My "audiophile" LM3886 approach

Android connected to 2-way Crossover (100Hz)
Lowpass output buffer has a gain of 2 (~6db)
Myref_C coupled via 1uF input cap.
FullRangers: Philips HQ48414P 8 ohms (x 2)
Subs: JBL GTO1214D (x 2) 8 ohms housed in a 1 cu ft. sealed box.

I have a spare crossover which I think is 200Hz. Not sure if JBL can handle that high-end.

Since this amp is never going to be used with another pair, I can add the required boost to the subs. JBL rates them at 93 but at 2.83V which is equivalent to 86.71 W/1M as per WinISD (thus the 6db boost in the xover for subs).
 

Attachments

  • IMG_20200727_000055960-1.jpg
    IMG_20200727_000055960-1.jpg
    437 KB · Views: 909
  • IMG_20200726_234324980-1.jpg
    IMG_20200726_234324980-1.jpg
    786.5 KB · Views: 898
Given that the LM318n is running at +/-12Vdc, does it mean I can use an input signal that swings at a theoretical max of +/-12Vdc to hit the full blast?
NO. The LM318 will only swing to within 2 volts of the supply rail, according to the datasheet. IMO, it is a poor choice for a crossover IC. These days, an LM4562 is only ~ $1.40 (qty 10) and is orders of magnitude a better opamp.
 
In a myref_C LM318-N is not used for crossover, neither is it a point of debate anymore. Please refer to the initial pages of this thread for the discussion.

My crossover uses LME49720 (removed OPA2134).

My suspicion is:
A. The room is too big for JBL GTO1214D. I'm in a 12' x 15' hall connected to a 10' x 20' gallery. My kinds on the top floor complaint of things shaking at 40Hz. However 2m away from the sub, there's little to write back home about.
B. The Philips HQ48414P doesn't go down to 100Hz.
C. Android headphone-out has a poor low-end. Because coupled with a Bluetooth received hooked to myref_C the bass is clearly audible. Thus my assumption of a pre-amp to bring up the levels at least to a reasonable level before feeding it to myref_C.
 
Hello,
I am a late newbie bird who is starting to build the legendary My_ReffC. Kudo to the inventor, Maruo, who sadly has departed.
I have read partially the forum content but still not able to find the answer: can I replace the zener shunt regulator diode for LM318 with any zener having 12V and 1W ratings?
Thanks!

Chandra
 
Hello, finally I have completed building it. After 2 days running, my impression for this amp, is a very good sounding. Almost similar, on par, with my HK-990. I use Marantz CD6005 cd player and JBL HP520 floor stander. The bass is very good, mid is good, but treble is a bit wee harsh to my liking. I use a cheap MKP for C13 input cap at 1 uF. I don't use black gate at C9 220 uF. I use also usual metallics resistors. If I would like to tame the harness a little bit, what parts should I upgrade?
Thanks.
 
It's your choice, but I'd go with the Rev C rather than Rev A - it has better subjective audible sonics.

MELFs and COG ceramics are OK. I've used NP0/C0G in my builds at some locations, and they work.

You don't need a preamp for a MyRef, unless you want to incorporate a tone-control. A pre- will add another layer of opacity to the signal path, which should be avoided if your main goal is transparency.
 
Hi Linux guru, we can mod the RevA board to RevC? I saw a Mauro’s pdf showing changes

If I m wrong let me know what to change.

Also I needed source selection and wireless volume control.

I m not having very good ear’s ( tested ) due to work environment so don’t know how what would sound ( not being rude )

I have been using raspberry pi with Volumio / Moode and I enjoy that. It has a JustBoom Amp Hat which is basically TAS5657M 24bit /192khz dac + Amp 30 + 30 in 8ohms in one chip.
 
i think careful matching of devider ratio with 1%resistor will give samesonics.0.1% is manufacturing tolerance.
This is one of the worst uninformed suggestion I've heard about the My_Ref.


Before questioning the designer specifications some research should be done, at least...


The My_Ref output stage is an Howland Current Pump, which NEEDS precision resistors for best stability and performance (see, for instance: https://www.ti.com/lit/an/sboa437/sboa437.pdf?ts=1630053071234).


In addtion hand selecting 1% resistors does not give you any guarantee they will not derate with time.


0.1% resistors are not simply selected for an initial measured value, they are actually DIFFERENT, often TCR is differetn too.



And obviously this very same topic have been explored several times on this and other threads about the My_Ref and variants.