Hi Denis,
Is a highly modified version of Rockna RD2.
http://www.audiokit.ro/cart/index.php?productID=118
DSD1792 instead of PCM1796 and modified i/v stage to cope with new DAC
Toshiba low Cob output devices at high idle current (single ended output)
Low noise shunt regulator for clock CXO, separated for other digital circuits.
OSCONs and silvered mica in key positions
Low radiation r-core trafo for analog stages.
CRCRC power supply in front of CCS+shunt onboard regulators for I/V stages.
and many many other small but important changes
Cheers,
Mihai
Is a highly modified version of Rockna RD2.
http://www.audiokit.ro/cart/index.php?productID=118
DSD1792 instead of PCM1796 and modified i/v stage to cope with new DAC
Toshiba low Cob output devices at high idle current (single ended output)
Low noise shunt regulator for clock CXO, separated for other digital circuits.
OSCONs and silvered mica in key positions
Low radiation r-core trafo for analog stages.
CRCRC power supply in front of CCS+shunt onboard regulators for I/V stages.
and many many other small but important changes
Cheers,
Mihai
Wow....
Good job.
Rcore transfo are they the same type of "split type' ? http://media.digikey.com/photos/Tamura Photos/3FL34-170.jpg
Good job.
Rcore transfo are they the same type of "split type' ? http://media.digikey.com/photos/Tamura Photos/3FL34-170.jpg
DRZ1 said:Wow....
Good job.
Rcore transfo are they the same type of "split type' ? http://media.digikey.com/photos/Tamura Photos/3FL34-170.jpg
No, that's an E-I type split bobbin. Those are the ones I use for low power applications. Good, but not as good as the R-core.
Re: Re: Why a carefully designed project can fail.
Thanks Salas for the assistance.
I tried already this solution with copper foil of 0,5mm thick arround toroid (which is made in Giatras - 500VA encapsulated = 75 euros), but... nothing. Cables are comming from the bottom of toroid thus for mounting it a sub-chassis is neccessary. I have made a plate from 3mm thick alu of 145mm width (which is exactly the diameter of x-former) which is joined between the front and back pannel. The mistake was clearly mine: I trusted Giatras for this custom ordered "supposevelly super low noise" toroid. Bearing in mind this, i designed the whole amplifier. I got my lesson. The only cure exists, it is the redrawing of amp pcbs so as can placed on heatsinks thus vertically in relation with toroid and FAR AWAY from it.
BTW, take a look in the DAC project of "roender" to get an idea how is placed the R type xformer of Selectronic which is absolute noiseless (i have tried one 500VA before 1 year).
Regs
Fotios
salas said:
Thanks Salas for the assistance.
I tried already this solution with copper foil of 0,5mm thick arround toroid (which is made in Giatras - 500VA encapsulated = 75 euros), but... nothing. Cables are comming from the bottom of toroid thus for mounting it a sub-chassis is neccessary. I have made a plate from 3mm thick alu of 145mm width (which is exactly the diameter of x-former) which is joined between the front and back pannel. The mistake was clearly mine: I trusted Giatras for this custom ordered "supposevelly super low noise" toroid. Bearing in mind this, i designed the whole amplifier. I got my lesson. The only cure exists, it is the redrawing of amp pcbs so as can placed on heatsinks thus vertically in relation with toroid and FAR AWAY from it.
BTW, take a look in the DAC project of "roender" to get an idea how is placed the R type xformer of Selectronic which is absolute noiseless (i have tried one 500VA before 1 year).
Regs
Fotios
P.S. On that amplifier pic I posted with the DIY copper circumferential grounded sleeve, the main buzz pick up lines were the input stage ground returns to the buss bar that were very near the trafo. 90% of the problem I solved by using shielded coaxial for them with only one end of the shield connected to the ground bar. The last 10% to total silence on 95dB speakers, was the toroid added copper sleeve. Maybe you can use that method too for the wires that run under your toroid?
