matjans said:you're pretty stubborn!
the datasheet tells you the absolute maximum voltage is 84v (thus, +/- 42 volts and YES, this matters, a lot). Either go on frying chips or use a psu with a lower voltage.
the by many people preferred (or agreed upon) schematics are found all around this forum. One good thread to start is this one (greg's ultimate gainclone) and a *very* good site to get started on building gainclones is nuuk's site . start by reading the basic gainclone pages.
but seriously, lower the psu voltage...
I dont really have a way to lower the voltage. I have that psu from a fried amp, and I dont want to mess with the windings of the transformer or use voltage regulators.
A solution that works is just to use diodes. A diode drops the voltage (correct me if i am wrong) 1.2V per diode.
So, chain like 8 or so of them in line with the + and - supply to lower the voltage.
I used the same schematic you used, but I did it on a PCB.
If you really want a nice amp, I suggest going to the post LM3886 Pcb, and downloading the pcb design that bigparsnip kindly created for me. That isn't the one I used; i used a hand designed one, but that will definately work better.
The only difference is i used a
So, chain like 8 or so of them in line with the + and - supply to lower the voltage.
I used the same schematic you used, but I did it on a PCB.
If you really want a nice amp, I suggest going to the post LM3886 Pcb, and downloading the pcb design that bigparsnip kindly created for me. That isn't the one I used; i used a hand designed one, but that will definately work better.
The only difference is i used a
i just changed the feedback resistor from 33k to 470k, and it's so much louder now. I calculated that the gain increased from 4.3 to 48. I have the 10k pot turned up about halfway, and I'm not sure if it's clipping or if the speaker is making that noise. The speaker is a 6x9 that I bottomed out too much, and now the voice coil scrapes against the magnet. I'll hook up my subs to it when I have time and test it again.
don't do it
ahhhhh... the caps would blow up, either before or after your chip blew up. You see the 35 volt caps don't give you 35 volts, they provide smoothing of the voltage you already have, but they are only designed to smooth 35 volts max. Most people would never use 50 V rated caps at 50 volts, this is much too close to the failure point, and they will wear out faster, if not immediately.
So if you MUST do what you are doing here are my suggestions:
- COPY the design of someone else from the links above, don't use your own design. The design you posted is messed up.
- connect about 10 diodes from your pos supply to the positive input of the chip, and 10 from your negative supply to the negative input of the chip--- make sure you orient these the right way! This will give you about .7 volts *10 = 7 volts drop and make your 50 volts into a more safe 43 volts, but I would use even more (.7 volt drop each) to drop it to 40 volts or 37 volts. Make sure you read up on diodes so you have foward current flow (if you put them backwards you just wont get any current flow and wont hurt anything except maybe your speaker.) Also use diodes rated for 5 amps and 100 volts (or more) (foward and reverse voltage) so you don't blow these up either. Buy these diodes at radio shack if you have to. They also have the other components you need.
-keep using with the heat gunk, good stuff.
-Don't touch anything.
You are still scaring me, you are way over your head.
ahhhhh... the caps would blow up, either before or after your chip blew up. You see the 35 volt caps don't give you 35 volts, they provide smoothing of the voltage you already have, but they are only designed to smooth 35 volts max. Most people would never use 50 V rated caps at 50 volts, this is much too close to the failure point, and they will wear out faster, if not immediately.
So if you MUST do what you are doing here are my suggestions:
- COPY the design of someone else from the links above, don't use your own design. The design you posted is messed up.
- connect about 10 diodes from your pos supply to the positive input of the chip, and 10 from your negative supply to the negative input of the chip--- make sure you orient these the right way! This will give you about .7 volts *10 = 7 volts drop and make your 50 volts into a more safe 43 volts, but I would use even more (.7 volt drop each) to drop it to 40 volts or 37 volts. Make sure you read up on diodes so you have foward current flow (if you put them backwards you just wont get any current flow and wont hurt anything except maybe your speaker.) Also use diodes rated for 5 amps and 100 volts (or more) (foward and reverse voltage) so you don't blow these up either. Buy these diodes at radio shack if you have to. They also have the other components you need.
-keep using with the heat gunk, good stuff.
-Don't touch anything.
You are still scaring me, you are way over your head.
Zplod, are you for real or just winding us all up?
I'm not looking for a sainthood but I'll tell you, it took a lot of time and effort to create my Gainclone FAQ page precisely to help people in your situation. And I'm not the only one by a long way.
People on this thread are desperately trying to help you so do your self a favour, listen to what is being said and do some homework before you take further advantage of the good nature of people on this forum.
I'm not looking for a sainthood but I'll tell you, it took a lot of time and effort to create my Gainclone FAQ page precisely to help people in your situation. And I'm not the only one by a long way.
People on this thread are desperately trying to help you so do your self a favour, listen to what is being said and do some homework before you take further advantage of the good nature of people on this forum.
