|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
24.98.146.44
Is there a prefered order of regulation after a raw B+ ?
I want to use a tube based voltage regulator and I want to use a cascode CCS.
Is the CCS ahead of the Vreg into the tube plate or the Vreg ahead of the CCS into the tube plate the way it should be?I would think ( with my limited experience) that the CCS would come after the Vreg so that there is less of a voltage drop across it than if it were before.
Raw B+ 195 Volts, Vregulated 150 Volts. Current after CCS 5ma.
The tube would be the venerable #26 DHTAdvise?
Thanks
JD
Edits: 05/30/12 05/30/12 05/30/12Follow Ups:
CCS first, feeding the VR tube. Voltage drop across the CCS is a good thing.It lowers the capacitance of the MosFet. Just be sure to use big heat sinks.
Set the CCS for the current you expect your circuit to draw plus about 15ma. for the VR tube.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 05/30/12
Thanks Tre.I'm not using the VR tube, I'll be using a regulator based on Mike Van Evers ECL82
Should I assume that the same placement of the CCS is still appropriate in your example for the same reasons?
Jeff
Edits: 05/30/12
I have read the text in the link regarding the ECL82 regulator, but I am not able to find the max current possible with this regulator. Any thoughts?
about 15ma
JD
If you are going to place the CCS after the series regulator you might as will use it as the plate load on the 26.Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 05/30/12
plan is also to use an output transformer (Lundahl stepdown) for output
JD
The feed side of the OPT should be driven by a voltage source, not a current source. The CCS has no place there.
--------------------------
Buy Chinese. Bury freedom.
Jeff, in this case CCS is not necessary.Naz
Edits: 05/31/12
ok chefs.... now I'm more confused than when I first asked the question.I want the voltage regulator to regulate the voltage from 195 volts or so down to around 150. I want the CCS to control current to 4ma-5ma.
I need the OPT to step down the output (4.5:1).I've been "all over" the #26 thread at diyAudio and have seen many variations of the placement of the CCS and Vreg using an OPT. I do not want to use a single solid state based Vreg, but I do want to use the Ixys based CCS, that's why I'm breaking down the two functions to the Van Evers regulator and the separate CCS.
Some on the aforementioned thread have opted for a single shunt regulator that accomplishes control of both current and voltage. I have seen examples of regulator tubes with a CCS before, after and before and after the regulator.So the more advice I get, the more it contradicts previous advice and what others have done that worked.
JD
Edits: 05/31/12 05/31/12
OK, let's start from scratch by defining the goal. I see two contradictions here. On the one hand you want the best result but on the other you seem to want to use all of the components on hand and this is causing some confusing, leading to the different responses, most of which are valid in their own way. Personally, I think too many components can lead to overkill.
Summary as I see it. If you want to use all of the components on hand then Dan's advice isn't bad. It can lead to a few other complications if you are not careful but a good solution nonetheless.
If you just want a good sounding buffer then here are some options (my opinion only):
(a) you don't really need a regulated supply but sound wise, all tube is good
(b) CCS on the plate with cap OP coupling will give you a nice result, fairly neutral top to bottom but the CCS can add its own flavour
(c) Serial connected TX OP (no CCS) will give you slightly better mids but less top end extension (which IME is already marginal with 26 tubes),
(d) Parafeed TX OP has the potential to be tuned to provide better bass performance than Serial, not much different to serial otherwise.
Cheers,
Naz
This is what I was thinking (at the moment):
How will the OPT make power if its primary current is not allowed to vary?
--------------------------
Buy Chinese. Bury freedom.
tweek:
If you're loading with the CCS OK, what is the input impedance that you're trying to match ?...the LL1660 secondaries need to accommodate. I like your approach...in think:> ). At first look, I would be concerned with the output impedance. Could work out great...try it out...
Stuben
not trying to match an impedence per se, just want to keep output impedence low for driving longer cables and the step down OPT will keep the output volts within the range of my amps input sensitiviteies, All this in relationship to the output levels of my sources.
