Ready to build my Gainclone

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello everybody,

It took me some time to start building.
Reason is first i wanted to compleet my VSPS phono pre.

http://www.geocities.com/rjm003.geo/rjmaudio/diy_pho5.html

I want to build a Regulated Valve Bufferd Non Inverting Gainclone.
In audio paths i use Riken resistors and the capacitors are BlackGate.
I want to have a custom made transformer.
A know a dutch company who makes very good transformers.

http://www.ae-europe.nl/

The idea is to run this amp with a regulated supply.
For the chips i want negative rail and possitive rail of about 40 volts 5 amps and regulated them to 30 volts with the lm338.
For the valve buffer i want a negative and possitive rail of about 40volts 1.5 amps and regulated them to 35 volts with the lm317 and the lm337.
For the filement i want 10 volts 1.5 amps and regulated this to 6.3 volts with a lm317.
I know AE-europe can build this transformer but before i order i would like everyones input on my plan.
Maybe some suggestions on this design.

Pleace feel free to comment on this.

Thanks
 
My only comment would be, if you are having a transformer built to your specs, why drop 10 volts across the LM338's? Keep it to 4 or 5 volts and you will need smaller heatsinks and waste less power. ;)

Good luck anyway - I am listening to mine now and it was worth the effort of building it and then some! :) (And that's without the regulated supply :eek: )
 
Nuuk said:
My only comment would be, if you are having a transformer built to your specs, why drop 10 volts across the LM338's? Keep it to 4 or 5 volts and you will need smaller heatsinks and waste less power. ;)

Thanks for the reply Nuuk.

The reason is that Pedja also drops 10 volts, thought that would be best.

Transformer output is +/- 30 volts AC so thats about +/- 40 volts DC.
Then regulate this back to +/- 30 volts DC.
So the voltage will stay +/- 30 volts DC with every load.

Correct me if i am wrong.
 
Output voltage regulation of the transformer basically depends on it's 'power capacity', the higher it's rated, the smaller the voltage drop under a certain load.

Now that you want to get the transformer custom made anyway, why don't you talk to the 'winder' about getting it done so that it fulfills your needs: low output voltage drop (less than 5V) with the load you specify (e.g. 5A for more than 1ms, >5A for <1ms).

They know how to make the transformer so that it drops the output voltage much less than, say, 7V under load.

That leads to another idea: what's the minimum voltage differential the LM338 needs?
I think e.g. the LT1083/84 is suited better regarding dropout voltage. It's specified for <1.5V drop at <7A load. And the part is not much more expensive than the LM338.

With a 30V secondary, rectified (-1V) and filtered (x1.414) voltage will be >41V. Constantly dropping around 10V under load will lead to serious heating of the regulators.

Just a thought ;)

Sebastian.
 
Nuuk said:
I am listening to mine now and it was worth the effort of building it and then some! :) (And that's without the regulated supply :eek: )

Hello Nuuk,

I thought you where using a regulated supply to.
Or at least you tried.

What did you think of the regulated gainclone sound?

Carlos states that it sounds way better then without regulation.

Thats the main reason for me to build this amp with regulation.
 
Nuuk said:
But I wonder if he would have dropped that 10 volts if he had the choice? What do you say Pedja? :)
Uncle Nick,

The schematic shows 24V secondary windings though in fact I have used 22V. It was like my insurance about possible lack of the regulation, and the supply was assumed to be able to drive both channels. Under 5A load, LM338 will drop 3V. Add at least 2V, and yet better 3V for the ripple you will have at the reservoir caps and you need rectified 30V for 24V output. Add Volt or two for rectifying diodes’ drop, it is 32V peak, i.e. about 23V RMS. (Now I should remember why I put 24V, probably “just for case”.) In practice, two Volts less is acceptable, but I anyway wouldn’t, under any circumstances, go with less than 5V of the headroom, and that assuming each channel has its own regs.

BTW, at the moment I posted that supply to the site, I couldn’t even dream about this number of GC supplies made that way. Though I do not have problems with anything written then, if I could be that long-sighted, I’d be more careful/detailed/precise about some things. Yet, during the time, plenty of info on the topic occurred elsewhere around (and you made nice contribution), so it is covered anyway.

