Hi,
I have been looking for low powered (e.g. ~1w, 9v battery supply) mono chip amp that can be used in an inverting configuration so that I can sum the stereo inputs without an additional summing stage.
Most of the low powered chips seem to have fixed gain set by internal feedback resistors and therefore, I don't think, could be used in the way I want to use them.
I had identified the TDA2003 as a suitable candidate as the feedback network is external to the chip, however, I am having difficulty understanding the feedback circuit which has a lp RC filter paralleled with Rf (R1) and a 470uf decoupling cap in the feedback path.
Datasheet
The datasheet includes some equations for working out the RC filter values based on bandwidth. Can anyone recommend some further reading which would explain this arrangement and the equations? Can I invert this amp simply by moving R2 to the inverting input?
It seems more common for a single cap to be used paralleled with RF and for the decoupling cap to be much smaller and go to ground in series with R2.
I am aware of the national boomer chips, but would prefer something easy to solder. Lm1877 is also a stereo option.
Any advice/suggestions would be much appreciated.
Thanks,
Ross
I have been looking for low powered (e.g. ~1w, 9v battery supply) mono chip amp that can be used in an inverting configuration so that I can sum the stereo inputs without an additional summing stage.
Most of the low powered chips seem to have fixed gain set by internal feedback resistors and therefore, I don't think, could be used in the way I want to use them.
I had identified the TDA2003 as a suitable candidate as the feedback network is external to the chip, however, I am having difficulty understanding the feedback circuit which has a lp RC filter paralleled with Rf (R1) and a 470uf decoupling cap in the feedback path.
Datasheet
The datasheet includes some equations for working out the RC filter values based on bandwidth. Can anyone recommend some further reading which would explain this arrangement and the equations? Can I invert this amp simply by moving R2 to the inverting input?
It seems more common for a single cap to be used paralleled with RF and for the decoupling cap to be much smaller and go to ground in series with R2.
I am aware of the national boomer chips, but would prefer something easy to solder. Lm1877 is also a stereo option.
Any advice/suggestions would be much appreciated.
Thanks,
Ross
Hi,
this chipsheet shows a minimum supply of 8V.
A new 9V battery under load will show about this voltage.
The chip draws ~40mA quiescent current.
The battery will not last long.
Have a look for high efficiently digital power chips designed specifically for battery operation.
this chipsheet shows a minimum supply of 8V.
A new 9V battery under load will show about this voltage.
The chip draws ~40mA quiescent current.
The battery will not last long.
Have a look for high efficiently digital power chips designed specifically for battery operation.
Thanks Andrew - good point (wasn't thinking!).
Unless I am mistaken, all the class D chips are small pitch leads and surface mount chips. This is for a class project for students who will have only basic soldering experience.
It seems the likes of lm38X, lm390, tba820M and TDA7052 are the only real options for me. None of these can be inverted, so looks like there is not an obvious single chip solution to this. Are you able to recommend any low current op amps for a summing that could run off the same battery supply as the amp?
From an ipod source, would I loose too much from passive summing?
Thanks,
Ross
Unless I am mistaken, all the class D chips are small pitch leads and surface mount chips. This is for a class project for students who will have only basic soldering experience.
It seems the likes of lm38X, lm390, tba820M and TDA7052 are the only real options for me. None of these can be inverted, so looks like there is not an obvious single chip solution to this. Are you able to recommend any low current op amps for a summing that could run off the same battery supply as the amp?
From an ipod source, would I loose too much from passive summing?
Thanks,
Ross
hello.
the lm 386 opamps can be " battery powered"...........and used inverted i think (datasheet).
greetings.............
the lm 386 opamps can be " battery powered"...........and used inverted i think (datasheet).
greetings.............
Hi
As far as I can see the gain of the lm386 can be altered with an additional cap and resistor, but the feedback network is fixed within the chip - so I don't see how it can be used inverted. The other chips I mentioned have similar set ups.
I checked the national boomer chips again and some of them are soic and could be used in a DIP adapter. The feedback resistors are external, so these would probably do the job if I can get the adapters cheap enough.
Ross
As far as I can see the gain of the lm386 can be altered with an additional cap and resistor, but the feedback network is fixed within the chip - so I don't see how it can be used inverted. The other chips I mentioned have similar set ups.
I checked the national boomer chips again and some of them are soic and could be used in a DIP adapter. The feedback resistors are external, so these would probably do the job if I can get the adapters cheap enough.
