Modelling PSU?!

[flyinglemur]

Hi,

Just wondering - is duncanamps' PSU designer odd?

I've tried to model with a 5V4 regulator, 7H choke and 15uF initial smoothing cap. Now, how do we proceed?

It won't let me put in a current tap with tank cap right after the inductor, even though thats how boozhound lists it to be, also will tell me that most things are an 'illegal combination!!'

Prolly an easy one, me having an off night!!

12SN7 driver and 6V6 outputs


L

Trying to model this

[IMGDEAD]http://www.boozhoundlabs.com/howto/images/schemo-6.gif[/IMGDEAD]

 

[Yvesm]

You cannot insert two similar elements (2 caps, 2 current taps, nor a R and a CT) in parallel.
Nor remove an element that provoque this situation.
For exemple, an LC filter between two loads (either R or CT) can't be removed unless you remove one of the loads before.
Sometimes, you cannot insert a CT after a cap, but only before it.
The same result, obviusly

Yves.

 

[flyinglemur]

Done!!


Cheers for the help, trying to bring the ripple below .05v


L

 

[gootee]

I like using LTspice. I have a few examples, here:

http://www.fullnet.com/~tomg/gooteesp.htm

Tube models are available, from other sources (including Duncan's site, I think).

 

[flyinglemur]

I've managed to extract some sense, cheers for the help, it's just a matter of getting used to it! What's an acceptable ripple?

 

[Yvesm]

. . .
What's an acceptable ripple?

It depends on the amplifier design/topology

A push pull can be completly quiet with up to 20 V peak to peak on the plate supply.
An SE penthode may accept up to 5 V on the plate supply if the screen is "clean".
An SE triode may produce unnaceptable hum in hi efficiency speaker with just 1V p to p

What are you building ?

Yves.

 

[flyinglemur]

SE. I have 50mV P-P



L

Tap for the 6V6s
[IMGDEAD]http://web.1asphost.com/Skateloughrea/I1.JPG[/IMGDEAD]

And for the 12SN7
[IMGDEAD]http://web.1asphost.com/Skateloughrea/I2.JPG[/IMGDEAD]

 

[flyinglemur]

I've got the ripple to even more acceptable levels on b+ since...

These are only preliminary, when I've got the rectifier and set up the circuit, I'll do more precise modelling

L

 

[flyinglemur]

with a little fiddling, 50mV became 2mV.

Those are acutlaly 90mA and 10mA draws...but sequentially kicking in after a 1sec


[IMGDEAD]https://img.photobucket.com/albums/v715/Psychobiker/traces.jpg[/IMGDEAD]

 

[ErikdeBest]

Hi Psychobiker

Maybe you won't need it, but it is interesting literature anyway:
'lowering SE amplifier noise' (in cathode biased amplifiers) by Broskie

http://www.tubecad.com/april99/page2.html
http://www.tubecad.com/may99/page3.html

 

[flyinglemur]

Cheers for the link, which is essentially what I'm doing - building God's own PSU . Brute force and ignorance ftw

L

 

[Yvesm]

But, is it really necessary at the plate of the 6V6 ?

I don't think so.

Yves.

 

[flyinglemur]

That level of overkill? No, it's not necessary...but what's the harm

L

 

[rdf]

Use a realistic value of DCR for L1 and find out.

 

[flyinglemur]

Heh, that's a moot point - remains to be seen once I fire it up, couldn't find a datasheet anywhere for the life of me :S. Just tried a DCR of 50ohm, bugger all changed really. But again, as I say, will report on the 'real world' functioning of the thing. Radionics (who made it), discontinued the thing aaaages ago (<2000)


L

 

[CarlyBoy]

Why I don't like PSU

What I hate about PSU II is you never seem to get the same results twice. Here's what it tells me when I plug in the same values Psychobiker:

[IMGDEAD]http://img358.imageshack.us/img358/104/psuhe6.th.gif[/IMGDEAD]

 

[rdf]

Did you step the load currents as per post #9 above?

 

[CarlyBoy]

I'm afraid you got me on that one - I don't know. In fact I may not know what the h*ll I'm doing. Should I have done anything different than selecting the same values..? I'm still learning this program...

 

[rdf]

Right click on the current source load and select 'Edit'. The dialogue below appears (minus red comments) allowing the current load to switch values between 1 and 2 after 'seconds'. The reason you see no ripple is probably a very low 2ma load.

 

[flyinglemur]

Cheers, RDF...it slipped my mind the first time that it was acutally a sequential load - but even the slightest deviaton from immediate full load sorts out any over current warnings.


CarlyBoy - You can watch the ripple get a little larger when the load kicks in...but it settles to a healthy <5mV in mine

L