You're going to get more distortion on high current loads, its a given unless a separate current buffer is added into the loop - but there's a lot of headroom since the distortion starts very low. 100mA is the current limiting, the usuable output current will be less. 600 ohm load at 15V peak is 25mA peak. This suggests upto 25mA can be drawn from the opamp output without much compromise, so 0.8V peak at 32 ohm should be pretty clean. Not clear at what current level things breakdown though.
I've used AD8656 CMOS opamp which has 220mA output capability for earphones to good effect - its only a 5V supply part though. A 5V opamp can afford much higher current limits than a 36V part without thermal issues of course. Its a good fit for low-impedance earphones for mobile devices, 1V rms is extremely loud in fact, but larger headphones want more voltage headroom as well as more current.
I've used AD8656 CMOS opamp which has 220mA output capability for earphones to good effect - its only a 5V supply part though. A 5V opamp can afford much higher current limits than a 36V part without thermal issues of course. Its a good fit for low-impedance earphones for mobile devices, 1V rms is extremely loud in fact, but larger headphones want more voltage headroom as well as more current.
Hi John
I used previously in IV convertor (current DAC) the OPA1612 with good tested results ..
However I am considering a new DAC and its IV stage (low current output) and I am looking at OPA1656 (fet input is an advantage)
My question is , what do you consider the most important parameters on IV convertors , slew rate , settling time , bandwidth.
I used previously in IV convertor (current DAC) the OPA1612 with good tested results ..
However I am considering a new DAC and its IV stage (low current output) and I am looking at OPA1656 (fet input is an advantage)
My question is , what do you consider the most important parameters on IV convertors , slew rate , settling time , bandwidth.
Can we know
what is the output stage bias current?
I would like to maintain class A operation.
Hi johnc124,
I'm interested to test this OPA1656 as a phono stage with a MC cartridge.
Is it possible to adapt the suggested application in the TI datasheet (dedicated to a MM cartridge) just by adding an extra 20dB voltage gain to the same schematic ?
I ask this question because I prefer the design simplicity of a single stage amplification for a phono stage, in order to keep the noise figure low.
Thanks,
Fred.
But you would never use this open loop, so that open loop output impedance is irrelevant.
At 90mA Iout you loose 6V on the output voltage range, so if you want to look at that as a resistance (which it is not), you could consider it as about 70 ohms. But, again, it has little or no relation to the OL Rout.
The Rout is not a physical resistor.
Jan
Last edited:
I have a question for the good TI folks in this discussion:
When driven from very high source Z, from 20k up to 100k, would the OPA1641 or OPA1656 be a better choice?
The 1641 data sheet mentions the unique IP stage with very low voltage modulated IP capacitance and I'm wondering if 1656 has anything similar.
Thanks
TCD
When driven from very high source Z, from 20k up to 100k, would the OPA1641 or OPA1656 be a better choice?
The 1641 data sheet mentions the unique IP stage with very low voltage modulated IP capacitance and I'm wondering if 1656 has anything similar.
Thanks
TCD
The OPA1641 would be slightly better. It has the most stable input capacitance vs. common mode voltage that I've measured on any op amp. When we developed the OPA1656, this was also a major design target for us, and the designer and myself spent quite awhile in the lab with previous CMOS audio op amps (specifically OPA1652 and OPA1688) hunting for sources of input capacitance variation. The end result was slightly better than the previous parts, but we still met some limitations from the CMOS process itself. The dielectrically isolated JFETs on the input of the OPA1641 are superior in this regard.
This article shows the measurements I took on the OPA164x family. The OPA827 measured just as good in this same test and has slightly better AC performance specs.
The performance of the CMOS audio op amps is actually fairly close and maybe slightly better than the OPA627, which was somewhat famous for its performance in this use case.
OPA1656, and most modern CMOS op amps with rail-to-rail output stages, use a biasing scheme often referred as a Monticelli bias, named after Dennis Monticelli who published the topology while at National Semiconductor. In a Monticelli circuit, the output devices are biased with a translinear loop, which has the end result of making them act like a constant current source when they are not conducting load current. For that reason they never enter cutoff like would happen in a more traditional Class-AB output stage.
