The M2 Output Stage in Class A/B, and maybe a Power WHAMMY?

[EUVL]

Oh, almost forgotten ....
C2 is only required when the NFB loop is closed at the opamp output.
i.e. M2 OPS is outside the NFB loop.

As shown above, C2 is not required and can be short-circuited.


Sorry for that,
Patrick

 

[Mark Johnson]

Be sure to use the correct LM385 with the correct part number suffix; if you accidentally use the incorrect LM385, the M2 amp will misbehave when supplying large output currents. For example, when playing loud into 4 ohm loads. See "Official M2 Schematic" thread post #13 for the correct LM385.

 

[EUVL]

Thanks for pointing out, my mistake.
It should be LM385 1.2V.
The Alpha tester has the correct version in draft BoM.


Patrick

 

[EUVL]

Revised version, but still draft.

Value of Rc2 to be determined in Alpha test.


Patrick

.

 

[Zen Mod]

why not using OP to set output offset, if you already have it

 

[EUVL]

Because the PCB is designed to accommodate an alternative feedback loop excluding the M2 OPS.

Patrick

.

 

[EUVL]

Well, it is making music.

Opamp is OPA551, in local feedback (excluding the M2OPS, post #26).
Yet DC offset is totally stable.
R13,14 should be at least 0R56, maybe even 0R62.

Patrick

 

[Mark Johnson]

Closed loop gain is significantly greater than 5x so you can swap in the decompensated (higher bandwidth) OPA552. That's the opamp in the Marauder front end for VFET, people seem to like its sound.

 

[EUVL]

OPA551 is chosen because it is more stable with capacitive load when the M2OPS is included in the loop.


Patrick

 

[Vunce]

Yep, The M2OPS is making music .
The pcb was easy to work with and solder. Initial power is fed from my lab bench supply (+/-24vdc). Input is configured for SE. Bias current was a bit high (1.7A) with 0R47 source resistors at R13, R14. Added 0R10 in series with 0R47 dropped current to 1.4A, probably should use 0R62 as Patrick suggested if heatsinks are borderline adequate. DC offset trimmed to 0mV and only drifted +/-5 over the next 60 minutes. No thumps/pops or noise of any kind is heard from the speaker when powered on/off. M2OPS is driving a small 2-way 4R speaker (Speedster's) and sounds very good. (subjective alert )

 

[EUVL]

Sounds like the alpha tester is happy.


Patrick

 

[Extreme_Boky]

Very nice. Well done EUVL. I always greatly appreciate and value your posts here on forums.

Great input from everyone else (Mark Johson / Vunce).

I really like the option of a balanced-in, straightforward minimalist design/approach.

 

[EUVL]

Nelson has kindly approved publishing the Gerber files.
But I shall wait till Vunce has finished first, including power supplies.


Patrick

 

[brian.goodman]

Is it possible to add the provision for through hole parts in the final Gerber for those of us with waning hand-eye dexterity?

Thanks in advance,

Brian

 

[EUVL]

The layout does not allow easy conversion to through hole.
Butf course you can always design your own PCB.
Schematics is published.


Regards,
Patrick

 

[EUVL]

Oh, one more option is to use the DIYA Store M2X board, and just use the output stage there.
Should be simple enough to bypass the transformer, and make an adaptor board for the opamp.
But Mr. Johnson knows it much better than me, of course.

Patrick

 

[Mark Johnson]

All of the existing opamp-based M2x input cards are unity gain buffers. None of them can be reconfigured for higher gain by changing component values or increasing bias currents.

However, if members want to lay out a new M2x compatible input PCB, the physical dimensions are shown in (post #56 of the M2x thread). You will quickly discover it has only four I/O signals, corresponding to the four attachment bolts: input, output, pos_23v, neg_23v . But Patrick's version in this thread needs one more "I/O bolt" , namely: signal ground. See resistor Rs in the schematic of post #24 in this thread. So you'll need to devise a way to add another connection between mother board and daughter board. The Melbourne project (link) shows one possible approach.

 

[Vunce]

Nelson has kindly approved publishing the Gerber files.
But I shall wait till Vunce has finished first, including power supplies.


Patrick

Stereo M2OPS Plank Amp is alive!
Trafo is an Antek AS-4220 and psu is Jkuetemann's "Simplified" capMx that was already built and sitting on a shelf. simplified-mrevil-pmi-capacitance-multiplier.241263
Trimpots adjusted to drop 2V5 across capMx for +/-24V M2OPS input.
Channel#1 still running 0R57 at R13 & R14 for 1.4A bias current, heatsink 47C.
Channel#2 has 0R62 at R13 & R14 for 1.25A bias current, heatsink 45C.
ZM's Iron Pre is driving the M2OPS and I'm enjoying this combo

I have a pair of MeanWell SMPS units on the way to try out......

 

[EUVL]

Photos ?

 

[dBel84]

Looking forward to trying this out. Thanks @Vunce for bearing the load of being a labrat and thanks Patrick for always thinking of ways to keep us out of trouble.. dB