A bunch of random questions for the DIY Guru's.

[Voodoochile]

Transformer (magnetic flux) shielding has to be ferrous to be effective. Aluminum and copper don't work well at all. I don't know much about Mu metal other than it's also effective and light, but expensive. Copper and other metals may work well for RF shielding, but not for magflux, IME.

Small steel Hammond chassis bases make decent transformer covers. In the case of a toroid, you must not thread a bolt through the cover/toroid/base, as it will short the windings. Use a normal mounting bolt/washer, then have a cover that does not contact the mounting bolt. For laminated core transformers, you just fit a cover..

 

[philodox]

Quote:

Originally Posted by Voodoochile Transformer (magnetic flux) shielding has to be ferrous to be effective. Aluminum and copper don't work well at all. I don't know much about Mu metal other than it's also effective and light, but expensive. Copper and other metals may work well for RF shielding, but not for magflux, IME.

So an iron box around the transformer is the way to go then I guess, thanks.

[Thanks for the tips about mounting as well]

 

[randytsuch]

In a 4 pin xlr, you would have L-, L+, R- and R+. Just like hooking up speakers. You would need to run another wire, or use the shield to run ground.

For some questions, like having a seperate PS, is an opinion. There are advantages to having you power supply transformer far from everything else, but there is also advantages to having the power supply close, so the wires between the power supply and the circuit are short. You have to choose your evils.

The squeezebox2 has digital volume control, and it is supposed to sound very good when modded. I know a guy that has a modded boulder cable one, and he is very impressed with it.

Randy

 

[philodox]

Hmmm... from reading Strohmie's thread that details his build for his CD player, it sounds like I2S might not be ideal if it is going from a transport to a DAC [each in seperate chassis]. It would be pretty hard to keep the path under 1 ft when you are throwing an IC in there.

EDIT: By the way, what about EIAJ? Is that a viable format to use for sending the signal between a two box transport/DAC?

 

[gerG]

It depends, to some extent, whether we are talking about a headphone cable, or a signal cable. Running 2 signal cables together in the same jacket (without individual shielding) will result in crosstalk. It is simply bad design practice, imho. Look around the industry and notice the effort that is put into even the most common of cables to keep the 2 sides isolated and shielded from one another.

Headphone cable is probably not as sensitive, but could still a problem. I would be shocked if a "high end" cable had the sides unshielded from one another. otoh some listeners might perceive crosstalk as pleasant. Personally I intend to err on the conservative side and keep my channels separated on future projects.

As for the plug, I don't see that as a compromise. Not enough length to cause a problem, and the pins are well spaced apart. A twin isolated lead running to a single 4 pin (like the K1000) is not a problem. I still would not do it upstream of an amp, though.


gerG



Quote:

Originally Posted by Edwood 4 Pin is not a major compromise. In fact, the whole using two separate 3 Pin XLR's is just silly. Make sure the shields are grounded to the shell of the XLR. Most (namely Neutrik) have a nice tab to solder to. Headroom uses Cardas cables that run both channels in the same jacket. I'd have to cut one open to be sure, but I doubt both channels are individually shielded. -Ed

 

[Nisbeth]

Quote:

Originally Posted by skyskraper all my xlr gear is wired pin 1 shield, pin 2 hot, pin 3 cold.

AFAIK XLR = eXternal (screen), Line (hot), Return (cold) in that order. Not sure if this is correct, but it works for me as a little "memory aid"

/U.

PS: Just change the mains inlets on my DIY-power amps to PowerCon. These are very nice indeed

 

[rickcr42]

1-XLR is the best balanced connector if compatibility is important but better for heaphone specific use would be a multipin LEMO for contact integrity or an Amphenol four conductor 1/4 inch phone for ease of use and multiple insertions and the ability to take abuse.

1b-LEMO again.Small contact area and solid gas tight connection plus self wiping contacts.Even a computer serial connector would be far better than the RCA jack and plug

2-I2S has many good features because it totally eliminates the jitter prone interface of modulate (transmitter)/digital interconnect/demodulate (receiver) by having all data and clock lines as individual connections.the SPDIF connection was never meant for widespread use but for testing purposes on the bench and like with the RCA connector becoming the loony standard for line level audio when never meant for such a piss poor interface.AES3 can be better but still being a SPDIF transmission not the best choice even though the industry standard.USB may change all that but maybe not

3-no reason the volume control can not be (and in many cases is) on the DAC output if selected for impedance compatibility with the next stage.what do you think a "passive preamp" is ?
Having it be an electronic control is a variable thing with there being both good and bad methods.

