back to normal ?
Aug 14, 2001 at 11:29 PM Post #16 of 37
believe it or not guys

pssst..tell no one ..be vewy vewy quiet

ssssh

if you have a toslink output you already have the reqs for spdif

the same driver less the optical transmitter current draw !!!!!

GAWD !!!

tel me more rickster !

OK , easy kids

both are serial , both operate essentially the same , one draws ZIP !!! NADA !!!


the other just happens to be the standard (huh ? where did I hear that before ?)

disable the tos , connect a coax jack and

WALLLA !

SPDIF

Rickmonster
 
Aug 16, 2001 at 4:55 AM Post #17 of 37
Ok, I measured the draw of the whole digital board. It's my design (no wonder it didn't work when I first turned it on
wink.gif
), and has 8414, 3 7805 voltage regulators (TO-92) and a fully-utilized 74ALS04 plus 3 diagnostics LEDs. It drawed about 25mA but with one LED on (which draws about 3.4mA). ALS04 is chosen to do reset duty and to power the LEDs. Without LEDs maybe even a HCT can be chosen which would reduce power further, probably below 20mA. I can't measure the draw of 8414 by itself.

This is with device being idle mind you. Let me turn the music on
smily_headphones1.gif
...

Less than 21mA (and no LEDs are on). The funny thing is how the power supply is huge - it could probably handle over 1A easily.

Is <20mA good?
 
Aug 16, 2001 at 5:41 AM Post #18 of 37
20 ma is seriously good

some quick questions

1-DAC chip ? (I am guessing a sima delta chip due to the low current draw)

2-digital filter chip ?

3-have you tried a battery just after the bridge rectifier (AC unplugged of course)

of course adding an optical receiver would add about 40 ma more (if i remember right)

wonder how the portable CD players and MD players do it for way less

Rick
 
Aug 16, 2001 at 5:44 AM Post #19 of 37
hmm, that seems very good, but what do i know
smily_headphones1.gif
(just wondering, should we use a cmos hex inverter instead?)

I thought it would be much higher, closer to 100mA for the digital section alone. But now, it seems like the whole DAC will be less than 100mA, probably closer to 50mA?

The DAC i (ie rick (
evil_smiley.gif
)) was looking at is the AD1868 18bit r2r, its very expensive, i dunno if you'd rather use a cheaper delta-sigma... The 1868 draws only about 10mA...

add an analog stage that draws another 10mA, + the toslink receiver, and you're set. I'm guessing the AD823 and BUF634 would be a good pair. (the El2001s look great but they arent spec'd below +/-5V) I don't think toshiba makes a datasheet for the toslink reciever modules? but the transmitters draw about 15mA. I'm guessing the recievers would be even less. Or you could go with rick's suggestion and use electrical s/pdif for 0 current draw.

soooooooooooo it the total draws 50mA, which would give us about 30 hours of battery life on a set of 8 NimH AAs... or you could use 16 AAs and the elantecs, but it wouldn't be too portable....

Or since the current draw is so low, maybe we could use a single 9V, good for 4h...
 
Aug 16, 2001 at 7:48 AM Post #20 of 37
Bear in mind, this is not including the DAC chip. My analog board with DAC draws too much power (using shunt regulators) to be useful for measuring. My second DAC (SDS) might be a good one to test but I don't think there is an easy way for me to test short of breaking the links on PCB... SDS schematics are available, if anyone has a suggestion on how to measure DAC current, I'm listening.

Anyway I tried digital board alone with a 9V battery (thank God for the idea to have separate power connectors and supplies for everything; it was so easy to do this). I've put a 15.3 Ohm in series and got:

0.30V / 15.3 Ohm = 20mA.

Rick: I am perfectly happy with sigma-delta DACs. Since they call for simpler schematics and takes less space, the only reason not to use it in a portable would be if it uses more power than R-2R. Of course, Apheared-style portable has no size, weight, or power limitations whatsoever, but I personally prefer all-SMD design for this one.
 
Aug 16, 2001 at 8:52 AM Post #21 of 37
Actually, looking at my schematics, it doesn't need xx04 chip at all - 4 gates are used for diagnostics anyway and 2 are used for reset. I think a transistor can be used instead. That eliminates need for one voltage regulator as well. CS8414 really doesn't need much help. So we are looking for well below 20mA.
 
Aug 16, 2001 at 1:11 PM Post #22 of 37
as with my pocket amps , the DAC is a total re-do

I have the AD1861 ready to go
+/- 5 volts at 10 ma per polarity

passive IV resistor , value not settled yet but between 25 and 100 ohm

output section option A : passive transformer expensive as hell at $65 per,but no drain and performs both gain and filtration. But since we are talking about using this particular DAC with a headphone amp the output voltage need not be high.I may even get away with a cheaper 1:2 step up

output 2 : LCL filter , again passive , to a simple op-amp gain stage , again with a gain of about 2

since I already have some headphone amps I see no reason to add that stage to this version so I am looking at a two box solution

