help needed dc coupling ,is there a better way?

[stephensank]

Don't forget that any really serious subsonic content will be blocked by the caps needed on the dac output, if there was ever actually a genuine concern about that. But I have done many, many dacs/players where I've eliminated all caps from dac chips to output jacks, and never a problem.
My feeling is that we simply have not yet figured out everything that needs to be measured, nor how to measure, to explain the large gap between what we measure & what we hear.
I am a big fan of Relcaps, and will always use them when I have the physical space & the budget to use them. In lytics, I will ONLY use Nichicon and Nippon Chemicon since 35 years of building/service experience tells me nothing else comes close to their reliability, and upper level grades of each perform at least as well as any lytic out there, usually better. But I never depend on lytics alone for any serious audio. Always do film bypasses, even if all I can squeeze in is a polyester.
If you want to really hear the difference between rectifiers, do a/b testing with a good quality phono stage, which will be most sensitive to supply noise. Have not yet detected a difference between IRC Hexfreds and IXYS DSEI-series soft recoveries, but I definitely am quite sure they make for smoother sound with blacker background than schottky's, and they are a hell of a lot more bulletproof.
I did not mean to say that a physically larger cap will have better bass than a smaller cap of same value, but rather that I always use a larger *value* cap than is technically needed, which I do think yields better bass.

[GLENZWORLD]

thanks for your help and honesty

now I just need to get a good and reasonable dc multimeter

so now I have to try and get some film on those dac out

I also just changed the large can caps in my acurus a150 amp and noticed what looks to be an input and output caps on each channel but I can`t seem to get a schematic for that and it has 2 tiny lytics on I think input and the rest are wima film

you like the nichicon and I`v heard they can sound bright
also heard that silmics can sound colored and that wima`s can sound screechy

when will they make a 100uf polypropylene that fits 5mm spacing lol

we would still want more

still searching for the holy grail
Glen

[stephensank]

Can't generalise about the sound of an entire brand of caps. But, generally, all but the lowest grades(i.e., not made for audio) definitely don't "sound bright". Be careful reading opinions and a/b test results. I don't know that anyone takes into account that electrolytic caps, in particular, take a solid 400 hours to come close to fully breaking in. When the lytic is doing signal path, this means 400 hours of rigorous passing of signal, so a lytic doing, for instance, power amp input coupling, where it gets way less than line level signal unless the amp is blasting, it can take a hell of a lot longer. Before break-in, even the best lytics typically sound coarse on top and mids, and people tend to rush to judgement, picking the lytic that sounds better right away, usually ending up with the cap that sounds far less than best after break-in.
Ideally, on your dac chip output lines, space allowing, I would use a 33uf/25v KZ Muse with a 0.22uf Rel PPFA + 0.01uf RT. Will sound beautiful immediately, and would get damn near as transparent as a (quality) jumper wire after break in. Really doubt you have that kind of space available, though. And I can't say I would consider that player worth investing in such nice caps.
The Acurus & Aragon amps, which are nicely built, but strangely sensitive to speaker load(can downright squawk with certain speakers), have an input/driver stage design that can be easily thrown off by small dc offsets at the input, so I would be a little hesitant to dc-couple the input. So, I would keep the input caps(the existing little Nichicon is bypassed by the 0.1uf mediocre film cap next to it). But at least you have ample room to put in a cap combo such as exactly what I suggest above.
BTW, my experience with very well-broken-in Wima metallised polypropylenes is that they sound hazy & veiled, not screechy.

[Torchwood421]

Hi, Have you ever considered using transformer output? It can be done. I am using Lundahl LL1690 and Edcor XSM 600/600. The trick is to get rid of any DC offset into the primary of the transformer. So with a TDA1541A based dac I used a Pedja AD844 based I/V and that has a current source to adjust the offset of the TDA to zero. The other dac is a CS4398 and by it's design has no real offset. Makes for a very simple output stage a single parallel cap across the primary to control high frequency roll off. Maybe not true DC coupling however transformers can sound quite good and I find that they beat most capacitor solutions or extra opamps. For power decoupling I am fond of Nichicon electrolytics. I use lots of Wima caps where necessary. I discovered PPS can be quite nice too.

