First Watt B1 Buffer Preamp

[TinTopHack]

sir tony namisslook nyo ata... full wave voltage doubler yong input nya.

No need for a doubler because your zener voltage is only 18V so I suggest you follow Tito Tony's suggestion. The original ESP circuit was using 30V zener thus the need for a doubler circuit.

[Engr. Dacuycoy]

ah ok.. baka kako namisslook lang ni sir tony kaya nya nasabi nyang mali yong caps  

[tony]

sir tony namisslook nyo ata... full wave voltage doubler yong input nya.

that is not the way to hookup a full-wave doubler:

this is the full-wave doubler:

[TinTopHack]

there is a mistake in the drawing, C1 and C2 is just 1 cap and connects to the + and - terminals of the bridged diodes....
16volts ac input means an unloaded rail of 22.4 volts, as it is, zener current is about 22mA without anything connected to the psu terminals, actually less if we consider the dc resistance of hte power traffo...

D3,R4,Q1 can be replaced with just a single 5watt zener diode.....

In this voltage doubler by ESP, D2 and D4 are useless. They should be removed. Once removed the circuit becomes what Tito Tony presented. Looks like ESP missed that...

[efren]

In this voltage doubler by ESP, D2 and D4 are useless. They should be removed. Once removed the circuit becomes what Tito Tony presented. Looks like ESP missed that...

Lessons learned: Di porke dinesenyo ng puti wala ng mali. Pairalin parin ang sentido kumon at pag-aralan ng mabuti ang mga napupulot natin sa internet

salamat sir tintop, pogi points sa inyo ni sir tony

[n0th1n6]

Mas nagiging interesting ito ah

[Engr. Dacuycoy]

dito pala nakuha ng ESP ang circuit.. use for regulating turbine

At low speed the circuit comprising D1, D2 and CI, C2 functions as a voltage doubler to gain enough voltage for charging, at speeds above the normal turbine cut in speed diodes D3, D4 switch in and with D 1, D2 function as a full wave rectifier. In full charging mode the capacitors still assist current production by allowing a larger portion of the turbine produced sine wave to be utilized.

http://www.astrosa.com/n03/n0803.htm

[tony]

ah ok.. baka kako namisslook lang ni sir tony kaya nya nasabi nyang mali yong caps  

i remember that the rectifier circuit also used a switch between the capacitor junction and one side of the bridge so that in atx psu's you are able to select between 220 and 110volt sources to get the same dc rails....otherwise if you are sourcing from a traffo secondary where no such need for choosing voltages exists, then a simple arrangement is enough, saves on parts as well.....



newbies should be able to know the difference..... the switch is put there for increased dc rail...

there is ambiguity here..........

[01110011]

pa subs madami ako matutunan kasi nandito lahat ng mga masters

[Born2BeWired]

good advise......

napasin ko ang title ng thread ay "First Watt B1 buffer preamp" i know this was a Nelson Pass design so this is the proper link: http://www.enjoythemusic.com/diy/0708/first_watt_b1_preamplifier.htm

and scheme:

i have no idea what "first watt b1" was, so i simply assumed beavishhifi link was the circuit that was under discussion

Nonetheless, while we are at it, might as well make some quick comparison.
Nelson Pass circuit is way better than beavish. In as far power supply decoupling is concerned, nelson pass made no bones about it, 15000uF! It may be an "overkill" but as i am learning right now, no such word exists on as far as high end audio is concerned. In fact, the author even mentioned that the circuit is essentially insensitive to jfet dc parameters, you can just plug in just about any jfet. This is another mark of a good circuit design.

But there is a little secret the author is not telling us -how he get an exceptional thd figure, hence, why he can put a 1 grand price in his preamp.

In audio work, a thd of 0.1 is a walk in the park. 0.01 requires some real work, and 0.001 requires a mighty skill and effort! At this level of performance, even what seems to be innocous placement of components will have a big say. And to think nelson achieve a 0.0007 without feedback is simply astounding.

I can imagine him starting his tests with a bagful of jfets, meticolously testing each jfet to select the golden part that is up to the job. It is not uncommon to get just a pair of "good parts" from a lot consisting of several dozens. That's why high end equipment costs so much. 

[Born2BeWired]

The human side of Born2BeWired talking... hmm you have a carbon-based side after all.

Waaaahah

Carbon based pero antimatter? 

[fantom_mayonaise]

Sir Born : Is it even possible to measure 0.0007THD ?

[Born2BeWired]

Audio precision build analyzers, such as model 2700, that can measure thd+noise down to better than 112db (that's around 0.00025%).

As you can see, nelson pass circuit is not very far from the residual distortion of the best distortion measurement equipment your money can buy.

[zrehtiek/keitherz]

must get that pcb design!

[tony]

looking at the 2SK170 datasheet, http://www.alldatasheet.com/datasheet-pdf/pdf/30581/TOSHIBA/2SK170.html

we can see that this device has indeed low noise levels.....the Pass Circuit is a source follower with a CCS bottom fet as load and sets the operating current of the preamp....at 5 to 10ma current, gm or transconductance is really higher compared to tubes, jfets are devices that are likened to tubes...

the 50 ohm output impedance did indeed suggested that gm is around 20mmhos... these kinds of impedances are really hard to attain with triodes....so the desiderata is choosing a device with high transconductance....

The stock parts are 2SK170's, LSK170's or 2SK370's, and you can use substitutes having Idss between than 5 and 10 milli-Amps and transconductance numbers from 5 to 30 milli-Siemens.

imho, very low levels of distortion was attained by using a very quiet device to begin with....and matching the top and bottom jfets ensures that the rail is divided equally between the two active devices....

[Born2BeWired]

Another thing i find interesting with the pass circuit is the hefty capacitance he used with the bias provider: 15000uF. This, in combination with the 10k + 10k voltage divider resistor, will result in an rc time constant of 75 secs! (5000 x 15000e-6)

Simply put, the circuit will require about 2minutes from turn on for the bias to settle to its design value. Its like waiting for a windows pc to boot before you can use it to its full. But this wont probably matter if used with valves amp, as it will take roughly the same amount of time to cold start these amps.

[tony]

i am one of those in favor of huge capacitance in power supplies, yes even for a tube amp, that is why you see them in my builds.....the quality of today's caps are hugely better than that of yesteryears, and cost is good too...

[efren]

... and I always advocate for the use of cap multipliers in any audio related project

[tony]

Another thing i find interesting with the pass circuit is the hefty capacitance he used with the bias provider: 15000uF. This, in combination with the 10k + 10k voltage divider resistor, will result in an rc time constant of 75 secs! (5000 x 15000e-6)

Simply put, the circuit will require about 2minutes from turn on for the bias to settle to its design value. Its like waiting for a windows pc to boot before you can use it to its full. But this wont probably matter if used with valves amp, as it will take roughly the same amount of time to cold start these amps.

as far as charging the big 15000ufd cap goes, the rc time constant is formed by the impedance of the transformer/rectifier circuit plus the   1.5k resistor leading to that cap....tube psu circuits are made deliberately with added resistors in the power traffo secondaries so as not to exceed tube ratings and avoid flash-overs....

[efren]

@ Eng.r Dacs,
ready for pik up na sa kanila ni sir tony yun parts, punta din yata sina sir Dandy at Noli mamya dun mga 3pm 2day, may kausap kasi ako kaya d ako maka-stay para sa mini EB 
____________________________________________________________________________________________________________________

Guys don't forget to take pics for "souvenir"