i have built lm3876tf based amp
it is wonderful.good power and great sound reproduction.but achilles heel is that the ic becomes very hot.load is 4ohm but the prob. continues even eith 8ohm.ihav used it in n.i mode with no cap at i/p signal.the feedback res. is 47k.voltage is very safe 26v at 5amps(power supply is totally clear with fast recifiers and adequate capacitance)
big heat sink that remains cool i.e the ic becomes very hot very soon.is it normal for it?my tda1554q and8563 remained warm during full power
but never as hot(untouchable) as it gets.one more,the ic is tf version i.e insulated version.
thanks in advance for help
it is wonderful.good power and great sound reproduction.but achilles heel is that the ic becomes very hot.load is 4ohm but the prob. continues even eith 8ohm.ihav used it in n.i mode with no cap at i/p signal.the feedback res. is 47k.voltage is very safe 26v at 5amps(power supply is totally clear with fast recifiers and adequate capacitance)
big heat sink that remains cool i.e the ic becomes very hot very soon.is it normal for it?my tda1554q and8563 remained warm during full power
but never as hot(untouchable) as it gets.one more,the ic is tf version i.e insulated version.
thanks in advance for help
Is the heatsink surface smooth? Do you have "sweat paste" (white silicone grease) between the IC and the heatsink?
Ic heat up!!!!
Hi sagar,
Normally the amp will heat up when fully driven, even the Tf series, why don't you check them if the amp is heating up in no load condition.
If its hot in no load , then might be your amp is Oscillating.
why don't you post the schematics .
thanks
Arasuk🙂
Hi sagar,
Normally the amp will heat up when fully driven, even the Tf series, why don't you check them if the amp is heating up in no load condition.
If its hot in no load , then might be your amp is Oscillating.
why don't you post the schematics .
thanks
Arasuk🙂
probably u dont have good contact between chip and heatsink, check both surfaces, try to put thermal grease...
Member
Joined 2003
Also chech to make sure that the IC is flush to the heatsink. If there is even the smallest air gap between the IC and the heatsink there will be tremendous loss of heat transfer.
Initially, I did have the same problem when I use veroboard to built my LM3876TF amp. When i soldered all the components directly to the IC, the problem sloved. May be the terminal between the IC and the components are not properly contacted. Did you use PCB?
combined reply to all
1)i have used pcb made with photographic process.
2)soldering is perfect
3)i have made it sure that ic is totally in contact with sink.initially i used heatsink compound(si grease) so that there is no air gap and noticed that the ic becomes too hot too soon.so removed all grease and reconnected the ic to sink anticipating that grease was barring the conduction but the problem persists.
4)at no load,the amp is cool as a cucumber.the feedback res is 47k.
5)heat sink surface is perfectly smooth.the overheating persists with or without grease.
the ic becomes too hot too soon .after a min or two of operation the heat sink also warms up(it a big one) but the ic temp remains remains too high for the ic to remain untouchable even at medium power.
what i feel is that gain resistance might be the cause.does anyone has experience on that?can it be problem?shuld gain be lowered?
heats sink worked well for tda series?is tf version problematic in terms of heat dissipiation??
four ohm or 8 ohm,prob persists.
check schematics i m posting the link:
http://www.generalguitargadgets.com/pdf/ggg_lm3886_amp.pdf
see the circuit and tell me
changes in circuit from me.
no 10k trimmer.
r1 changed to 47k
r4 changed to 47k
10ohm 5watts res. used
12 turns of wire
thanks in advace for help.
-------------------------------------------------------------------------------
listened at full power only for10sec in 3days
1)i have used pcb made with photographic process.
2)soldering is perfect
3)i have made it sure that ic is totally in contact with sink.initially i used heatsink compound(si grease) so that there is no air gap and noticed that the ic becomes too hot too soon.so removed all grease and reconnected the ic to sink anticipating that grease was barring the conduction but the problem persists.
4)at no load,the amp is cool as a cucumber.the feedback res is 47k.
5)heat sink surface is perfectly smooth.the overheating persists with or without grease.
the ic becomes too hot too soon .after a min or two of operation the heat sink also warms up(it a big one) but the ic temp remains remains too high for the ic to remain untouchable even at medium power.
what i feel is that gain resistance might be the cause.does anyone has experience on that?can it be problem?shuld gain be lowered?
heats sink worked well for tda series?is tf version problematic in terms of heat dissipiation??
four ohm or 8 ohm,prob persists.
check schematics i m posting the link:
http://www.generalguitargadgets.com/pdf/ggg_lm3886_amp.pdf
see the circuit and tell me
changes in circuit from me.
no 10k trimmer.
r1 changed to 47k
r4 changed to 47k
10ohm 5watts res. used
12 turns of wire
thanks in advace for help.
-------------------------------------------------------------------------------
listened at full power only for10sec in 3days
You not applying the thermal paste a bit too thick?
Remember it doesn't so much transfer heat as displacing microscopic air pockets....