If you want to guarantee Class-A operation, the simple answer is that you'd need a current source which always sinks or sources more current than needs to be delivered to the load.
Greetings TI folks and thank you for all this good products avaliable for diy people.
Sorry for my noob, I have please a question without knowing if it's possible to acheive. SIC transistors seems to have some sucess in some amps in Japan or USA (Neslson Path).
Is it feasible in an opa size for the low power needs for the dac, headphone amp,,,
Sorry if stupid question, I have no understanding about transistors world, just a fan of some old products as opa861 for I/v, opa 2604 for buffers.
Sorry for my noob, I have please a question without knowing if it's possible to acheive. SIC transistors seems to have some sucess in some amps in Japan or USA (Neslson Path).
Is it feasible in an opa size for the low power needs for the dac, headphone amp,,,
Sorry if stupid question, I have no understanding about transistors world, just a fan of some old products as opa861 for I/v, opa 2604 for buffers.
You definitely could, reduce R4 by a factor of 10 and increase C4 by a factor of 10 to try to maintain low frequency performance. But for MC phono pre-amps a low noise bipolar input op amp like the OPA1612 would give the best noise performance.
Although, I have to think that all of our hard work achieving ultra low noise in phono preamps goes out the window as soon as the needle touches the vinyl, and circuit noise is swamped by mechanical noise in the system. Give it a try!
Most (or all) of the silicon carbide processes available today are targeting very high power applications of MOSFETs, such as for motor drives in electric vehicles. These same processes probably wouldn't make a really good op amp or other analog IC, just because the strengths of SiC aren't needed, and the performance you do care about (transistor matching, noise, linearity, good passive components) probably weren't optimized in a process targeting power FETs.
And that op amp would be really expensive!
Last edited:
Hi John,
Thank you for your kind answers. I have couple suggestions: 1) can you please somewhere mention that the XXX audio opamp has the YYY equivalent with low offset (e.g.OPA164x - OPAx140). Sometimes it is better to use opamp with low offset instead "audio" opamp with servo; 2) can you please add to the opamp selection table the first column - technology (BJT, JFET or CMOS input) and 3) there are so many opamps in your portfolio and it makes hard to make a right selection. Once you introduce new opamp can you please specifically mention the target application - e.g. headphone amplification in advanced handheld products, NS part replacement, etc.
BR, Dimitri (princ. eng. at Samsung/Harman)
Thank you for your kind answers. I have couple suggestions: 1) can you please somewhere mention that the XXX audio opamp has the YYY equivalent with low offset (e.g.OPA164x - OPAx140). Sometimes it is better to use opamp with low offset instead "audio" opamp with servo; 2) can you please add to the opamp selection table the first column - technology (BJT, JFET or CMOS input) and 3) there are so many opamps in your portfolio and it makes hard to make a right selection. Once you introduce new opamp can you please specifically mention the target application - e.g. headphone amplification in advanced handheld products, NS part replacement, etc.
BR, Dimitri (princ. eng. at Samsung/Harman)
Hi Dimitri,
Thanks for the great feedback. I'll pass it along to the TI team in charge of these products now.
Thank you for your reply and also for your advices.
Is the OPA1656 unity-gain satble ? I think that it should be when observing the test condition datas but as it's not clearly mentioned in the datasheet features, I prefer you to confirm it.
This circuit combines a lot of advantages that only several different Op-Amps can offer individually in the same level (Low Ib/Cin/Vn/Vos and High Io/BW/OLgain) all with a RRout !!
Yes, OPA1656 is unity gain stable. Usually (in the last 8-10 years) we only say when an amplifier is NOT unity gain stable (decompensated). The majority of op amps TI is releasing are unity gain stable, although my team did put out the OPA607 recently, which is decompensated.
Unfortunately, we've found decompensated amplifiers scare some customers, and so now I feel like we sell them like the amplifier equivalents of cigarettes (Warning! Only stable in gains greater than 5!). Personally, I love decompensated amplifiers.