Bob it's '' The Rite of spring'' and that has to be one of the most complex most
difficult music for any amp.Also check Dorati's ''Firebird'' 24 bit.The best recorded music i've heard. Keep up the good work!
Agreed, A real obstacle course. My ''Firebird'' is just 44K but I'll hunt down a better one.
Thanks
All My_Ref versions (both official and unofficial like the one in this thread) can drive 4 Ohm speakers but a 22V (instead of 25V) transformer should be selected.
You can contact LinuxGuru for Rev C boards and kits or wait for the Fremen Edition upcoming group buy (hopefully in a month or two).
Regarding BOMs LinuxGuru have his one, the Fremen Edition one is still under construction (you can see FE Beta BOM here).
Last edited:
Here are my impressions of the Rinken 390 ohm resistors at R10 instead of the stock Vishay/Dale 392 as listed in the BOM.
Foremost, there appears to be a definite enhancement of the overall 3D qualities of the music. That's both the stage in the room and inside the sounds themselves. On the end of "The Right of Spring" there are places where the tympani and the bass drum are in unison. With the stock resistor I hear more of the head of the tympani, with the Rinkens there is more response of the internal volume of both drums. This imparts a more physical sense as well as deeper/richer and more realistic tonality.
There is tighter control of the low end on everything - which can be misleading as the bass production level is still very much program dependent. I had to do the swaps 4 or 5 times to define what I was hearing. There seems to be a slight lessening of the amp's overall brightness/shimmer, but I have yet to determine if I'm mistaking that with the removal of a distortion element.
I should note that I did these tests without installing the 12K Rinkens, hopefully to be able to judge each change. I'll do the 12Ks next. I also bought a pair of ZNs for C13 from a forum member that should be here tomorrow.
I'm depending on my ears for everything now, but I'm getting curious about adding some sort of scope equipment to find out if I can see what I'm hearing. Does anyone have any experience/recommendations with the PC based approaches that don't require a second mortgage?
Foremost, there appears to be a definite enhancement of the overall 3D qualities of the music. That's both the stage in the room and inside the sounds themselves. On the end of "The Right of Spring" there are places where the tympani and the bass drum are in unison. With the stock resistor I hear more of the head of the tympani, with the Rinkens there is more response of the internal volume of both drums. This imparts a more physical sense as well as deeper/richer and more realistic tonality.
There is tighter control of the low end on everything - which can be misleading as the bass production level is still very much program dependent. I had to do the swaps 4 or 5 times to define what I was hearing. There seems to be a slight lessening of the amp's overall brightness/shimmer, but I have yet to determine if I'm mistaking that with the removal of a distortion element.
I should note that I did these tests without installing the 12K Rinkens, hopefully to be able to judge each change. I'll do the 12Ks next. I also bought a pair of ZNs for C13 from a forum member that should be here tomorrow.
I'm depending on my ears for everything now, but I'm getting curious about adding some sort of scope equipment to find out if I can see what I'm hearing. Does anyone have any experience/recommendations with the PC based approaches that don't require a second mortgage?
Yes, Rikens are incredible resistors that, like Black Gates, are no longer produced, sadly.
You should take in account that an all Rikens config would probably be colored, usually one or two of them in a circuit is plenty.
You've been wise...IMHO, it's useful to try a component a time so that it's possible to evaluate what each single components do.
It will be the next surprise...
Using diagnostic tools (like oscilloscopes) you can hardly see any difference, their resolution is too low.
You would need analysis tools like distortion analyzers, spectrum analyzers and so on... and, sadly, they're all out of hobbysts budget.
SoundEasy has a built in spectrum analyzer. But for testing electronics, one might consider Audio Tester. It seem pretty capable. I intend to try it myself.
http://www.audiotester.de/
http://www.audiotester.de/
OOps - I lied. I found that in the swap fest yesterday the Rinken 12Ks were actually installed when I wrote my last impressions. So today I did a real 12K comparison. It appears these may actually be the piece that enhances the 3D effect. Maybe Dario can confirm. The stock resistors also add some graininess that I wouldn't exactly call distortion but is definitely hearable. In any event I'm keeping the Rinkens in both positions.
I installed the M-Cap ZNs at C13 less than an hour ago. So far I'm not impressed. I'm hearing a dryer less-colorful presentation with reduced depth when compared to the Sonicaps. Even though these are not brand new, I'll reserve my final judgement till the whole amp has had a chance to adjust and ferment. I still think Tom E made a good call with the Sonicaps - though at this level, we are dealing with personal preferences.
