[rookster]

I will need 2 mono units for my application. 

I am leaning towards the 45 step Shallco's with Holcos in most spots, with Caddocks added into the common listening settings.  Or 24 step Elmas.  In both cases I will install a regular pot first and then detemine the resistance at my range of listening levels and select the pots and step sizes from this.

I figure if people, including myself are upgrading caps and resistors, then it makes sense to convert the pot to 2 high quality resistors (I will go ladder or maybe shunt) and implement a stepped attenuator.

[jules]

rookster,

A 45 step Shallco is a seriously amazing looking monster but it also comsumes huge volumes of resistors. particularly if you you're paying for more a 50c each!

I've used an Elma full ladder with Rodersteins but if you do a shunt, you can put the worlds most exotic resistor in the signal path and use lesser types elsewhere [unless of course the fact that a circuit is circular might be an issue  ].

23 steps with a well balanced choice of values is totally fine as far as changes in volume go.With Aksas changes in volume equate more to changes in amount of information rather than volume [if that makes sense] so that feeling that a setting is too loud/soft isn't nearly as pronounced.

[rookster]

Perhaps my question should be this;

When using a GK-1, with an Aksa 55N+ and Aksonics, would 1 db steps in a volume control be OK?  Or are bigger steps OK? 

My desire for a 45 step Shallco is slightly offset by the cost and time required to populate it with resistors.  But it would give a lot of 1 dB steps over the predicted range.

It is going around and around in my head and I need to get back to the reality of actually building my 55N+ 's and GK-1.

[AKSA]

Michael,

Most audiophiles are pushed to notice 1dB increments.  1.5dB is acknowledged as just discenible, 2dB is fairly apparent, 3dB is obvious.

A properly designed 24 step SA will have coarse steps for the first two, fine thereafter until the last three, which are loud, incredibly loud and overload and normally carry 3-6dB intervals.

I personally don't believe a 45 step is necessary, and it's certainly a much more expensive and complex solution - and very fiddly to make.  Go for the 24 step system.

Cheers,

Hugh

[DSK]

Michael,
My TKD has 40 or 41 steps that are 1db increments virtually across the board. I've never struck the situation where 1 step, or even 2 steps, was too coarse. Usually, to turn it up just a fraction, I use 2 or 3 steps.

So, I'm sure you will be fine with 2 to 2.5db increments in the main range.

[stvnharr]

The 2 decibel increments on both my attenuators are just about right.  And in my system w/o preamp, the 2 decibel steps go all the way to the top, which is nice as I have a couple recordings done at low recording level and I can even play one without attenuation.

My old BAT preamp had .5 decibel increments and I normally adjusted things in units of five, as the volume on the led display was in numbers.

[rookster]

Hugh, DSK & stvnharr,

Thankyou for your replies and for a reality check.  This was what I was thinking about at the start.  Perhaps I should have asked this at the start.

Costwise and timewise, a 24 step ladder is more practical.  I will look more into this as a practical solution.

Thankyou all for the help.

[franky]

Hello DSK,

I e-mailed you personally for more info on the Bob Crump TKD mod. Don't know if you visit the forum often, so I gave it another go through this thread. If anyone else could help, that'll be great too. You can contact me through the forum. Thanks in advance.

Interesting thread btw.

Regards

[DSK]

Hi Franky,will dig out the info and send it to you later today. Cheers!

UPDATE: Franky, you have mail! 

[SamL]

For those who looking for a remote valume control with source selection. Just found that Jos van Eijndhoven release a Audio Volume Relay Attenuator with IR Control. A 64-step logarithmic stereo attenuator. The IR recever can control multiple board to provide multi-channel volume control if required. All at a very competitive pricing.
http://jos.vaneijndhoven.net/switchr/design.html

Have fun,
Sam

[AKSA]

Thanks Sam!!

This fellow is a truly impressive designer.  This is top flight technology, really good.

6 relays for 64 steps is extremely economical and well sorted.  1.1dB steps are all you would ever need.

I notice he's a confederate of Guido Tent, another clever man.

Recommended.....

Cheers,

Hugh

[Felipe]

Hugh,

It really is a very apealing design, and surely looks a nice option to include in the GK-1.

Jos warns about the output resistance. With default resistor values it shows 22K CONSTANT output . Remember me, is this OK with the AKSA's ?

quote from Jos site :

Traditional potentiometer usage provides a constant input resistance, and variable output
resistance. The switched-attenuator in this design in principle works the other way around: it
has variable input resistance and constant output resistance. (However, due a specific
design trick, the default attenuator resistor values provides a lower output resistance for the
lower half of the volume steps.) You could use this attenuator as a complete passive preamplifier.
It greatly depends upon your power amplifier whether the relatively high output
resistance is OK for your system.

Felipe

[AKSA]

Felipe,

This would be just fine, because the output impedance of the SS section of the GK1 is 30R, so any load more than 300R would be fine;  in this case it would be of the order of several kilohms.

