diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Pass Labs (http://www.diyaudio.com/forums/pass-labs/)
-   -   Aleph ONO working - belated thanks to pquadrat (http://www.diyaudio.com/forums/pass-labs/27318-aleph-ono-working-belated-thanks-pquadrat.html)

GaryB 1st February 2004 01:53 PM

Aleph ONO working - belated thanks to pquadrat
 
2 Attachment(s)
Hi all,
I finally got my Aleph ONO phono stage working. I used the pcb boards that pquadrat laid out and offered up to the DIY community. More details on the boards are found here:
http://www.diyaudio.com/forums/showt...threadid=10326

I did change the design slightly and didn't build in all the functionality that is possible. I didn't bother with the relays since I intend to leave it on all the time. I also didn't need balanced operation, so I eliminated those parts. I also got rid of the output coupling cap per the modification suggested by Peter Daniel.

Other than that I built it as designed. After a bit of listening I decided that I didn't quite like the MC stage and went to the simpler Pearl type 1st stage. This means taking the output directly from R40 and bypassing the last 2SK170 and the electrolytic coupling caps. This cleaned up the mids but slightly hurt the highs. Thinking that this might be related to the moderately high output impedance of the stripped down MC stage, I tried converting the input of the MM stage from cascode to bootstrap cascode.
To do this I removed R17 (10k to ground) and inserted a low noise zener (5v LM336, bypassed with 0.1uf pp cap) from the bases of Q1,2 to the sources of Q5. To get the biases right, I changed R3 to 29Kohm and R7 to 332ohms. This puts about 0.9ma through the zener and keeps the bias of the input transistors unchanged. This change worked very well and restored the nice highs of the original circuit and left the midrange clarity of the simplified MC stage and its what I'm listening to now.

Oh one last hint - I originally used a silver mica cap for C3, which is critical for the high frequency stabilization of the circuit. This didn't seem to cause a problem with the original circuit but with the change to bootstrap cascode this was problematic. Changing to a ceramic returned the circuit to its original well behaved state, suggesting that silver micas don't do too well in the Mhz range.

---Gary

GaryB 1st February 2004 01:55 PM

2 Attachment(s)
Here's one more picture of the ONO stage in-situ. Obviously I need to build it into a nice box for tweaking purposes its just sitting on an old plinth that I used to use for my turntable.
---Gary

GaryB 15th February 2004 03:12 PM

Schematic
 
2 Attachment(s)
Quote:

Originally posted by GaryB
I tried converting the input of the MM stage from cascode to bootstrap cascode.
To do this I removed R17 (10k to ground) and inserted a low noise zener (5v LM336, bypassed with 0.1uf pp cap) from the bases of Q1,2 to the sources of Q5. To get the biases right, I changed R3 to 29Kohm and R7 to 332ohms. This puts about 0.9ma through the zener and keeps the bias of the input transistors unchanged.
---Gary

There hasn't been much response to this, but for completeness I thought I'd include a schematic to make the details of the MM input stage modification clearer.

---Gary

Stefanoo 6th December 2008 12:08 PM

this post is old...but very interesting.

Are you still using your ONO?
Did you make any other change to it?

analog_sa 6th December 2008 03:03 PM

Interesting indeed. Sadly, the technical part does not make a lot of sense, at least to me.

Stefanoo 6th December 2008 04:49 PM

What do you mean analog?

I thought of posting this thing in accordance to what we were saying on the other 3d.
He did the same thing but interesting he found that changing the cascode configuration changed the sound.
I personally don't see how a bootstrap cascode might fix a impedance problem though, if there was an impedance problem...but it is certainly my lack of knowledge so I hope Gary will post again.

Sadly that an interesting subject with this didn't catch any attention.

GaryB 7th December 2008 12:06 AM

2 Attachment(s)
Quote:

Originally posted by Stefanoo
this post is old...but very interesting.

Are you still using your ONO?
Did you make any other change to it?

Stefano,
Thanks for your interest. It's been almost 5 years since I built the ONO so it's surprising that someone found this old message. I'll try to answer your questions as best I can.

First, I still have the ONO and it works fine and I use it from time to time. My main phono stage is a highly modified NYAL Super It tube phone stage. It uses an FET/tube input stage, passive RIAA, followed by another FET/tube gain stage, and a MOSFET source follower output stage. It sounds slightly better than the ONO but the difference is small.

The only other tweaks I made to the ONO were to upgrade the RIAA network parts. I changed the resistors to TaN film SMD resistors from IRC. You can see them here:

http://www.mouser.com/catalog/636/621.pdf

I also managed to find teflon film caps and use of these caps made a big difference in sound quality. I don't have a good source to recommend. I found the 0.01uf cap from someone here in diyaudio and the other caps on ebay.

Quote:

Originally posted by analog_sa
Interesting indeed. Sadly, the technical part does not make a lot of sense, at least to me.

Quote:

Originally posted by Stefanoo
What do you mean analog?

I'm also not sure I understand the comment from Analog-sa. He probably thinks that the input impedance of the MM stage is set by the 47k resistor at the input to the 2SK389 and the output impedance of the MC stage doesn't matter. Perhaps he's right but I don't think so. If you look at the datasheet for the 2SK289, you'll see that the input capacitance changes as Vds changes. I've attached this figure below. The bootstrap cascode configuration keeps Vds constant and eliminates this tendency for Ci to change with signal. I think this effect is more noticeable when you drive this with a high output impedance.
Even if the technical explanation is flawed, the sound quality is definitely improved by this change. It's not too hard so why not try it out.

---Gary

analog_sa 7th December 2008 06:53 AM

Quote:

Originally posted by GaryB

He probably thinks that the input impedance of the MM stage is set by the 47k resistor at the input to the 2SK389 and the output impedance of the MC stage doesn't matter.


Hi Gary


Your suspicion is correct. You may be right with your explanation; it's just impossible to prove you are not compensating for the different distorion spectrum after the removal of the NFB stage. It would be interesting to listen to a MM cartridge with and without bootstrapping.

Stefanoo 7th December 2008 08:10 AM

thank you Gary.
This is such a good technical comment.
Thanks a lot.

I will try this out for sure.


With regard to resistors i am planning on using Naked Vishay S102 that are premium resistors.
For cap, still premium one, with Auricap Film Caps.
Electrolytic will be Black gates STD for the smoothing caps and F for the RIAA 2 X 220uF (which i will replace with just one 470uF).

I will use the monolithic for the cascoding of the @SK381 instead of the 2 transistors and same thing for the current mirror.

I am planning on using the 5V reference voltage AD586 (it's 2ppm with an extra low residual noise) on both current mirror and bootstrap.
- if you want to comment this last modification i don't know if you tried to replace the led with a good voltage reference-

One question:

you said you had to change the bias resistor R3 (ok) and R7 (? did you change this one? did you mean to say R67? this sets the bias on the differential amp along with the voltage reference. which by the way what does R7 do? i find it on the main diagram for active RIAA so i guess is part of the equalization network, right?)


Thank you Gary for your relevant post.

analog_sa 7th December 2008 08:59 AM

I find the Auricaps not so amazing. At low voltages BG-N are a very viable alternative and much easier to accomodate on a board. Won't even mention the price.


All times are GMT. The time now is 10:21 PM.


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