diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Pass Labs (http://www.diyaudio.com/forums/pass-labs/)
-   -   Yet Another Ono Thread (http://www.diyaudio.com/forums/pass-labs/110209-yet-another-ono-thread.html)

wayne325 17th October 2007 02:46 AM

Yet Another Ono Thread
 
Hello all,

I'm beginning the road to an Ono amp. I have a few ideas and
questions though, I haven't seen these really answered,
though one can certainly read about Onos for hours on diyaudio
(ask me how I know).

So far I've made the power supplies and I've got separate
chassiseseses for the PS/amps. I added a lot more filtering to
the PS, total supply filtering is 120,000 uF. Maybe takes a
while to get going...

1) has anyone tried to use surface mount components? Really
I'm only thinking of the resistors. Should be better due to
lower parasitics and shorter traces but maybe they sound
"bad", or perhaps not as good as the RN60s that are often used
and that are often used by Pass Labs.

I'm thinking perhaps Vishay/Dale CRCW series, 0805 or 1206.
Mouser has them in unit quantities.

There are only, what, 50 or 60 resistors per board, I can solder
that by hand. I have access to a good magnifying glass
intended for such purpose where I work.

2) If going with thru-hole resistors, Mouser doesn't have a
large selection of RN60. THere are lots of values in the Ono
schematic that are not stocked. BUt there are Xicon and KOA
SPEER 1% metal film resistors. Also IRC 0.1% metal film,
might be good for the RIAA. Anyone have an opinion on those
pro/con vs the RN60?

3) For the small value caps - the 100, 220, 330, 680 pF caps,
what do people like? Silvered Mica? Ceramic?
Tantalum is right out, they have a bad rap where I work for
realiability reasons.

4) To get 2 matched quads of the 2SK170s, how many should
I buy?

5) I'm planning to add 10 uF of film cap in parallel with C12, C20,
and C36. ANyone else do this? Would it be worth it to put
some relatively inexpensive film caps in those positions?

6) The famous LED that sets the voltage across R53... does
anyone have a Mouser or Digikey part number? It seems there
has been a lot of discussion about it but I can't find anywhere
that a part number is posted.

Thanks for any ideas / help...

W.

luvdunhill 19th October 2007 12:52 AM

Re: Yet Another Ono Thread
 

1.) I'm ordering some Koa RK73A SMD for audio for another project (check their website) but haven't tried them yet, but they are supposedly excellent... I think Goldpoint uses them in their steppers and they supplanted the Vishay-Dale SMD resistors in their lineup.

2.) I used PRP and Holco through-hole resistors with Roederstein Resista for the RIAA section. I purchased the former from PartsConnexion and the latter from Michael Percy Audio. I took plenty of time to carefully match these where it made sense, using some fancy equipment at work that could do four-wire measurements.

3.) I tried a lot of different caps and liked polystyrene (Xicon and some NOS from Ebay) as well as the Panasonic ECQP(Z) the best. I tried micas and liked them the least. In the end I went with the Panasonics. For the RIAA section I used Relcap RTE (Polystyrene Film and Foil) a MultiCap Film & Foil Polypropylene (PPFX) and a silver mica. I probably still have all the needed values in a few of these options around here

4.) I purchased 75 and got two well matched quads. My actual measured values are posted in one of the Ono threads. Also, there is another pair that should be matched as well, here I tried a bunch of transistors and decided that the BC550C sounded better to my ears. If I had to do it again, I'd also look at matching the IRF devices in the balanced section, as the actual ones in my finished Ono very wildly and I think this is why they run so hot (maybe not, just a guess).

5.) I use Black Gate NX capacitors and Black Gate standard grade and didn't at all like the sound of them bypassed. As for the 10uF output capacitors I used Clarity Caps bypassed by Roederstein MKP 1837. These sounds quite nice. I tried some of the cheaper Mundorfs and some Wima 10uF and prefered the above configuration. Maybe one day I'll try Auricaps, but the entry price was too steep for me.

6.) This guy is hard to find, I purchased 100s of LEDs from Mouser and Digikey and all were around Vf = 1.7. IMHO, don't bother looking, as you won't find them. Better instead to tweak the resistor values then mess with finding the right LED. However, if you want to go down this path, the final LED I used was pulled from a computer and was a red T1 sized from a 3.5" floppy drive. One idea might be to use two infrared LEDs in series. IMHO, the correct value helps the sound tremendously. Here a .1 V makes a huge difference in the heat and the sound and makes sure that the gain switches work (I had this problem before I found the right LED and this fixed it right up).

