[amb]

Quote:

Originally Posted by skyskraper ok it was late and id had a long day. run it before the amp section but run it from the output of the pot.

The input impedance of the VU meter should be very high or you'd be introducing an undesirable load at that point in the circuit. If the meter is passive (i.e., containing nothing more than rectifier diodes and the meter coil itself) then it would definitely not be suitable. A separately amp'ed meter drive circuit would be ok, but be aware that the node just after the volume pot in is the most sensitive point in the amp (to noise interference and stereo crosstalk), so be very carefully about adding extra wiring or components there.

 

[Jazper]

I couldn't help but tinker.. it's in my nature..

Today I pulled the C2 caps from my M3 (which were 1uF wima MKS)

A few things to report:

1. the offset has dropped (from 10mV on both channels to about 7.3mV) on the otherwise high offset ths4631D chips, which still run hot
2. the noise floor has dropped (the hiss as heard from the headphones) -there is more "blackground"
3. The soundstage has greatly opened up and the imaging is better
4. The amp sounds more neutral, and has more bass impact.
5. The amp seems faster than it was.
6. The amp sounds clearer, as if it had a cold before and doesn't have it anymore.

Now before everyone goes and pulls their C2 caps, the main reason I thought to do so is because Morsel did say a while ago that she couldn't hear the difference, and I've heard lots of "bad things" about having caps in the signal path.

I don't think I'm in the "golden ears" club, but I know what sounds good, and this sounds good.

So I pulled them, and I'm happy with the result.

Setup in my sig.

 

[morsel]

Wima MKS caps are polyester film. If you are going to populate C2, try using MKP polypropylene film. .1µF should be adequate, so you can use the same caps as for C6. While you are at it, you can use them for C8, if you use C8 at all.

amb, any reason why polyprolylenes are not listed for C2 and C8 on the parts list page? It seems worthy of a note. Also, C6 can be polyester instead of polypropylene.

 

[Jazper]

Quote:

Originally Posted by morsel Wima MKS caps are polyester film. If you are going to populate C2, try using MKP polypropylene film. .1µF should be adequate, so you can use the same caps as for C6. While you are at it, you can use them for C8, if you use C8 at all. amb, any reason why polyprolylenes are not listed for C2 and C8 on the parts list page? It seems worthy of a note. Also, C6 can be polyester instead of polyprolylene.

Well on the original parts list by nate, and on the page, polyester film were recommended. In my amp I just dropped in 1uF MKS wimas for C8 and C2. I don't think there'd be THAT much difference with C8 being polyprop over polyester, however after what I'm seeing here now if you're going to populate C2, then I'd say you should use something high quality(polyprop or better). Right now I don't see the point in putting a cap there..

Isn't general theory that caps in the signal path = bad?

 

[morsel]

C2 spans the Vbe multiplier, theoretically assisting it in maintaining a constant voltage between the MOSFET gates. To the extent that the Vbe multiplier is doing the job, C2 is superfluous, since they are in parallel. Is C2 in the signal path? Sort of, but not in the same way that an input coupling cap would be in the signal path. C2 has a very minor influence on the amp, since the Vbe multiplier does a pretty good job.

 

[amb]

Quote:

Originally Posted by morsel amb, any reason why polyprolylenes are not listed for C2 and C8 on the parts list page? It seems worthy of a note. Also, C6 can be polyester instead of polypropylene.

C2 and C8 are specified to be MKT or MKS2 because the Wima MKP2 series metallized polypropylene box caps only go up to 0.33µF. For these purposes a metallized polyester is just fine.

Jazper, if there was a DC offset and noise floor change by pulling C2, then something is not right with your amp. That cap should have absolutely no influence on DC offset or noise. While RMAA test shows no difference with or without it, and I did not observe a discernable difference in square wave response at the output, nevertheless my tests are limited and are not likely to show all the intricacies. It is common practice to bridge the bias network with a capacitor to improve its performance at high frequencies, and I do not recommend removing C2.

 

[morsel]

amb, .1µF should be fine for C2 and C8. I think it's worth mentioning the polypropylene option on your web site.

 

[Jazper]

well all the voltages, resistances and capacitances measure out ok, don't know what could be wrong here...

 

[silvervarg]

Quote:

Jazper: well all the voltages, resistances and capacitances measure out ok, don't know what could be wrong here... ...ths4631D chips, which still run hot...

