[jcx]

i don't understamd mosfets for this app, particularly ones rated at 100X current

but you should really reconsider the assumption that mosfets have a self-stabilizing positive (Vth, Rds) tempco before getting rid of the source resistors - just look at the IRFZ24 data sheet fig 1&2, the 4.5 Vgs curve shows a positive Ids tempco in the linear region, this is because Vth has negative tempco, only at high currents or near saturation does the positive Rds tempco dominate in vertical mosfets

 

[DCameronMauch]

digi01: I believe zobel networks help the amplifier deal with complex load impedances and back emf. Since headphones don't usually have crossovers, the load impedance is not that complex. And back emf is relatively small.

jcx: Current rating is not that important, so long as it is more than enough. Voltage rating and power dissipation are of primary concern. Hum. Interesting. It looks there are some conditions where the tempco for mosfets is positive after all. News to me. Still, with a push-pull single n/p pair topology, there is no real danger of runaway. Bias current may drift a little, but that is not a big worry here either. That said, I don't think source resistance is a bad thing. Might be wise to take the better safe than sorry approach.

 

[Nisbeth]

Quote:

Originally Posted by DCameronMauch digi01: I believe zobel networks help the amplifier deal with complex load impedances and back emf. Since headphones don't usually have crossovers, the load impedance is not that complex. And back emf is relatively small.

Actually, I believe the purpose of the Zobel is to keep the amp from oscillating if the load is capacitive. Either way, it shouldn't be necessary here.


/U.

 

[morsel]

Let's talk cases and board size. Are people happy with the 12×8×3" PAR-Metals 20 Series case? Can anyone envision wanting a larger case than that? The board could be up to 6×8" and still only take 1/2 the case, leaving the other 1/2 for the power supply, separated by an internal metal plate. I think the PAR-Metals might be a wee bit too big, but that is the smallest size listed. Does anyone have a superior case suggestion? For those on a budget, a smaller economy case could be used, along with a separately boxed STEPS or ~500mA wall wart. Probably the nicest case work I have seen is that done by Larry of LaRocco Audio. Perhaps he would be interested in making cases for us. Finally, there is Front Panel Express for designing custom panels, but they do not make cases as far as I know.

 

[Super-Gonzo]

Quote:

Originally Posted by morsel I think the PAR-Metals might be a wee bit too big, but that is the smallest size listed.

Par-Metals does custom sized enclosures, if several were purchased at once, perhaps we could get a smaller spec'd 20 series case for a reasonable price?

Also, I think it would be a good idea leave enough room inside the enclosure for a stereo ladder attenuator, and maybe a rotary input switch.

 

[Voodoochile]

I like the Par chassis well enough. It is fairly important to request anodizing though, as their paintwork is not good at all. Also, request that they countersink the screws, or you get round-head by default.

Backing up a little bit, I had mentioned wanting pads for the Dact attenuator within Alps Blue pads. If it is practical, I'd like to see the pads at the rear of the board, or perhaps front and rear (as the input pads on the PPA are).

Since this will not be built in the Hammond cases, having the pot right near the input jacks is ideal (IMO), and the shaft extensions and bearings are easy to work with.

Linky to Handmade. Welborne also has them.

 

[DCameronMauch]

voodoochili: Both excellent recommendations. How much of a premium did you have to pay with Par-Metals for having the entire enclosure anozided and countersunk screws? For the pot, shouldn't it be positioned such that the leads from the pot to the AD8610s are short as possible? To avoid RC lowpass effects. A few inches difference from the audio source to the pot shouldn't make much difference.

 

[morsel]

Gonzo: I was also thinking that custom smaller PAR-Metal cases might be good, as long as we don't have to get into complex group buy issues. We will definitely have room for a stepped attenuator.

Elma Type A

Elma Type R

Elma Type N

I prefer the surface mount Type N series attenuator. It is easy to board mount and has a short signal path with good stereo separation.

VC: AMB and I talked a bit about using a shaft extension, but we were a bit unsure how others would feel about this. How much of a pain is it to install a shaft extension and panel bearing?

I know about the need to order the case anodized, not painted, but what's this about countersinking the screws, can you give us more detail? I assumed no one keeps the stock screws.

Are you OK with the 12×8×3" case size?

 

[Super-Gonzo]

Quote:

Originally Posted by Voodoochile Since this will not be built in the Hammond cases, having the pot right near the input jacks is ideal (IMO), and the shaft extensions and bearings are easy to work with.

