My_Ref Fremen Edition RC - Build thread

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Put my paint comment in context.
What way around does the painting machine orient the unpainted resistor?

If you and the machine are synchronised then maybe you get the resistor/s the same way around.

What effect do the different pigments have on the resistance and the parasitics attached to that resistance?
 
All good points Andrew. I can only offer the possibility of a consistency within the reports of many socket users describing what they hear as validation.

As much as I hate bringing up the subject, it is starting to look like many of the changes in sound could be emulated with a graphic equalizer. Even if someone took that path (which I strongly resist) it IMHO, is still worth the trial process at the RC stage to develop the "best" base unit possible.

I think we are all anticipating some good information from RC builders who have propper lab equipment to present a more technical picture/explanation of what is changing. Though not highly sophisticated, I hope to make some progress later this week on the Visual Analyzer setup mentioned earlier in this thread.
 
Last edited:
I think the beta pieces Dario did for his brother were socket free. Maybe he can report on the significance here.

When socket are removed it always sound better, it was so for the Rev C and it's so for the FE.

What way around does the painting machine orient the unpainted resistor?

If you and the machine are synchronised then maybe you get the resistor/s the same way around.

So far markings seems consistent (orientation wise) between batches but, in general, we dont have any guarantee on it.

What effect do the different pigments have on the resistance and the parasitics attached to that resistance?

I know, Andrew, it seem absurd considering resistors orientation and pigments on resistors could be the next 'logical' step... :D

I refused to trust it for years but I've had to accept it (at least in my reality... ;) )
 
Knee-O-Bee-Ummmmm

I am in the process of comparing unusual choices of capacitors for C9, 220 uF. Good news, Soongsc's CapXon NP caps arrived today. I'll give them time to burn in, but the first impression was promising.

The other cap I am playing with is an AVX Oxi-Cap, low ESR, Niobium Oxide. This is a larger than most, surface mount capacitor that could be soldered directly to the through hole pads at C9 on the FE RC board. Since I want to compare them to the Black Gate PK (early favorites) and to the CapXon, I am soldering on leads that will work with sockets.

Is it worth the trouble? I think so. My impression is that the Niobium cap is competitive with the BG PK. I won't say any more at this point, but would like to invite other builders to try them and give their impressions. For obvious reasons, that especially means you, Dario and Bob (check your PM please).

A short note on Niobium Oxide caps. Niobium is a cousin of tantalum. In oxide form, the cap is made of sintered material in a plastic case. The sintered material can absorb humidity. This causes a problem as the water can expand quickly during soldering and cause the cap to fail. For that reason, the caps are shipped in a sealed bag at low humidity. They should be soldered soon after opening the bag. Alternatively, they can be baked to dry them out.

Also, they have a polarity with the anode or plus denoted by a black stripe on one end of the cap. Here is the Mouser page;

NOME227M006R0040 AVX | Mouser

And here is how I added leads for socket comparison. First, using tacky solder and normal surface soldering technique, solder on a lead.



Then, cut out the lead in the middle to separate the two side of the cap.



After filing the leads to fit, this is the cap with leads on the board.



Jac
 
Last edited:
I have attached picture of my 15nf polistirene caps without yellow dot. Where should be the dot ?

I have tested my new MyRef with some low sensitivity speakers small standmounter 8ohm 85db (Spendor SA1) and Spike protection was activated very soon. The same is happening with some test speakers that i have home also not so sensitive (Sony Sb1000) also 8ohm. Unfortunately i dont have anymore (sold) the speakers used when testing RevC amplifier.
Should i be worried about this Spike protection that activates so soon ?
 

Attachments

  • 15nfpolisirene.jpg
    15nfpolisirene.jpg
    763.3 KB · Views: 254
Atupi,

The yellow dot is on the top of the capacitor. Look closely at the pictures in post #25 of this thread. The yellow dot looks like a messy paint mark on the end.

Also, I have 4 ohm speakers (22V transformers) with about the same sensitivity. The heatsinks get slightly warmer than hand warm, but no SPIKE protection, even after some loud playing. You may have an issue. Did you use thermal grease between the 3886 and the heatsink? I tried it briefly without grease and the chip got hot very quickly.

