My_Ref Fremen Edition RC - Build thread

[lehmanhill]

Originally Posted by madisonears

My application and evaluation of these caps was in the MyRef C, so perhaps the sound of the amp itself contributed to it.

Very nice evaluation, but focused almost entirely on broad tonal qualities. You make scant mention of the spatial characteristics, and that's where I found the True Coppers to excel beyond any others I've tried, in addition to providing a warm, dynamic sound with realistic midrange tonality and depth. You found the bass excessive, but I didn't notice that because my amps drive only the mids/tweets.

Perhaps I should give those K71-4's a try. Are there any commercial dealers, or must they be purchased from private parties at places such as eBay?

Perhaps you should give the TC's another listen for the layering of depth that I detected: subtle spatial and tonal characteristics of the music that no other cap provided. That, to me, differentiates the illusion of real music from an electronic reproduction. If the bass is too distracting, try moving your speakers.

You also failed to mention whether the caps you tested are the same value. We know how much influenece that has on the bass response of the MyRef amps.

Personally, I thought the DC coupled amp sounded pale and thin, and I preferred the musicality of almost any cap to the sound of no cap. That goes against prevailing opinion about caps in general, but, perhaps in this case, the amp was designed to sound best with some cap in the circuit, and listening without one gives a false reference. I suggest comparing caps to the sound of high quality commercial products or live acoustic music.

I never heard anyone mention a roll-off of Dynaudio tweeters. You certainly have some nice drivers there. Do you think there is much information above 14k to roll-off, and can you even hear it? I suspect the additional brightness you're hearing is much lower than that. What I have found is that brightness, good or bad, is often produced by the upper limits of a midrange driver, not the tweeter. I agree that the ZN cap does not have the extension to reproduce adequate detail in the highest frequencies, and the midrange is somewhat forward.

As you say, these caps provide the final flavoring to these wonderful little amps, and each person will optimize for their system and taste.

Peace,
Tom E

Tom,

I tried to stick to tonal qualities because it seemed easier to relate to speaker character.

I agree with much of what you say about the True Coppers. They had a warm sound and the sound stage was wide and deep. I'm not sure about the subtle spatial and tonal characteristics you mention. Perhaps I found them the same as the K71 or, more likely my speakers didn't show those differences.

At any rate, the K71 were cooler (warm enough for my system) than the TC and have a sound stage that matches the TC. If anything, the K71 gave a little more space between instruments than the TC, but I am talking very small differences.

The DC coupled amp is certainly detailed, but as you say a little thin. Overall, the FE sounds pretty good DC coupled, but I have to agree that the right cap makes the sound easier to listen to. I mainly used the DC coupled to better understand how the cap changed the sound.

I do like the Dynaudio drivers. The D28 are 28 mm soft domes, so they don't remain flat to 20k Hz. Your comment about high mid-range maybe being the difference is very interesting. The crossover is about 2k Hz, so maybe there could be improvement in the integration of the drivers in the crossover. I do want to clarify that what I am hearing is not so much an increase in brightness as an increase in detail and clarity.

As for the cap values, I would be interested in the history of sensitivity to C13 in MyRefs. In answer to your question, the K71 are 1 uF, the True Coppers are 1.2 uF, and the ZN are 1.5 uF.

[lehmanhill]

Originally Posted by ClaveFremen

The strange thing is that you heard such 'boost' only with True Copper...in your tests did you try the different orientation of caps?

Dario,

Orientation is good point. No I hadn't thought to try that. As I still have Bob's caps, maybe I should see if that would make a difference. In my tests so far, I oriented all caps so that the signal was flowing from the beginning of the name to the end (from the A in Audyn as input to the n in Audyn as output of the cap and input of the FE). Of course, there is no reason for this other than consistency.

Although there was some increase in bass from DC coupled with all caps, it was biggest in the case of the TC. There was one song on my evaluation list, Herbie Hancock, Possibilities, "I just called to say I love you". This song showed the difference very clearly.

The song has orchestra, piano, voice, and electric bass guitar. At one point, the bass guitar line is awkward, doesn't fit the rest of the music. With the ZN and K71, the bass guitar was low enough to be in the background. You could tell that the recording engineer turned down the volume on the bass guitar for that period because it was awkward. On the True Copper alone, the bass guitar was loud enough to realize how it didn't fit the rest of the music. This may also be a fault of my speakers. There may be some room interaction or boosted range of frequencies in the speakers that were emphasized by the TCs.

I want to emphasize that I don't criticize the True Coppers, just say that they don't fit well with my system. Too many other respected member have rave reviews of the True Coppers, so I assume the fault is with my system.

Jac

[bcmbob]

IMHO, its all getting into an area better described as "peripheral" to the actual FE development. As I said this morning, I hooked up my former go-to speakers and the difference was so dramatic one would have thought I was using a completely different amp. Not better or worse, but an entirely new character.

I have trashed my idea to use both sets of speakers to try to determine a baseline for the FE. The Sunflowers with the OB mids are naturally refined and expansive. The MTMs highly dynamic and penetrating with a much narrower soundstage. Getting them both to do what they were designed for (MTMs as part of a 7.1 HT system) appears too much to ask of a single set of amp components. Though doubtful/skeptical at first, Jac's claim that the True Copper or any quality cap - though stellar, may not be appropriate for all installations, has been demonstrated by what I heard today. I think he described it as "capacitor conflict". I didn't need the other caps on hand for direct comparison to understand I'll have to pick just one speaker design and tune to that. Tom's comments on speaker placement and room treatment will have me moving the furniture and visiting the local fabric store soon.

