thedoctor

which one is easier to find, and which is cheaper. the LND150 are the toughest of the bunch, but might be hard to source for. i can't find them in Malaysia (even Farnell and RS don't stock them). you guys up south might have better luck. the CRD (1N5297) is the easiest to work with, and should suffice the voltage used in the amp. though CRDs tend to be costly. but taken that CRDs are easily replaceable with other compatible parts if should the 1N5297 is nowhere to be found, then i would opt for them. and one should most probably have easier time working with CRD than MOSFET (LND150). CRD is reverse polarity protected, and you won't burn them by plugging them in the wrong direction (well theoritically anyway). as for MOSFETs, most should survive, but i've burned (exploded really) one MOSFET by plugging it the wrong way.

the JFET are somewhat too fragile to be used. at least the J113 are. 2SK30A are not pin compatible with the board though those are a lot sturdier. and thus i would suggest the LND150 MOSFET or the 1N5297 CRD.

QSilver, are those Rubycon ZLs in there? you're using all ZLs issit? if so, how much do they cost?

first thing to take into consideration is the voltage to be supplied. the Nokia batteries have only 3.6V of output. that barely runs anything other than the cheapo National chip used in the PA2V2. well maybe the AD8397 on the PINT could also be used. but judging that those needs 3V of minimum voltage, then i have to say that you have an awfully short run time. well maybe one could use a DC booster. there's a few DC switching chips from Linear, Maxim and Texas that has high efficiency and could work all the way down to 1.5V. some are operating at 200KHz switching range that would most probably won't be audible when properly implemented. as in no switching noise. even Xin's portable amp are using a DC booster chip to bump the operating voltage to 12V from 5.5V and 3V. or you could run them in series and get higher voltage. but by then you couldn't use the normal headphone charger anymore. or paralel them and stick to a DC booster and have better run times. but all in all, Lithium Ions are not exactly easy to work with. you need an overcharge protection and take into consideration of the ultra low impedence of the batteries itself. put too much capacitance in the power supply and most probably something will blow from the initial power-on rush. needs careful consideration and design. for the hassles, i'd rather buy me a rechargeable NiMh 9V. but that's just me..

thanks for the nice compliments. i'm still trying to finalize the circuit. what i'm calling the Portable Reference V2 have 5 different prototypes till now, and i'm still squeezing the best sound from it before concluding on a final design. er, you guys wouldn't mind if i start a new thread with each new project? i have a few ongoing projects, and the Gainclone & an M3 next in line (after i finalise the PRV2).

electronics shops, perhaps? Sim Lim Square, maybe? i have no knowledge about singapore, but these : sells at RM12-RM15 a piece over in Malaysia. not exactly uber l33t pretty looking case, but does the work quite marvellously. and i doubt a big enough hammond will be cheap. those that i've used are 3"X6"X9" cases, and fit the board just enough. big enough for the trafo to go in. one could mount the trafo outside, but you need 3 wires + 3 prog plug. and you definitely need some space. a cramped inside with flying wires sitting too close to each other might induce the ever dreadful hum. and yes, those are all SOHAs in the pic.

ouch.. think i've made a mistake there. i believe what i've seen is a post of a schematic on another forum of a crossfeeder that has no caps in it. but from what i've searched (for about 30 mins), each and every one of them does have a cap. either i've made a mistake or i can't seem to find it. i believe what i've seen is the Simpson Simulator, but those also have a cap (yet low value) in it. maybe my brain ain't working correctly.

hum is a somewhat peculiar issue with the SOHA. some get them, some don't. i can't even get any after 8 tries. and not even after doing this, to 3 different transformers : and pardon my ignorance, but what is exactly a hum pot?

er urm.. not to dissuade you from your first adventure in DIY. well DIY is a good thing, but unless you're experienced enough, i'd advice you to build something else rather than a Crossfeeder. unless you're thinking of using polypropylenes for C1 and C2, better drop the Modified Linkwitz altogether. even a polyester in there sounds rather sucky compared to the Crossfeeder that comes pre-built with foobar2000. there's another crossfeeder that's simpler than the Modified Linkwitz. one that don't have a capacitor in the signal path/circuit. try to find that instead. i'm not cap-fetish, but i've built a few Modified Linkwitz before only to find that the internal foobar plugin to be better. yeah sure, maybe it's just me. but there's another simpler crossfeeder available.

two Opamp, with what looks like a single JFET biased Class A. two voltage regulators, with what looks like a dual rail power supply. lots of Wima MKP 0.1uF/160V caps with what looks like a crossfeed circuit, or a CRC filter, perhaps? yet i have idea the the use of the 8 TO-220 parts below the RCA jacks. mind to explain what exactly did you do? looks like an unbuffered headamp.

i particularly like the Discreet Diamond Buffer. don't matter if it's operating in Class A or AB, but surely beats the BUF634 in transparency and speed. the BUF634 sounds veiled. even a simplified Diamond Buffer or the Push-Pull Discreet Buffer sounds better than a BUF634. yet i've never tried the Intersil nor the Elantec buffers, but did a quick comparison between a simplified Diamond Buffer against the BUF634 and the DB is the clear winner. never used a BUF634 ever since. tho for simplicity, a BUF634 might be easier to implement/use.

