M³ Project Announcement

[DCameronMauch]

AOS: Actually, I have been in electronics for a very long time. Any electronics project includes mechanical work. There is no way around it. And it has been made more than clear that the parts are easily gotten. You don't need a workshop to use a plumbing copper cutter. It couldn't be easier. You guys sound like a bunch of girlie men.

[aos]

I don't know why is it so hard to understand that having to buy extra tools (which many will never use again) and cutting metal on the carpet of your living room is not appealing to many people. For the record I have a drill press, sanding paper, files and I use them all on my living room floor and I can do my own board etching if necessary. But that attitude will eliminate a lot of builders from the project. If the goal is that only rich people and US citizens with access to home hardware can build it, then ok.

[amb]

ppl: Thanks for the links. The Ocean State Electronics shaft extenders and couplers look very promising. I also just ordered a sample of the HHSmith panel bearing to see if it might fit our needs (The Newark online catalog page does not show a picture, and it's not in their printed catalog).

aos: Yes, you have a good point, and that's why I asked. I do my own cutting and drilling in the house even though I don't have a workshop (my dining table is my workbench ), but I think I'm the minority here. Even though I don't think this amp should be a first-time DIYer's project, I do want to make this amp as easy to build as possible, so these considerations are very important. Tangent does a great job with his projects and I hope to be able to emulate him in some fashion for the M³.

BTW, Morsel's DSL connection is down (she is switching services), so she is offline for a day or two.

-Ti

[Earwax]

Originally Posted by DCameronMauch

AOS: Actually, I have been in electronics for a very long time. Any electronics project includes mechanical work. There is no way around it. And it has been made more than clear that the parts are easily gotten. You don't need a workshop to use a plumbing copper cutter. It couldn't be easier. You guys sound like a bunch of girlie men.

Hey, I'm no girlie man, I'm a real girl!

Soldering is ton's easier than casework. And I worked my way through college as a bicycle mechanic and I still feel that way.

It's not a big deal to me, but if the KISS principle is being applied to the electronic aspects, the expectation might reasonably that the case work would also be kept more or less basic. But us wimps can always run wires to a front panel mounted pot, can't we?

[amb]

Just for laughs, here is a picture of my humble dining table "workbench", currently with M³ breadboarded prototype in test



-Ti

[kevin gilmore]

where is the dead pizza??

[DCameronMauch]

AOS: Uh, okay. The pipe should be readily available even in Canada for like $2. You talk like it's a third world country. And the circular cutter tool is like $5. Rich Americans? What kind of salaries do you guys make up there? Someone could provide the parts for those who don't feel they are upto it. Like Tangent does for the PPA. Everyone who did that project had to make their own case and drill those. Where you thinking this wasn't going to be a DIY project, but some kind of everything supplied minimal work kinda kit?

Earwax: Sweet.

[aos]

All I'm saying is that in my experience you'll get a considerably higher successful completion rate if you provide these "insignificant" things. I HAVE seen people never finishing projects because of lack of a few trinklets. And I HAVE seen Home Hardware and Radio Shack not carrying a lot of things that their namesakes carry in US (we are a much smaller market). And I know about aversion of electronics people for doing mechanical work. Please don't mistake what I'm personally capable of and what I believe a lot of potential builders are capable of (which was my concern) - and this is a project for a wide audience. I've already built amps like this years ago so I'm not desperately seeking to get a piece of pipe pre-cut for myself, that's for sure (though it'd be nice).

amb: that's pretty innocent actually. I recall a similar thread way back on headwize (that's probably why Kevin was asking about dead pizza - your workspace is "too neat").

And now on-topic comment: Why the 10 Ohm resistors in series with opamp rails? Both here and on PPA version 1.1? I hope you're aware that the resistor's function was actually a "poor man's" substitute (or I should
say approximation) for a current source like FET - which you already use? It looks like unnecessary duplication, and
it mandates good bypassing of the opamp, otherwise you'd calling for trouble with high speed chips for obvious reasons.

[comabereni]

RE: Shaft extensions

I'm just catching up on this thread and haven't read all six pages yet, but I did note that there seems to be a need for shaft extensions.

The recommendation to buy commercially available shaft extenders and cut them to length is excellent, but for those averse to learning how to use a hacksaw , I could turn some extensions to exact length on my lathe from stock aluminum, Delrin, or 6/6 nylon. I'd only need dimensions.

Only hitch, this would have to be in a couple months or so--I'm in the middle of a major home remodeling project and my lathe is buried deep in my shop behind all my furniture.

I'm no pro machinist, but my other DIY hobby is making telescopes and telescope drive systems, and this is a pretty simple part. PM me if you are interested. I'd do them for cost, or even for just the cost of shipping--I have lots of cut-off ends and scrap lying around in my shop.

Just trying to help the DIY effort so you guys will be generous with your knowledge once I start asking lots of questions about circuits and various amp components .

