I also am enjoying and learning from the in depth conversation. However, I'm still stuck a bit back on the target. I have a couple of Siva's class A oriented modules and two highly regarded Pass amps - F5 and BA-3. They IMHO, are simply different from the MyRef family. When I listen to Class A I'm really impressed with the cleanliness and an impression that almost shouts "I am distortion free". It's a nice base for complex classical as well as intimate folk groups and maybe string quartets.
When I turn back to the MyRef FE it's - "OK, I'm actually in the room with these musicians".😀 I think the shorthand for that is "coloration". Nelson Pass sometimes describes it as ,paraphrasing, 'warmth adjusted by introduction and/or removal of second order distortion'. I take from that, at least in his view - the concept that distortion (variance from pure technical Class A operation) is not in itself a bad thing - more correctly an element of tuning to preference and system capabilities. There are often trimmers in the circuit specifically intended to facilitate adjusting that second order distortion.
So my question - particularly to Siva who has done a lot of ClassA/ClassAB integration work, and Tom who has owned many top shelf amps including some ClassA tube products - what is the target and is it realistically obtainable? Ninety percent of what I listen to is split between large jazz groups an symphonic works. That "edgy saxophone" - or better said saxophone edge cutting through is what gives me that "in the room" character I love about the MyRefs. I don't know that I would like to "tweak" that out of the FE. Someone described one of discrete amps I built as a "Concert Amp". That is quite accurate with the F5 and BA giving a no drift middle of the auditorium presentation. No way am I trying to dampen or diminish the FE tuning activities, but the old Sesame Street song "One of These Things is Not Like the Other" keeps playing in my head.😉
Dario, did you ever have a conversation with Mauro dealing with the A/AB topic? So often in so many threads Class A is touted as "optimum" or more pure, but I'd be interested to learn if he also considered the introduction of small amounts of controlled distortion to be advantageous to the overall performance of his design, and specifically the EVO.
When I turn back to the MyRef FE it's - "OK, I'm actually in the room with these musicians".😀 I think the shorthand for that is "coloration". Nelson Pass sometimes describes it as ,paraphrasing, 'warmth adjusted by introduction and/or removal of second order distortion'. I take from that, at least in his view - the concept that distortion (variance from pure technical Class A operation) is not in itself a bad thing - more correctly an element of tuning to preference and system capabilities. There are often trimmers in the circuit specifically intended to facilitate adjusting that second order distortion.
So my question - particularly to Siva who has done a lot of ClassA/ClassAB integration work, and Tom who has owned many top shelf amps including some ClassA tube products - what is the target and is it realistically obtainable? Ninety percent of what I listen to is split between large jazz groups an symphonic works. That "edgy saxophone" - or better said saxophone edge cutting through is what gives me that "in the room" character I love about the MyRefs. I don't know that I would like to "tweak" that out of the FE. Someone described one of discrete amps I built as a "Concert Amp". That is quite accurate with the F5 and BA giving a no drift middle of the auditorium presentation. No way am I trying to dampen or diminish the FE tuning activities, but the old Sesame Street song "One of These Things is Not Like the Other" keeps playing in my head.😉
Dario, did you ever have a conversation with Mauro dealing with the A/AB topic? So often in so many threads Class A is touted as "optimum" or more pure, but I'd be interested to learn if he also considered the introduction of small amounts of controlled distortion to be advantageous to the overall performance of his design, and specifically the EVO.
Last edited:
Possibly until we create at least two reference amps with identical components and as homogenous as possible the other parts of the chain like alps dect pots and so on it would be difficult to get it sorted 🙂
No
Class A is supposed to be better simply for the lack of crossover distortion.
Amps like the MF A370 or Quad 405 should achieve similar performance using current pumps, if I've undestood correctly.
I'm simply hoping to be able to optimize the performance of the FE without seriously changing it's unique character. Just trying to better define the goal or target of potential mods. Don't want to fix something that's not broke.🙂
Last edited:
Dario, I do remember your system when I visited you. It was the first time I was hearing a "My_ref" amplifier and I've been impressed for the wide stage and the very good attitude to distinguish instruments tone and position.
Of course, I experienced no ear bleeding.
Bob, you are right: in the technical presentation of the My_ref (from Mauro Penasa), it was explicitly written that the topology of My_ref (in brief: voltage gain given by an opamp + a "current stage" + total negative feedback) has unique characteristics. They can be seen as an "evolution" of MF A370, at most.
I do not think that it will never be possible to push the My_ref design to be like class A amplifiers, that usually have no negative feedback at all, or just local one. In few words: in class A amplifiers the performances are obtained with "linearity": everything must be as linear as possible. While Penasa's design is taking components that has non-linearities (call them THD or whatever) and it is hiding them by a very smart use of phase compensations and circuit configurations.
Penasa wrote that he was explicitly trying to obtain a sonic signature as the one of SE tube amplifiers (at least for rev.A). I maybe wrong but I guess that he worked on phase compensations and feedback+feedforward techniques just to implement the "picture" he has in his head.
Regards,
Daniele
Thanks Daniele. I don't have the technical depth of you and others but it sounds like we are on the same page. I'll just repeat that I encourage and want to participate in further refinements, but just hope we can do that while maintaining the "signature" Mr. Penasa worked so hard to reveal.😉
On another topic - hum, I transplanted the beta builds into the monoblock chassis. I was getting hum that made the amps unusable with the attachment of SPDIF and USB shield on a direct connected DAC with a temp build on a wooden platform . I had even added long jumpers on the power grounds (not in photo now) to attach at various places to hunt down the problem. Nothing worked.