Remember it doesn't so much transfer heat as displacing microscopic air pockets....
if no load=no heat
dude , your amp is oscilating , try to connect the output directly to the speaker (before the snubber network)
dude , your amp is oscilating , try to connect the output directly to the speaker (before the snubber network)
Have you checked with the AN-1192 and the design guide how much power at used supply voltage you can expect? The TF package has it's limitations. High power application requires the T package.
http://www.national.com/appinfo/audio/files/Overture_Design_Guide13.xls
http://www.national.com/appinfo/audio/files/Using_Overture_Design_Guide.pdf
http://www.national.com/an/AN/AN-1192.pdf
http://www.national.com/appinfo/audio/files/Overture_Design_Guide13.xls
http://www.national.com/appinfo/audio/files/Using_Overture_Design_Guide.pdf
http://www.national.com/an/AN/AN-1192.pdf
i used good qty. of grease.
sss:😛lz. explain properly how to remove oscillations(ic is jus warm at no load ,i suppose u gettin me)
peranders:::i m using safe +-26v.so voltage no concern.saw an1192 before making pcb
sss:😛lz. explain properly how to remove oscillations(ic is jus warm at no load ,i suppose u gettin me)
peranders:::i m using safe +-26v.so voltage no concern.saw an1192 before making pcb
Can't you get continuous output power or will the overheat proctection trigger? Hot chip is normal as long as you'll get proper sound out of it.
hi peranders,
the o/p is brilliant,no clipping or distortion etc.the thermal shutdown never occured bec i lower the volume as soon as the chip hots up too much( i touch it to see if things' getting hot)
the o/p is brilliant,no clipping or distortion etc.the thermal shutdown never occured bec i lower the volume as soon as the chip hots up too much( i touch it to see if things' getting hot)
peranders::
have u seen the circuit that i have posted.i have used high value of gain.can high gain be bad?
have u seen the circuit that i have posted.i have used high value of gain.can high gain be bad?
How high gain do you have really? If you can rule out that you have no oscillations I see no special disadvantage with high gain.
Have you checked with an oscilloscope that you don't have any unstability?
Do you have any output filter, zobel network etc.?
Still, if the chip is MUCh hotter than the heatsink you may have a heat conducting problem. Can't you take a photo of the amp?
Have you checked with an oscilloscope that you don't have any unstability?
Do you have any output filter, zobel network etc.?
Still, if the chip is MUCh hotter than the heatsink you may have a heat conducting problem. Can't you take a photo of the amp?
If you've followed the schematic to the letter you have no DC blocking. If there's DC at the input that would appear on the output in proportion to the setting of the volume knob, which might cause excessive heat.
Feedback resistance does not affect heat level, as long as the amp doesn't oscillate. Heat is simply a function of the voltage and current, regardless of amplifier type.
The TF package has a harder time getting rid of the heat, but remember that what feels hot to the fingers isn't necessarily particularly high temperature. 60C feels really hot, but is no problem for the chip. And remember the chip is protected internally, so it will cut the volume if necessary. Let it sing and see what happens!
Rune
Feedback resistance does not affect heat level, as long as the amp doesn't oscillate. Heat is simply a function of the voltage and current, regardless of amplifier type.
The TF package has a harder time getting rid of the heat, but remember that what feels hot to the fingers isn't necessarily particularly high temperature. 60C feels really hot, but is no problem for the chip. And remember the chip is protected internally, so it will cut the volume if necessary. Let it sing and see what happens!
Rune
Good point Rune, DC level at the outout especially if the gain his high.
sagarverma, show as the whole circuit including how much DC you have from your signal source. Only interesting if the amp has no input coupling cap.
sagarverma, show as the whole circuit including how much DC you have from your signal source. Only interesting if the amp has no input coupling cap.
No sure I'll understand you correctly but do you mean that you trigger the overtemp pretection?sagarverma said:
You should only use a very small amount of heat sink grease - it's really just to ensure that tiny inperfections in the surfaces are filled so you get the maximum surface area conducting heat and to make sure there are no air gaps.
I have built about 150 PCs and servers and always do this when seating a heatsink or water block.
Put a grain of rice sized bit of thermal compound on each surface, use a credit card or similar to make a very thin layer or compound on each surface, place the surfaces together and gently move the chip / heatsink about a bit, then attach it.
Any more then a really thin layer of compound actually reduces the thermal transfer to the heatsink.
Also ensure both surfaces and completely flat and free from any manufacturing defects.
I have built about 150 PCs and servers and always do this when seating a heatsink or water block.
Put a grain of rice sized bit of thermal compound on each surface, use a credit card or similar to make a very thin layer or compound on each surface, place the surfaces together and gently move the chip / heatsink about a bit, then attach it.
Any more then a really thin layer of compound actually reduces the thermal transfer to the heatsink.
Also ensure both surfaces and completely flat and free from any manufacturing defects.
- Status
- Not open for further replies.