via use of a low-impedance load does not mean you are pushing each individual speaker that much harder, since, after all, that power is now divided up amongst many speakers.

Another thing to consider is that many of these chips will actually deliver more power if you up the supply voltage.  For instance, the TDA2050, a favourite of Fender and Crate and probably a bunch of other companies for low-to-medium power guitar and bass amps, will work down to +/-4.5v but will also work with up to +/-25v supplies, delivering more output power for the same load with a higher supply voltage.  That has two implications: one is that you can take a chip nominally rated at some wattage for 12vdc and get more from it with 15v or 18, 24, etc., and another is that you can take a nominally higher-powered chip and still get reasonable power out of it at something other than an ideal supply voltage.

You should note the following:

1) You normally need to multiply the wattage by a factor of 10 to effectively double the perceived loudness.  In other words, a 20w amp pushed full tilt into the same speaker as a 2w amp will only sound twice as loud.  Indeed, there are many instances where a 2w amp with decent preamping, going into a speaker with good efficiency (e.g., 100db output with 1w input) would kick the ass of a *nominally* 20w amp with only modest preamp gain, going into a modestly efficient speaker.  Do not equate wattage directly with volume.  They ARE rrelated, but lots of other things mediate that wattage/volume relationship.

2) Ten watts will almost invariably require a) more than one amp of power, and b) heatsinking of the chip.  I mention the one-amp figure because 3-pin regulators in the standard flavours are of the 100ma and 1A variety.   You can get other types, but they will require heatsinking.  Alternatively, you can push more than 1A of current to your power amp by paralleling several unsinked regulators.  Finally, there is always the route of not using a regulator, although for best hum performance, you probably want to use something more than just a bunch of diodes and caps in the supply.  Naturally, heatsinking either regulators or the power-amp chip will require planning out the layout carefully so as to allow room for the heatsink and a suitable position for everything.

3) There is nothing that says you can't use a bunch of lower-power chips into multiple speakers.  For instance, a quartet of LM380's or even LM386-4's, going into a quartet of 4" speakers, would probably be plenty loud.  These could be all fed from the same preamp output, or if you felt like it you could dedicate specific amps/speakers to specific preamp outputs; a bit like using a couple of guitar amps ganged together, only in one easy-to-carry package.  Examples of the two-in-one approach include the classic JC120 and amps from Magnatone, Gibson, and more recent entries where the reverb signal was allocated to one amp/speaker and the normal signal to another.  These can sound extremely rich and full of character.  Again, consider that a quad of LM386's may not be that much quieter than something like a TDA2003/LM383 or even a TDA2030.

[spongebob]

Finally, there is always the route of not using a regulator, although for best hum performance, you probably want to use something more than just a bunch of diodes and caps in the supply.  

A regulated power supply is not needed for chips like the TDA2050 or LM3886, these chips have quite good power supply ripple rejection, I used one 4700uF filter cap per rail on my TDA2050 amp, and it's dead silent.

Check out the chip amp sub-forum at diyaudio.com, enough stuff to keep you busy reading for days!

[Prive]

The TDA2030 is in the Peavey Rage 158 (the new ones) and in the Marshall 8010, it uses a TDA2030 and just a 4558 give it a try or just to the power section.

http://www.drtube.com/schematics/marshall/8010.pdf

Saludos, Marcelo.

[cove]

These just came in my email today ...check em out

Audio

LM4780 stereo audio amplifier capable of delivering 60W per channel of continuous average output power  http://www.national.com/pf/LM/LM4780.html

     
LM4781 three channel audio amplifier capable of delivering 35W per channel of continuous average output power into an 8 Ohm load  
http://www.national.com/pf/LM/LM4781.html
   
LM4782 three channel audio channel capable of delivering 25W per channel of continuous average output power into an 8 Ohm  http://www.national.com/pf/LM/LM4782.html  


LM4908 10kV ESD Rated, Dual 120mW Headphone Amplifier
http://www.national.com/pf/LM/LM4908.html

[Brian Marshall]

hmmmmm.  me thinks im going to build an amp that will run off of 18 volt.....

1 watt should be ok for practicing

only one prolem... when the heck am i going to do it.

[Ge_Whiz]

I've gone for the TDA2050 (but I'm still building!).

Incidentally, MOSFET output stages do NOT need an output transformer.