This is silent with only the earth ground connected to the chassis. It was so good that I had to double check if the PCB had a ground connection to the mounting pads/holes. I see metal vias (without continuity) in the beta but none in the final boards. Is there any significance to that Dario?
Anyway, the other thing that was interesting was the small space between the signal out wires and the transformer. These posts are long for use on speaker cabinets. I can detect no change due to that and attaching the transformer shield wire also has no effect - good or bad. It appears very close proximity component mounting is well tolerated by these amps. Somewhere down the road I'd like to build a 6" X 6" x 6" cube as the ultimate compact chassis - I think it will work just fine.😀
This is silent with only the earth ground connected to the chassis. It was so good that I had to double check if the PCB had a ground connection to the mounting pads/holes. I see metal vias (without continuity) in the beta but none in the final boards. Is there any significance to that Dario?
Anyway, the other thing that was interesting was the small space between the signal out wires and the transformer. These posts are long for use on speaker cabinets. I can detect no change due to that and attaching the transformer shield wire also has no effect - good or bad. It appears very close proximity component mounting is well tolerated by these amps. Somewhere down the road I'd like to build a 6" X 6" x 6" cube as the ultimate compact chassis - I think it will work just fine.😀
Attachments
Last edited:
Lookin good Bob! you have a small production line going there 😀
Can anyone tell me if these Amtrans are the ones listed in the BOM for C12, C30 & C32 the tolerance seems wrong.
Amtrans AMCH Capacitors homepage
Can anyone tell me if these Amtrans are the ones listed in the BOM for C12, C30 & C32 the tolerance seems wrong.
Amtrans AMCH Capacitors homepage
Thanks, Daniele 😉
So, finally, you managed to get your beta build hum-less?
Fine 🙂
No, I don't think plated mounting holes makes any difference.
Yes, they're the correct ones.
Obviously the more strict is tolerance better it is but 5% if more than fair and usually they measure within 2%.
BTW...No one stops you from buying them in excess and match them to 1% or less... 😉 😀
Thanks Dario, think i'll just use what they send me though 😀
Those Mundorf 10,000uF PSU caps are expensive! glad I'm not building a class A monster using them 😱
Those Mundorf 10,000uF PSU caps are expensive! glad I'm not building a class A monster using them 😱
If they fit, 63V caps will generally have lower Dissipation Factor than 40V, other things being equal. I'd go with 63V.
A Precautionary Tale
This won't apply to most builders but I thought I'd share it anyway. The new beta installation was starting to distort on one side. I disassembled to reheat everything and check for other problems. No help. I decided to just replace the LM3886, and as I started discovered pin #11 wasn't attached at the junction (bottom) of the chip. That's a NC pin, but I realized over the months I had often flexed all the pins excessively by leaving the small heatsink attached (added leverage) while doing some sort of mod or adjustment. So, no doubt one or more of the active pins had lost contact.
The LM3886 is a rugged device but being flexed that much - that many times had to cause a problem sooner or later. I was able to successfully remount a pulled LM3886 and the sound is fine now. But I wanted to share the story cause so many of those little shortcuts (read - lazy) we gamble with usually get us eventually. The time saved rarely balances out in the long run.
This won't apply to most builders but I thought I'd share it anyway. The new beta installation was starting to distort on one side. I disassembled to reheat everything and check for other problems. No help. I decided to just replace the LM3886, and as I started discovered pin #11 wasn't attached at the junction (bottom) of the chip. That's a NC pin, but I realized over the months I had often flexed all the pins excessively by leaving the small heatsink attached (added leverage) while doing some sort of mod or adjustment. So, no doubt one or more of the active pins had lost contact.
The LM3886 is a rugged device but being flexed that much - that many times had to cause a problem sooner or later. I was able to successfully remount a pulled LM3886 and the sound is fine now. But I wanted to share the story cause so many of those little shortcuts (read - lazy) we gamble with usually get us eventually. The time saved rarely balances out in the long run.
Attachments
What a nice looking heatsink. Copper to spread the heat and nice ally fins. It's even better looking than Dario's prototype heat sink. Would you care to share the source?
Jac
Jac
I wish I could. I bought about eight on eBay several years ago for a video render farm. Others have asked and I've searched several times for the vendor unsuccessfully. It was a surplus PC parts store in California but they disappeared. You might find something similar on eBay with keyword AMD Opteron or socket F cooler/heatsink. The laminated design helped keep the profile down for 1U server blades. If I find some more I'll post.
Made it through the snowbank!
The RC build arrived in tact in Denver today. Jim will give us an independent review and maybe a few scope readings if we ask nice.😀
The 2-day delivery took 6 due to weather but the mail gorillas left it in good shape.
stay tuned.......
The RC build arrived in tact in Denver today. Jim will give us an independent review and maybe a few scope readings if we ask nice.😀
The 2-day delivery took 6 due to weather but the mail gorillas left it in good shape.
stay tuned.......
Last edited:
correct.
A passive multi-way speaker is one that has the passive crossover components between the Power amplifier/s and the speaker drivers.
An active multi-way speaker is one that has the crossover components before the separate Power Amplifiers feed the respective speaker drivers.
As an aside, an active speaker can have a passive crossover! The amplifiers directly feed the respective drivers.
My k71-4 impression: sound is good. Top range is emphasized, mid range is lean and lows are not bad but lack some punch and grip. But overall sound is not as good as audyn TC. Price quality is great anyway.
Noticed this capacitor gives faster sound than TC which I really like.
Noticed this capacitor gives faster sound than TC which I really like.