I have this highly unusual problem. I am trying to build a simple lm3886 amp no GC or anything just the standard no inverting mode. My power supply is made from 25-0-25 torid feeding two rectifiers one per rail and four 68000uf 50V caps two per rail. I am having this strange problem with the supply. I am getting the expected 35v DC per rail but I am also getting 77VAC per rail this is with no load. This is measured accorss the capacitor terminals. When I disconnect the mains I am still getting the ac voltage! I have an only basic understanding of electronics but I don’t think that it is possible to have 77VAC coming out of a power supply that is connected to no mains or loaded. This problem comes and goes. I experienced this problem some weeks ago with the ac and shelved the amp cause it was wrecking my head. The picture is here http://www.pundip.com/amp/powersupply.jpg I couldn’t get it to attach to this post. The grey block in the picture is a AC filter for the mains. I have bypassed it with no effect. I don’t think that this is an instrument error as the ac steadily declines as the caps discharge on their own. This picture was taken with the mains unplugged. The only way I can explain this is alien intervention. Anybody have a better idea?
PS When the supply seemed to work ok I was getting 35V dc on the output. The chip was getting both rails. There was 100MV dc on the input pin.
Thanks in advance
Pundip
PS When the supply seemed to work ok I was getting 35V dc on the output. The chip was getting both rails. There was 100MV dc on the input pin.
Thanks in advance
Pundip
Ask me a hard one...
Hmm. If you connect a resistor of a few k across the caps, does the level drop? Assuming it does, the following may be happening:
(Edit: I see you write it DOES drop. So there we are).
This meter measures AC by single-polarity rectification; the lowest AC scale being 600V gives it away. It will also assume it sees a sine wave (actually, it will be calibrated to a sine wave RMS factor, it's a cheapo meter, sorry).
What I suspect happens is that on the AC scale, there is a rectifier that rectifies one peak voltage and scales that to what it thinks was the original AC voltage level. But you put DC on it. No matter, the rectifier dutifully lets it through and the meter is confused in thinking this DC has been produced by an AC level of 150V. I would expect half of this, but then again I don't know how the internal scaling works.
Try reversing the leads on the cap, there's no danger, on AC scale the meter can handle either polarity right? I bet you get zero or close to it. Or maybe you get the same, that would indicate dual-polarity rectification. But I doubt it.
dhaen: a 150VAC stray field from mains across a supply cap??? tsk, tsk.
Jan Didden
Hmm. If you connect a resistor of a few k across the caps, does the level drop? Assuming it does, the following may be happening:
(Edit: I see you write it DOES drop. So there we are).
This meter measures AC by single-polarity rectification; the lowest AC scale being 600V gives it away. It will also assume it sees a sine wave (actually, it will be calibrated to a sine wave RMS factor, it's a cheapo meter, sorry).
What I suspect happens is that on the AC scale, there is a rectifier that rectifies one peak voltage and scales that to what it thinks was the original AC voltage level. But you put DC on it. No matter, the rectifier dutifully lets it through and the meter is confused in thinking this DC has been produced by an AC level of 150V. I would expect half of this, but then again I don't know how the internal scaling works.
Try reversing the leads on the cap, there's no danger, on AC scale the meter can handle either polarity right? I bet you get zero or close to it. Or maybe you get the same, that would indicate dual-polarity rectification. But I doubt it.
dhaen: a 150VAC stray field from mains across a supply cap??? tsk, tsk.
Jan Didden
Yep you were right
Thanks it worked I changed polarity and the ac was zero. That also explains why some times I got no ac. You have no idea how many times I banged my head over this. Perhaps you can help me with another problem. I am getting 35V DC on the output of the amp. I am using that standard nsc schematic the only difference is that the 10k pot is not connected and I have put an input cap. The chip is getting both rails. I put a 1k resistor input and ground the dc goes and nothing comes out the output. This could be a muting problem but the DC come on as soon as I plug the amp in. Does this mean the delayed turn on is not working? Well any advice on the DC on the output would be greatly appreciated.
Thanks
Thanks it worked I changed polarity and the ac was zero. That also explains why some times I got no ac. You have no idea how many times I banged my head over this. Perhaps you can help me with another problem. I am getting 35V DC on the output of the amp. I am using that standard nsc schematic the only difference is that the 10k pot is not connected and I have put an input cap. The chip is getting both rails. I put a 1k resistor input and ground the dc goes and nothing comes out the output. This could be a muting problem but the DC come on as soon as I plug the amp in. Does this mean the delayed turn on is not working? Well any advice on the DC on the output would be greatly appreciated.
