My "audiophile" LM3886 approach

I changed LM318s and the transistors (BC546A & BC639) today and noticed some strange behaviour:

When turning the power on with LM318 pulled out, everything seems to be ok. Relay switches on and LED is shining.
But when the LM318 is in its socket, the problems I described above occur... 😡

It's really weird...
 
Hawaii said:
I changed LM318s and the transistors (BC546A & BC639) today and noticed some strange behaviour:

When turning the power on with LM318 pulled out, everything seems to be ok. Relay switches on and LED is shining.
But when the LM318 is in its socket, the problems I described above occur... 😡

It's really weird...


I would start to look for open or shorted passive components. or possibly a component in the wrong spot?

Hope that helps.
 
Precision Resistors

Guys,
The price of the precision resistors is INSANE! Wow. I saw some for as much as $23. Craziness.
I thought maybe I will buy 0805 devices and put them in series and parallel. 4 for each resistor needed. Brings the price down a lot.
The resistors I am talking about are the 22k and 47k .1% resistors. .1% of 22k is 22 and my DMM wont measure down that far. I have a bunch of Dale 1% metal films that fit the bill for normal tolerances at 22.1k but its probably nowhere close to .1% if my DMM would measure it.
So, whats the deal with these resistors? Why the precision?
Also, any purchasing leads? What P/N and distributor did anyone who found cheap source use?
Oh, and where did you find your non-inductive wirewound .47ohm resistors? Any decent prices?
Many thanks
Uriah
 
.1% resistors are not strictly required.

A slight mismatch will raise the output impedance a bit.

I got perfectly fine results using 1% resistors which I matched as close as I could. they only need to match each other, not the indicated value. You are after precise matching more than precise value.

Also, the output resistor can be inductive, its not a problem.

Cheers!
Russ
 
Thanks guys. I appreciate then answers.
Glad the non inductive is not necessary.
I was considering using regular wirewound, but in series and parallel with each resistor facing the opposite direction of the previous resistor to reduce inductance AND be able to use lower wattage resistors therefore lower cost.

Sci-Tech Dictionary:
Ayrton-Perry winding

(¦er·tən ¦per·ē ′wind·iŋ)

(electricity) Winding of two wires in parallel but opposite directions to give better cancellation of magnetic fields than is obtained with a single winding.


Uriah
 
Suitable for 2 ohm??

Hi members,

I like to build this amplifier and use it for the Mid/ High Ribbon of my apogee Duetta. It would be driven active, so amp will "see" arround 2 ohm.
It works beginning from 400 Hz (24 dB), so the amp doesn't need to deliver the more energetic low frequencies.
What's your opinion about this; Is this chipamp capable of serving the ribbon?

Thanx for you input.

Olaf
 
Re: Suitable for 2 ohm??

taotao said:

What's your opinion about this; Is this chipamp capable of serving the ribbon?

The "my_ref" works best with >= 8 ohm speakers. The max current output it can deliver (limited by the power op-amp chip capabilities) is about 10A. That's not much for a ribbon.

You may consider the "full version" (two parallel 3886 per channel) of Mauro's "My_Ref Evolution" (AKA "My_Evolution" or "My_Evo" for brevity) that, apart from being able to handle harder loads, does sound much better than the original my_ref.
 
Hi,
the 3886 specification quotes the maximum output current as 7A min to 11.5A max for a 10mS pulse.
This is for a chip with Ta=Tc=25degC.
Spike will reduce these maxima.
National are careful to not tell us how much Spike reduces the maximum current as Tc rises.
 
Hi Mike,
We expect the kits to be $65 for dual mono. This includes the boards and all components. We may may may consider selling just PCBs but we are not keen on the idea at the moment and have not asked the permission of Twisted Pear or Mauro to sell just PCBs. I dont think they would have an issue but its not been brought up and so far we would prefer to just do the whole kit , that way we can bring down costs for those who are purchasing the entire kit.
BTW here are those who are now in for kits:
bacood_2000 x2
warp_foo x2
Coreyk78 x2
SRMcGee x2
troystg x4
samsagaz x2
widman x2
billyk x2
korben69 x2
tube-arama x2
bandol83 x2
erikdebier x2
Joshua Adams x6
helfred x2
vglaso x2
Dxvideo x2
nvrgdenuf x4
Mindcrime x2
ben_lesmana x2
Daveze x2
Fredlock x2
argofanatic x2
GiovanniM x2
bhm x2
jERiCOh x2
TonyX x2
madisonears x2

=========
62 kits.... 38 to go