Ross
Hi,
If its just to demonstrate the principal you could try a power opamp such as the L272 from ST. This will work down to 4V and deliver upto 1A. However its distortion is nothing special at 0.5%.
Regards,
Andrew
If its just to demonstrate the principal you could try a power opamp such as the L272 from ST. This will work down to 4V and deliver upto 1A. However its distortion is nothing special at 0.5%.
Regards,
Andrew
Hi,
Thanks for your replies.
TEA2025 and TDA1517 both draw more current than the the TDA2003. TDA1517 only takes non inverting input.
The boomer chips are ideal - most of the application circuits are inverted anyway. I have ordered an Lm4890 to try, and might be able to get some soic/dil adapters made. Off the shelf converters are far too expensive.
The plan was that the students would do their own layout on veroboard, but time is an issue anyway so we will perhaps make up boards for them with the soic chips presoldered.
The L272 looks promising, I will see if I can order one for comparison - you are right, it is the principal that is important.
Any more ideas?
Ross
Thanks for your replies.
TEA2025 and TDA1517 both draw more current than the the TDA2003. TDA1517 only takes non inverting input.
The boomer chips are ideal - most of the application circuits are inverted anyway. I have ordered an Lm4890 to try, and might be able to get some soic/dil adapters made. Off the shelf converters are far too expensive.
The plan was that the students would do their own layout on veroboard, but time is an issue anyway so we will perhaps make up boards for them with the soic chips presoldered.
The L272 looks promising, I will see if I can order one for comparison - you are right, it is the principal that is important.
Any more ideas?
Ross
After checking the ST site (and local suppliers) found their TDA2040 looks like a better candidate (do no confuse it with the tripath chips). From the design sheet, looks like slightly lower current draw, and simpler circuit design.
After looking at the data sheet for the 2003, it appears that the reason for the low value feedback resistors and the large coupling cap is because they use emitter of the input transistor for the inverting input, rather than a long tail pair (Differential pair) like most amplifier and op amp circuits. The RC circuit is probably for high frequency compensation - so it won't oscillate.
I have used a TDA7240 which is a bridged amp (Very simple and low parts count, plus a bridge doesn't require a coupling cap when run off a single supply, but is not suitable for headphones, and doesn't sound like it would satisfy your requirements).
After looking at the data sheet for the 2003, it appears that the reason for the low value feedback resistors and the large coupling cap is because they use emitter of the input transistor for the inverting input, rather than a long tail pair (Differential pair) like most amplifier and op amp circuits. The RC circuit is probably for high frequency compensation - so it won't oscillate.
I have used a TDA7240 which is a bridged amp (Very simple and low parts count, plus a bridge doesn't require a coupling cap when run off a single supply, but is not suitable for headphones, and doesn't sound like it would satisfy your requirements).
Thanks for the explanation on the 2003. Although I have basically abandoned the idea of using it, I was still interested to know the reason for the circuit configuration.
The 2040 does have a lower current draw, but is still high compared to the other chips intended for battery operation. The boomer chips have stated max 10ma, and the likes of lm386 are similar.
I may have a source for cheap soic adapters - If I get those the boomers are all systems go!
By the way, it will drive a small speaker, not headphones.
Thanks,
Ross
The 2040 does have a lower current draw, but is still high compared to the other chips intended for battery operation. The boomer chips have stated max 10ma, and the likes of lm386 are similar.
I may have a source for cheap soic adapters - If I get those the boomers are all systems go!
By the way, it will drive a small speaker, not headphones.
Thanks,
Ross
Another thought. If you drop the requirement to do it in a single stage you could easly make something using two cheap dip parts. Such as a LM386 0.5W amp chip and almost any cheap opamp.
This would be more reprisentative of a real world situation where you would normally use an opamp as a preamp stage and use a seperate power amplifer. If you are goign to do a little PCB the difference in complexity to solder two dips agains one is almost nothing. Could work out cheaper than making a SMT carrier and then a seperate PCB for the circuit.
Regards,
Andrew
This would be more reprisentative of a real world situation where you would normally use an opamp as a preamp stage and use a seperate power amplifer. If you are goign to do a little PCB the difference in complexity to solder two dips agains one is almost nothing. Could work out cheaper than making a SMT carrier and then a seperate PCB for the circuit.
Regards,
Andrew
- Status
- Not open for further replies.