Beyond that, I'm no help. Hopefully Dario and others can help.

Jac
 
Regarding if C9 capacitors change the sound level.

I did a little experiment. I measured the sound level with a cheap Radio Shack sound level meter. Playing the same music, but different caps, the sound level was the same within my ability to read it. So far, Dario is right. The sound may seem louder, but it is probably something like improved bass rather than an overall sound level change.

Jac
 
I have attached picture of my 15nf polistirene caps without yellow dot. Where should be the dot ?

They're for C7, right?

If they're 15nF they don't have any yellow dot.

The KP1834 marking shoud go toward the LM3886 but I would try it using sockets (the right side is the sweeter and fuller one)

Should i be worried about this Spike protection that activates so soon ?

It depends... sensitivity is important (better >=87db) but it the dinamic impedance goes under 4 ohm (it could be the case) it could be a worse problem.
 
I will try also to change the termal paste to a more advanced one as i'm using a generic one now. The heatsink and chip are heated but to too much as i can keep my fingers on them. My heatsink is Heat Sink A4291-60 :: Heat Sinks :: Electromechanical :: Electronic Parts :: Banzai Music 60x165x35mm

I read that the Spike protection is related to temperature but i think a hungry (ineficient) speaker can heat up too much the chip.

What confuses me is that on RevC i used two computer heatsinks with almost no termal paste and had no problems with Spike.
 
Last edited:
atupi, when you remove the heatsink can you see a full footprint from the LM3886? This is what happens when a bit of old compound residue remains between the chip and the heatsink. Do you have any means to measure the temperature other than touching?
 

Attachments

  • FPnt.jpg
    FPnt.jpg
    127.8 KB · Views: 326
I have changed today the C32-150pf cap from Amtrans to NOS polistirene recommended by DArio and the difference was amazing even more noticeable than rolling different caps in C7 position.
For C9 i played around with BG PK and Nichicon Finegold and i'm not decided yet. Nichicon sound a bit fuller that BG wich is very detailed but i'm not decided yet.
 
Beta build is operational again.

Component mix: ApexJr 120 VA toroids, M-Cap Zn, Black Gate STD, C101/201 Generics, Amtrans, Caddocks and Renkins. Still sounds great and very close to the premium RC build. I'll do a little more swapping and twirling and then remove all sockets for a final solid build.
 

Attachments

  • Beta1.jpg
    Beta1.jpg
    270.4 KB · Views: 297
  • Beta2.jpg
    Beta2.jpg
    200.1 KB · Views: 290
Yes, the full impression is louder by about 10- 12% - not a big bump up. What is more significant is the sense of reserve power or headroom. All amps I've heard have a point when it sounds like they are struggling to keep up. The larger transformer moves that point farther away/up the volume scale. Remember this is closely related to the size and efficiency of the speakers being driven. Bookshelf speakers might be just fine with smaller transformers.

The low end is both augmented and better controlled - particularly when combined with the Mundorf AG caps , but that improvement actually extends well into the mid frequencies. I'm not sorry I bought the "Big Daddys" but anything between 125 and 200 would have been completely adequate. I get the impression from Dario that R-cores push a bit more power per transformer rating. That might be something to take into consideration.
 
Last edited:
Thanks a lot. Now I'm using two 160VA from my GC, so I'll wait to end other project (about room's acoustic) and I'll evaluate an upgrade toroidal or r-core.
I've also read, Dario experience with r-core and what is striking is the difference in efficiency: 120VA for r-core and at least 250VA for toroid!

Roberto
 
CapXon - Courtesy of Soongsc

I have finished my evaluation of the CapXon 220uF non-polarized cap in C9 position. I made several evaluations in hopes of not fooling myself. My main comparison cap was the BlackGate PK which has been my early favorite.

Overall, I found the CapXon NP to be equal to the BG PK, but a little different. The image between the two was very similar, but the CapXon seemed to have slightly more detail. In contrast, the BG PK had slightly more bass and bit rounder sound. The choice between the two would depend on your taste and your system, but they are both very nice sounding caps. I am especially happy that Soongsc found the CapXon because it is a good alterantive and they aren't making BlackGates anymore.

Jac
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.