This sound like the OP-amps are oscillating to me, but its just a guess.
I would turn bass boast to zero or bridge Sbb and then check C1 and R4.
What type of cap did you use for C1 and what value?
Have you tried with other OP-amp brands? Do they also get hot?

 

[Jazper]

Quote:

Originally Posted by silvervarg This sound like the OP-amps are oscillating to me, but its just a guess. I would turn bass boast to zero or bridge Sbb and then check C1 and R4. What type of cap did you use for C1 and what value? Have you tried with other OP-amp brands? Do they also get hot?

urm, even amb noticed that when he tried the ths4631Ds that they run hot, they draw about 11.5mA quiescent - there is also a reason that there is a version with a "powerpad" - heat seems likely.

I don't think they're oscillating

 

[fyleow]

This might seem like a trivial question but what kind of standoffs are suitable for the board? I checked AMB's site for parts list but it doesn't list any specific item numbers.

If someone could list some item numbers that would work I'd really appreciate it. Thanks

 

[amb]

Quote:

Originally Posted by fyleow This might seem like a trivial question but what kind of standoffs are suitable for the board? I checked AMB's site for parts list but it doesn't list any specific item numbers. If someone could list some item numbers that would work I'd really appreciate it. Thanks

A lot of different standoffs are appropriate, it depends on what kind of enclosure you're going to mount the board in and how high you want to elevate the board off the bottom. I see that you're in Cupertino, the best place for something like this is HSC Electronics (aka Halted) in Sunnyvale. They have a large selection of these and various associated hardware at cheap prices. Just go there and get what you need.

 

[fyleow]

Quote:

Originally Posted by amb A lot of different standoffs are appropriate, it depends on what kind of enclosure you're going to mount the board in and how high you want to elevate the board off the bottom. I see that you're in Cupertino, the best place for something like this is HSC Electronics (aka Halted) in Sunnyvale. They have a large selection of these and various associated hardware at cheap prices. Just go there and get what you need.

I've actually moved north to Hayward now unfortunately

I miss the central location of Cupertino (3 Fry's within ~15 minutes) but housing here is cheaper.

I have a bunch of standoffs used for motherboards so that might work, but it's good to know I can get them locally too. Thanks for the info.

 

[GregVDS]

Hi,

I had the (practically) last pack from Mouser, and began to build.

Just to be sure I can do it, I began the STEPS first rather than the M³. I cut legs before solder. The components fit thightly enough for this, and this permit to heat very rapidly the component lug and the pcb hole. Very nice finish.

For the ones that have built it, in the STEPS, the Tantalum C7 is polarized, right? Ther is a + hole on the pcb, but no longer leg on C7. On the yellow drop, there is a kind of upside down L sign. This one points down to the + leg?

How much of silicone grease should I put, practically nothing, the less the best, or there is need for at least a drop of it?

The wire I usually use tend to brake just after the solder joint after some movement. Why is that so? Is this all the time the case, or my wire is a bad quality one? I read it depends also how you remove the pvc sheet (I pay attention, but I touch the copper with my cutter, is it enough to damage the wire?).

Thanks,

All the best,

GregVDS

 

[amb]

Quote:

Originally Posted by GregVDS (STEPS) the Tantalum C7 is polarized, right? Ther is a + hole on the pcb, but no longer leg on C7. On the yellow drop, there is a kind of upside down L sign. This one points down to the + leg?

Yes tantalum caps are polarized. Usually there should be a '+' mark next to the positive lead, but depending on the brand of the capacitor the actual marking might vary (e.g., possibly the negative lead is marked instead with a '-').

Quote:

How much of silicone grease should I put, practically nothing, the less the best, or there is need for at least a drop of it?

Just a little is enough. Too much and you will have it oozing out the sides when you screw the part on the heatsink and it's messy.

Quote:

The wire I usually use tend to brake just after the solder joint after some movement. Why is that so? Is this all the time the case, or my wire is a bad quality one? I read it depends also how you remove the pvc sheet (I pay attention, but I touch the copper with my cutter, is it enough to damage the wire?).

If you are not using a proper wire stripper tool and you cut too hard, the copper conductor of the wire becomes nicked and becomes easy to break. Also, solid core wires become stressed when you bend them around and will eventually break, so avoid doing that.