New Posts New Posts RSS Feed - 1970s Design Indulgence
  FAQ FAQ  Forum Search   Events   Register Register  Login Login

Welcome to the Graham Slee Audio Products Owners Forum

 

Open to all owners plus those contemplating the purchase of a Graham Slee HiFi System Components audio product and wishing to use this forum's loaner program: join here (Rules on posting can be found here)

This website along with trade marks Graham Slee and HiFi System Components are owned by Cadman Enterprises Ltd


1970s Design Indulgence

 Post Reply Post Reply Page  <1 185186187188189 345>
Author
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (0) Thanks(0)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 07 Jan 2022 at 6:24am
There is always signal on the output of a power amplifier. It is either playing music or its own noise.

The noise, hiss and hum, is multifaceted, comprising resistor noise amplified by the voltage stage, semiconductor noise for which the first transistor is mostly to blame, controlled oscillations meaning they're incidental to all the efforts to prevent them (stability capacitors have resistance and inductance all the same), induced noise from electromagnetic devices and charging devices (transformers, capacitors and any wire carrying signal or noise) etc.

Low frequency noise is the hardest to attenuate, but at around one thousandth or better of the signal, is relatively benign.

Upper and high frequency noise still produces hiss, but only audible within a few inches of the speaker cones - the sound energy halving each doubling of distance away from the loudspeaker.

But the point here is that you can hear it, and it can be measured, and right in the mid frequencies it is, at the most, one ten thousandth of the signal (1/10,000), and getting close to -120dB.

Are we not told class-B distorts badly because it chops off the smallest of signals?

Well, noise especially at -120dB is the smallest of signals, isn't it?

Hopefully that's another nail in the coffin of the charlatan marketing departments!

That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
peterb View Drop Down
Senior Member
Senior Member
Avatar

Joined: 02 Feb 2017
Location: Cornwall
Status: Offline
Points: 332
Post Options Post Options   Thanks (0) Thanks(0)   Quote peterb Quote  Post ReplyReply Direct Link To This Post Posted: 07 Jan 2022 at 2:02pm
Congratulations, those are remarkable figures.
But the 'Slee' question is, does it sound better/correct/clean/right ? Smile
Peter
---------------------------------------
Dual 505-1, Cyrus CD T, DIY 80W MosFet amp and PreAmp, 2xKEF 103.2
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (0) Thanks(0)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 07 Jan 2022 at 3:47pm
It's marvellous what you can learn about the giant transformers we rely on for electricity distribution.

Obviously none of it applies to audio power supply transformers (this is a sarcastic comment).

You may remember circulating currents in parallel secondaries?

But what about non-linear loads increasing magnetising current? (yes, you read that right)

What could be described as a non-linear load? Perhaps a capacitor that only receives charge current part way through each half cycle? Yes, that counts as a non-linear load.

It results in harmonics in the windings, which waste power, and that is why it is a concern in electricity distribution. But when it comes to audio power supplies, hey, you don't need to know - you're over analysing the situation.

No I'm not!

You measure the magnetising current. You measure the a.c. load current. When taken together 1+1 = 2? Except, the meter tells you 1+1 = 3 or anything but 2! (these are illustrative values).

So, signal fluctuations which load the non-linear capacitive input rectifier filter, can have no influence on the amplifier? Well, pardon me, I think they must do.

Perhaps a magnetising current of only a few milliamps might be subtly increased by a few more milliamps as the amplifier starts to swing signal, and then a little more on signal transients - get the picture?

But if this happens to the signal symmetrically, it might be of little concern. But what if the load is asymmetrical about the signal?

Was it worth analysing the output capacitor action? Well, if I hadn't, we'd not know.

So how do you limit the magnetising current delta? You make the fluctuations comparably small. How? by increasing the magnetising current! 
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (1) Thanks(1)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 08 Jan 2022 at 10:13am
An experiment you can try at home as long as you're competent and have an oscilloscope and a multimeter, and a hi-fi toroidal transformer.

Other equipment: a home made current transformer comprising a Talema AC1005 "secondary" and 10 turns of mains rated insulated wire through it. The ends of the wire go to live feed to the transformer, and the other to live supply. The secondary of the hi-fi toroidal transformer is open circuited.

Solder a 100 ohm resistor across the two end pins of the AC1005 (ignore the middle pin). Clip the scope leads to either end and parallel with it a multimeter set to a 2 or 3 volts range.

You'll also need an automatic washing machine, to do your laundry at the same time!

With everything connected, switch on and observe the current waveform. The upper and lower should be equal like a reflection each half cycle.

Now, do your laundry.

While your automatic washer runs through its program (I tend to use "baby hygiene") observe the upper and lower waveforms. See how it goes lopsided from time to time. Notice that on some cycles one half of the waveform completely disappears, while the other half "grows" in size.

You may also hear the transformer grumble during these odd cycles. The internet will tell you this is DC on the mains. You will know what to think of the internet.

If magnetising current is the "admission fee" then if you don't pay it, you don't enter.

Bearing in mind that in proper use the secondary is connected to a harmonic generator otherwise known as a capacitor smoothed rectifier, then harmonic current flows in the secondary, and therefore must flow in the primary.

And they call that hi-fi? Huh!
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (1) Thanks(1)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 08 Jan 2022 at 1:42pm
Originally posted by peterb peterb wrote:

Congratulations, those are remarkable figures.
But the 'Slee' question is, does it sound better/correct/clean/right ?