Re: Re: AMB Laboratories' ?4: A discrete, cascoded, fully-differential power amplifier
No, they're 625VA toroidal transformers, an overkill.
Moreover, there is a 100VA toroidal transformer for the front end on each side of the heatsinks in the front. Altogether I have 4 transformers in the case.
fotios said:
No, they're 625VA toroidal transformers, an overkill.
Moreover, there is a 100VA toroidal transformer for the front end on each side of the heatsinks in the front. Altogether I have 4 transformers in the case.
salas said:
But that is just as bad with an EI core as with a toroid. Flux encirling the secondary exits will be the same in either case - but you're not as likely to put them close to the inputs/feedback with an EI trafo.
I've had secondary exits indcung hum into the input stages of an adjacent amplifier. The hum got worse in the amp on the bottom the harder the top amp in the rack was pushed. They were EI core.
wg_ski said:
Hum noise caused from electromagnetic flux (Ö) radiated from the core of transformer in any direction. Of course its stronger part radiated in the same direction as it flows in the core. Another one source that can cause hum noise, it is a high magnetizing current... e.g. if the core of toroid is made from single steel band which has 6 times less magnetic permeability (ì) from electric steel (please see in wikipedia for details) then it needs 6 times more current to magnetized to produce the appropriate "Ö". An ideal core, could be made from Mu-metal, which presents the higher "ì" but its price is prohibitive. Electric steel is the usual alloy of cores and has 5 times less "ì" from Mu-metal.
A good core "Ö" shielding can be done only from alloys with high "ì" because only these can absorbe "Ö". Copper and Alu have 700 times less "ì" from steel and 4000 times from electric steel. Thus in practice they can't offer any good shielding except than have enormous thicknes. It is prefferable the use of electric steel for "Ö" shielding.
Permalloy is the second alloy in the table of "ì" and has 2 times bigger "ì" from electrical steel. It is less expensive from Mu-metal, and there are laminates from permalloy for constructing EI xformers for very sensitive measuring instruments like medicals. Of course in audio power xformers its cost is prohibitive.
The vast majority of toroid xformer constructors, they based on the shielding of core by the copper windings, in the momment that copper offers the same "Ö" absorption like water. Alu, Copper, water, vacuum have the same low "ì". Also, they do a second mistake which is the tightening of the spiral formed core from the copper windings. How much strain force can accept a thin copper wire?
IMHO a good core must be formed very tightly, and dipped in a bath of epoxy varnish under vacuum to become still like a rock. The same make the good constructors of EI xformers. Thus any mechanical noise will be minimized. After the core has dried thoroughly, the primary winding can be done. After, a bath again in epoxy varnish it is good. After insulating the primary winding with (usually) mylar tape, an auxiliary internal belly band from copper tape it is good. A terminal wire soldered on this belly band it is appropriate to earthed later. Above the internal belly band, a second winding from mylar tape it is indispensable for insulation. Now it is time for the secondary winding. After its completion, again a bath in epoxy varnish it is good. Then it follows the insulation of the secondary winding with mylar tape. Now we have a toroid with the ordinary appearance. At this point, an external belly band around the circumference of toroid from electrical steel, can offer a good absorption of "Ö" radiated from core. Alu and copper thin foils, are good conductors of "Ö", consequently are useless. Instead a bowl made from 8-10mm thick Alu, can be a good screen. If we have a such made transformer, we can place it near and in any direction with the sensitive pcbs.
In different case, the only that we can do, it is the place of sensitive pcbs vertically in relation with the circumference of toroid... and as far as possible from it.
Mechanical noise is a negligible part of hum noise injected in the audio path. The 99% of hum noise is caused from "Ö". Top and bottom covers of amplifier case, they share the "Ö" of xformer generated in unpredictable directions inside the case... like mirrors or transmitters (i don't know exactly).
A good example of pcbs orientation in relation with xformers, appears in the attached picture of a Mark Levinson.