4-Nuetrik Powercons

5-No

6-a good idea if you don't like fires.Could you sleep at night if your house did not have a breaker or fuse panel ? I could not !

7-An external power supply is always in every case preferable to an internal supply except in the cost area.Part two would be full coverage Mu-Metal shielding (expensive)

8-There is no ideal and would be impossible to attempt.You make system decisions based on individual matchups not some theory of "best" for everything.
Line level,speaker level,low level,high Z,low Z combined with point in signal path and gain level.........all have a place and a funbction based on function

 

[philodox]

Quote:

Originally Posted by rickcr42 1-XLR is the best balanced connector if compatibility is important but better for heaphone specific use would be a multipin LEMO for contact integrity or an Amphenol four conductor 1/4 inch phone for ease of use and multiple insertions and the ability to take abuse. 1b-LEMO again.Small contact area and solid gas tight connection plus self wiping contacts.Even a computer serial connector would be far better than the RCA jack and plug

Those LEMO connectors sound like a cool idea... any one in particular? I went to their site and it says they have over 55000 different connectors.

Quote:

Originally Posted by rickcr42 2-I2S has many good features because it totally eliminates the jitter prone interface of modulate (transmitter)/digital interconnect/demodulate (receiver) by having all data and clock lines as individual connections.the SPDIF connection was never meant for widespread use but for testing purposes on the bench and like with the RCA connector becoming the loony standard for line level audio when never meant for such a piss poor interface.

Yeah, I am thinking that I2S or EIAJ may be the way to go, but am worried about what I've read in Strohmies thread about the signal paths needing to stay under 1 ft. This makes using an IC between the transport and DAC [two box setup] quite difficult.

Quote:

Originally Posted by rickcr42 3-no reason the volume control can not be (and in many cases is) on the DAC output if selected for impedance compatibility with the next stage.what do you think a "passive preamp" is?

Ok, so this would be on the analog stage then, not the digital domain right? If I did this and set up my headphone amp in a poweramp like configuration should I be good to go or are there things that I should worry about? Quote:

Originally Posted by rickcr42 6-a good idea if you don't like fires.Could you sleep at night if your house did not have a breaker or fuse panel ? I could not !

Good point.

Quote:

Originally Posted by rickcr42 8-There is no ideal and would be impossible to attempt.You make system decisions based on individual matchups not some theory of "best" for everything. Line level,speaker level,low level,high Z,low Z combined with point in signal path and gain level.........all have a place and a funbction based on function

Fair enough, but what if you are designing a transport, DAC and amp that are meant to be used together. With the volume control residing on the DAC and the headphone amp set up like a power amp with no attenuation [or perhaps just a gain switch]? Just a hypothetical question of course... at least until I learn some more and gain some DIY experience.

[I'm thinking about taking some courses]

 

[rickcr42]

1-two conductor with plastic shell.Punch in your requirments and you will get a part number and source.The RCA is all "pretty" metal combined with piss poor electrical contact never meant for anything other than connecting the phonograph section of an all in one "phono/radio/amplifier/TV" console to the amplifier section in the 1930s hence the "phono jack" designator which still clings as a name.

2-there are workarounds (buffered lines) but even if not I have my DAC right at the transport so 1ft no impediment

3-digital domain would mean its implementation would have to be as a codec/dsp chip between the reciever and DAC chip and unless you have those skills ( and i do not) better left alone.Analog domain CAN be a digital control but be away 99.95 have serious flaws that will degrade the sound though the Wellborne design looks to be well thought out.
A 10K pot following a DAC and going to a "straight" headphone amp stage having a fixed hi-Z input resistance should work fine unless you want to use interconnects over ten feet in lenght then a simple unity gain buffer/voltage follower if active or line output transformer if passive should drive the cable (assuming good drive capability of the DAC output stage)

4-Safety first man ! ALL else is secondary

5-take everything one step at a time so you know your requirments rather than pre-decide what is "best" at all stages.Begin with the "source" which like the mountain top that spawns the river is the beginning of all things.
Transport/DAC first then depending on the particular requirements this ppalces on the next stage will determing which way to jump.
Any audio stage can have an "impedance matcher" added but when there is a choice always best to avoid any non essential intervening circuitry so Basic transport -to-DA conversion-to-controls/switching-to gain stages if needed-to-transducer drivers (speaker or headphone amps or maybe even simple current pumps if you already have enough overall gain for good volume drive)-to-final transducer (headphones/speakers)

Quote:

[I'm thinking about taking some courses]

Plenty of "101" style informational links in the DIY links database.Free access,easy to ingest at your own pace and in the privacy of your own space.Zero pressure and at your own pace but if jammed up with this forum for "clarity" of the highly technical areas.Always someone here willing to help.

