Link for future reference to another nested/composite amplifier using the LM3886 chipamp:
http://www.diyaudio.com/forums/chip...mposite-amplifier-achieving-0-0004-thd-n.html
http://www.diyaudio.com/forums/chip...mposite-amplifier-achieving-0-0004-thd-n.html
Fifth run
I've started a new GB for My_Ref FE boards:
http://www.diyaudio.com/forums/group-buys/262319-my_ref-fremen-edition-gb-fifth-run.html#post4065553
I've started a new GB for My_Ref FE boards:
http://www.diyaudio.com/forums/group-buys/262319-my_ref-fremen-edition-gb-fifth-run.html#post4065553
I just wanted to humbly make available my single side "REV C" monobloc design which is suitable for REV-A/B/C (with a little simple re-arrangement).
PDFs
The PDFs are suitable for the toner transfer method as the copper side is not mirrored, and the component side is.
Please enjoy it if you are capable of etching your own boards.
If anyone makes a good amp with it then this journey has been worth it for me.
Thanks to everyone who has made this a lot of fun (and very educational) for me. The results are largely thanks to you. Thank you Mauro most of all.
Cheers!
Russ
EDIT: fixed the link.
The link is broken.
Does anyone have this file?
Thank You
Question about using 318.
First, to the OP I would like to say very nice implementation and congratulations.
It is curious that you chose the 318 instead of the more popular op amps. I haven't seen one since the 80s. I realize that it is very fast and has good DC performance, both probably useful for your circuit.
Did you use any overcompensation on the 318? It was often recommended for audio circuits, and the couple of times I used one I did it. Did you consider this in your design and prototype?
I'm working on a design for a nested 3886 amplifier. It will be much more modest than yours. After looking at your circuit, I was thinking about trying the 318 for the front end. For a simple nested, buffered inverting op amp type circuit, is their any advantage to using the 318 over say one of the fancy Burr Brown devices?
I'm going to Miller the front end out around 300 kHz f3 for a start. The 3886 will be a simple non inverting 20 dB nested gain amp used as an output buffer for the input stage. So would a 318 be overkill or maybe even a liability for this simpler configuration?
Thank you and thanks to anyone that comments on my questions.
First, to the OP I would like to say very nice implementation and congratulations.
It is curious that you chose the 318 instead of the more popular op amps. I haven't seen one since the 80s. I realize that it is very fast and has good DC performance, both probably useful for your circuit.
Did you use any overcompensation on the 318? It was often recommended for audio circuits, and the couple of times I used one I did it. Did you consider this in your design and prototype?
I'm working on a design for a nested 3886 amplifier. It will be much more modest than yours. After looking at your circuit, I was thinking about trying the 318 for the front end. For a simple nested, buffered inverting op amp type circuit, is their any advantage to using the 318 over say one of the fancy Burr Brown devices?
I'm going to Miller the front end out around 300 kHz f3 for a start. The 3886 will be a simple non inverting 20 dB nested gain amp used as an output buffer for the input stage. So would a 318 be overkill or maybe even a liability for this simpler configuration?
Thank you and thanks to anyone that comments on my questions.
Mauro (the OP) has left the premises, though he's still around in the real world and usually replies to email.
However, I'll try to answer some of the questions from the information generally known about the MyRef Rev C.
1) Compensation: The Rev C uses a fairly elaborate compensation schema, of which feed-forward plays an important role.
Mauro liked the sonics of the LM318 from the Musical Fidelity A-370 developed by Tim Da Paravinci and others. He wanted to replicate the sonics of the A-370, but with a chipamp power stage, and the MyRef is the result. Several other opamps are usable as the outer-loop opamp in a MyRef, but the sonics are mostly worse.
If you want to use a wide variety of other opamps in a composite amp, check out the MiniRef:
http://www.diyaudio.com/forums/chip-amps/184165-miniref-schematic-pcb-layout.html
Not a direct answer, but the Miller configuration would be closer to the MyRef Rev A (except the LM3886, which is still a Howland Current Pump in the MyRef). Generally, the sonics of the Rev C have surpassed the Rev A across the board. If you want see a voltage-voltage composite amp with the LM3886, there are a few (perhaps including the Modulus).
For fresh designs, I would suggest looking at other opamps than the LM318 for the outer loop. As mentioned earlier, it is required for the MyRef because the compensation is very specific to it, and gives excellent audible sonics. However, I have used LT1361, LM6172 and other opamps in composite designs (not the MyRef) with excellent audible sonics.
Thank you. Back in the early 80s there weren't a lot of op-amp choices for good audio circuits. The 318 was one of the choices. Now it's so much easier. and cheaper too.