Xplod,
Don't take this personally, but you really need to do more homework on the chip amp before you try to build one, especially a high-powered one.
I would suggest building a LM386 amp first. You can design it on a radioshack pre-etched ic board, and will run off a 9V battery.
It won't be able to power a sub, but it will get you into the theory behind chip amp operation.
Thats how I started, then I moved up to the LM383 (or something like that, it was an 8W chip, but is now discontinued). Then I moved up to the 3886. Nothing blew up, and everything worked as expected.
It sounds to me like this is a first attempt at an amp. Really, DON"T START WITH A HIGH POWER AMP, especially not the 4 3886 in parallel.
You have been warned, so don't come complaining to us when your new chip frys from 100V
Again, don't take anything above personal, were just warning you.
-Mike
Don't take this personally, but you really need to do more homework on the chip amp before you try to build one, especially a high-powered one.
I would suggest building a LM386 amp first. You can design it on a radioshack pre-etched ic board, and will run off a 9V battery.
It won't be able to power a sub, but it will get you into the theory behind chip amp operation.
Thats how I started, then I moved up to the LM383 (or something like that, it was an 8W chip, but is now discontinued). Then I moved up to the 3886. Nothing blew up, and everything worked as expected.
It sounds to me like this is a first attempt at an amp. Really, DON"T START WITH A HIGH POWER AMP, especially not the 4 3886 in parallel.
You have been warned, so don't come complaining to us when your new chip frys from 100V
Again, don't take anything above personal, were just warning you.
-Mike
lgreen said:
The transformer puts out 35v, but after the diodes, the voltage is 50v. The caps in the psu are rated 6800 uF 50v.
What is wrong with my design?
Nuuk said:
All this is for real.
I switched to a 25v psu. I'm not sure if it will put out enough power to run 4 lm3886s in parallel. There is a 2 amp fuse before and after the transformer, and the caps are 3300 uF 25v.
All I have to do now is tweak the value of the feedback resistor, to get the most power out of the chip. Then, I'll parallel 4 of them.
Do you have full control over the relation
load impedance/supply voltage/parallel conntection/bridging?
4 LM3886 in parallel seems to be a waste.
Check the guide to determine the best combination.
http://www.national.com/appinfo/audio/files/Overture_Design_Guide13.xls
http://www.national.com/appinfo/audio/files/Using_Overture_Design_Guide.pdf
You have read the datasheet and seen that the max voltage is 42V? No? 35 V AC is a little bit too much. 2 x 24-28 VAC is more suitable.
load impedance/supply voltage/parallel conntection/bridging?
4 LM3886 in parallel seems to be a waste.
Check the guide to determine the best combination.
http://www.national.com/appinfo/audio/files/Overture_Design_Guide13.xls
http://www.national.com/appinfo/audio/files/Using_Overture_Design_Guide.pdf
You have read the datasheet and seen that the max voltage is 42V? No? 35 V AC is a little bit too much. 2 x 24-28 VAC is more suitable.
ok, you'll need some way to get a lower supply voltage. the max is 84V and you have 100V. an impportant second Q is how much current the transformer is rated at?
for that config the gain needs to be 10. ignoring the capacitor, this means the(resistor attached to the input / resistor attached to ground) +1 >= 10.
next, the lower limit will be given by the RC circuit of the cap and resistor that go to ground in the "feedback path". it is a good idea to make them large, but realize that the other resistor needs to be 10x more.
for that config the gain needs to be 10. ignoring the capacitor, this means the(resistor attached to the input / resistor attached to ground) +1 >= 10.
next, the lower limit will be given by the RC circuit of the cap and resistor that go to ground in the "feedback path". it is a good idea to make them large, but realize that the other resistor needs to be 10x more.
tbla said:
are you still alive - give us a sign.........
Yes, I'm still alive.
I got the amp to work pretty good. The gain is set at 48. One problem that I have with it is when I turn it up more than half-way, it sounds really bad. I can hear cracking, and it puts out less power than when it's turned up half-way. It only puts out 12v to the speaker. And the speaker is not the source of the problem. Is my gain set too high?does the amp get hot fast?
i missed the 100V PSU down to 25V PSU. for the parallel chips you'll need a higher current transformer.
a high gain may be an issue.
do you have a complete design, including the PSU, and ratings on all parts? is the 25V PSU +-25V or just 25V? if it is just 25V, then you need to use the single supply schematic.
are you using the servos in the feedback network, or just resitors? what size resistor do you have on the ouptut of each chip?
i missed the 100V PSU down to 25V PSU. for the parallel chips you'll need a higher current transformer.
a high gain may be an issue.
do you have a complete design, including the PSU, and ratings on all parts? is the 25V PSU +-25V or just 25V? if it is just 25V, then you need to use the single supply schematic.
are you using the servos in the feedback network, or just resitors? what size resistor do you have on the ouptut of each chip?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.