That's exactly why that statement is completely bogus. It's a weasel way to try to market stuff.Jan,
Measures better is science.
Sounds better is emotion.
Emotion is subjective, which, by definition, you cannot dispute.
Patrick
jan
Strong opinions this evening, Jan.
I am sure that emotion is a large part in the hobby. You can see it now and then when presenting a new DIY device, many will have an predefined opinion when they see something glow. A fact that I have misused a few times
When seeing an SMPS of any quality I have the same. "Needs to go ASAP". Probably the best human instruments to judge audio are ears and not eyes.
I am sure that emotion is a large part in the hobby. You can see it now and then when presenting a new DIY device, many will have an predefined opinion when they see something glow. A fact that I have misused a few times
When seeing an SMPS of any quality I have the same. "Needs to go ASAP". Probably the best human instruments to judge audio are ears and not eyes.
Last edited:
Ceramic caps (X7R etc) are absolutely fine for decoupling DC like supply rails of reg outputs or reference voltages. They aren’t very good for coupling AC. Big difference in requirements.
So the fact that a reg needs a ceramic decoupling cap on the output will really not cause any issues and you cannot point to that as a problem.
Don’t use ceramic caps (X7R type dielectric) for compensation (can cause oscillation) or anywhere you are likely to see substantial AC voltages like EQ circuits for example. The capacitance changes substantially with applied voltage and with temperature.
Note these comments do not apply to COG/NPO which are very stable with applied voltage and temp.
I use an LM4562 for low power low noise regulator duty, but some of these new low noise high PSRR regs strike me as being pretty good and worlds away from old work horses like the 78 and 317/337 type regs - which btw still remain great for bulk +-15V PSU duty.
So the fact that a reg needs a ceramic decoupling cap on the output will really not cause any issues and you cannot point to that as a problem.
Don’t use ceramic caps (X7R type dielectric) for compensation (can cause oscillation) or anywhere you are likely to see substantial AC voltages like EQ circuits for example. The capacitance changes substantially with applied voltage and with temperature.
Note these comments do not apply to COG/NPO which are very stable with applied voltage and temp.
I use an LM4562 for low power low noise regulator duty, but some of these new low noise high PSRR regs strike me as being pretty good and worlds away from old work horses like the 78 and 317/337 type regs - which btw still remain great for bulk +-15V PSU duty.
There's no bandwidth with a large capacitive load ...
You don't need to drive a large capacitance when offering a refference to a dac...nor to offer any bandwidth or feedback speed , actually you don't want any bandwidth at all...
I had this discussion before with the low drop regulators used in the 80's low noise nakamichi preamps where they used 10 000uf caps on their regulator's output slowing down the regulator right from its comparator loop.Completely opposite philosophy to actual trends, where they used to slow the electronic regulator to simply regulate the input voltage and the output caps to offer enough filtering and current for transient surge.You can use low esr and low esl capacitors for high slew rate circuits, but i doubt the refference input of a dac presents any such issue...All this discussion is purposely drifted to hilarious pathways...
I used to callibrate and fix 4.5kWatts smps with 0.02V precision at 150v / 30 amps continuous 24/7 output .Switching F =80khz , pfc= 40khz .refference done by 10v high precision ref ic + opa2227.The whole smps pcb had about 20 x opa2227 on it for all the precision work in a fantastically noisy environment yet the power supply output was rock solid with any load and working regime. I remember old days motherboards using lt1084...1085 with 10...15 paralleled low esr output capacitors to supply the microprocessors and bridges + video operating at 200Mhz...If you're interested in supplying high slew rate demanding circuits you should definitely look on computer motherboards for inspiration...If supplying the ref input of a damn dac is such a complex problem then i should quit simple electronics and start playing the piano...
Last edited:
I have that with piano, at least with most classical piano music. I particularly dislike hearing baroque music that's meant for harpsichord played on a piano.
No - never but I’ve read there’s been complaints. I used it in the last phono stage (with a JFET front end) and had no problems. I’ve sort of transitioned now to the OPA164x - the JFET input offers a lot of advantages.
(I may be wrong but the 4562 still offers the best overall PSRR of any GP opamp - about 80 dB at 100kHz - so very nice for low power regulator applications)
Last edited:
what i have learned from the past is that there is not such thing as noise, there is that impedance curve that works with dual polarity Power supplys. That curve is dependent on the load and wiring, abb(a be back), igbt i gonna be terminator, and capacitors, types, wire corners, material, environment in total, temperature, gold, silver, opamp load and so on.
So, i have used many LM317LZ and 337LZ series and in right configuration, they better everything out, super good sound control out of them.
It also depends on dynamic range on the input, all they need is long lasting battery and skynet is burn and we die all. vs out voltage and, diodes used there and here... i mean....if LM317 works as it suppose to, it can be configured.
But yes, PSU is everything, most important.
So, i have used many LM317LZ and 337LZ series and in right configuration, they better everything out, super good sound control out of them.
It also depends on dynamic range on the input, all they need is long lasting battery and skynet is burn and we die all. vs out voltage and, diodes used there and here... i mean....if LM317 works as it suppose to, it can be configured.
But yes, PSU is everything, most important.
How would an EE explain the difference?Furthermore OPA164x has "Superior Sound Quality" while LM4562 is only "Optimized for Superior Audio Signal Fidelity"
TI didn’t bother past that marketing statements, that would be better aimed at audiophool consumption. It’s not that being an EE makes one absolutely exempt from audiophoolery.
Yes I also went to OPA1642. Still a little unsure about OPA1656/2156 although they do perform very good till now. Must be the CMOS stamp that makes one judge a book by the cover