I'm with Voodoo on this one, I like the idea of putting the pot/att. at the back and using a shaft extension. (IF it results in a shorter lead to the opamp.)

 

[Super-Gonzo]

Quote:

Originally Posted by morsel Elma Type N I prefer the surface mount Type N series attenuator. It is easy to board mount and has a short signal path with good stereo separation.

ooh, the Type N is pretty. And believe it or not, I prefer soldering SMD resistors to the through-hole variety. Is there a ladder version of this attenuator? I didn't see one on the elma site.

On a related note, what are the thoughts about using SMD components for other amp components? For items where it could shorten signal path and improve performance, perhaps it should be considered?

 

[Nisbeth]

Quote:

Originally Posted by morsel VC: AMB and I talked a bit about using a shaft extension, but we were a bit unsure how others would feel about this. How much of a pain is it to install a shaft extension and panel bearing?

It's not much of a problem really. The extender can be any piece of 1/4" or 6 mm tube or rod (with a tube it's possible to run LED wires inside for an illuminated knob). The hole for the panel bearing is just a bit larger than the thread on an Alps blue (9 mm. vs 8 mm. IIRC) and the flange around the bearings I have used have been quite wide so drilling the hole slightly larger (say 10 or 11 mm.) will give a little slack to aid in allignment if your measuring skills are less than perfect.


/U.

 

[Voodoochile]

Quote:

Originally Posted by morsel VC: AMB and I talked a bit about using a shaft extension, but we were a bit unsure how others would feel about this. How much of a pain is it to install a shaft extension and panel bearing? I know about the need to order the case anodized, not painted, but what's this about countersinking the screws, can you give us more detail? I assumed no one keeps the stock screws. Are you OK with the 12×8×3" case size?

The extension is a fairly simple to use, requires only a *little* more insight than a panel-mounted pot. Mainly calculating the height for the hole from the panel base, typically lower than the chassis base due to overhang. A table could be made giving the offset based on 1/8" standoffs under the PCB, 1/4", 3/8", etc. To this you would add the overhang dimension of the panel.

As for the countersinking, you have to ask Jeannette or John at Par very specifically. They will furnish screws then. And instead of painted zinc screws, they then furnish proper black screws, and stainless for the faceplate. You may opt to not have the faceplate countersunk if you wish to use Allen-style machine heads, or button head screws. Having the entire chassis anodized in black, and the faceplate in silver, with countersunk holes throughout runs about $55 shipped for a 12"x8"x1.75"/2" (1.75 box/2"faceplate). That is also including the fist thickness upgrade on the faceplate, which is well worth it, I think. I believe it is 1/8" thick.

As for the case size, that's tough. The case I describe above is perfect for a PPA, is almost exclusively what I have used. But it is tight vertically. The next (stock) height is the 3" case you describe. This seems tall to me, at least in the 12"x8" size. I think I would do one of the following:

Either keep the 3" height, and get the 16"x8" or 16"x12" case to keep proportions nice, or...

stay with 12"x8", and ask for a custom 2.5 or 2.625" height as a custom job.

I don't believe they gouge you on price, but you will wait for it. As it is they take about 10 days to 2 weeks for a chassis or two in stock sizes, since they do make them to order, even when 'stock' sizes. I think by stock it just means that they have jigged settings or routing packages saved for these sizes. They don;t actually have them truly 'in stock'.

They (Par) are very nice people, and the work is pretty good, but it is imperative that you are very explicit with your preferences.

 

[Nisbeth]

Quote:

Originally Posted by Super-Gonzo On a related note, what are the thoughts about using SMD components for other amp components? For items where it could shorten signal path and improve performance, perhaps it should be considered?

I would prefer through-hole but SMD is OK with me as long as they're 0805 or larger

In my opinion SMD's should only be used in positions where they yield significant benefits.


/U.

 

[Voodoochile]

Gonzo- the ladder is available, but it is *big*, and the price is proportionate to the size. I made one on a 5-deck Elma attenuator switch from Mike Percy, and it was modestly obscene, with a pile of Holcos to solder. The attenuator is the size of a bottle of Tabasco sauce.

From Goldpoint it's about $250, from Dact, closer to $325. Brutal. But I do love mine. The kit from Percy was about $180 IIRC. At the time, a series Dact was $181 from DIY cable.

So, block out a lot of soldering time if you really want one!

 

[Super-Gonzo]

Quote:

Originally Posted by Voodoochile The attenuator is the size of a bottle of Tabasco sauce.

Thats half the appeal!