AudioSector-chip amp kits, dacs, chassis

[Peter Daniel]

Those are pretty expensive.

From what I've read so far, it doesn't neccessarily has to be 0.1 ohm value. In case of bigger mismatch between chips, even 0.5ohm is quite acceptable (and actually beneficial), so 0.47, 3W Panasonics from Digi Key can be used here as well. It will just generate bigger voltage drop, but with higher impedance loads, and average power levels, it shouldn't be a problem.

[Peter Daniel]

Here we go, a version with dual output resistor footprints. Caddocks MP-820 and MP-930 can be used as well.
Thanks for suggestion, pinkmouse

[janneman]

Originally posted by Peter Daniel
I'm slowly running out of LM4780 boards, and i was considering a revision, with updated layout, accomodating all 3 options: parallel, balanced and stereo.

If balanced would be to implemented, the board does not use inverted method (like per AL-1192 notes), but both halves are non inverting and additional circuit is required.

I will be comming with balanced active line stage, so that could be used there, or I might include a small board, implementing LM6172 for buffering and phase splitting.

Below is the proposed layout I was working on yesterday. As you see, the signal traces lengths are substantially reduced, there is more separation (in layout) between both sections and the chip can be also used in stereo mode, providing same funsctionality as two LM3886 chips.

If stereo or balanced operation is looked for, the balast resistors are not required and output is taken directly before those output resistors. In case of parallel operation, resistors should be in place and there is one output only.

The signal grounding is very symetrical for both channels and also simplified.

I plan to do mirrored boards (2 amp board per one main board, as usually). So if someone uses boards for balanced operation, they can be put back to back with PS connections aligned for paralleling. A multiple number of such parallel modules could be connected in parallel (with bus wires for supply) providing simple power distribution scheme and high power.

The accompanying PS board will feature all presently popular options (large caps and snubbers)

As usually, all comments and suggestions are welcome. It will be at least a month before those boards become available.

Peter,

I like your layout. I very much like your thinking ahead for multiple uses, if you can pull that off without getting an awkward board, that'll be great.
You've come a long way, buddy!

Jan Didden

[pinkmouse]

Originally posted by Peter Daniel
Those are pretty expensive.

Welcome to UK overpricing!

I have frequently used 0.22 with other GCs, and they do work well, as you say, with the average load, it makes little difference.

Panasonic resistors? Not here mate...

[Peter Daniel]

This board will be offered at the same price as previous version, yet it offers 4 independant channels. So if someone has already LM3875 (or LM4780) based 2 channel amp, with just one kit can upgrade to 6 channel HT setup, and save substantialy (and not neccessarily in UK )

Besides, it seems like LM4780 chip has been neglected somewhat, I'll try to revive some more interest with various interesting implementations.

[Hanzwillem]

Peter, what are the (approximate) dimensions of the DAC PCB?
Then, I can start hunting for an enclosure.

[Peter Daniel]

It's 7.35 x 2.7". It fits inside 2 x 3 aluminum square tubing. I didn't call the place yet, but those boards should be ready for tomorrow.

[Colin]

Originally posted by Peter Daniel
I will be comming with balanced active line stage.

Hi Peter

This sounds interesting. Will this be a stand alone project or intended for use with the power amp boards?

One of the aspects of the Gainclone that appeals to me is that the size and price make it attractive for use around the home to power speakers in different rooms, all fed from a central, high-quality source. A balanced line stage is just what is needed for this.

Colin

[Peter Daniel]

I spent some time modifying ML380S preamp; it is a pretty good sounding preamp, and apparently one of the most successful products from Madrigal.

I know it's configuration pretty well, there is nothing really propriatory in that preamp (besides volume control), so I was hoping to come up with a design that would be partially based on the building blocks used in that preamp, yet having separate character.


So this would go like that:

Power supply based on MUR860 rectifiers, a pair of 1000u caps (with possibility of upgrading to BG STD), LM2990/2937 regulators which to me are as good as most discreet regulators I tried so far.

Input gain stage based on OPA627 and with switchable gain (2 choices available).

Phase splitting circuit, based on LM6172, allowing conversion from single ended to balanced and vice versa.

Next would be a place for volume control, and here we would have a choice to use either a switching attenuator, matched pots or volume attenuator based on IC (like WM8816 for instance).

The discreet output buffer (modified diamond buffer) would be the most interesting part of a circuit, and I have to experiment here a bit more to fully optimise it (basically make it looking less like ML circuit )

The rectifiers, regulators and main caps would be close to a buffer, additonal pair of regulators would be placed at the input circuit. Both sections could be used separately, with buffer being the main attraction.

Another approach to preamp, that I'm also considering, would be using a Jensen transformer at the input (providing buffering and phase splitting), an IC based volume control, and AD815 at the output, basically the configuration used in Rowland Coherence preamp.

[motherone]

Peter,

There's quite a number of threads regarding diamond buffers on the head-fi forums. You may want to check there for other examples. I think most diamond buffers look pretty similar, though (based on my limited knowledge).