No doubt about that
I think you know that I meant a decrease in performance measurement- and specificationwise.
Not really... matching the apparent input resistance (option a) is probably the preferred (and cheaper) option as it leaves the system DC coupled, and doesn't put any capacitors in your signal chain if you're fussy about that.
Otherwise, performance is still very good even with SE input. Chances are your SE source will be the limiting factor in this setup. If you keep your signal runs short, well shielded, and properly grounded, then you should not see any significant increase in noise.
Overall, eliminating the output common ground by using the BAL-BAL with SE input still puts it above the performance of an SE-SE, so you should be good to go.
Regards,
Owen
I'd like to address the concerns about TI EOL'ing some of the key parts for The Wire.
Although I'm generally quite disappointed that TI chose to EOL these parts, the reversal of the LME49600 EOL means there's actually not much impact to these designs. I had been recommending the LME49610 for these amps, but that was only because it was the same price as the LME49600 at the time, and offered the flexibility of running at +/-22V rails if people needed more voltage swing for whatever reason. The original designs, however, were all done with the LME49600, and all measurements shown here were also with this part. As a result, the simple answer here is to just continue to use the LME49600. As of today, it is in-stock at pretty much every major retailer in healthy numbers.
The LME49990 on the other hand is now pretty much entirely gone. I have a stash of them here, but not enough to be shipping them out with orders. The good news on this one is that we have many choices for replacements, and overall performance of The Wire is not really limited by the op-amp.
For replacements, I would suggest the following in order of my personal preference:
OPA1611
LME49870 (also EOL)
AD797
LT1028/LT1128
LME49710
To be honest, you're not likely to notice a difference between any of these parts. The differences in a typical setup cannot be measured, and output noise levels and distortion remain unchanged. If you are running very high gain (eg, >20dB) then you should probably weight your selection towards lower noise (LT1028/LT1128). If you are running a typical gain of 0dB or 6dB then you would be better served by the low distortion options like the OPA1611. If you need higher voltage swing, and still have some LME49610 parts around, then opt for the LME49870 if you can find it.
There are countless other options, and if anyone finds something they like, or something they have measured and tested, then please do post it here. It's important to note that not all op-amps are unity gain stable, and since most people will be running with unity gain, I would not suggest using anything which is not stable at unity gain in these circuits.
Overall, I will miss the LME49990 as a great "go to" all-rounder. It was the kind of op-amp you could drop into pretty much any circuit and get great performance out of. The above op-amps will need to be selected with a bit more prudence, but performance will remain excellent with any of them.
Regards,
Owen
Although I'm generally quite disappointed that TI chose to EOL these parts, the reversal of the LME49600 EOL means there's actually not much impact to these designs. I had been recommending the LME49610 for these amps, but that was only because it was the same price as the LME49600 at the time, and offered the flexibility of running at +/-22V rails if people needed more voltage swing for whatever reason. The original designs, however, were all done with the LME49600, and all measurements shown here were also with this part. As a result, the simple answer here is to just continue to use the LME49600. As of today, it is in-stock at pretty much every major retailer in healthy numbers.
The LME49990 on the other hand is now pretty much entirely gone. I have a stash of them here, but not enough to be shipping them out with orders. The good news on this one is that we have many choices for replacements, and overall performance of The Wire is not really limited by the op-amp.
For replacements, I would suggest the following in order of my personal preference:
OPA1611
LME49870 (also EOL)
AD797
LT1028/LT1128
LME49710
To be honest, you're not likely to notice a difference between any of these parts. The differences in a typical setup cannot be measured, and output noise levels and distortion remain unchanged. If you are running very high gain (eg, >20dB) then you should probably weight your selection towards lower noise (LT1028/LT1128). If you are running a typical gain of 0dB or 6dB then you would be better served by the low distortion options like the OPA1611. If you need higher voltage swing, and still have some LME49610 parts around, then opt for the LME49870 if you can find it.
There are countless other options, and if anyone finds something they like, or something they have measured and tested, then please do post it here. It's important to note that not all op-amps are unity gain stable, and since most people will be running with unity gain, I would not suggest using anything which is not stable at unity gain in these circuits.
Overall, I will miss the LME49990 as a great "go to" all-rounder. It was the kind of op-amp you could drop into pretty much any circuit and get great performance out of. The above op-amps will need to be selected with a bit more prudence, but performance will remain excellent with any of them.
Regards,
Owen
Hi Owen.
Have you had the time to send out the payment requests for The Wire boards? I haven't received any PM from you and I do have my name on the ordering sheet...
I already got the bom parts from Digikey for a SE-SE and power supply boards. I choose the LME49600 and OPA1611 as you recommended.
Have you had the time to send out the payment requests for The Wire boards? I haven't received any PM from you and I do have my name on the ordering sheet...
I already got the bom parts from Digikey for a SE-SE and power supply boards. I choose the LME49600 and OPA1611 as you recommended.
Most headphones will not take more than 5Vac and some should be limited to ~5Vpk
5Vac is available from a 15Vdc supply.
±12Vdc allows upto 8Vac
If one adopts a differential output from the same supply rails, then the maximum signals roughly doubles.
That means ±12Vac could give a maximum differential output of ~16Vac
I cannot see why anyone would need voltages this high !
5Vac is available from a 15Vdc supply.
±12Vdc allows upto 8Vac
If one adopts a differential output from the same supply rails, then the maximum signals roughly doubles.
That means ±12Vac could give a maximum differential output of ~16Vac
I cannot see why anyone would need voltages this high !
From my limited knowledge the THD+N vs output voltage plots of most op-amps starts to rump-up after the output voltage crosses 1/3(for older ones) to 1/2(for newer models) of the supply voltage value (for bipolar supplies). For an example see page 6 of LME49990 datasheet. Thus it is probably safe to assume that if the output voltage is within proper bounds there is no need to further increase the supply voltage. For the more difficult loads (e.g. driving low impedance loads) the THD+N starts to rise earlier which may explain the need for an increased supply voltage in some cases.
Hi Owen,
Been a long time since I've been here. I run across an original Se-Se Headphone PCB. I'm in the process of building the amplifier and would like to add a stepped attenuator on the input side. I've searched around this thread but can't seem to find the post describing the hookup for a level pot of if any of the input resistor values need to be changed and what their value should be changed to. The attenuator I have is 10k in value.
Thanks,
rondr
Been a long time since I've been here. I run across an original Se-Se Headphone PCB. I'm in the process of building the amplifier and would like to add a stepped attenuator on the input side. I've searched around this thread but can't seem to find the post describing the hookup for a level pot of if any of the input resistor values need to be changed and what their value should be changed to. The attenuator I have is 10k in value.
Thanks,
rondr