Pedja
 
sek said:
That leads to another idea: what's the minimum voltage differential the LM338 needs?
I think e.g. the LT1083/84 is suited better regarding dropout voltage. It's specified for <1.5V drop at <7A load. And the part is not much more expensive than the LM338. Sebastian.

Hello Sebastian,

Thanks for the reply, so it looks like it's better to build it with the LT1084.
But the LT1084 are more expensive than the LM338.
LM338 is 4.32 euro and the LT1084 is 13.04 euro.
Not that this is to expensive.
So if the transformer is +/- 25 volts AC the DC wil be about +/- 35 volts.
Than regulated this with the LT1084 to +/- 30 volts.
Is this fine or do i need a higher transformer output voltage?
Can i use one LT1084 for the positive and one for the negative rail for both channels?
 
Thanks Uncle Pedja for clearing that up. So it is a good idea to leave a bit more headroom for the regulators and accept that we need adequate heatsinks!

ebijma - the 30 volts output from the regulators isn't written in stone ;) A few volts less won't affect the operation of the GC.

So, if you have +/-35 volt rails before the regulators, you could regulate down to +/-27 volts, still leaving enough headroom as described above by Pedja.

But if you are having your transformer built, why not specifiy 27 VAC secondaries (although this will be a bit high if for any reason you decide not to stick with the regulated supply later)?
 
Nuuk said:
So, if you have +/-35 volt rails before the regulators, you could regulate down to +/-27 volts, still leaving enough headroom as described above by Pedja.

But if you are having your transformer built, why not specifiy 27 VAC secondaries (although this will be a bit high if for any reason you decide not to stick with the regulated supply later)?

Thanks Nuuk,

Transformer with +/- 27 volts is an good idea.
This gives me about +/- 38 volts than regulated this to 30 +/- volts.


carlosfm said:
One thing is using what you have.
But if I was going to order a custom-made trafo for using at +/-30v with LM338s, I would order 2*27v.
The ideal situation is between 5 and 10v voltage drop.
Point to the middle, that should be fine.:D

Thanks Carlosfm,

Good idea.
But can i use one lm338 for the positive and one for the negative rail and connect both channels?

So what regulator is the best?
I have two lm338 but if the lt1083/84 are better i would rather order these.
I think i will stick with the lm338 for now.
 
Nuuk said:
ebijma - you will gain a lot from reading the previous threads on regulated supplies for GC's (on this forum).

We talked about things like adequate heatsinking, use of LM338T compared to LM338K and most of the other issues that you will need to know. ;)

Hello Nuuk,

I know i've bin reading allot about it on this forum.
But my questions still remained.
So i placed a thread of my own.
My question wasn't about heatsinking and not about the use of LM338T or the LM338K.
But thanks anyway.
 
yldouright said:
ebijima

I would like to know how you built the phono preamp for your gainclone? I am thinking about integrating a phono section into a gainclone and running the whole thing in an old Kenwood amp case. Are you putting the phono stage in a separate case?

Hello yldouright,

The phone pre is not specialy for the gainclone.
I build it in a separate housing.
Read all about it here....

http://www.diyaudio.com/forums/showthread.php?threadid=33555

Its a simple but very good sounding pre.
 
ebijma

The link "has exceeded its bandwidth" but I will check it again. In the meantime, why did you decide to build it in a separate case instead of making it a stage within a larger gainclone? From reading Kuei Yang Wang and others, it seems that the shorter the signal paths the better so I am assuming that integrating the phono stage would have resulted in better sound. Was this a power supply issue?
 
yldouright said:
ebijma
In the meantime, why did you decide to build it in a separate case instead of making it a stage within a larger gainclone? From reading Kuei Yang Wang and others, it seems that the shorter the signal paths the better so I am assuming that integrating the phono stage would have resulted in better sound. Was this a power supply issue?

Hello yldouright,

I build mine for use in the living room, i have a jvc set there and i wanted to connect a turntable to it.
The gainclone im going to use in a other room.
Thats a kind of listing room.
In the future im going to build another phono pre and a non oversampling DA converter for use in my listing room.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.