Though I still feel the Fremen Edition is the best base I have owned to work from, it looks like guaranteed audio nirvana is still elusive, selective and personal as ever.

[madisonears]

Pardon me for butting in again on a thread where I have no business.

Regarding value of C13: waaaay back in the MyRef C ultimate BOM thread at post 247, regiregi posted three graphs which show FR and phase response of three sizes of cap. The smallest, .68uF, shows decent response but some phase distortion (if that's what it's called) at 50hz. Of course, both factors got lower as the caps increased to 1.0 and 2.2uF. That's when the group decided to use 2.2uF in the "ultimate" kits. After some listening, we determined that value might be too big, and backed off some to 1.2 and 1.5uF. As Dario states, the original design value was 1.0uF.

Regarding DC coupling: if your source does not output DC when idling, that's no guarantee that it will not produce DC when active or operating the volume control. If your source output is AC coupled, then you have a lot less to worry about. In essence, you will still be hearing the MyRef AC coupled, but now you have no control over what cap and what value. I prefer the source DC coupled and I get to choose the DC blocker at C13.

Jac, thanks for making the distinction between brightness and clarity. They are, indeed, two very distinct qualities, but often confused though they're not really related. I can't tolerate brightness, but I cannot get enough clarity! It's too easy to make bright equipment and speakers, but very difficult to extract clarity. It's not mere semantics, and it's not at all a fine distinction. If cymbals sound tinny and tinkly, that's brightness. If they resonate with a bronze shimmer, that's clarity.

Clarity also extends well into the midrange, as in the detection of what a vocalist means, not just what words they're singing. I hope no one objects to my boring them with an example. On The Band's eponymous album, there is a song called "Rag Mama Rag." I have heard this song hundreds of times on a wide range of systems, and I know all the lyrics by heart. I was not prepared for the revelation that took place the first time I listened to it through the MyRef's and I heard Levon Helm not merely singing, but by the inflection in his voice I knew he was telling a joke. I have had similar experiences with classical music through the MyRef's, where, on familiar recordings, I could detect an emotion in the music I had never heard before. Maybe it's all just DIY delusion, or perhaps it's clarity.

Peace,
Tom E

[AndrewT]

Originally Posted by bcmbob

That safety factor has stopped me from trying no-cap as well. I might try some experimenting with the TP V1.2 builds - though I'll need some time to get my finger on the on/off switch to react faster than electricity.

No cap may test out as OK and probably work.
But why would it test out as OK?
Because there is probably a DC blocking capacitor somewhere earlier in the system.

How can a "no cap" be relevant, if there is an earlier DC blocking cap in the system?
Please explain that to me.

[AndrewT]

Originally Posted by madisonears

...................... making the distinction between brightness and clarity. .............

the following will be pure conjecture based on my understanding of other published test reports.

Excessive brightness could easily be explained as amplifier gain being higher for fast signals than for the normal signal. Measuring the frequency response with a continuous sinewave can show a perfectly flat result until the designed roll-off of the input RF filter.
Input a deliberate HF signal with fast edges and you will see either flat amplification of the faster edges or lesser amplification, or greater amplification. Many amplifiers fall into that last category, especially when a little capacitance is added to the load.
If you need a clue to where to look: square wave testing and look for overshoot.

Now to clarity.
If the HF is equally amplified when as a sole low level signal as compared to riding on the high amplitude envelope of the lower frequencies then you will "clearly" hear the HF.
However if the HF varies depending on what it is mixed with then the "clarity" is muddied. Not a scientific word but I think it expresses my thought.

Madison is absolutely correct to bring up this issue and you should all take heed.
BTW,
the Members with test gear and the knowledge to apply it, would all be able to test/measure for this and eliminate the bad from the wanted or simply reject as "not good enough".

[bcmbob]

Andrew, you are reading my mind again.

I considered expressing the situation similar to the way you are, but just chose a different set of words. Thanks for again making it clear that DC must be dealt with somewhere along the signal path. At this point I'm more than comfortable hiding behind the shield of C13.

[bcmbob]

Tom, You belong on this thread as much as anyone. The FE is, as I'm sure Dario would acknowledge, an extension of all the work and contributions from a whole lot of folks like yourself. Being MyRef veterans, we are still hopeful that you, linuxguru, Andrew, and Uriah will at some point get your hands on a pair of these amps and hear them in your own environments. That's partially why I asked Dario about any major changes proposed for the final PCB.

From my point of view, there is nothing better for explaining technical factors than the use of real world musical examples experienced by device builders/designers. Far from being boring - it's the target we are, or should be focused on. As posted, the BoxNet folders I set up expands that idea to the actual sharing of examples between forum members. It is still mostly unused but that was expected to be the case until a larger number of FE builders complete the amps. I'll re-post the link to the folder and ground rules a bit later. Any MyRef version user is more than welcomed to contribute examples if they choose to.

Hoping to get you and the other "Old Dudes" onboard - hands-on soon.

[Marra]

Thanks for the reply Dario.
I connected things again and the voltages on c102/202 were +- 14.5v give or take a little but the relay clicked on/off a few times before not locking on.The output offset was fluctuating between a few mv and 100mv and would not give a steady reading.Input was initially 1.5v but was slowly falling.
The LM318 was changed and unfortunately I don't have any spares of the bc639/640's.

[kroyin]

I had the same issue with myref d...lm318 was getting 24V in between pin 4&7.
Nothing across relay diode..nervous offset,relay led switches on & off & then stays off.
Linuxguru told me that this was a problem with the compensation circuit.
If you have 1 working channel debugging would be easier (as for me I had missed a jumper :-D)