I installed the M-Cap ZNs at C13 less than an hour ago. So far I'm not impressed. I'm hearing a dryer less-colorful presentation with reduced depth when compared to the Sonicaps. Even though these are not brand new, I'll reserve my final judgement till the whole amp has had a chance to adjust and ferment. I still think Tom E made a good call with the Sonicaps - though at this level, we are dealing with personal preferences.
Last edited:
Never tried Rikens on feedback...I'll do when they arrive.
I know, Sonicaps sounds delightful, but it's their signature (AKA sound coloration) you're hearing.
But you can have the same delightful sound from the My_Ref along with the finesse and details that Zns let pass and Sonicaps simply mask...
Strictly IHMO, of course...
Some news
Today I've tried Susumus RG32/RG20 on both input and feedback and KOAs RK73H on shunt positions.
They all sound GREAT.
Susumus are not too distant from naked Vishays, quite same detail, a bit less impact, a bit more full... considering their price there's no match, Susumus are a real bargain!
RK73H sound very similar to Caddock MK132s and they sound very very good also on R10 (390R) position.
I'm more and more convinced that the way to go for the final PCB version is all SMDs for the signal path.
I've yet to try Susumus and IRC PFC (tantalum) for input shunt (100K) and IRC PFC for R10 (390R).
Tried also for C102/202 Nichicon FGs, FWs, Black Gate PKs and no caps.
Black Gates wins but FGs are good seconds and I think they'll be part of the final BOM, unless someone can confirm that those caps are not necessary.
Today I've tried Susumus RG32/RG20 on both input and feedback and KOAs RK73H on shunt positions.
They all sound GREAT.
Susumus are not too distant from naked Vishays, quite same detail, a bit less impact, a bit more full... considering their price there's no match, Susumus are a real bargain!
RK73H sound very similar to Caddock MK132s and they sound very very good also on R10 (390R) position.
I'm more and more convinced that the way to go for the final PCB version is all SMDs for the signal path.
I've yet to try Susumus and IRC PFC (tantalum) for input shunt (100K) and IRC PFC for R10 (390R).
Tried also for C102/202 Nichicon FGs, FWs, Black Gate PKs and no caps.
Black Gates wins but FGs are good seconds and I think they'll be part of the final BOM, unless someone can confirm that those caps are not necessary.
Well, after some weeks of experiments BOM is near the final version.
All direct signal path resistors will be Susumus RG32 and shunt ones KOAs RK73H.
Shunt's PS caps will be Nichicon FGs 100uF 63V and C9 also Nichicon FG (Black Gates are better but FGs are great available alternatives).
Beta-testers can find the latest through hole version on the shared Google Docs folder.
I've yet to do some tests with R101/R201, with SMD film caps (Panasonic ECHU) and compensation resistors (probably going from MELF to Susumus).
Since tests results the final PCB will have more SMD components on the amplifier section (LM318's compensation caps, resistors), some resistors and capacitors (LM318 compensation) will be a bit smaller (0805).
I've attached an updated layout which includes the changes above, larger isolation from ground planes and hopefully easier to solder thermals.
All direct signal path resistors will be Susumus RG32 and shunt ones KOAs RK73H.
Shunt's PS caps will be Nichicon FGs 100uF 63V and C9 also Nichicon FG (Black Gates are better but FGs are great available alternatives).
Beta-testers can find the latest through hole version on the shared Google Docs folder.
I've yet to do some tests with R101/R201, with SMD film caps (Panasonic ECHU) and compensation resistors (probably going from MELF to Susumus).
Since tests results the final PCB will have more SMD components on the amplifier section (LM318's compensation caps, resistors), some resistors and capacitors (LM318 compensation) will be a bit smaller (0805).
I've attached an updated layout which includes the changes above, larger isolation from ground planes and hopefully easier to solder thermals.
Attachments
I did the C13 comparisons over the last week using Dario's suggester rig for easy swaps. It is really close and both caps have their own good qualities.
The Mundorf ZNs sounded more precise with noticeable separation between voices and instruments. They definitely have a dryer sound in my system. The side to side stage seems about the same but the front to back depth is reduced.