i do think the PPA will suit Senns much more than a Millet will. the transparency that you will get from the PPA could NEVER be matched by a hybrid tube amp. i've built quite a number of amps, mainly the YAHA, MHHA, SOHA, META, PIMETA, and the PPA outclasses them by far when it comes to transparency and speed. some might swear that the Millet with Diamond Buffers are close sounding to the PPA while retaining the speed and transparency. but i wouldn't trust reviews too much. i can't get hold of some of those rare 12FK6/12AE6/12FM6 radio tubes, if not i would definitely have a Millet + Diamond Buffer built in a jiffy. up till now, i've built a 2 maxed out YAHA with Diamond Buffers, 3 MHHAs, 8 SOHAs, and 5 PPA v.2. the PPAv2 currently holds the throne. currently i hold a PPA Demo program for Malaysians, here since i can't draw interest from people to look at an RM600 fully built PPAv2. even when i comes with 1 year full warranty. ^ that is about what, SGD260-something, right? original sales thread here. and sorry if i made a mistake by doing this. just so that everyone knows, that there's someone quite actively building headamps in this part of the world. and not putting insane prices like US based builders.

er erm.. the truth is, all the schematics are all in my head. i've built quite a number of new amps, and the ones that you see above is one of the completely new design. i'm still trying to learn Eagle, so please bear with me. but even with a PCB, i doubt we could make it as small. since the amp is cramped into a very small space, use of jumpers cannot be avoided. well unless we go with 2 layer PCBs that is.

maybe some unlucky ones. because some do built it with good soldering and all and still it goes haywire. tried plugging all sorts of different impedence to it? some find it acting quite differently when driving different loads. and sometimes it goes sideways being given different voltage. for the instability issues it might have with different cans and voltages, i'm just not *very* interested in it. i reckon the AD8397 is THE lowest Opamp that could be used with low voltage. something like 3V minimum i think. and if you're looking for even smaller case and AAA batteries, then you can't use AD8066 and AD8620 since those needs around 9V to work. one way to solve the oscillating ground channel is to swap the ground channel Opamp with some other chip. though i would say the difference from doing that ranges somewhere from NONE to MINUATURE, yet some golden ears seems to say otherwise. some seems to say it's even more musical and the such. so YMMV. but since your unit seems to have no stability issues, i'd say don't bother with it. the current draw is about 27mA when driving my 32 Ohm cans. idles at 25-26mA. yet there's a few miliamp difference when using different chips. the power output of the amp are limited by the buffer section. the use of the discreet diamond buffers allows for something about 800mA max. the Opamp used have NO effect to max current output since it's a buffered design.

using what chip? each Opamp draws different current. if it's the AD8620 and AD8066, it should draw about a quarter of the AD8397. well, theoritically. but i do think it will be about 1/4. the PINT when running with 2 X AD8397 draws 50-55mA.

not exactly. there's always a replacement fuse that could be used. the CD800a have about a max of 200mA, and something like 0.2A of 0.25A fuse could be used. the voltage rating could scale from a few volts to hundreds of volt if you're not paranoid about it. the CD800a fuse are somewhat easy to burn. get a few and stock it.

man, thedoctor, you are right!! then maybe something like this, perhaps? not exactly a PIMETA, and even closer resembling a PPA, implementing a discreet diamond buffer at each output. all channels, including the ground channel are buffered, and biased into Class A with Toshiba Low Noise JFETs. have an integrated TREAD power filter/regulator within the amp itself. and still, portable enough to be run with batteries. sucking less than half the current drain of the PINT/Mini3. but really, the amp does draw less than half the current. which should means twice the runtime. i would really suggest otherwise. the AD8397 is used in the first place because it has huge output current and don't require an output buffer to work wonders. this allows the whole buffer section to be dropped since the single AD8397 could output 200mA of current, which is about the same as what the BUF634 is capable of. swapping the AD8397 to AD8620 or AD8066 means that you're back to a maximum of 40mA-something of current output. there's been talks about swapping the AD8397 with LM6172. and it helps in stabilizing the whole circuit. and that it sounds better than the AD8397. but by then the LM6172 only have a maximum of 50mA of output current. ^ which still negates the original purpose for the use af AD8397 in the first place.

and i still can't understand why people would wanna mess with the unstable AD8397.. you guys are so dying for a portable, good sounding amp i presume?

maybe because it's so transparent that the sound is coming like from all over the place? sometimes seems like it's coming from right in front of you, before moving to the side and fading below to your feet. and in my experience, the crossfeeder circuit could be much replicated by the Crossfeeder plugin in foobar. at least not the stock plugin, and there's another one that could do a much better job than the stock one. unless the crossfeeder uses only polypropylene caps (if the crossfeeder circuit requires them), then i would really suggest sticking to the crossfeeder plugin.

the NOS Mullard 12AU7 are somewhat raved as being the best. yet i've never tried them out. i'm in favor of the Ei Elite gold pin 12AU7, though some class them as being bright. the JJ ECC82 sounds somewhat harsh to me, and the Electro-Harmonix as being too laid back. the SOHA is somewhat sensitive to the tube used. which is a good thing in the sense that one could try all sorts of different tubes to match their preference. even more sensitive to tube rolling compared to the Millet Hybrid/YAHA/MHHA. and so that everyone knows, i've tried the Diamond Buffer build a long time ago for the SOHA. and with about 5% unmatch (if my memory recalls correctly) between the NPN and PNP pairs, the amp outputs 6V of DC offset. and from experience working with Diamond Buffers, finding matching NPN and PNP pairs could be somewhat a pain.