It's a funny saying/joke he has because he only listens to and designs devices tubes 🙂
Well there are now possibilities.
I digress to draw attention to relatively recent advances from Ti.
There are no longer barriers to considering Fets or Jfets.
One of the problems with Fets -Jfets is variations of gate capacitance with voltage Right. You then need to include surrounding circuits to stabilise that.
Refer to the link below.
https://www.ti.com/lit/an/slyt595/slyt595.pdf?ts=1744788928133&ref_url=https%3A%2F%2Fwww.google.com%2F
Ti have developed a die fabrication process that gets around this problemo in the OPA1644. The OPA 1656 also has advanced technology including feed forward to improve performance under a wide range of conditions.
I have discussed elsewhere on the forums how diy people can now achieve SOA measurements without resorting to complex discrete stages. You can also parallel opamps and class A SE buffers if required for lower noise with reduced feedback impedances.
For example a quad OPA1644 can used in parallel with 10R load sharing resisters to drive lower impedances like the NE5532.
If you have followed Douglas Self 2 part opamp distortion tests on the AudioExpress publication he provides all the applicable conditions with applications like Sallen Key filter where common mode voltage is a primary source of distortion.
Ti offer full support with detailed application notes and tools like Tina Ti software for free. You do have to use Surface Mount packages but there are immense advantages for optimum layout so your implementation works just like your simulation.
The graphs Ti provides on source impedance versus noise clearly outlines when to use either bi polar or the Fet-JFet counterparts.
These monolithic parts have inherently low temperature drift which means stable temperatures offsets that normally plague discrete through hole implementations.
Some of these opamps have very low noise too.
If you look at the Hifisonix website he has examples of these newer opamps in fully fleshed out projects.
No affiliation with Ti.
For awareness and education only
https://www.ti.com/lit/ds/symlink/opa1644.pdf?ts=1744784091115
https://www.ti.com/lit/ds/symlink/opa1656.pdf?
Attachments
Last edited:
Hi Brijac,
Yup. I figured. I chose the Carver as an example as the supply voltage is about 250 VDC rail to rail. I wasn't kidding, it can kill you easily - but it runs cool normally. 375 wpc into 8R.
Hi macka,
Thanks for posting the links. The input common mode issue was a concern for JFet input op amps.
Yup. I figured. I chose the Carver as an example as the supply voltage is about 250 VDC rail to rail. I wasn't kidding, it can kill you easily - but it runs cool normally. 375 wpc into 8R.
Hi macka,
Thanks for posting the links. The input common mode issue was a concern for JFet input op amps.
I have seen more speaker systems (woofers and crossovers) absolutely ended without clipping from high powered Carver amps. Because they are light, people don't believe they will put out the power they do. (HINT: they do!)
AC/DC - yeah! My grandson's favorite band. "Grandpa's stereo is loud!".
AC/DC - yeah! My grandson's favorite band. "Grandpa's stereo is loud!".
I use Borbely’s direct coupled jfet/ fet balanced DAC , phono, line stage, and power amp. All the low level stages are powered by Erno's shunt regulators and use TI OPA 189 for DC servos, all less than 1 mv offset. The line stage and power amp have DC protect circuits, and to be sure I test the DC protection every once and a while to make sure it operates. The actual circuits have never failed and some are well over 20 years old.
Carlmart- where in Brazil are you? I visited RIo, Sao Paulo, and Belo many times for work. I still visit Rio, if you are interested in Borbely circuits I can help with Semiconductors
Carlmart- where in Brazil are you? I visited RIo, Sao Paulo, and Belo many times for work. I still visit Rio, if you are interested in Borbely circuits I can help with Semiconductors
Lol!
Ya got me! I eat other things too!
You are a very good host. One of the better meets, better than most seminar's I have attended.
Ya got me! I eat other things too!
You are a very good host. One of the better meets, better than most seminar's I have attended.
Another interesting find.
In one of Bob Cordell’s power amp design YouTubes he referred to the typical large value capacitor to ground in the feedback loop to control the gain at DC.
Bob Cordell said that capacitor (electro) is quite important. He recommends loudspeakers crossover electro capacitors which have a high voltage rating because they measure lower distortion. Not a low voltage capacitor that is often used.