[CingKrab]

Originally Posted by morsel

If there is enough interest in board mounted Elma Series N stepped attenuators, perhaps they could be ordered with a shaft of the proper length and sold along with the M³ pcb and panel bearings.

I don't know if this is relevant, but like morsel says, why don't you guys just do a group order from Elma with the desired shaft length? I found this quote in this datasheet: http://www.elma.com/files/products/r...ries_n_002.pdf

Originally Posted by http://www.elma.com/files/products/rotary_components/pdfs/audio_switches_series_n_002.pdf

Special shaft length
To order, state the shaft length AL as shown in diagram, measured
from mounting face.
Specify dimensions on page 150.

[amb]

where is the dead pizza??

Hehe, if you were here you'd see a couple of pizza boxes on the kitchen counter.

aos: Re: the power supply 10 ohm rail resistors: this area of the circuit is definitely not cast in stone and subject to change. Morsel took it from the PPA design but we will most likely make some changes. So far we have focused more on the amp itself. In fact all my testing so far have been done without the TLE2426 rail splitter nor the capacitance multiplier. I am using just a standard split power supply with dual IC regulators and some pretty big reservoir caps. I do like the rail splitter idea because it synthesizes a signal ground from the regulated and filtered supply rails, and should be cleaner than the more ordinary approach.

comabereni: Thanks for your offer, we should keep that in mind. I hope though, that you understand what you might be signing up to do .

-Ti

[amb]

Originally Posted by CingKrab

I don't know if this is relevant, but like morsel says, why don't you guys just do a group order from Elma with the desired shaft length?

Yeah, if there is enough people who want to go that route then this is definitely something we ought to look into.

-Ti

[pabbi1]

I apologize in advance for not being able to read a schematic.... but guys, I'm a CPA.

I'm so interested in this project, but can someone describe, in lay terms, how this will be a superior design to the SDS (v1.1)? Is Sheldon at all involved?

For the record, my SDS still does the job, but I did have to increase the R11/61 values for the HD600 - will this be designed for all impedences or will it be switched?

[DCameronMauch]

The whole shaft extension idea is ONLY important if it allows us shorten the leads from the pot to the first opamps significantly. Careful board layout is another way to do it. So all this may just be academic.

aos: I have been told that sometimes I can come off as being argumentative and confrontational. Just incase I did in this case, I appologize. Didn't mean to be.

[amb]

Originally Posted by pabbi1

I'm so interested in this project, but can someone describe, in lay terms, how this will be a superior design to the SDS (v1.1)? Is Sheldon at all involved?

For the record, my SDS still does the job, but I did have to increase the R11/61 values for the HD600 - will this be designed for all impedences or will it be switched?

Hi pabbi1,

Sheldon is not involved in this project. I think Morsel had contacted him many moons ago but he wasn't interested in a project like this.

As stated in the initial announcement, this amp is inspired by the SDS Labs MOSFET amp as well as the PPA. As such it bears some resemblance to both.

The basic M³ amplifier is similar to the SDS Labs amp (which is a text-book opamp + discrete output buffer design), but improves upon it in a few areas:

1. The biasing of the Vbe multiplier and the drive from the op amp to the output MOSFETs now employ a cascoded JFET current source. This has a couple of important benefits:

   
   - Improved power supply rejection ratio (PSRR) because there is no longer any resistors going to the power rails.
   - The cascoded JFET current source pulls current out of the op amp to bias the op amp into class A. The amount of current is also adjustable to suit differing op amps and tastes. In the stock SDS Labs MOSFET amp, a small amount of current is fed into the opamp's output instead, but in reality it's not sufficient to really bias the op amp into class A, and also that causes the PNP side of the op amp's output stage to be on, rather than the NPN like the M³ design. The NPN device is typically more linear.
   - Having a constant current source provides a more robust drive to the MOSFETs. This is good because the gates of the MOSFETs are capacitive.
   

2. The three-channel architecture is inspired by the PPA and absent in the SDS Labs amp. I won't go into the benefits here because that's been written up already in the announcement as well as on the PPA web site.
3. Sheldon's PCB layout has the power transformer located too close to the op amp and as such induces AC mains hum (and its harmonics) into the audio signal. This is audible with sensitive headphones and is clearly visible on a spectrum analyzer. The M³ PCB will not have this problem, and will be a more deluxe two-layer plate-through board with silkscreening and soldermask.
4. Our parts selection differs from that of Sheldon's, in particular, the op amps (using singles instead of dual units for better channel separation), and the MOSFETs.
5. The amp will be suitable for use with both low impedance as well as high impedance phones. The gain of the M³ is set at 11 but builders can change this.
6. The power supply will be quite different as well, and other smaller details too.

I know the above is not really a "lay person's" description, but what we are doing will hopefully result in an amp that would sound superb. Your ears will be the judge when we're done!

-Ti