Neither
I figure that if the gang comes up with how it works on their own, they will learn more. Yes, I'll guide them if they get stuck. But I think several of the people here could dream it up if they think about it long enough.
Hell, I did. And I'm and idiot. Or is that "E-E-E-E-D-I-O-T"?
Anyway.........ftorres has the right idea. See, I knew someone could figure it out. I just like common-base stuff. Comes from years of designing RF preamps.
NO. I didn't compare squat. I just built one, laid out a board, and started shipping them.
And "Ren"........I think I know who you are. I don't have time now to address your "stinky" comments.
Jocko
I figure that if the gang comes up with how it works on their own, they will learn more. Yes, I'll guide them if they get stuck. But I think several of the people here could dream it up if they think about it long enough.
Hell, I did. And I'm and idiot. Or is that "E-E-E-E-D-I-O-T"?
Anyway.........ftorres has the right idea. See, I knew someone could figure it out. I just like common-base stuff. Comes from years of designing RF preamps.
NO. I didn't compare squat. I just built one, laid out a board, and started shipping them.
And "Ren"........I think I know who you are. I don't have time now to address your "stinky" comments.
Jocko
OK "Ren", as that is your real name.
Let's see.......stinky parts, unwashed lederhosen, "cool, man", and finally DVD mods.......
Now I know for sure who you are, Charlie. Bet you and "Toslink" dreamt this one up just to annoy me even more than the non o/s crowd. And you probably have Harry in on the joke. Ha, ha, ha, funny, funny. "Let's get Jocko wound up and let him go off like that insane Ren cartoon character. He even screams like him."
OK, you got what you wanted. But here it is again:
E-E-E-E-D-I-O-T-S! I am surrounded by stupidity! Why are you all so stupid?
There, happy now.
No, I am not going to modify DVDs, SACDs, or any other of that delta-sigma stuff. No matter how much you pester me.
Business must really be slow if you two have time on your hands just to annoy me. Back to work. Go pester Harry next time.
Jocko
Let's see.......stinky parts, unwashed lederhosen, "cool, man", and finally DVD mods.......
Now I know for sure who you are, Charlie. Bet you and "Toslink" dreamt this one up just to annoy me even more than the non o/s crowd. And you probably have Harry in on the joke. Ha, ha, ha, funny, funny. "Let's get Jocko wound up and let him go off like that insane Ren cartoon character. He even screams like him."
OK, you got what you wanted. But here it is again:
E-E-E-E-D-I-O-T-S! I am surrounded by stupidity! Why are you all so stupid?
There, happy now.
No, I am not going to modify DVDs, SACDs, or any other of that delta-sigma stuff. No matter how much you pester me.
Business must really be slow if you two have time on your hands just to annoy me. Back to work. Go pester Harry next time.
Jocko
BJT vs. MOSFET for I/V
Hi everyone,
I am new to this thread but thought I'd throw in some ideas.
Firstly I have just performed some comparisons of BJTs vs. FETs for I/V conversions. I used a sound card to perform the measurements.
The attached zip file shows the result when driving the sound card with a function generator at 1 KHz as a reference. - generator.jpg
IV.jpg is the output of a MOSFET (IRFD9110) I/V converter at full drive, the second harmonic is 70dB down.
IVIII.jpg is the output of a BJT (BC560) I/V converter. IVIIIa.jpg is the other channel. The second harmonic is now around -90dB!!
dacchipput.jpg is the spectrum the DAC chip itself sees at the source of the MOSFET. dacchipoutIII.jpg is the spectrum it sees at the emitter of the BC560.
Does the BJT sound better? Yes
I did buffer the base of the BJT with the original MOSFET.
The DAC I was using is the one I designed on my site - This is the one HPotter did such a lovely job of building. The DAC chips are PCM1704s.
Like Jocko I am not a fan of using old technology. I am currently running a CS8420, DF1704 and PCM1704s. I have yet to try an SM5847.
I am also not a fan of zero-oversampling DACs. I think any "improvements" brought about through zero-oversampling are probably the result of eliminating a bad digital section.
Hi everyone,
I am new to this thread but thought I'd throw in some ideas.
Firstly I have just performed some comparisons of BJTs vs. FETs for I/V conversions. I used a sound card to perform the measurements.