In one of Bob Cordell’s power amp design YouTubes he referred to the typical large value capacitor to ground in the feedback loop to control the gain at DC.
Bob Cordell said that capacitor (electro) is quite important. He recommends loudspeakers crossover electro capacitors which have a high voltage rating because they measure lower distortion. Not a low voltage capacitor that is often used.
Douglas Self made a similar point in his books. He measured distortion as a function of signal across the elcap.
The higher the voltage rating of the cap, the lower the distortion.
Since a cap's impedance rises with falling frequency, in the situation with an elcap in the ground leg of a feedback loop, signal voltage across the elcap rises with frequency going to the bass range. Taking a cap with a high voltage rating can largely prevent an increase in THD at lower frequencies.
Jan
The higher the voltage rating of the cap, the lower the distortion.
Since a cap's impedance rises with falling frequency, in the situation with an elcap in the ground leg of a feedback loop, signal voltage across the elcap rises with frequency going to the bass range. Taking a cap with a high voltage rating can largely prevent an increase in THD at lower frequencies.
Jan
Agreed. I have seen this in practice and measured to confirm. Size is an issue of course, electrolytic capacitors tend to get better and level off between 63 VDC and 100 VDC for most brands. Many times 85°C caps perform better than 125 °C capacitors. Sometimes not, you have to test them to see for sure.
Mind you, reality. Look at the situation. At low bass frequencies most speaker systems generate a great deal of distortion on their own. So it may not even matter (within reason) depending on your system. Many speaker systems struggle below 50 Hz anyway, so worrying about distortion at 20 Hz in your system may be a waste. The system actually sounds better if you roll off the lows below the cutoff of your speaker. Sealed speakers work better when the response isn't very low.
I guess some amplifiers may show higher distortion at higher frequencies with very low frequency information as well. I haven't tested that, interesting idea.
Mind you, reality. Look at the situation. At low bass frequencies most speaker systems generate a great deal of distortion on their own. So it may not even matter (within reason) depending on your system. Many speaker systems struggle below 50 Hz anyway, so worrying about distortion at 20 Hz in your system may be a waste. The system actually sounds better if you roll off the lows below the cutoff of your speaker. Sealed speakers work better when the response isn't very low.
I guess some amplifiers may show higher distortion at higher frequencies with very low frequency information as well. I haven't tested that, interesting idea.
Agreed Jan.
I'm just trying to bring in some awareness of what people are doing so they think about what is important. For my audio system, it is important to have low distortion at lower frequencies. For anyone with bookshelf speakers, the importance of distortion in their power amp at 20 Hz might be purely academic. It is common for people to pursue perfection at great expense. They may also compromise other factors, such as noise pickup from that large capacitor. It's all a balance that depends on many things. In the end, what matters more, and to what degree?
Engineering. Most questions come down to "it depends, what else is going on?".
I'm just trying to bring in some awareness of what people are doing so they think about what is important. For my audio system, it is important to have low distortion at lower frequencies. For anyone with bookshelf speakers, the importance of distortion in their power amp at 20 Hz might be purely academic. It is common for people to pursue perfection at great expense. They may also compromise other factors, such as noise pickup from that large capacitor. It's all a balance that depends on many things. In the end, what matters more, and to what degree?
Engineering. Most questions come down to "it depends, what else is going on?".
I think sometimes diy people don’t consider impact on the phase shift collectively though all their equipment
Well, let's look at phase shift.
We determine where things are from a difference in phase between our ears. If both channels are equal, they can be out of phase from 1 KHz by a lot without making any difference in what we hear. Having said that, only an all pass filter will shift phase without a level change. If the level is falling in your speakers, welcome to a phase shift. unavoidable.
We determine where things are from a difference in phase between our ears. If both channels are equal, they can be out of phase from 1 KHz by a lot without making any difference in what we hear. Having said that, only an all pass filter will shift phase without a level change. If the level is falling in your speakers, welcome to a phase shift. unavoidable.
I don't know if that is always true these days Dave. Bateman's work was on early capacitors, new ones are improved.
Tests I have done do not always support that. Depends on signal levels too.
Tests I have done do not always support that. Depends on signal levels too.