One thing I like about using the 318 for a front end is the offset trim function. But I prefer simplicity if possible for this design; a dual op amp in a DIP-8 is ideal.
The compensation I was talking about was from the datasheet. You can slow it down and/or control the settling time with a couple caps and resistor. I wanted to know if this was a factor in the design process at all. But I will probably look at some other devices for the front end. My design will not require any elaborate compensation scheme or feedforward compensation; it will rely on one Miller cap, brute force style. I want simplicity and predictability for this design.
So thank you for your comments.
One thing I like about using the 318 for a front end is the offset trim function. But I prefer simplicity if possible for this design; a dual op amp in a DIP-8 is ideal.
The compensation I was talking about was from the datasheet. You can slow it down and/or control the settling time with a couple caps and resistor. I wanted to know if this was a factor in the design process at all. But I will probably look at some other devices for the front end. My design will not require any elaborate compensation scheme or feedforward compensation; it will rely on one Miller cap, brute force style. I want simplicity and predictability for this design.
So thank you for your comments.
Again I thank you, linixguru. The MiniRef link you provided has a good article about stability and grounding with the 3886. Although I'm not going the MiniRef route, this article provides some good pointers.
http://www.neurochrome.com/audio/?page_id=941
I'm hoping to come up with a flexible scheme that can also be configured as a nested bridge amp (with a differential output op amp front end) and bridge parallel amp, scalable up to 200 watts. But first I will build a simple 50 watt version. I have all the expensive stuff anyway and only need a few chips and capacitors to complete the project.
http://www.neurochrome.com/audio/?page_id=941
I'm hoping to come up with a flexible scheme that can also be configured as a nested bridge amp (with a differential output op amp front end) and bridge parallel amp, scalable up to 200 watts. But first I will build a simple 50 watt version. I have all the expensive stuff anyway and only need a few chips and capacitors to complete the project.
Last edited:
You mean like a BPA200? Check out the 3886 application notes...
digi01 did some boards a while back.
BTW: do you know that Mauro's "my_evo" (which was originally based on the "my_ref" Rev.C) have received a (minor) later update?
I had a chance to listen to both original and new configuration, and there are some little but significant improvements in sonics.
The interesting thing is that (for the relevant parts) that update can be applied to the old "my_ref" as well... it may be worth trying.
Last edited:
BTW: do you know that Mauro's "my_evo" (which was originally based on the "my_ref" Rev.C) have received a (minor) later update?
...
The interesting thing is that (for the relevant parts) that update can be applied to the old "my_ref" as well... it may be worth trying.
If the my_evo is the same as the Evolution, I have seen the outline schematic that was published in the Italian audio magazine (excerpts of which have been posted here earlier).
IIRC, the main changes were:
1. addition of TL071 DC servo and associated filter components;
2. addition of series-shunt regulator based on LM317 for the opamp rails;
3. addition of an optional second Howland current pump (LM3886);
4. Changed input DC blocking capacitor value to 220 nF (only possible with DC servo installed);
5. zener diode limiter/clamp in feedback path;
6. Slightly modified compensation values.
I believe that portions of 2 & 6 can be retrofitted on a Rev C board. I have tried various values of C10, C34, C32 and R43 - with modest to significant change in sonics and soundstage.
The interesting thing is that (for the relevant parts) that update can be applied to the old "my_ref" as well... it may be worth trying.
It was already tried with success with the Fremen Edition, the My_Evo (Evolution) Rev. A mods can be applied to Rev C. (where applicable) with success and sonic improvements.
BTW... you missed the GB, again
ciao,
Paolo had been talking about updates to the evo project :
ARCHIVIO SCHEMI E PROGETTI AUDIO
For the point 2; - no, it's not an lm317 series shunt;
3; not necessarily, principally it's still a single current pump, though expandable
4; had been changed in the 'evo rev a' upgrade; to 1~ uf
All this, and a little tweaking with parts - for me real great sound.
Ciao, george
Paolo had been talking about updates to the evo project :
ARCHIVIO SCHEMI E PROGETTI AUDIO
For the point 2; - no, it's not an lm317 series shunt;
3; not necessarily, principally it's still a single current pump, though expandable
4; had been changed in the 'evo rev a' upgrade; to 1~ uf
All this, and a little tweaking with parts - for me real great sound.
Ciao, george
No. I would use a true differential output op amp for the input stage, with each leg of the bridge output stages nested into the input stage.
Thanks for the link - it appears that all the Evolution-related pdf files have gone offline, though I had downloaded some of them a few years ago. Perhaps the only one that I may have missed is Evolution Rev A - which apparently has several changes, including the 1uF input coupling caps.