Without wishing to get all protective about the LM3886, there is no reason why these devices shouldn't be considered a valid part of an audiophile system. In fact several "audiophile" makes use these parts as output devices in their amplifiers.
What is definitely not audiophile is the protection - any chance of this goes on, and these devices get relegated a couple of divisions.
JoeBob, from experience, I'd suggest you keep things simple to start with. You can always improve later if you thing it's necessary. One transformer should be fine.
Project 19 PCB's draw 60mA quiescent per channel in my system. I've got a single 1kVA transformer (Toroid Corp) for 10 channels (though only the rears are full range), and the regulation seems pretty good with all channels connected, though I haven't tested yet at full output. You say that you're not looking to raise the roof - the transformer sizes you are talking about go way beyond most commercial designs.
What is definitely not audiophile is the protection - any chance of this goes on, and these devices get relegated a couple of divisions.
JoeBob, from experience, I'd suggest you keep things simple to start with. You can always improve later if you thing it's necessary. One transformer should be fine.
Project 19 PCB's draw 60mA quiescent per channel in my system. I've got a single 1kVA transformer (Toroid Corp) for 10 channels (though only the rears are full range), and the regulation seems pretty good with all channels connected, though I haven't tested yet at full output. You say that you're not looking to raise the roof - the transformer sizes you are talking about go way beyond most commercial designs.
Most commercial 'hifi' designs are also only 'mid-fi' at best and there transformers are so poorly rated that the supply rails jump about so much they shouldnt even be considered as usable.
For 10 channels @ 50W you should have at very least a 2kVA transformer as it will be forced to work hard at full power which will heat it up causing more power loss within the transformer and decreasing efficiency....
Your 1kVA transformer will suffer about 5% voltage troop @ full power and you house wiring will drop about another 5% of the voltage you also lose around 3V in the power supply bridge and caps which at this voltage is close to 10% and if you are using a capacitance multiplier circuit this will drop about 5% of the voltage also. Your transformer should only be used continuously at half its current rating or it will heat up and more power will get lost.
So, 1kVA drops to 800VA drops to 400VA at best. Half of that 400VA will go on chip power dissipation which lowers total amp power down to 200W rms or 20W per channel you could strech this to 25W per chip or maybe even 30W but the transformer will now be running quite warm if not hot and the extra voltage drop will bring you to the verge of the protection circuit activating ... you can probably get more out of such a transformer if you have plenty of capacitance but the gains just arent worth the troubles involved.
Extra transformer power is almost always a good thing and going from 2 x 625VA transformers for the front channels (5 amps) to 2 x 750VA transformers cost about $2 so it is more than worth it given that the transformers will now be running considerably cooler and hence more efficiently as well as the higher rated transformers also being better regulated. 2100VA for 7 x 50W amps seems rather high but when you look at the numbers above you will see that it is generous but certainly not extravagant.
For 10 channels @ 50W you should have at very least a 2kVA transformer as it will be forced to work hard at full power which will heat it up causing more power loss within the transformer and decreasing efficiency....
Your 1kVA transformer will suffer about 5% voltage troop @ full power and you house wiring will drop about another 5% of the voltage you also lose around 3V in the power supply bridge and caps which at this voltage is close to 10% and if you are using a capacitance multiplier circuit this will drop about 5% of the voltage also. Your transformer should only be used continuously at half its current rating or it will heat up and more power will get lost.
So, 1kVA drops to 800VA drops to 400VA at best. Half of that 400VA will go on chip power dissipation which lowers total amp power down to 200W rms or 20W per channel you could strech this to 25W per chip or maybe even 30W but the transformer will now be running quite warm if not hot and the extra voltage drop will bring you to the verge of the protection circuit activating ... you can probably get more out of such a transformer if you have plenty of capacitance but the gains just arent worth the troubles involved.
Extra transformer power is almost always a good thing and going from 2 x 625VA transformers for the front channels (5 amps) to 2 x 750VA transformers cost about $2 so it is more than worth it given that the transformers will now be running considerably cooler and hence more efficiently as well as the higher rated transformers also being better regulated. 2100VA for 7 x 50W amps seems rather high but when you look at the numbers above you will see that it is generous but certainly not extravagant.
why not buy secound hand?
I have hear in denmark , put up a message at some internet page's for secound hand stuff, writting that im interested in buying damage power amp's, As long as the trafo is good.
I have been aple to buy some amp that i even repair for small money.