Take care!

edit: I noticed you're up in Dallas. If you're ever in Austin, holler at me and you can take a listen to my Xono ;)

wayne325 19th October 2007 03:44 AM

Thanks for the reply... I've got a lot to look into now.
I looked up the KOAs - where does one buy the small qtys a
person like you or I would buy?

luvdunhill 19th October 2007 01:14 PM

Quote:

Originally posted by wayne325
Thanks for the reply... I've got a lot to look into now.
I looked up the KOAs - where does one buy the small qtys a
person like you or I would buy?


I couldn't find anywhere as well, so I used this as justification to request samples from KOA ;) They have not arrived yet, and that was 2 weeks ago, but I did get some sort of order confirmation.

wayne325 19th October 2007 01:31 PM

niiiiiiiice

wayne325 27th October 2007 10:53 PM

Update:
I've ordered all the silicon bits needed except the stuff I can
get from Digikey easily. I'll have 2SA929 instead of 2SA991.

The rest is being entered into spreadsheet as a mix of
PRP, Holco, as the preferred resistors and RN60 where I can't
get the right values. RIAA is only exception since they seem
to be little used values. I'll look into the Rodersteins for the
RIAA but there are KOA as I wrote before. I decided to go
with thru-hole resistors as they are easier to handle.

I'm still a bit confused about that pesky LED that biases
Q16 and Q17. I thought the Veb of a BJT operating in linear
region is a diode drop = about 0.7V. If we want 1.0V across
R67 and R53, then it should be fine given the base current
through R22 and R55+R54 should be near 0, and so the
voltage drop across those resistors is also near 0. So what
am I missing since the 1.0V + 0.7V is equal to 1.7V which
is the Vf of the LEDs you measured? It seems to me like
1.7V is what we want for the LED. I'm going wrong somewhere.
Does anyone know where my error is? More to the point, what
Vf do I want in the LED D5 in order to get 1.0V across R67 and
R53?

wayne325 1st November 2007 12:31 AM

I'm hoping someone can help me out with the LED part and the
associated circuitry.

I believe that Q16 is acting as a current source for the 2SK389
differential pair. The biasing is done by R67 which is 475 ohms.
If we want 1.00V across R67, that means the current source
is drawing 2.10mA. If I look at the curves for 2SC1844 (Q16),
I see that Vbe for Ic = 6.0V (the only curve available) is right
around 0.6V.

If we assume that Ib = 0 then the drop across R22, the base
resistor of Q16, is 0. So to have 1.0V across R67, I need
to have a LED with Vf = 1.0 + 0.6 = 1.6V.

And that takes me to the problem. The Mouser and Digikey
are full of LEDs that have Vf as low as 1.7V, but none as low
as 1.6V.

So.

What I have to do is get a LED with the lowest Vf I can find,
and then change out R67 so that the current through it
is 2.10mA. Assuming a 1.7V LED, I'd need to increase the
voltage across R67 by 0.1V in order to keep the bias current
at 2.10mA, and so I'd need 475 + 47.5 ohms = 522 ohms.

The point is, I have to bias the differential pair to 2.10 mA.
If the collector voltage of Q16 is up 0.1V or so, that doesn't
much matter at all.

The Q17 side gets the same treatment only I change R53 to
about 165 ohms or so.

THe other thing I can do is adjust the current of the LED by
changing R5 but I have no curves for any LEDs so I don't know
how much I can affect things there and I don't know if lowering
the current will allow the voltage to drift or be less stable.
I guess that's what C27 is there for though.

Did I get that about right ?

(Dammit Jim, I'm a digital guy, not an analog guy)
:D

spzzzzkt 6th December 2007 03:39 AM

has anyone tried this LED?

HLMP-K155 Vf=1.6 typ. , 1.8 max.

digikey pn# 516-1312-ND

wayne325 6th December 2007 05:38 AM

I've bought and measured from Mouser this part:

696-LX5093SRC/DV

But I haven't built the amps yet and I don't know what current
I need. With a target of 1.7Vf, I need about 12k resistor to
bias it. I'll see what happens. Oh ya, I'm going to layout a
trimmer so I can play with the level current.

If I want after it's set I can measure and replace by a fixed resistor.

Thanks for the part#... if I'm way off I'll try it.

W.

spzzzzkt 6th December 2007 10:42 AM

good to hear. just thought that point six in post two indicated that the 1.6V LED was a unicorn when mouser, digikey, rs-components et al do stock a LED that should be close to that figure.

I've had a set of Ralph Stens' (X)Ono boards since early january this year and have been thinking about finally pulling the finger out and building them up. Still need to pick up the psu board, but if i wait til the 26th I can order on the 12 month anniversary of ordering the channel boards ;)

Still checking out parts sources - at this stage I think I'll go for Vishay VSH-J bulk foils for the RIAA resistors and PRPs for the rest.


All times are GMT. The time now is 03:47 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2