[GLENZWORLD]

Originally Posted by stephensank

Can't generalise about the sound of an entire brand of caps. But, generally, all but the lowest grades(i.e., not made for audio) definitely don't "sound bright". Be careful reading opinions and a/b test results. I don't know that anyone takes into account that electrolytic caps, in particular, take a solid 400 hours to come close to fully breaking in. When the lytic is doing signal path, this means 400 hours of rigorous passing of signal, so a lytic doing, for instance, power amp input coupling, where it gets way less than line level signal unless the amp is blasting, it can take a hell of a lot longer. Before break-in, even the best lytics typically sound coarse on top and mids, and people tend to rush to judgement, picking the lytic that sounds better right away, usually ending up with the cap that sounds far less than best after break-in.
Ideally, on your dac chip output lines, space allowing, I would use a 33uf/25v KZ Muse with a 0.22uf Rel PPFA + 0.01uf RT. Will sound beautiful immediately, and would get damn near as transparent as a (quality) jumper wire after break in. Really doubt you have that kind of space available, though. And I can't say I would consider that player worth investing in such nice caps.
The Acurus & Aragon amps, which are nicely built, but strangely sensitive to speaker load(can downright squawk with certain speakers), have an input/driver stage design that can be easily thrown off by small dc offsets at the input, so I would be a little hesitant to dc-couple the input. So, I would keep the input caps(the existing little Nichicon is bypassed by the 0.1uf mediocre film cap next to it). But at least you have ample room to put in a cap combo such as exactly what I suggest above.
BTW, my experience with very well-broken-in Wima metallised polypropylenes is that they sound hazy & veiled, not screechy.

great info and thanks again for your time

my gear is
Akai ap100 but looking at Origin Live Aurora MK2
Yamaha rxv1000 for surround and phono temporarily (best phono stage under 1000?)
Marantz cc4000ose 80 bucks in caps and regulators and getting better as we speak
also looking at Cambridge azur 751bd for video and hopefully ref cd sacd dvda
Acurus rl11 upgraded caps soon to be dampened
Acurus a150 new cans and dampened
Dynaudio audience 70 tower as yet untouched
Ixos speaker 9 gauge and interconnects nicest I can afford
Yarbo 10 gauge mains still breaking in

pics here, lol kinda smoky



http://i1041.photobucket.com/albums/...R/DSCF3000.jpg

I fully agree that ALL electronics need LOTS of break in and you know what they say about opinions ha.

thanks for your input on caps because experience in this respect I think what matters more than specs (I used to laugh at guys looking for specs on gear "buy it by the pound")

questions for you about the amp

should I leave input and output caps in and just up the quality?

how hard is it to remove that pcb?

it has solder lugs from pcb to speaker terminals

and if i undo screws to transistors on the heat sink will they just pull away?

just redo heat sink paste when installing?

last thing is some mods to speakers
dampening
wire
cap grade

I would really like to hear your opinion on that and I guess I will try some bypassing as you seem pretty adamant it makes a difference, as do others

I will start checking dc offset next anything you recommend for meter other than fluke?($200)

by the by I do like to blow my hair back (loud) and I listen to everything from classical, piano, some opera rock house punk and metal

it all really comes to dollars in the end 50 bucks for caps or 1000 dollars starting for a new player and the 50 bucks for cap for it.

thanks man I appreciate the help and opinion
Glen

[GLENZWORLD]

Originally Posted by Torchwood421

Hi, Have you ever considered using transformer output? It can be done. I am using Lundahl LL1690 and Edcor XSM 600/600. The trick is to get rid of any DC offset into the primary of the transformer. So with a TDA1541A based dac I used a Pedja AD844 based I/V and that has a current source to adjust the offset of the TDA to zero. The other dac is a CS4398 and by it's design has no real offset. Makes for a very simple output stage a single parallel cap across the primary to control high frequency roll off. Maybe not true DC coupling however transformers can sound quite good and I find that they beat most capacitor solutions or extra opamps. For power decoupling I am fond of Nichicon electrolytics. I use lots of Wima caps where necessary. I discovered PPS can be quite nice too.

thanks for the info and I have heard of that and direct coupling with transistors neither of witch I understand much but will read more

also heard of guys putting a rheostat across the op amp to adjust bias or offset but not there yet for understanding

1 question I have is does the CS4398 swap with TDA 1549T

and does a parallel cap control dc and or noise?

[stephensank]

I would frankly replace your RL11 and RXV1000 with a stock Nakamichi CA-5(the first, NOT the CA-5II, nor the truly awful CA-7). It's typically $250-350 on ebay or canuckaudiomart, and it's a very, very neutral and considerably transparent sounding preamp, both for line & esp. nice phono, and it's built so well that there's nothing to upgrade except the rectifier diodes. I think it's the best SS preamp out there for under a few grand. And frankly, I would also replace the amp entirely with a mod'd Nak PA-5 stasis amp(or any of the four Nak stasis amps), although it should darn well beat the Acurus even stock.

Unless dynaudio used electrolytic caps, you'll make the biggest improvement by using foil inductors. It's a not at all subtle improvement versus ANY other coils.

Although I have three Fluke meters, I have had just as accurate results with "house brand" meters from MCM Electronics and such. Nowadays, a DMM is a DMM, and Fluke has very little advantage.