The 22K output impedance of the potentiometer would also be fine.  The input impedance of the self-biased cathode follower used on the GK1 is 4.3 megohms (calculated), and so any source of 430K output impedance OR LESS would be terrific.  In this case it's 22K, which gives a favourable impedance ratio of 196 - a rhapsodic conjunction, you might say.....

However, there are three issues.  They are, in order:

1.  cost
2.  cost
3.  cost

As you can see, the cost of a single potentiometer function using this dreamboat design is around 200 Euros, and this is not the sort of price to gladden the heart of DIYers.  My thinking on this, and Bjorn Kolbrek's relay driven pot was available when I designed the GK1, is that the better option is always to optimise the working environment of the pot so that its significant disadvantages can be minimised, and this was, and will remain, the solution I chose.  A good pot was, and remains, a LOT cheaper than this elegant relay solution!!

Cheers,

Hugh

[rookster]

Hugh,

The resistor/relay design is quite impressive.

As a simpler version of a stepped attenuator, could I use a resistor across the "pot input" and "pot output" on the GK-1 board and then a log pot to ground.  I think it is called a series shunt design.  I realise the impedance varies, but not over a great portion of the range.

I am trying to simplify the design so save as much time as possible.  I am having a severe motivation problem and am trying to cut down down my build time as I really don't have time to build what I would have liked.  I know I will be happy with the result but just desperately want finish it and then be able to listen to music.

[AKSA]

Hi Michael,

As a simpler version of a stepped attenuator, could I use a resistor across the "pot input" and "pot output" on the GK-1 board and then a log pot to ground.  I think it is called a series shunt design.  I realise the impedance varies, but not over a great portion of the range.

Yes, no problem at all.  I'd suggest a 15K series resistor with a 20K pot.  This will give a max throughput of 20/(15 + 20), or 0.57, which will reduce gain considerably, but you can replace the 47k shunt resistor in the feedback network with 27K to help here......

I am trying to simplify the design so save as much time as possible.  I am having a severe motivation problem and am trying to cut down down my build time as I really don't have time to build what I would have liked.  I know I will be happy with the result but just desperately want finish it and then be able to listen to music.

Don't get too hung up on 'improvements'.  This is a highly optimised design, and further fiddling could go wrong on you, but I certainly understand the impulse just as I understand the motivation issues.  Gratification is very important in DIY, and you want to have bite-sized chunks, not too onerous and involving.  A digitally controlled, relay switching pot, of course, is a very big project, not to be treated lightly, at least as big as the GK1.

Cheers,

Hugh

[JosvE]

Hi all,
I received a tip by email that my relay attenuator got mentioned in this thread.
(http://www.vaneijndhoven.net/jos/switchr/design.html)
Thanks for your positive remarks!
I see interesting and good quality discussions here!
Over my past few years in audio design I learned that using fewer electronic components in simpler designs typically lead to better result, together with the need for first-quality components.

My attenuator design (in my experience) can sound great in some set-ups as just passive pre-amp. In fact, I recently listened to a system where my 100Euro PCB sounded significantly better than a commercial 10,000Euro preamp. In general however, I agree with an earlier poster that the larger output impedance (of any passive) in combination with external audio cables and some amplifier loads can audibly spoil some quality (fine highs, sound staging).
So in general I would feel safer to recommend (simple but good quality) active buffering, and/or building your DIY integrated amp with shorter internal cabling.

Greetings,
Jos

[Rom]

Hi Jos,

Welcome to AKSA forum.

Very Interesting product you have there. I just sent you an inquiry regarding your product.

Maybe you could enlighten as well the forum on how we could make an order and mode of payment.


Hugh,

my apology for highjacking the thread.

Cheers
rom

[AKSA]

Jos,

Thank you for your post, and for your company.  You are amongst friends here.  I quickly read your website, and was impressed with your technology, and interestingly, your humanity as well.  I even read your brother's website, he too is an interesting, clever man!  I hope you return often;  we like to have open design discussions here, and it has proven very, very fruitful on countless ocassions.

Your design is an absolute corker.  Congratulations!   
 
Rom,

Don't mention it!  Highjacking threads is so much more beneficial than lowjacking - ah, we must keep our technical discussions as high as we can make them!!   

Cheers,

Hugh

[rookster]

Hugh,

Thanks for your reply and resistance values. 

I am not trying to alter the electrical design, more to simplify my mechanical design.  I had originally designed 2 dual mono chassis.  But the build time and space has confined this to a single chassis, but it will still look quite unique while matching my dual stereo chassis.

Between yourself and Bluesky, I should be back building soon.  Like most people I have so many projects that I have come to a standstill trying to think of what to do next and how to proceed.

[AKSA]

Ah, yes, Michael,

A case of PBA!!  (Paralysis by Analysis - I do love a nice TLA, don't you?)

At its lowest level, 'Which chocolate bar?', at it's highest, 'Which topology?'

Cheers,

Hugh