john curl said:
Actually, what I wrote so far has quite a bit of info in it. Although I realize I speak in a more conceptual form that throws a bit of a curve (but makes perfect sense to me). Any part of it can be decyphered.
1audio said:
You are right on both points. The main thought to ponder is the return relative to the reference. With the first you can't escape the effects of the traces and with the second you have to understand what you are referencing to and the requirements of the reference (to keep it as nuetral as possible).
I'm curious as to your derived ground. My approach, that has evolved over the years, is to create defined paths back to the power supply hard ground. A sort of bastardization of the star ground approach (which, in most implementations I see, I find to create excessive loop areas). I'm assuming your approach is a variation of a virtual ground. I haven't tried it but, offhand would think that a virtual ground could work as a reference but you'd have to work to keep it from getting contaminated if any of the return currents are involved (and I know that is vague but I don't have the words, it's 4 am).
Your right about the challenges of single ended designs.
I'll leave it at that for now. Mike.
john curl said:
M Curl,
If you would have to redesign your preamp now, what solution would you apply to solve the inevitable ground loop induced by connecting different equipments with their power supply.
Would you follow Bill Whitlock's view and use balancing between equipments. Would you use transformers or balanced amps.
Do you believe that what this solves is more than what it degrades (if any).
By the way, I am still interested in having your opinion on my questions about relay distortion.
Thanks
JPV
john curl said:
Hi John,
You are certainly right. The higher peak rate of change of 30,29 kHz will certainly make the test more sensitive. Its just that CCIF 19+20 kHz is sort of a standard, and it is convenient when we wish to compare different measurements. For example, I really like it that John Atkinson shows 19+20 kHz IM spectrum at usually near full equivalent power. It is the most useful and telling measurement he does in terms of revealing an amplifier's HF distortion performance.
Cheers,
Bob
Just ruminating out loud;
I wonder if we don’t all invite trouble when we build the common and conventional split supply? One diode bridge; a plus output, a minus output and a single ground.
I would guess a more correct answer would be two totally separate supplies, one plus and one minus with independent grounds.
In the later case, the grounds are tied together at either the load point or at the system star ground. In the paragraph one case, by using a center tapped transformer we have created a ground loop that can not be removed.
I wonder if we don’t all invite trouble when we build the common and conventional split supply? One diode bridge; a plus output, a minus output and a single ground.
I would guess a more correct answer would be two totally separate supplies, one plus and one minus with independent grounds.
In the later case, the grounds are tied together at either the load point or at the system star ground. In the paragraph one case, by using a center tapped transformer we have created a ground loop that can not be removed.
hermanv said:
In my experience yes we do. Two separate supplies is one way to try, with the grounds connected at the central star point, but I don't find the use of a single bridge to be a problem. Sonically I prefer it. It does require separate secondaries with the ground legs brought out to the central star.
And I know what's coming next, because it comes up every time I've had this conversation, and my reply is to take a current probe and actually look at what's going on on the centertap; not the single wire but look at the current on the return leg of the indvidual windings. Also look at the currents in your filter caps with the centertaps connected or not connected. With no centertap the results follow the theory so often repeated.
Try the same measurements on an amp at both low and high power while playing music... I have a 100watt amp module with the bridge and filters all connected so that I can easily swap power transformers and measure these. The differences I see show a strong correlation to the sonic quality of the transformer.
The rest of the approach to grounding builds on this.
John mentioned he hadn't found anything useful in what I wrote. Try using separate ground returns from the supply regulators and supply bypassing on the +/- supplies at the line stage rather than a single as was used in the Blowtorch. The two supplies and the regulation returns are unrelated and should not be mixed due to the reasons Demian stated. The bypass caps should not return through any signal reference. And description of the signal reference is a whole other item.
Mike.
Has anyone tried the circuit as detailed in "Development of Second Generation Series Power Supply Module Exceeding Battery-operated Power Supply" ?
Regards
James
Regards
James
you are one of the few setting a good example
by using separate supplies for different stages in amplifiers
sometimes when I look at other so called good designers work
.. I ask myself
who forces them ????
to use only one trafo
the same one for Power Output at Ampere Level
as for the sensitive & critical input at uA level
... or have they not understood what they are doing
Keeping things simple is good
but simplify very important things like power supply
is plain madness ... in my book.
Having separate supply for stages does simplify to make good stages!!
There is nothing gained but problems by this ONE TRANSFORMER stupidity.
Someday they will learn .. eventually.
From you John Curl and from me if they bother listen to me hehe
by using separate supplies for different stages in amplifiers
sometimes when I look at other so called good designers work
.. I ask myself
who forces them ????
to use only one trafo
the same one for Power Output at Ampere Level
as for the sensitive & critical input at uA level
... or have they not understood what they are doing
Keeping things simple is good
but simplify very important things like power supply
is plain madness ... in my book.
Having separate supply for stages does simplify to make good stages!!
There is nothing gained but problems by this ONE TRANSFORMER stupidity.
Someday they will learn .. eventually.
From you John Curl and from me if they bother listen to me
hehe
john curl said:My personal preamp has 4 separate power transformers feeding 8 separate discrete diode bridges and uses 4 individual ground lines from the separate power supply case. Isn't this enough?
John;
In your experience are separate transformers needed or will just multiple secondaries do?
Is this to do with interwinding capacitance, or some other form of internal transformer crosstalk?
If multiple transformers are necessary, do you further isolate primaries with an LC or some other network?
- Status
- Not open for further replies.