Hi Guys, I have been working on this Guitar amp useing a LM3886 as the power amp for a Long time now and I still am haveing the same problem with low output power....
At first I thought it was because my preamp wasn"t putting out enough signal to drive the LM3886 but now I suspect that is not the case now....
I used the Data sheet example for building the Power Amp stage with the LM3886 and I know for absolute certainty that i built it correctly as I have built several revisions and still have the same problem....
I am powering it with a 100vA 2x18v Toroid with 8000uF of filtering on each rail...I am driveing an 8 Ohm load and the Preamp has a total Gain of 500, But I also tried bypassing the preamp and going directly from a distortion pedal right into the LM3886 and still don"t get very much volume....
At full volume the volume is about Listening level for a TV set ,There is no Noise of hum just very little volume....
I also noticed that the Heatsink gets VERY hot very Fast, The heatsink is about 6 in wide and 3.5in Tall and about 3 in Deep but it is made from thin amuminum.....
I don"t know what the problem could be and I have rebuilt this amp from top to bottom 3 times allready with the same problem every time and I used different preamp designs every time and redesigned the PCB for the LM3886 several times allready......
The only thing I can think of that I haven"t tried yet was I haven"t tried a different Transformer...Do you think If I treid running it at a higher voltage like 30v I would get better results??
Any help would be great....
Thanx
At first I thought it was because my preamp wasn"t putting out enough signal to drive the LM3886 but now I suspect that is not the case now....
I used the Data sheet example for building the Power Amp stage with the LM3886 and I know for absolute certainty that i built it correctly as I have built several revisions and still have the same problem....
I am powering it with a 100vA 2x18v Toroid with 8000uF of filtering on each rail...I am driveing an 8 Ohm load and the Preamp has a total Gain of 500, But I also tried bypassing the preamp and going directly from a distortion pedal right into the LM3886 and still don"t get very much volume....
At full volume the volume is about Listening level for a TV set ,There is no Noise of hum just very little volume....
I also noticed that the Heatsink gets VERY hot very Fast, The heatsink is about 6 in wide and 3.5in Tall and about 3 in Deep but it is made from thin amuminum.....
I don"t know what the problem could be and I have rebuilt this amp from top to bottom 3 times allready with the same problem every time and I used different preamp designs every time and redesigned the PCB for the LM3886 several times allready......
The only thing I can think of that I haven"t tried yet was I haven"t tried a different Transformer...Do you think If I treid running it at a higher voltage like 30v I would get better results??
Any help would be great....
Thanx
I don't know much about the protection circuitry used on these devices, but it sounds like the amp may be oscillating or you may have enough of a short between the output and ground causing the amp to source a lot of current, heat up a lot, and go slightly into thermal/overcurrent protection. Have you measured the impedance of the load you use with the amp or the impedance between the output terminal and ground?
Also, a gain of 500 for a preamp is incredibly high. What is the reason for that and what gain do you have the LM3886 configured for?
Also, a gain of 500 for a preamp is incredibly high. What is the reason for that and what gain do you have the LM3886 configured for?
I built the Preamp with a Gain of 500 because each Preamp I tried didn"t seem to have enought to drive the LM3886....
I have no Idea what the gain of the LM3886 is configured for as I used the Data sheet schematic so it is whatever that one is whatever that is.....
I noticed that the schematic for the LM3886 doesn"t have an Input cap, Could that cause this??
The impedance of the Signal going into the LM3886 is pretty low as I have a NPN Transistor buffer before the LM3886 so the impedance in probably under 1k......
Would a Low input impedance cause a Low output??
I also tried Connecting a CD Player to it and it also had very low output.....
I am really Really stumped and a Little pissed off as I have been at this working allmost every day for over a month and I still can not get very much power.....
I don"t think it is Oscilateing because I know what that sounds like and that isn"t what it sounds like.....
It actually sounds very good it just has absolutly no power, my headphones sitting on that table are louder than the amp cranked to it"s Max, I"m probably getting less than 1 watt of output power......
Thanx
I have no Idea what the gain of the LM3886 is configured for as I used the Data sheet schematic so it is whatever that one is whatever that is.....
I noticed that the schematic for the LM3886 doesn"t have an Input cap, Could that cause this??
The impedance of the Signal going into the LM3886 is pretty low as I have a NPN Transistor buffer before the LM3886 so the impedance in probably under 1k......
Would a Low input impedance cause a Low output??
I also tried Connecting a CD Player to it and it also had very low output.....