DAC/Amp

instead of output jacks ,soldering the output cables directly to the board

Digital section :

input receiver CS8412 (the hungry devil), to some "glue logic" -couple of D-flip flops and some nand gates (no digital filter,1X sampling rate) and if cmos the draw is really zip.So the whole digital section draws say another 40 something ma

VA+ 20ma
VD+ 20 ma

so with a SPDIF input we have about 40 ma or so per polarity,50 worst case

I can live with it,just need time to get it done

also going with shunt regs,little smd buggers

of course the AD1868 solves many problems (but no current out ,maybe I should only do that for home DAC V5) being a low current single supply device , would make life simple

BUT THE CHALLENGE !!

gotta lovit man , shoot high and even if you fall short you are still up there

dacman
 
Aug 16, 2001 at 11:49 PM Post #24 of 37
Hmm, i always thought using R2R would be easier, if you go with the 0 oversampling/filtering design concept, then you can take the signal right out of the reciever, put in the DAC (using v-out), put in a pot, and put in an output stage, and you're set... right? I don't think stuff like auto muting/demphisis is needed...

It seems just as simple to use as a delta-sigma, only it costs more and draws less current...I can't comment on sound, i haven't actually built anything yet.

The 1861 looks nice, i actually have a pair sitting on my desk right now, but it'll probably go into my home DAC instead...

Or i'm thinking of using AD1896 to upsample to 192khz, and put it through PCM1704 for >= DVD-A/SACD quality sound? but i guess i should finish a simple DAC first before messing around with those complex designs...

I dunno, probably the best method would be for me to shut up and let you guys come up with a sweet design, which i can then copy
very_evil_smiley.gif
 
Aug 17, 2001 at 12:32 AM Post #25 of 37
one thing to remember Thomas

never ever socket the input receiver or DAC chip

BAD,REAL BAD

but if you want to fool around with a digital filter you CAN socket that,as well as the glue logic

that way you can use DIP headers as replacements for experimentation

the main problem with ALL modern DACs is the totally ****ty built in op-amp stage , all cmos op-amps

have you ever tried one of these suckers as a headphone amp ?

not pretty man

so other than the IV output of a DAC , you can not get around the output stage already built in

also--certain folks beleive that the ONLY sane output for a sigma delta DAC is a simple coupling capacitor on the + output , I have not tried this myself

so you have the AD1861 ?

cool !

I get the time I will do this DAC , tweak the bitch then send you the verbatum plans

cool or what ?

rick
 
Aug 17, 2001 at 1:36 AM Post #26 of 37
I dunno Rick, both my DACs sounds pretty damn sweet even though one has all chips in DIP and all are socketed while the other has both receiver and DAC chips in SMD. I do see where you're coming from though. I'm not sure if receiver, being digital, really needs to be directly soldered, but DAC would probably benefit. However, all Crystal receivers worth considering today are SMD (older DIPs are not only obsolete but hard to find too).

I don't see a problem with sigma delta DAC chip output being connected to a proper line amp. One of my DACs has here a full-blown balanced discrete class A line amp which drives all my headphones beautifully. Or did you refer to the DAC chip *itself* driving headphones?

HOLY CRAP!! Digikey is out of PCM-63! Just who the hell is buying that stuff? They had couple of dozen just a month ago and now estimated time is november????? I was ordering them from Insight which is much cheaper but those guys are now not even replying the online orders, much less shipping them. What gives??
 
Aug 17, 2001 at 4:38 AM Post #27 of 37
SMD , way of the future-no more part swapping unless on daughter boards

and of course you are correct about the digital chips,but I would never socket a DAC or ADC chip,never,just me

and of coarse a discrete output stage would sound fine ,but what it really is -a preamp man !

you could follow that with an attenuator straight to the amp

My personal preference is to choose my own poison,as in current output to whatever I decide is best at the time

FET,OPA604,or resistor IV stage

then follow it with whatever gainstage I like at the time,and it changes

but it is a blast to get digital right,headphon amps are easy compared to this

parts : this **** is getting WAY out of hannd
mad.gif


I mean ,c'mon,every popular part is understocked

not good business practice
 
Aug 18, 2001 at 1:55 AM Post #28 of 37
Where can I find info about DIY DACs? Not necessarily portable, but for home systems, since when I go portable, it's the bare minimum that I use.
 
Aug 18, 2001 at 3:21 PM Post #29 of 37
rick, sounds cool...

i have AD861, burr brown 1702, 69 (delta sigma with I-out???) i'll try and use them all with the same digital section...


Hmm, i guess i won't work with sockets, though i'm worried about heat/static damaging the DACs...How about those machine pin sockets with integrated decoupling caps?

and are there any ways of working with SMT without making a PCB? I was thinking of gluing them "dead bug" style on a board and soldering wires onto the pins, but that takes too much time...
 
Aug 25, 2001 at 10:31 AM Post #30 of 37
SMD is actually quite nice to use if you have the proper pads. SOIC, 0805 and larger is quite easy to work with, anything smaller is a bloody nightmare. Dead bug style is also a nightmare.

I like to use surfboards (without the SIP pins) which are inexpensive and easy to obtain (digikey).
 

Users who are viewing this thread

Back
Top