If new to this building then an entire system from the start may be a bit much unless you already know your equirements.Start with the beginning,the source,and go from there to match each stage to the previous as you comne to it would be my recommendation

 

[philodox]

Quote:

Originally Posted by rickcr42 2-there are workarounds (buffered lines) but even if not I have my DAC right at the transport so 1ft no impediment

Yeah, but once you count the distance inside each component that eats up a lot of the length. Maybe I should look into buffered lines, thanks.

Quote:

Originally Posted by rickcr42 3-digital domain would mean its implementation would have to be as a codec/dsp chip between the reciever and DAC chip and unless you have those skills ( and i do not) better left alone.Analog domain CAN be a digital control but be away 99.95 have serious flaws that will degrade the sound though the Wellborne design looks to be well thought out. A 10K pot following a DAC and going to a "straight" headphone amp stage having a fixed hi-Z input resistance should work fine unless you want to use interconnects over ten feet in lenght then a simple unity gain buffer/voltage follower if active or line output transformer if passive should drive the cable (assuming good drive capability of the DAC output stage)

That makes sense, thanks Rick.

Quote:

Originally Posted by rickcr42 5-take everything one step at a time so you know your requirments rather than pre-decide what is "best" at all stages.Begin with the "source" which like the mountain top that spawns the river is the beginning of all things.

That is a very philosophical way to look at it.

Quote:

Originally Posted by rickcr42 Plenty of "101" style informational links in the DIY links database.Free access,easy to ingest at your own pace and in the privacy of your own space.Zero pressure and at your own pace but if jammed up with this forum for "clarity" of the highly technical areas.Always someone here willing to help. If new to this building then an entire system from the start may be a bit much unless you already know your equirements.Start with the beginning,the source,and go from there to match each stage to the previous as you comne to it would be my recommendation

I have been reading a lot of those links, but some of them are over my head.

I know that building an entire system will be challenging, but I do have a complete system that I quite like, so I can do each new component in its own time and have no need to rush.

 

[Jeff Wong]

Jay - Kal Rubinson wrote 2 articles on a DIY I2S circuit in The Audio Amateur (3/95 and 1/97) available from AudioXpress. While I2S is mostly an internal connection some companies have made connectors for external use. There are the only 2 external I2S "standards" I can think of. Audio Alchemy and Perpetual Technologies used the inferior 5 pin mini DIN plug. As I recall, Sonic Frontiers (?) made an I2S connection using a computer cable with 10 leads. Unfortunately, I2S never took off as a connection, even though I2S is a far superior format as compared to AES/EBU or S/PDIF. Keep in mind that the performance of I2S degrades as the cable gets longer. About 1/2 meter is ideal with an upper limit of only 1.5 meters, and that's pushing it.

 

[philodox]

Thanks for the info Jeff, I'll check out that Sonic Frontiers cable... sounds like what I'm looking for.

 

[Tim D]

Not a DIYer but just reaffirming what everyone is saying for 7.

I prefer external seperate power supply. I have done real-time analysis with RMAA to actually see 60hz hum and other noise creep into the output the closer I move an external power supply to the actual amp, etc (or if I move unbalanced cabling close to the transformer box of my MAD Ear).

Its unfortunate but a good deal of even high-end commercial products still stick everything in one box. I'd imagine not only because of cost, but because of wife factor and aesthetics. On the otherhand that doesn't mean a wallwart with the transformer at the power plug and long cabling to the unit is ideal either. So my vote would be IEC cord to external PSU with about 3 feet cabling to unit. Typically it would be such a hassle/headache in terms of placement which is probably why commercial units don't do this, just think if you had to place every component and its PSU 3-4 feet away (lol you could have a DIY rack, one for components and one for PSU's right behind it). So yes external power supplies (and lets not forget component placement) can easily make both audible and measurable differences in your final results.