I am really Really stumped and a Little pissed off as I have been at this working allmost every day for over a month and I still can not get very much power.....
I don"t think it is Oscilateing because I know what that sounds like and that isn"t what it sounds like.....
It actually sounds very good it just has absolutly no power, my headphones sitting on that table are louder than the amp cranked to it"s Max, I"m probably getting less than 1 watt of output power......
Thanx
Minion said:
This could be very important. What you're saying is the amp you made is DC coupled. That means any DC at the input will show up at the output multiplied by the gain. This could cause a large DC offset, which will force a lot of current to flow through the chip and the load. Try placing a coupling cap between your volume pot and the input resistor.
Just out of curiosity, when you power up the amp is there a very large turn on thump? And what is the DC offset of your preamp?
Well there is an Ouput Cap on the Preamp (o.22uF) so I guess that should block DC from the Input......
There is a Little crackle I guess when you turn it on but no more than when turning on any amp ....
Well the DC offset of the Output Chip on the preamp is something like 1mV (OPA134)...There is actually a Overdrive circuit between the Preamp and Power amp.....
I believe the Gain for the LM3886 is configured for a Gain of 20 (1k Input resistor and 20k feedback resistor).....
I am going to try a few different things, I"ll try putting a Input Cap on the Power amp stage....I am also going to try increaseing the Feedback resistor to get more gain (maybe switch from a 20k to a 33k)... I am also going to try a Bigger Higher Voltage Transformer, I have a 24v,0v,24v 200Va EI Transformer that I might try to get a higher input Voltage.....
Would putting a Cap on the output also help with DC Offset??
Thanx
There is a Little crackle I guess when you turn it on but no more than when turning on any amp ....
Well the DC offset of the Output Chip on the preamp is something like 1mV (OPA134)...There is actually a Overdrive circuit between the Preamp and Power amp.....
I believe the Gain for the LM3886 is configured for a Gain of 20 (1k Input resistor and 20k feedback resistor).....
I am going to try a few different things, I"ll try putting a Input Cap on the Power amp stage....I am also going to try increaseing the Feedback resistor to get more gain (maybe switch from a 20k to a 33k)... I am also going to try a Bigger Higher Voltage Transformer, I have a 24v,0v,24v 200Va EI Transformer that I might try to get a higher input Voltage.....
Would putting a Cap on the output also help with DC Offset??
Thanx
Hmm... DC offset no longer looks to be the culprit. I'm out of ideas if you don't have a semi-short at the output or it's oscillating badly at a frequency you can't hear.
The gain for a noninverting config is (1+Rf/Ri), so 21 for your case.
Putting a cap on the output will help DC offset, but will also form a high pass filter with the load. You would need to use a rather large value electrolytic cap (of the appropriate voltage rating) to obtain a low corner frequency.
Good luck.
The gain for a noninverting config is (1+Rf/Ri), so 21 for your case.
Putting a cap on the output will help DC offset, but will also form a high pass filter with the load. You would need to use a rather large value electrolytic cap (of the appropriate voltage rating) to obtain a low corner frequency.
Good luck.
Well here is the PCB design I used,...
Can you see something wrong with this layout that would cause this problem??
Also with my Grounding scheme I have all off the Audio grounds come together to one point and I have AC ground connected to the audio ground with 2 opposite faceing 1N4007 Diodes.....
Anyways, I"m off to try to implement some of my changes...
Cheers
PS: I also just realized that my preamp is a Bit screwed but even going from my Guitar through the Overdrive circuit and into the LM3886 I still get low output....
Can you see something wrong with this layout that would cause this problem??
Also with my Grounding scheme I have all off the Audio grounds come together to one point and I have AC ground connected to the audio ground with 2 opposite faceing 1N4007 Diodes.....
Anyways, I"m off to try to implement some of my changes...
Cheers
PS: I also just realized that my preamp is a Bit screwed but even going from my Guitar through the Overdrive circuit and into the LM3886 I still get low output....
Where is the layout?
I did a silly thing when I designed mine. I misinterpreted the use of the muting pin and didn't wire it to V-. I just had it connected to a cap and that was that. I simply put a 22k resistor between that pin and V- and everytihng worked straight away. Yoy MIGHT have this problem, but I'm also leaning towards oscillation. You can't take a picture of what you have done to make thing a little clearer?
Cheers
Gert
I did a silly thing when I designed mine. I misinterpreted the use of the muting pin and didn't wire it to V-. I just had it connected to a cap and that was that. I simply put a 22k resistor between that pin and V- and everytihng worked straight away. Yoy MIGHT have this problem, but I'm also leaning towards oscillation. You can't take a picture of what you have done to make thing a little clearer?