JD
You need a coupling cap from plate to OP tranny in this config (parafeed). For the best outcome all the values should be carefully chosen and I'd even go as far as taking into account the coupling cap value from the previous stage if you are using a grid choke, but that's me.
As you know I love Schottkies but you largely lose their value when you string so many together. A good resistor will suffice and because you are using a CCS there is no need to bypass.
Naz
Naz, both the DAC I'm using and the phono stage use output caps at 2.2uf. So the cap to the output tranny would need to be calc'd. And won't have to worry about diferent sources being different uf's.Resistor is an option, just need to calc the appropruate value.
Many who have done a 26 like a 9v Alkaline on the grid, but have not seen it done with a grid choke.
For some reason a battery doesn't appeal to me as it has to be watched over over time as its not rechargable in circuit, but if the sonics are as good as reported, it could be something to live with.
JD
Edits: 06/01/12
Jeff,
Firstly I agree with TK that if I was using a gapped tranny I'd keep it simple and substitute the CCS with the IT. However the major thing parafeed has the potential to provide is tailored LF boost IF you want to take advantage of that. In this case several factors come into play, including the value of each element and even the load, which is needed to dampen the resonant peak you will end up with.
From the specs I was able to get, (still not exactly sure about L for the GK and IT) assuming you go with parafeed then use something like 47K for your attenuator, start with 1K5 for your Cathode R and about 0.47uF for the OP coupling cap.
The load you are driving will play a part. If it’s around 50K you will end up with a useful 2db of boost centred around 25Hz. You can choose to further dampen this with a 47K (or lower) R or increase the value of OP coupling cap to 2.2uF to move the peak down to a frequency where the response is already falling off. You could tune the frequency and amplitude to whatever suits your system and even incorporate a variable control.
Cheers,
Naz
Isn't the Lundahl gapped? Is there a compelling reason for using it in parafeed? Wouldn't it be preferable to eliminate the CCS, remove the series cap, and use the transformer as the manufacturer intended?
--------------------------
Buy Chinese. Bury freedom.
TK
I have heard countless accounts of gapped transformers used in parafeed and always they sounded better in parafeed.
I see no reason it would be different here.
DanL
"Countless accounts" doesn't qualify as a compelling reason. I thought maybe there was data regarding the Lundhall's performance with and without DC current.
--------------------------
Buy Chinese. Bury freedom.
TK
I consider it quite compelling especially since it
has never been reported it to sound worse in parafeed.
Unless you need someone to try THAT specific interstage
in THAT specific circuit with THAT specific power supply
to even consider it compelling to try it?
Come'on - be reasonable here.
DanL
"Unless you need someone to try THAT specific interstage
in THAT specific circuit with THAT specific power supply
to even consider it compelling to try it?"
Of course, testing the circuit as proposed (including the always critical coupling cap) against the standard feed is the right way to do it. Even then, it needs to be scientifically conducted as a double-blind test and requires more than a single listener. There's no validity otherwise, no matter how many people offer a generic opinion that parafeed doesn't make it worse.
--------------------------
Buy Chinese. Bury freedom.
TK
Logic dictates that the less stress
on a componant the better it will perform.
Experience confirms this to be true.
The major stress on a SE transformer
is the DC current it has to handle.
Removing that stress will be a plus
on the performance of the transformer.
Reports on every time it is tried
has also confirmed this to be true.
Whatever the major lack in the performance
as a SE transformer when used in parafeed
that is the most improved characteristic
whether transparency, bass or treble.
Add to that, Dave Slagle always maintains
that some gap on a transformer makes the
transformer perform better even in parafeed.
From the preponderance of evidence, one must
concede that there is something to this.
If you still maintain that you need a
double blind test before you can accept this
then it just is preposterous.
This leads me to wonder what are you doing on this site?
What happened to you over the years?
You used to be a reasonable guy.
You used to be open minded and adventurous.
Now you are so closed off and cynical.
I miss the old TK 8^(
DanL
In my experiences. A properly designed and implemented direct feed transformer sounds better than that same transformer parafeed.