With the latest toroidal the answer is I don't know yet.

The difficult part is understanding why, and that is the research part of R&D.

There are problems with power supplies, and these are exaggerated by the asymmetry of the type of amplifier I chose to develop (my choosing - can't blame anybody else).

The "burn-in" problem is due to the dielectric absorption of large electrolytics. I've tried to show before this phenomenon by suggesting the experimenter charges a large electrolytic, and then discharges it (using a suitable resistor to prevent damage), and then storing it on a shelf for a few weeks, then measuring the voltage between its terminals. Anyway, it is in numerous textbooks, so it's nothing new.

This time I requested the toroidal transformer man make the toroidal have something like 57mA mag current at rated primary voltage. The new transformer measured 13mA, and so you might imagine my frustration? Nevertheless I installed it, and tried it for nearly 72 hours, but could hear that the sound was heading toward bright. Rather than renting my frustration at him, I decided to try the next lower tap which causes an increase in mag current. It also results in a higher HT, which the transformer man insists is the problem - that I have insufficient voltage - regardless of the fact that the frame transformer does that "insufficient voltage" and everything works well.

Yesterday the sound was obviously (to me) artificial, but today is more correct, which I can only put down to the long time dielectric absorption of the smoothers.

The frame transformer, on the other hand, has something like 60 - 70 mA mag current, and after a week the amp sounded "better/correct/clean/right" to use your words.

The problem with the frame transformer is the mechanical noise it, and the metal case, causes. As, I've previously explained, this is because of lateral magnetostriction (also covered in textbooks). The transformer has to be bolted down flat for it to fit. It also takes up an amidships position for sensible centre of gravity. The result is the magnetostriction (following the eddy current paths) is in the lateral axis of the case. Magnetism can be induced into non-ferrous metals too - the case being alumnium. This is also covered in textbooks, and the reader might wish to investigate how a mechanical speedometer works.

On a wooden test bench top, the frame transformer is inaudible, but inside the case structure, the magnetostriction lines are "amplified" and I doubt anybody would entertain a mechanical buzzing amplifier. It might be possible to attenuate the noise by rubber mounting the frame transformer, in addition to chassis slots which break the magnetostriction path. This is another something I will need to try.

So yes, it can be made to sound better/correct/clean/right, as well as having remarkable specification figures, and so is well worth continuing with.
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (1) Thanks(1)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 09 Jan 2022 at 10:59am
Floating the transformer

Funny how I remember some of the pre-HFA-days reviews, and one description of the sound of the first toroidal transformed amplifier - it was "steely."

Snippets like that make me research to the extremes one peer calls "over-analysing."

The "non-americano" (and non-Japanese) mains voltages vary from 200V to 254V, but yet we can buy at least one commercial amplifier with no voltage selector, "designed" to run on 230V (the made up voltage of the EU).

It has a toroidal transformer. Over the 200 - 254V range the transformer will have virtually no mag current, right up to saturation. But, it makes money.

Nevertheless, such amplifiers sound "perfect" to all their customers on all those voltages, so what's the secret? Is it a regulated power supply? No, they're unregulated. The secret is there is no secret, except from a "Chum Bucket" wearing crowd of owners who have been convinced they sound good through marketing hubris, and a hi-fi press which (I'd better stop right there).

"Steely" describes a distortion - plain and simple, and that's the sound I get from this amplifier using toroidal after toroidal after toroidal transformer. I put them all in a heavy duty stores tub mounted on a dolly used to wheel the tubs around industrial works. It took two of us to lift the tub off the dolly onto a low shelf of a heavy duty rack which is used in the "heavy stores." Half were off the shelf, the others custom made. Would anybody like them?

It looks as though I'm stuck with the frame transformer. The problem is its mechanical noise. It isn't much, and probably the same as that from a valve amp, but this being a transistor amp, the noise will not be forgiven.

The remedy is to break the magnetostriction and eddy-currents path. Slotting of the chassis can only go part way. There are all kinds of methods which will, and are (were) used to reduce chassis noise. But here, the design has gone along with the false promise of toroidals, and complete redevelopment is a costly and time consuming thing.

When the "shops open" I will attempt to float the frame transformer on rubber bushings as illustrated below.

anti-vibration



Edited by Graham Slee - 09 Jan 2022 at 6:04pm
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
Graham Slee View Drop Down
Admin Group
Admin Group
Avatar
Retired

Joined: 11 Jan 2008
Location: South Yorkshire
Status: Offline
Points: 16298
Post Options Post Options   Thanks (0) Thanks(0)   Quote Graham Slee Quote  Post ReplyReply Direct Link To This Post Posted: 16 Jan 2022 at 6:17pm
I'm afraid to say I must be totally wrong.

Only the current for the positive excursion passes in the upper transistor.

The energy in the energy store for the lower transistor to drive the loudspeaker negative - the output capacitor - is always there. It was put there at switch on. It is never depleted by the load, and is therefore magic.

I would give you the link, but why bother.
That none should be able to buy or sell without a smartphone and the knowledge in how to use apps
Back to Top
 Post Reply Post Reply Page  <1 185186187188189 345>
  Share Topic   

Forum Jump Forum Permissions View Drop Down

Forum Software by Web Wiz Forums® version 12.01
Copyright ©2001-2018 Web Wiz Ltd.

This page was generated in 0.109 seconds.