There is just no getting around the fact that the Sonicaps just sound more "musical" to me. I put that in quotes cause it is such a subjective area/definition. The 3D sensation is clearly more engulfing and I can't tell if that approaches something slightly "gimmicky" as Dario might suggest. It is so difficult to know and to reproduce what sounds "natural" and preferable to each individual.
So my non-answer answer is - the jury is still out. As the Fremen Edition developes there may be a combination of parts that brings out the best in each, or eraces the differences. I'll put the Mundorfs back in again later in the stream.
The Mundorf ZNs sounded more precise with noticeable separation between voices and instruments. They definitely have a dryer sound in my system. The side to side stage seems about the same but the front to back depth is reduced.
There is just no getting around the fact that the Sonicaps just sound more "musical" to me. I put that in quotes cause it is such a subjective area/definition. The 3D sensation is clearly more engulfing and I can't tell if that approaches something slightly "gimmicky" as Dario might suggest. It is so difficult to know and to reproduce what sounds "natural" and preferable to each individual.
So my non-answer answer is - the jury is still out. As the Fremen Edition developes there may be a combination of parts that brings out the best in each, or eraces the differences. I'll put the Mundorfs back in again later in the stream.
Attachments
Thanks Bob for testing and reporting.
I do agree, the Sonicap is more "musical" but the question, IMHO, is:
Which one is more transparent?
Also, did you tried to DC-couple for comparison?
It seems you're still using the Illinois Capacitor caps for C102/C202, am I right?
If so the dry sound you're earing is possibly due to them.
If you can, mount sockets for C102/C202 and try the amp leaving those positions unpopulated.
If you can try Nichicons FGs there, they're pretty musical too.
Input caps and resistors are the most critical, IMHO, since they carry a low level signal, I think here transparence is utmost important.
I do agree, the Sonicap is more "musical" but the question, IMHO, is:
Which one is more transparent?
Also, did you tried to DC-couple for comparison?
It seems you're still using the Illinois Capacitor caps for C102/C202, am I right?
If so the dry sound you're earing is possibly due to them.
If you can, mount sockets for C102/C202 and try the amp leaving those positions unpopulated.
If you can try Nichicons FGs there, they're pretty musical too.
Input caps and resistors are the most critical, IMHO, since they carry a low level signal, I think here transparence is utmost important.
Here's something to throw up the flagpole. I went from Flim and the BBs with lots of electronic bass and jazz voicing, to Pachelbels Canon in D played by a chamber orchestra. The Canon realy sounded muddy and indistinct with the Sonicaps installed. The Mundorfs made a dramatic improvement in seperation of the instruments/sections.
The difference was so apparent the idea of installing both with an A/B switch came to mind. As stated, I usually lean toward the Sonicaps with "modern" music, but the ZNs show a clear advantage with a lot of Deutsche Grammophon type classical recordings.
I know much of this "tuning" can be done with DSP software but I would like to keep the signal as pure as possible. I am assuming such a toggle would be done with the power off, but maybe Andrew and/or others can shed some light here.
Does anyone have comments/experience with this type of arrangement. Switching between caps is so easy and immediate with the fastons it would seem to be a simple option.
The difference was so apparent the idea of installing both with an A/B switch came to mind. As stated, I usually lean toward the Sonicaps with "modern" music, but the ZNs show a clear advantage with a lot of Deutsche Grammophon type classical recordings.
I know much of this "tuning" can be done with DSP software but I would like to keep the signal as pure as possible. I am assuming such a toggle would be done with the power off, but maybe Andrew and/or others can shed some light here.
Does anyone have comments/experience with this type of arrangement. Switching between caps is so easy and immediate with the fastons it would seem to be a simple option.
Last edited:
It's the reason I try to avoid any strong coloration, it may sounds wonderful with some type of music and awful with others.
Yes, it's better if you power off if you make the swap manually.
But probably if you use a switch it's possible to do the swap on the fly, IMHO.
Let'see if someone else can confirm it or not.
I was afraid that you would suggest more sockets, but I thought it a possibility, and will do so.
Fine Joe!
A bit of time has passed...
By memory the four tiny should be BAV99W diodes (tiny three legs), the two bigger ones are the wrong 27V zeners (you will not use them) and two tiny 27V zeners.
Attachments
Last edited:
D1 and D2 are the BAV99s you can see in the attached picture.
D3 is the 27V zener (not still mounted in the attached picture)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.