Else you have the box and trafo/heat sink for small money.
just an advice. maney people have things that they don't wan't to repair because it's better to buy a new, insted of sending it for a pro's to look at it.
best regards jc.
some link for lm3886 applications.
http://www.dckits.com/advise.htm
http://www.saunalahti.fi/elepal/vahvistin/vahvistin1.html
http://www.nextra.ro/rofarom/tsy/index.html
I have hear in denmark , put up a message at some internet page's for secound hand stuff, writting that im interested in buying damage power amp's, As long as the trafo is good.
I have been aple to buy some amp that i even repair for small money.
Else you have the box and trafo/heat sink for small money.
just an advice. maney people have things that they don't wan't to repair because it's better to buy a new, insted of sending it for a pro's to look at it.
best regards jc.
some link for lm3886 applications.
http://www.dckits.com/advise.htm
http://www.saunalahti.fi/elepal/vahvistin/vahvistin1.html
http://www.nextra.ro/rofarom/tsy/index.html
JoeBob,
One advantage to going with a number of smaller transformers is that you can build just 2 (or 4) channels and try it out. I would consider building the rear channels first - lower power requirements and lower cost. If you like the sound, continue on with the rest, and you'll be more qualified to make decisions about big vs. small transformers, etc. Otherwise, you haven't wasted too much money.
One advantage to going with a number of smaller transformers is that you can build just 2 (or 4) channels and try it out. I would consider building the rear channels first - lower power requirements and lower cost. If you like the sound, continue on with the rest, and you'll be more qualified to make decisions about big vs. small transformers, etc. Otherwise, you haven't wasted too much money.
one more thing... using the separate power box for the front amps will complicate the grounding scheme ... it wont be impossible to avoid ground loops but it will make the job a bit harder.
Well, I don't mind if the grounding is a bit of a challenge, that's what makes these things fun. As for using a smaller transformer like Jamie F sudjested, I'd rather have too much then too little, after all, the prices aren't that much when compared to smaller transformers from plitron.
I'd rather build the front channels before the rear, then I get to actually get some good use out of them when it's done to see if I like it, and the transformers for the front are only $90 more then the one for the back, so it's not that much more.
I'd rather build the front channels before the rear, then I get to actually get some good use out of them when it's done to see if I like it, and the transformers for the front are only $90 more then the one for the back, so it's not that much more.
My 10 channel amp is effectively only 4 full-range channels, which is why it will probably be usable, but I will do some tests to see whether it struggles. Also I am only driving voice coils, so the speakers present an easy load and efficiency will be very high. No one needs to stretch a 50W amplifier driving a tweeter.
I also meant to say... don't under-estimate the hassle of grounding and noise - I've been trying to locate a noise source for a couple of months now, with access to some pretty good test gear. I have also gone through a lot of time and expensive re-wiring to address ground loops that are inevitable on multi-amp systems unless you are balanced. On this point, construct the thing so you can easily take it apart and re-assemble it. Maybe you'll be lucky, but with a complex set-up, unless you're really experienced, you're probably in for more fun than you're looking for 🙂
I also meant to say... don't under-estimate the hassle of grounding and noise - I've been trying to locate a noise source for a couple of months now, with access to some pretty good test gear. I have also gone through a lot of time and expensive re-wiring to address ground loops that are inevitable on multi-amp systems unless you are balanced. On this point, construct the thing so you can easily take it apart and re-assemble it. Maybe you'll be lucky, but with a complex set-up, unless you're really experienced, you're probably in for more fun than you're looking for 🙂
I'll keep in ind to make it easy to disassemble/reassemble, for I will most likely not get it anywhere near perfect the first try. More fun then I bargaine for? Bah, no such thing, it might take me a while, but I truly doubt I'd ever give up...
Well, I've reached another question... When I hear people talking about the value of caps for the power suply, most say more is better with lower values added up, and to buy as much as your budget allows, but that there is a point of diminishing returns. My question, what is this point of diminishing returns? Electrolytic caps are something I can get for free, so my budget isn't an option. Since I've got five amp modules, and will put caps on the + & - of each module, what values should I use?
Use around 10 000uf per rail per amp preferably made from 10 x 1000uf caps with high ripple current rating and low esr. Much above 10000uF wont give you much gain.... although if you've got them for nothing and dont mind having to wire a soft start circuit I'd use about 20,000-50,000uF per amp you'll also want to bypass each electrolytic cap with a 2uF film cap.
- Status
- Not open for further replies.