The attached zip file shows the result when driving the sound card with a function generator at 1 KHz as a reference. - generator.jpg
IV.jpg is the output of a MOSFET (IRFD9110) I/V converter at full drive, the second harmonic is 70dB down.
IVIII.jpg is the output of a BJT (BC560) I/V converter. IVIIIa.jpg is the other channel. The second harmonic is now around -90dB!!
dacchipput.jpg is the spectrum the DAC chip itself sees at the source of the MOSFET. dacchipoutIII.jpg is the spectrum it sees at the emitter of the BC560.
Does the BJT sound better? Yes
I did buffer the base of the BJT with the original MOSFET.
The DAC I was using is the one I designed on my site - This is the one HPotter did such a lovely job of building. The DAC chips are PCM1704s.
Like Jocko I am not a fan of using old technology. I am currently running a CS8420, DF1704 and PCM1704s. I have yet to try an SM5847.
I am also not a fan of zero-oversampling DACs. I think any "improvements" brought about through zero-oversampling are probably the result of eliminating a bad digital section.
Dave, I think better numbers are possible than you have achieved with your mosfet circuit. Comparing a bad mosfet implementation to a decent BJT one doesn't prove much.
I'm not sure it is possible to demonstrate something as basic as "BJTs are better than MOSFETs"...a great deal depends on topology, the distortion characteristics of the DAC in question, personal preference, etc. Some designs will always be more suitable than others in certain applications, but absolutes are hard to come by.
Various persons who shall remain unnamed are certainly welcome to their own opinions as to what is best.
I'm not sure it is possible to demonstrate something as basic as "BJTs are better than MOSFETs"...a great deal depends on topology, the distortion characteristics of the DAC in question, personal preference, etc. Some designs will always be more suitable than others in certain applications, but absolutes are hard to come by.
Various persons who shall remain unnamed are certainly welcome to their own opinions as to what is best.
tiroth,
The two implementations were very very similar. The MOSFET set-up was an IRFD9110 connected as a common base amplifier to the output of the PCM1704. A resistor was connected between the drain and - supply and the measurments taken between the drain and ground. The MOSFET was biased at 20mA by a constant current source.
Personally I think that was a pretty good MOSFET set-up.
The BJT set-up was exactly the same expect it was biased at 10mA just to get dissipation down. I have since moved back to the 20mA bias with the BJT and performance improved still further.
Now don't get me wrong, I'm not saying BJTs are in general better but in this application I think they are. Both in subjective and objective performance.
It's easy to see why, a BJT present a lower impedance to the output of the DAC chip compared to a MOSFET.
The two implementations were very very similar. The MOSFET set-up was an IRFD9110 connected as a common base amplifier to the output of the PCM1704. A resistor was connected between the drain and - supply and the measurments taken between the drain and ground. The MOSFET was biased at 20mA by a constant current source.
Personally I think that was a pretty good MOSFET set-up.
The BJT set-up was exactly the same expect it was biased at 10mA just to get dissipation down. I have since moved back to the 20mA bias with the BJT and performance improved still further.
Now don't get me wrong, I'm not saying BJTs are in general better but in this application I think they are. Both in subjective and objective performance.
It's easy to see why, a BJT present a lower impedance to the output of the DAC chip compared to a MOSFET.
In some testing with an IRF610 setup patterned after the Pass D1 I've achieved better performance than you quote despite having an opamp voltage2current converter in the circuit.
It's true that the BJT is better at providing the virtual short than the MOSFET. How much the DAC cares, though, about 1ohm or 7ohm impedence depends a lot on the DAC in question. The DAC I am working with (AD1853) is spec'd to drive 1kOhm into 2nF...that is a lot of drive compared to some other IDACs.
Rest assured I will try a BJT implementation at some point though...like you, I'm after the best sound.
It's true that the BJT is better at providing the virtual short than the MOSFET. How much the DAC cares, though, about 1ohm or 7ohm impedence depends a lot on the DAC in question. The DAC I am working with (AD1853) is spec'd to drive 1kOhm into 2nF...that is a lot of drive compared to some other IDACs.
Rest assured I will try a BJT implementation at some point though...like you, I'm after the best sound.
The DAC chip probably has a lot to do with. The closer its output is to being an ideal current source the better it will perform with common base/source style I/V converters. The PCM1704 has an output impedance of 1K which is far from being an ideal current source. Hence it needs to see a very low impedance.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.