Cheers
Gert
Sorry , Here"s the Layout.....
Well I just tried it with a 200Va 30v,0v,30v AC transformer with a 4.7uF Silver Mica cap before the Power amp stage and I changed the Feedback resistor from a 33k to a 44k and I put a Bigger heatsink of the Chip ,One from a Socket A CPU....
Well it made absolutly no differance in Volume....
It isn"t the Muteing fuction as the Mute switch works.....
If it is Oscilateing how do I stop it from oscilateing??
I have experienced Oscilateing opamps and they made an awefull sound but this amp doesn"t sound awefull, It is actually quite quiet and makes a good crunchy guitar sound it just is at a low talking volume.....
It also heated up very quickly with the larger heatsink and goes into over heat protection in about 60 seconds, The heatsink didn"t even have time to heat up before the chips spike kicks in, I also have thermal compound on the Chip...
Does anyone have a Single sided PCB Design for a prooven working Layout?? and maybe a Good Guitar preamp to go with it??
Thanx
Well I just tried it with a 200Va 30v,0v,30v AC transformer with a 4.7uF Silver Mica cap before the Power amp stage and I changed the Feedback resistor from a 33k to a 44k and I put a Bigger heatsink of the Chip ,One from a Socket A CPU....
Well it made absolutly no differance in Volume....
It isn"t the Muteing fuction as the Mute switch works.....
If it is Oscilateing how do I stop it from oscilateing??
I have experienced Oscilateing opamps and they made an awefull sound but this amp doesn"t sound awefull, It is actually quite quiet and makes a good crunchy guitar sound it just is at a low talking volume.....
It also heated up very quickly with the larger heatsink and goes into over heat protection in about 60 seconds, The heatsink didn"t even have time to heat up before the chips spike kicks in, I also have thermal compound on the Chip...
Does anyone have a Single sided PCB Design for a prooven working Layout?? and maybe a Good Guitar preamp to go with it??
Thanx
Attachments
What LM3886 schematic you are using?
Do you have any 0.1UF on chip?
Also, I saw my friends using this chip as guitar amp but they use two gain stages. It run beautifully.
>>
If you decide to rebuild a new one, you can keep the amp but change the pre- to the following. Many ideas and technics are there as well.
http://sound.westhost.com/project27.htm
http://sound.westhost.com/project27b.htm
Do you have any 0.1UF on chip?
Also, I saw my friends using this chip as guitar amp but they use two gain stages. It run beautifully.
>>
If you decide to rebuild a new one, you can keep the amp but change the pre- to the following. Many ideas and technics are there as well.
http://sound.westhost.com/project27.htm
http://sound.westhost.com/project27b.htm
How big is your heat sink? If too small that could be a reason for getting hot.
Which version of the LM3886 did you use? perhaps there is a partial short to the heat sink. What rail voltages do you get with your 18V transformer - it should be around 25V. That should get you 22V peak output, roughly 15V RMS. That's only 25W into an 8 ohm driver.
Have you tried feeding a signal just to the chip amp input and measured the output voltage? You could use a test tone generated by software or in a pinch just an open E should be useful. Measure the amp section input voltage and output voltage. Is the output approximately 20 times the input?
Measure with and without the speaker - there may be a partial/intermittent short in the speaker wiring. If you swing more output voltage without the speaker you've found your problem.
Did you use a zoebel on the output? What values did you use? If the cap is bigger than 100 nf (0.1 uf) you are probably loading down the output with the zoebel.
With a gain of 500 you're going to seriously overdrive the amp. I use around 40 with my bass preamp.
Edit: It's hard to see, but it looks like you have a 10 ohm resistor in series with the output. 10 ohms/100 nf makes a good zoebel, but you don't want that in series with your speaker. You'll get less than half the voltage the chip amp swings on your speaker. You don't need a resistor in series with the output at all unless you are paralleling chips.
Which version of the LM3886 did you use? perhaps there is a partial short to the heat sink. What rail voltages do you get with your 18V transformer - it should be around 25V. That should get you 22V peak output, roughly 15V RMS. That's only 25W into an 8 ohm driver.
Have you tried feeding a signal just to the chip amp input and measured the output voltage? You could use a test tone generated by software or in a pinch just an open E should be useful. Measure the amp section input voltage and output voltage. Is the output approximately 20 times the input?