BTW can anyone comment on external PSU's and units that spit out VAC as opposed to VDC?

 

[philodox]

Well I have an all-in-one CD player and a dynahi in a single box and have heard zero hum with my rig, so I'm sure there must be ways to limit the interference to a point where a two box setup is not required. As you said, from a design and marketting standpoint you can see why companies tend to go with single box setups.

Here is an idea... what about just housing the transformers externally?

 

[rickcr42]

Quote:

Yeah, but once you count the distance inside each component that eats up a lot of the length. Maybe I should look into buffered lines, thanks.

Audio Alchemy was once the "champion" of the I2S digital connection and if they had their way would be the standard for all things digital.Having separate clock/data/control lines has a lot of merit in my personal opinion and any time you MUX a signal (multiplex) you are looking for trouble just by the addition of "more" to the path.The SPDIF connection in fact spawned a whole cottage industry of "reclockers" just to get the jitter added by the receiver and transmitter OUT of what never had it to begin with.

for some background intel on I2S :

Pinout-

http://db.audioasylum.com/cgi/m.mpl?...2s&r=&session=

actual circuits from Audio alchemy (print out and study)-

http://www.audioasylum.com/images/dtipro32.pdf

http://www.audioasylum.com/images/DDE3.pdf

USB also eliminates the MUX/DEMUX part of the digital audio interface but in a different way entirely

Quote:

That is a very philosophical way to look at it.

Born of practicality.Something that may whip ass when conbined with a certain other device may in fact flat out suck if mismatched in a systemic way.
Beginning at the "start" there is a set of knowns which make the unknowns easier once understood.The "all in one box" CD player on many levels superior than the two-box solution but what you usually are stuck with is corners cut for cost savings.It is here that you need to decide if your requirements are better served by separate DAC/CD Transport or if modifying the internals more practical (such as replacing everything from the IV stage outwards to the RCA jacks).Once you get to the output stage this will then determing what is needed at the next stage.
Why have 20dB gain (X10 gain) if you are already pushing 5 volts output at the CDP or DAC output jacks while having an amp that attains full volume with 1 volt ?!?!
Would it not make more sense to have an impedance converting stage here,if such is needed,rather than gain which adds noise by a factor of the gain and makes your volume control close to useless by having all the adjustment at the bottom end of the rotation where tracking sucks ?
that or the inclusion of inline attenuators ?
What about if your source is a phono stage,or tape deck,or radio tuner ?

In "the old days" everything had its own volume control PLUS input attenuaors at the point before the input selector PLUS and overall volume control PLUS interchannel balance control PLUS many times a power amp "sensitivity" control (no more than another volume control) and this all because they could no foresee what additonal hifi componantry would be used in what combination or what the levels would be.
so the control preamp would almost every time have excessive gain "just in case" with many points to toss off the excess.
good practice is you do most of your attenuation at the front end amd least amount at the top end (at the speaker or headphonbe amp) for best S/N ratio and sensitivity comination.
all DIY means you can select exactly what you need thus eliminating the need to throw gain away.


Do a search here in DIY on "interconnections" or similiar items or better yet go stright to Jensen Transformers read everything in the FAQ section and then bounce over to the Rane Library for extensive info in level and impedance matchups.
For me personally i like to see four total gain stages from source to output transducer.

1-Phono stage/DAC output stage/tape head preamp/microphone preamp/tuner output stage (whatever the source is)
2-Preamplifier/Control Stage (in reality a "post" amplifer since the "pre" ceeding stage is in reality the "pre" amplifying stage with every single gain stage being an "amplifier" in reality) which serves dual duty as the entire gain stage section for and headphones.It may require the addition of a follower or matching stage but not another gain stage
3-Power Amplifier "driver" stage with "passive" gain being a real option here if the requirements not exeeding +6dB.A "true" integrated amp is really a three stage amp though these days the amount of gain available from digital usually means a two stage amp has a simple input selector/volume control added and then "called" an integrated
4-Power amplifier "output" stage

Toss in maybe an MC cartridge and you get an additional stage .

at each point all you want is enough gain to drive the next stage to full volume with enough overhead for peaks while not imposing and undue strain on the previous stage in its ability to drive it PLUS drive the next stage.sounds simple but not so very