Thanks
tsk tsk indeed..
Who said 150v? Actually I've seen about 50v, but of course that was on a higher voltage secondary than Pundip was using...
My mistake to assume that he was measuring the AC on the transformer: hence my comment about finger trouble
Pundip, I don't know the circuit of your amp, but you probably do need a resistor from the input to ground (=0v on power supply), but 1k is far too low. Maybe 100k is better.
Does nsc mean National Semiconductors, and this is a chip amp?
Well diagnosed Jan
Who said 150v? Actually I've seen about 50v, but of course that was on a higher voltage secondary than Pundip was using...
My mistake to assume that he was measuring the AC on the transformer: hence my comment about finger trouble
Pundip, I don't know the circuit of your amp, but you probably do need a resistor from the input to ground (=0v on power supply), but 1k is far too low. Maybe 100k is better.
Does nsc mean National Semiconductors, and this is a chip amp?
so far so good
dhaen ,
did a bit more research and turns out that you are right the input needs to be grounded. I have put a 4k7k to ground straight from the input pin that is after the 1k input resistor and the input cap. The amp is working fine, no dc on the output no distortion it all seems pretty good its just that I cant get the volume very high just using a old discman(wont miss it of it got fried) headphone socket. I am using lm3886t from National Semiconductor. It’s a standard cuircuit on the first page of the datasheet. 1k input 20k feedback non inverting mode. –in is grounded via 1k resistor and 22uf cap in series. Does it matter if put the 47k on my +in before or after my input resistor. Here what the thing looks like at the moment http://www.pundip.com/amp/amp1.jpg maybe i need a preamp? My 15-0-15 secondary supply is almost ready.
Thanks for all the help so far.
THis is sort of my first amp project. Last time I picked up a soldering iron was 6 years ago and even then there was smoke fire explosions and once I was even resposible for blowing the main fuses for my entire dorm.
pundip
dhaen ,
did a bit more research and turns out that you are right the input needs to be grounded. I have put a 4k7k to ground straight from the input pin that is after the 1k input resistor and the input cap. The amp is working fine, no dc on the output no distortion it all seems pretty good its just that I cant get the volume very high just using a old discman(wont miss it of it got fried) headphone socket. I am using lm3886t from National Semiconductor. It’s a standard cuircuit on the first page of the datasheet. 1k input 20k feedback non inverting mode. –in is grounded via 1k resistor and 22uf cap in series. Does it matter if put the 47k on my +in before or after my input resistor. Here what the thing looks like at the moment http://www.pundip.com/amp/amp1.jpg maybe i need a preamp? My 15-0-15 secondary supply is almost ready.
Thanks for all the help so far.
THis is sort of my first amp project. Last time I picked up a soldering iron was 6 years ago and even then there was smoke fire explosions and once I was even resposible for blowing the main fuses for my entire dorm.
pundip
typo
sorry I meant to say 47k like you suggested. Do I need a preamp of some sort to drive this amp? I did a quick test and my amp does not go very loud. A friend of mine make a lm3886 and drove it with a laptop sound card and eveybody in the building knew when he was playing duke nukem 3d. I am using a 330va trany so that cant be it.
sorry I meant to say 47k like you suggested. Do I need a preamp of some sort to drive this amp? I did a quick test and my amp does not go very loud. A friend of mine make a lm3886 and drove it with a laptop sound card and eveybody in the building knew when he was playing duke nukem 3d. I am using a 330va trany so that cant be it.
I think you need more gain...pals...he needs more GAIN~!!!...you friend probably has plugged his amp into the headphones output...there's already an amp for the headphone...so another amp there...gainclone plus the headphone amp (internal)...adds to more gain and louder volume...it's not really recommended though...
Re: tsk tsk indeed..
I realised that, but couldn't resist the comment
I looked at the picture on the link he gave and there were the test probes, sitting on the caps.
Have a nice Sunday John,
Jan Didden
dhaen said:
[snip]My mistake to assume that he was measuring the AC on the transformer: hence my comment about finger trouble
[snip]
I realised that, but couldn't resist the comment
I looked at the picture on the link he gave and there were the test probes, sitting on the caps.
Have a nice Sunday John,
Jan Didden
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.