Measure with and without the speaker - there may be a partial/intermittent short in the speaker wiring. If you swing more output voltage without the speaker you've found your problem.
Did you use a zoebel on the output? What values did you use? If the cap is bigger than 100 nf (0.1 uf) you are probably loading down the output with the zoebel.
With a gain of 500 you're going to seriously overdrive the amp. I use around 40 with my bass preamp.
Edit: It's hard to see, but it looks like you have a 10 ohm resistor in series with the output. 10 ohms/100 nf makes a good zoebel, but you don't want that in series with your speaker. You'll get less than half the voltage the chip amp swings on your speaker. You don't need a resistor in series with the output at all unless you are paralleling chips.
The resistor is wrapped with wire as an Inductor.....
I know it isn"t the Speaker as I have a Crossover network and I have tried connecting directly to the speakers without the crossover and with the crossover....
as For a Zobel Network I don"t know what one is...I simply used the exact schematic from the Datasheet.....
The heatsink I am useing if from a socket A CPU so I thought it should be good enough and the one I used before that was used on a 120w stereo Chip so I thought that would be good enough....
The chips I am useing are the Isolated ones so they don"t transfer heat very well.....
I haven"t tested the output Voltage But I suspect it isn"t very much as if it were very high I would be able to hear something.....
I also tried the 30v,0v,30v 200vA Transformer without any increase in Volume......
Could it have something to do with my Power supply??
I am basicly useing a simple Bridge rectifier with 16000uF of total filtering going into the LM3886 chip, no other components between the Transformer and the LM3886, just Caps and a bridge rectifier......
well I"m totally stumped.....
Thanx
I know it isn"t the Speaker as I have a Crossover network and I have tried connecting directly to the speakers without the crossover and with the crossover....
as For a Zobel Network I don"t know what one is...I simply used the exact schematic from the Datasheet.....
The heatsink I am useing if from a socket A CPU so I thought it should be good enough and the one I used before that was used on a 120w stereo Chip so I thought that would be good enough....
The chips I am useing are the Isolated ones so they don"t transfer heat very well.....
I haven"t tested the output Voltage But I suspect it isn"t very much as if it were very high I would be able to hear something.....
I also tried the 30v,0v,30v 200vA Transformer without any increase in Volume......
Could it have something to do with my Power supply??
I am basicly useing a simple Bridge rectifier with 16000uF of total filtering going into the LM3886 chip, no other components between the Transformer and the LM3886, just Caps and a bridge rectifier......
well I"m totally stumped.....
Thanx
DOH! Check your feedback network resistors. You wouldn't be the first to swap their positions giving you virtually no gain. The LM3886 isn't unity gain stable, so there is the likely cause of no gain and oscillation.
Use your multimeter to measure. No multimeter? Pick up a cheapie at Radio Shack or a hardware store. Most useful for troubleshooting and we don't need real accuracy here, just a bit better than order of magnitude.
If that isn't it, we really need to see some voltages. Pictures of your wiring layout would help, too. If the main caps aren't on the board are there caps on the board from rails to ground? - 100 uf and a 1 uf film would be good.
Use your multimeter to measure. No multimeter? Pick up a cheapie at Radio Shack or a hardware store. Most useful for troubleshooting and we don't need real accuracy here, just a bit better than order of magnitude.
If that isn't it, we really need to see some voltages. Pictures of your wiring layout would help, too. If the main caps aren't on the board are there caps on the board from rails to ground? - 100 uf and a 1 uf film would be good.
Geesh, it is a tough one.
That Socket A sink is definately too small. I've used pretty big ones on mine, see here post #8 and they get pretty warm when I work them hard into 4ohms. But you aren't getting volume so you shouldn't be generating any heat.
Do you have access to a 'scope? If so, have a look at the output signal with the input shorted to ground. See if the chip is oscillating. Also (with a DMM) measure your DC output voltage and report back. This should be 0V. I've seen some guys actually seeing a couple of mV on the output, although I see none on mine.
That is the only next logical step you can take, I think.
That Socket A sink is definately too small. I've used pretty big ones on mine, see here post #8 and they get pretty warm when I work them hard into 4ohms. But you aren't getting volume so you shouldn't be generating any heat.
Do you have access to a 'scope? If so, have a look at the output signal with the input shorted to ground. See if the chip is oscillating. Also (with a DMM) measure your DC output voltage and report back. This should be 0V. I've seen some guys actually seeing a couple of mV on the output, although I see none on mine.
That is the only next logical step you can take, I think.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.