Exactly right on pulling the Acurus boards. The sink compound will have a lot of stick, so you just need to pop each xstor loose before yanking on the board. And, yes, I would keep it cap input, with mainly just a better film cap/caps than that cheap black one. The other lytic/film cap combo nearby would be the low frequency corner cap from the feedback loop, and needs to stay there to prevent amp from having DC gain, and what's there is about as good as will make an audible difference. Any other lytics on the board will be supply line secondary filtering for the drive/input stages, and can increased in value liberally with potentially excellent effect, and film capping is good there, too.

Wouldn't touch Cambridge with a ten foot pole. If you want great dvd video & excellent sacd sound, along with the BEST reliability, get a Pioneer DV-79avi or DV-59avi for hdmi, or the DV-47ai or DV-45a if hdmi is not needed. These are all really cheap now, having been replaced by Pioneers quite excellent & reliable blu-ray(no sacd) players and by successor DV models that have garbage mechs(i.e., stay away from DV-58avi, 46av, etc.). I've been running a DV-59avi literally 24/7 at my shop for over a year with perfect performance, which replaced, for the sake of hdmi out for led monitor, a DV-45a that was still perfect after 2 years or more of 24/7 play. Of course, if you get a DV-79avi, which comes stock with great Burr-Brown OPA2134 opamps after the PCM1738 dac, it might make you want to abandon the Marantz player upgrades.

Judging by opinions of one or two that I trust, the TDA1549 in your player is pretty certainly better than a CS4398(which I consider quite etched sounding), and there is no easy way to replace one with the other, without designing an adaptor board. There are very few dac chips that match pin for pin with any other, even from same maker.

It would be very complicated and expensive to try to use transformers for decoupling in that unit, and I would sure as heck not choose a Llundahl or, yecch, Edcor. I would choose an appropriate Cinemag(started by the designer of all of Jensen's transf) if I was going to do that sort of thing, but I would not even try it on a voltage out dac like yours, as there would not be any expectation of benefit, versus what could be more easily and cheaply achieved with caps as I suggested. Transformers only make sense on current output dacs, IMO.

[Torchwood421]

Hi GLENZWORLD, I would have to agree with stephensank. The TDA1549 isn't going to be easy to directly replace with a CS4398. It sounds best (CS4398) with the Lundahl LL1690 or other good quality transformer. I didn't get good sound from the CS4398 with opamps or capacitors. Your experience may no doubt be different. My experience with the transformers I mentioned including the Edcor is that they work best in a voltage situation. My TDA1541A dac doesn't have enough current output to drive any transformer that I am aware of. That's why I use an AD844 in a Pedja I/V setup. Modified for 4X oversampling. That drives the Edcor with no problems. The Pedja is also no negative feedback and has 150 mA driving capability. Cinemag or Jensens would be an excellent choice, yes better then an Edcor. Edcor's are cheap, behave pretty well but lack shielding and are large in the 2.5 Watt version. The parallel cap is used to adjust the frequency response of the transformer. In a dac situation you want the rolloff to be lower then the transformer is designed to be. I have seen values of 4 to 7 nF used with many different transformers. The best sounding dac of the 1 bit type that I have heard to date is a CS4397 with output transformers. That one benefits from using a 1:2 transformer ratio as the CS4397 doesn't have the output of the 98. I find that the CS4397 sounds closer to a properly applied TDA then anything else I have heard. YMMV.... I have tried many output filter types. Sallen-key is bad. GIC gyrators sound interesting however are hard to build and fussy about opamps and don't always perform as you'd expect. I'm still listening to a 7th order Bessel filter on one of my dac projects. I wanted to hate the transformer approach. After trying it I was surprised in the simplicity and ease of building with it. The clarity and musicality can be outstanding. If you try the transformer approach by all means go with the better transformers. If your budget is tight give the Edcor a try. I'm putting my Edcors in a box to shield them, so far I'm happy with them. Although now I'm curious to try the Cinemags or Jensens. A warning on the Lundahls.... The zoebel in the datsheet is not optional. They can sound very hard edged without that resistor capacitor across the secondary. With the zoebel they work well. A little exotic with the permalloy.... Not to mention the price is in the painful zone. Ouchhhhh! Dave

[stephensank]

The biggest problem with Edcor is that their core materials are terrible. This is most obvious with the ribbon mic trannies, which is my field, but clearly a problem with all units. The Llundahls are just poorly designed, and as expensive as the FAR better Cinemags.
Still, between the internal opamp and the extra bits needed to deal with the dc offset, I don't think it practical or, in the end, better to try transformer output on a 1549, nor on the majority of V-out dacs. My view is that if you can't get rid of feedback in i/v and output stages, what is the point of transformers?

[DF96]

Originally Posted by GLENZWORLD

is the only way to find out if there is intermodulation with a scope?

No, spectrum analyser.