I have on occasion found exceptions to this particularly when the direct feed transformer was being pushed too hard or was light in inductance for the tube it loads. (what I would call improperly designed or implemented) In the above cases, parafeed can help particularly in the bass since removal of the DC current (if operated at high DC flux) can place the AC in a more linear spot in the BH loop and as Naz pointed out the CL resonance can be tuned to extend the bass an octave lower.
dave
Dave
I am sorry if I mis-stated your position.
IIRC I read a thread on your Intact Forum
that you asserted quite strongly that
transformers sound better with a gap.
This was a few years ago (maybe more).
This always stuck with me.
I also remember someone on the Magnequest site
that used a Robin Hood series fed OPT
and liked the sound but then tried it parafeed
and was blown away with the improvement.
Also many posts on Hammond and Edcor iron
that regailed vast improvements in sound
when they were used in parafeed.
So I added that together and came up with
the conclusion as I have stated before.
Maybe "premium" iron is excluded from this,
as these were all budget irons being used.
DanL
I have maintained that a small intentional airgap is preferred to a design that goes for a minimal airgap even if there is no DC present.
Dave
Naz
A 9V battery would be good there too.
(Rechargeable, of course)
DanL
Dan, absolutely, and it seems an obvious choice but I'm not a fan for several reasons. Mainly, as Jeff pointed out the maintenance which still may have been worth it for a sonic improvement but IMO it's not as good as a simple R in this instance.
Naz
Naz
I have no experience with parafeed preamps
but I would think the unbypassed cathode resistor
would increase the tube's output impedance
well beyond any transformers ability to
convey 20-20KHz response across it.
DanL
If this were a power amp I'd be more concerned however it's only a line stage not pushing much load. In addition, since a CCS is being used the essentially constant current through the Cathode R means very little degeneration.
Naz
battery on the cathode is also an option as the current there would help keep it charged, or at lease keep it long lived.
Tho I'm a bit concerned about the sonics... different animal I know but I was biasing my input tube on my 437a / 320BXLS with a 2.4vNiMh and change to Cree's, the difference was amazing. The Schottkies had a not insignificant difference to the sound quality for the better.
JD
JD
But comparing one 9V battery to 12 Crees might be a different story. We can both agree it's better than an electrolytic.
Although I have been curious of the (polyester film) cap linked below.
DanL
Web site says sold out :(
But lookey what I found on the good old e(vil)Bay....
Only 6 left... hmmm I wonder where the other two just went ;^)
JD
supply the 1 Amp required by the OPT in this arrangement (roughly 150V/150ohm DCR of the primary)....
and this is where the Good Mr. DANLaudionut's suggestion is good...
the circuit will work, and sound quite good with a good coupling cap between the plate and the good 'parafeed' transformer.... ok
Continue having Fun, and Good Music to you...
Sincerely
-3db
JD
Use the regulator like you want.
Then use the CCS to load the 26.
Then cap couple the OPT(parafeed).
You have all the componants you want
used to their maximum benefits.
DanL
and generally sounds best...
the tube regulator you have shown is a good simple one...
the IXYS CCS is also simple and works...there are better CCS's out there,
however generally the CCS seems to sound better with any voltage regulator, and best of all when it is downstream...
IMHO...YMMV...try it and listen...
sincerely...have fun,
-3db
"Some on the aforementioned thread have opted for a single shunt regulator that accomplishes control of both current and voltage."The regulated supply that I showed regulates the total current draw from the supply but it does not make the amplifier stage constant current.
If your output transformer has high enough inductance and low enough shunt capacitance and if it's secondary is not loaded down then your amplifier stage will be constant current without the use of a CCS.
It's really all about load lines. If the load line is horizontal enough then the stage is said to be constant current even though nothing is totally constant current.
The idle current of the 26 is set by it's bias voltage (cathode resistor value).
If the OPT represents a high enough impedance the current through the 26 will not change much at all as you modulate the it.
In other words, the OPT IS the CCS.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 05/31/12
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: