The posts I was looking through were the Microsim SPICE simulations, are the measurements posted a bit later?
My measurements were done with a driver in series with the output, the simulation was applied at the negative end of the driver, which was the reference feedback, the data was taken from the driver positive terminal which is the output of MyRef. I did not short the inputs, and probably I should.
Perhaps I should also use sine wave instead of MLS.
Last edited:
Let me look through my crystal ball.I don't know how to tell which is better... any clue?
MyRef C Caddock R3 with LM3886 grounded at power GND
MyRef Dario Caddock R3 old C32 value
Isn't it interesting?
An externally hosted image should be here but it was not working when we last tested it.
MyRef Dario Caddock R3 old C32 value
An externally hosted image should be here but it was not working when we last tested it.
Isn't it interesting?
I am wondering, the most important thing is to look at the phase of current to see what is going on. This is where the SPICE sim come to play. Note that the SPL curve is the inverse shape of the impedance curve. What this seems to tell use that if the source impedance is high, the damping is not going to be good. However, with a current pump, the source impedance is supposed to come out high. It seems that a different topology may be necessary.
Sorry Soongsc,
I don't get it ...
This is Rev C compensation with LM3886 ground to GND
This is the alternate compensation with old C32 (C10=33pF, C34=47pF, C32=150pF)?
It would be nice to have also RevC with LM3886 to 0V
Idem.
Can't you comment a bit more?
I don't get it ...
This is Rev C compensation with LM3886 ground to GND
This is the alternate compensation with old C32 (C10=33pF, C34=47pF, C32=150pF)?
It would be nice to have also RevC with LM3886 to 0V
Idem.
Can't you comment a bit more?
But this could also be due to the different connection of the LM3886 ground.
Dario,
You got it right. I just don't have the original MyRef C data now. Need to open the thing up to get it back in place. Since I'm out of town tomorrow, it's going to be a couple of days. The PCB is not liking all this abuse. I'm using lead free solder.
Last edited:
Lots of liquid flux will keep that pcb in good shape for many solder/desolder operations.
About all I can see is that it's a bad idea to ground the LM3886 at power ground (e.g. at the junction of C1/C2). That 60Hz resonance peak was probably what was misinterpreted as "deeper bass".
I'll stay with the LM3886 grounding at the (lifted) signal ground.
Until we see Soongsc's measurents of a plain RevC I think it's a bit early to say that or dump this solution.
At the end it's the very same thing Mauro did on the My_Evo with RevA (along with other mods)...
And knowing how accurate and thorough is his work I doubt he wouldn't have noted such resonance...
I remember to have seen some, but can't remember where. Could be in some other place (possibly some italian forum).
NP's quote is balderdash.
If the numbers show that the amp could blow up, or the numbers show that the output cannot be a processed copy of what went in, then only a fool would disregard those numbers.
Just like a wine. If the numbers (chemical analysis) show that the wine contains some toxic to human contaminant, then only a fool would drink it.
Look at the numbers and decide whether they are of value and only after that examination can one decide to use the information in determining whether to buy the wine and/or the amplifier.
If the numbers show that the amp could blow up, or the numbers show that the output cannot be a processed copy of what went in, then only a fool would disregard those numbers.
Just like a wine. If the numbers (chemical analysis) show that the wine contains some toxic to human contaminant, then only a fool would drink it.
Look at the numbers and decide whether they are of value and only after that examination can one decide to use the information in determining whether to buy the wine and/or the amplifier.
quite!Isn't it interesting?
...now it would be great if those remaining "spikes" could be flatten out too!
clearly you haven't got the meaning of that quote.
As clearly stated, it refers to (sound) QUALITY, not mere functional evaluation!
Indeed a chemical analysis may tell you if a wine is badly altered (say, it become vinegar...) or - as you say - if it contains some (known) toxic contaminant. But the quote talks about «fine wines». No chemical analysis will tell you how they will taste. Let alone if one fine wine will taste better or worse than another one.
Basically, it will NOT tell you anything about the wine quality. Because no chemical analysis is able to process all the countless details of wine composition in exactly the same way your senses (and brain) does.
For audio it's exactly the same thing!
Numbers may tell you whether something will work or its badly broken, but they can NOT tell you HOW it will sound to your hears.
Our hearing system is not just a simple microphone or spectrum analyzer. And no available instrument works in much the same way our hearing system does. Thus all "numbers" (measurements and/or simulations) are basically meaningless WRT perceived sound quality. Sic et simpliciter.
Pretending to judge the sound quality (which means perceived sound quality) of any audio device just by looking at its "numbers" is naive (and stupid). Your hears are (and can be) the only arbiter. Because your hearing system is the only existing instrument that match the specs of the target.
Got it now?
BTW: do you really believe that any fine wine lover would ever buy any bottle based on chemical analysis?!
Last edited:
What an apt response, but it will not even register on the stone ears of objectivists. Many will obsess over miniscule percentages of distortion or amplitude shown by their precious graphs and simulations, but ruthlessly ridicule anyone who dares to detect differences by carefully listening to something.
I assume that NP is Nelson Pass, and I love that quote. I don't hold anyone in absolute authority when it comes to audio, but I would bet that very few forum members have as much or as thorough technical expertise and experience as NP. I surmise he has judged the sound of more amplifiers that he built or designed than most of us have ever heard. When he says hard data can't tell the whole story of how a component sounds, I give that pretty high priority in my belief system.
Numbers can tell you a lot, but your ears can tell you the truth.
Peace,
Tom E
I assume that NP is Nelson Pass, and I love that quote. I don't hold anyone in absolute authority when it comes to audio, but I would bet that very few forum members have as much or as thorough technical expertise and experience as NP. I surmise he has judged the sound of more amplifiers that he built or designed than most of us have ever heard. When he says hard data can't tell the whole story of how a component sounds, I give that pretty high priority in my belief system.
Numbers can tell you a lot, but your ears can tell you the truth.
Peace,
Tom E
Final Ultimate BOM
While making the BOM for my My_Ref version I've finally managed to complete the (hopefully) final version of my Ultimate My_Ref Rev C BOM.
This last version is more clear, simpler, it includes only two main suppliers.
I've posted it here to give it the largest audience and posted also a link on Tom's thread where it's born from both mine and Tom research for best sound (even if Tom don't agree on some components ).
While making the BOM for my My_Ref version I've finally managed to complete the (hopefully) final version of my Ultimate My_Ref Rev C BOM.
This last version is more clear, simpler, it includes only two main suppliers.
I've posted it here to give it the largest audience and posted also a link on Tom's thread where it's born from both mine and Tom research for best sound (even if Tom don't agree on some components
).
Last edited:
MyRef C Caddock R3 with LM3886 grounded at power GND
An externally hosted image should be here but it was not working when we last tested it.
MyRef Dario Caddock R3 old C32 value
An externally hosted image should be here but it was not working when we last tested it.
Isn't it interesting?
Now for the MyRef C with Caddock R3
An externally hosted image should be here but it was not working when we last tested it.
How interesting. The original MyRef C had the glitch, but the high frequency glitch is only in one location. Damping seems better, as well as the phase. In my listening tests, the MyRef has the most dynamics, slight coloration, best detail. Still looking for more improvements. I hope it is possible to make the mid frequency damp better, and not vary so much with impedance.
I think the next test I'm going to add snubbers to my PS as I have on the other amp.
Last edited: