A New Take on the Classic Pass Labs D1 with an ESS Dac

[qusp]

OK fair enough, well thats a pretty similar outlook to my own and is why i've been making so much noise.

yeah I wondered about that, I was just reading the 600R directly from the build manual, 1.2k makes more sense, why frame it as 600ohms? doesnt that just confuse matters since noone is ever going to be using the number to calculate a parallel cap value? its OK, I just didnt read carefully enough.

however going by my calcs a 22uf cap gives me a corner of 6.03Hz, with phase distortion up to 60Hz

F = 1/(2 * pi * C * R)
PI = 3.14
CAPS FORMULA 0
INPUT IMP FORMULA 1200
WORKING 0.03
WORKING 0.17
WORKING 6.03

soz looks as though you've mostly covered that above.

I may end up throwing an....electro (gasp, a bipolar) in there temporarily. they make a 20uf, but it adds 20ea to the price for 10uf.


you spoke of increasing the input Z for the headamps, what about doubling the input for the poweramp? making it 2.4K? then i'd be happy enough with 10uf. this is why I wasnt really thinking about my own caps for a while and why I was leaving this to Pinnochio, but seemed like a good opportunity to get the pricing down, i'm fine for the LPUHP

 

[pinnocchio]

Owen, Qusp,

I'll be using this on two different power amps. One with 5K and one with 10K input impedance. What would be the optimal value to satisfy both? Should I stick with the BoM value?

Thanks
Do

 

[opc]

Hi Do,

With an input impedance of 5k worst case, you would want roughly a 16uF cap to fulfill the 10x rule. That basically means you're targeting an Fc of 2Hz which means you won't get any phase shift until about 20Hz.

I would say that with a loudspeaker, that rule is overkill since the amplifier's contribution is going to be swamped by variations in the loudspeaker, but if you want it to be technically perfect, then stick with that.

I strongly doubt you could hear a difference between a 10uF and a 16uF cap in your application though, so you'd be safe with either.

As a side note, there is no "ideal" value. The capacitor just has to be "big enough" to work correctly with your next stage's input impedance. If you went ahead and put a 100uF in there, you'd be safe for almost any reasonable impedance in the next stage.

Cheers,
Owen

 

[Emphrygian]

Optimal depends on how particular you are about LF response. That said, 5k input impedance is "worse" than 10k, so that should probably be the target of your calculations.

So, using the formula for the cutoff (-3db) of a regular RC-filter:

Fc = 1/(2*pi*R*C)

R is 5000 in your case and C we haven't decided yet. Let's try out two values, 4.7uF and 10uF.

For 10uF Fc = 3.18Hz and for 4.7uF Fc = 6.77Hz. Bear in mind that in both cases, the roll off will start well before the cut-off, but only you can determine how low you *need* the amp to linear. For example, in the example Owen posted earlier, the -3db is somewhere between 6 and 7 Hz, but at 20Hz you do have a small attenuation too. If you can live with that, you should be fine with 4.7uF, if not, go for a higher value.

There are also phase properties to consider, but I leave that topic out for now.

Note: I don't mean you should make a decision based on the data I present here. I'm merely showing how to go about finding out what you need.

 

[qusp]

yeah i'm thinking we just stick with 10uF, i'll throw in a bipolar electro for a while, till I figure out what to do, maybe the auricap will end up being the bypass haha.

 

[qusp]

Optimal depends on how particular you are about LF response. That said, 5k input impedance is "worse" than 10k, so that should probably be the target of your calculations.

So, using the formula for the cutoff (-3db) of a regular RC-filter:

Fc = 1/(2*pi*R*C)

R is 5000 in your case and C we haven't decided yet. Let's try out two values, 4.7uF and 10uF.

For 10uF Fc = 3.18Hz and for 4.7uF Fc = 6.77Hz. Bear in mind that in both cases, the roll off will start well before the cut-off, but only you can determine how low you *need* the amp to linear. For example, in the example Owen posted earlier, the -3db is somewhere between 6 and 7 Hz, but at 20Hz you do have a small attenuation too. If you can live with that, you should be fine with 4.7uF, if not, go for a higher value.

There are also phase properties to consider, but I leave that topic out for now.

Note: I don't mean you should make a decision based on the data I present here. I'm merely showing how to go about finding out what you need.

no, I dont think we can completely disregard phase. I think maybe 10uf is the better choice for all concerned though, I already asked about their stock, it makes not much difference to the price, 20uf does.

i'm using a different dac and amp below 60Hz anyway

 

[Emphrygian]

no, I dont think we can completely disregard phase

I agree, I just wanted to show some kind of starting point.

 

[qusp]

I agree, I just wanted to show some kind of starting point.

fair call. but i'm concerned that there is still only the usual suspects contributing here....

its only of pretty temporary concern to me, my speakers will be 3db down at 60Hz while they are just 2 way, till I get my subs going. the Subs will use a different dac and amp channel, not this IV and then when theyre ready i'll probably push that up to 120Hz.

plus i'm using digital crossover with convolution, so to be all precious about phase response is pretty funny really.

 

[MisterRogers]

For my uses, 10uf is adequate (headamp).

 

[qusp]

ok i've started putting together the emails (wish we could send more PMs at a time, its a pain having to send it multiple times multiple times... i'll send the keratherm and chassis stuff separately.

i'm buggered and I havent heard back about the 10uf stock yet plus it would be good if more can poke there head in here and see the conversation. its nearly 5AM and I havent gotten the work I needed to get done last night done yet, so ive gotta go, i'll finish it off when I get up. but at least there was some conversation here.

 

[MisterRogers]

Thanks for your efforts qusp; it's usually a thankless task - but it's very much appreciated.

 

[Emphrygian]

you spoke of increasing the input Z for the headamps, what about doubling the input for the poweramp? making it 2.4K? then i'd be happy enough with 10uf. this is why I wasnt really thinking about my own caps for a while and why I was leaving this to Pinnochio, but seemed like a good opportunity to get the pricing down, i'm fine for the LPUHP

I'm also interested to know the implications of doing such a modification. It will most certainly increase the noise if I'm not mistaken, but by how much? Or, more importantly, is the tradeoff worth it for getting better LF behaviour (for a speaker setup that is. Head amp is already covered thanks to higher input z)?

(Not that I expect the second question to be an easy one to answer...)

Or, we just buy twice as many 10uF caps

 

[Emphrygian]

ZFoil

Qusp
Pinnocchio
MisterRogers
Rolle
BigPandahk
Reenberg
lwif
analog_sa
atom6422
crazikid
flocchini
vitalica
hirez69
lm1no2
Coolhead
ccliu
macfixer
tehrkn
Blitz
Slartibartfasst
necplusultra
Regland
woresseo
Malvin
Greg Stewart
Emphrygian

Capacitors - Auricap XO

Pinnocchio
Bigpandahk
reenberg
lwif
atom6422
crazikid
MisterRogers
hirez69
lm1no2
Coolhead
ccliu
macfixer
tehrkn
necplusultra
Regland
woresseo
vitalica
Greg Stewart
Emphrygian

Capacitors - Vcap CuTF for Bypass


Keratherm

Pinnocchio
bigpandahk
Reenberg
Rolle
hirez69
lwif
atom6422
crazikid
flocchini
MisterRogers
lm1no2
Coolhead
ccliu
macfixer
tehrkn
kasey197
Blitz
Slartibartfasst
necplusultra
woresseo
Malvin
Greg Stewart
Emphrygian

Enclosure

Pinnocchio
Rolle
crazikid
MisterRogers
ccliu
qusp
woresseo
vitalica
Emphrygian

 

[Emphrygian]

So, qusp, what are your thoughts on the enclosure? I read you suggested we get the sink tapped for 1xNTD1 + Salas or 2xNTD1. When you say Salas, are you thinking 4x Salas for the NTD1?

What about the back? I would like 2x XLR outputs, one USB input (pretty much the same as the XLR mounting holes as I will use Neutrik connectors), one toslink in and one IEC outlet. Somewhere I'd like a switch to turn the whole thing on; either a rocker at the back or an illuminated one on the front. For the former, a LED to show whether the whole thing is on would be nice on the front. A front display would be cool too, but I don't know when I'll have the time to get such a thing going so it's not necessary at all (for me).

 

[necplusultra]

board received

thanks

 

[pinnocchio]

So, qusp, what are your thoughts on the enclosure? I read you suggested we get the sink tapped for 1xNTD1 + Salas or 2xNTD1. When you say Salas, are you thinking 4x Salas for the NTD1?

What about the back? I would like 2x XLR outputs, one USB input (pretty much the same as the XLR mounting holes as I will use Neutrik connectors), one toslink in and one IEC outlet. Somewhere I'd like a switch to turn the whole thing on; either a rocker at the back or an illuminated one on the front. For the former, a LED to show whether the whole thing is on would be nice on the front. A front display would be cool too, but I don't know when I'll have the time to get such a thing going so it's not necessary at all (for me).

Same for me but I would use a momentary illuminated Bulgin switch with AMB switch driver to turn the whole thing on. I've been using this combination for the longest time and it is really good.

Here's the switch I was thinking about and the switch driver
The ε24 Power Switch Driver Circuit
MPI002/28/D4 Bulgin | 708-1449-ND | DigiKey


As for the LCD hole, what would be good is that the chassis be ready, drilled and tapped but the LCD window not fully removed. Maybe leave 1/32" - 1/16" metal thickness so it would be easy to knock the piece out and sand/grind/whatever it.

Let me know if you're interested

Do

 

[MisterRogers]

Those who ordered in the US - boards and FETS arrived safely yesterday. Over the next couple of days I'll be getting these packed for redistribution. I'll be in contact re: shipping fees. My hope is to have them all ready to go out Monday morning.

 

[pinnocchio]

Received boards and fets yesterday. Thanks Owen!!!

Really nice boards. Can't wait to get started on this

Do

 

[Jean-Charles]

Received my board Owen, thank you very much.

 

[opc]

Glad to hear the boards are slowly starting to arrive!

All of this reminds me:

Algar_emi - I don't have your address... I've sent you an email that got bounced back, and a several PM's on here, none of which have been answered.

I did get your payment, but you did not include your address in the notes, so your package is still sitting on my desk without an address.

Please send me a note and let me know where to send your stuff.

Regards,
Owen