John Curl's Blowtorch preamplifier part II

[Steve Eddy]

Hey Mike!

> But of course an ideal transformer has no reactances.

Are you sure about that ?

Quite sure. The key word being "ideal."

An ideal transformer simply reflects impedances.

If you take an ideal 1:1 transformer and put a purely resistive 1k load on its secondary, then when looking into the transformer's primary, you'll see a 1k, purely resistive load. No reactances.

The reactances seen in a realworld transformer are parasitic. Namely, the non-infinite primary inductance, leakage inductance (due to non-ideal coupling between primary and secondary) and winding capacitances.

se

 

[Steve Eddy]

There'll be another chance

Oh alright. But don't take too long.

se

 

[PMA]

An ideal transformer simply reflects impedances.

If you take an ideal 1:1 transformer and put a purely resistive 1k load on its secondary, then when looking into the transformer's primary, you'll see a 1k, purely resistive load. No reactances.

Even at DC, I would guess

 

[john curl]

For everyone who might what to understand Q in the context of transformers. Let me go through what I learned in found, so many decades ago, when at Ampex.
I was assigned to evaluate tape heads and transformers that were used with them.
It is easiest to start with a conventional analog magnetic tape head.
This tape pick-up is composed of some sort of mumetal like material, often slightly compromised toward more hardness to keep wear to a minimum. Then a coil is wound on the core of the head usually between 1mH and 1H. More than one H is exceptional, because the wire gets really tiny and the head will self resonate below 20KHZ typically.
1mH is only used for record heads, in general, because the output is so low, that a high ratio transformer is mandatory to get the level up for best S/N.
For this example, let us take a 1H head or equivalently a 10mH head and a 10:1 transformer. These two examples are remarkably equivalent.
Now, let us first concentrate on the 1H head, and let us say that we find an equivalent series resistance of 1Kohm. What is the Q of the head at 1KHz?
Well: 6.28K/1K = Q = 6.28. What about 10K? Then 62.8K/1K (ideally) = 62.8
20K? 125.6K/1K = 125+ Wow!
Now what do we usually measure? Instead of 125, we might get 5 on a good day, with a good head. What happened? Eddy current loss. AND what happened to the self noise of the tape head? It increased significantly with frequency proportional to the difference between the IDEAL Q and the REAL Q of the head.
Now, what IF you made a transformer of essentially the same material in the same way? After all, a transformer is made of 2 coils, the secondary usually being very much like the 1H head. Could there be a similarity in behavior? I should think so, and I measured it, long ago.

 

[simon7000]

Even at DC, I would guess

Theoretically DC doesn't exist as the life of the universe is finite. Theoretically on average if you have one foot on fire and the other frozen in ice you are comfortable.

Ain't theory wonderful! Shame reality intrudes, then we actually have to use knowledge and skill to get the results we aim for.

 

[hitsware]

"
Wattage 0.5W
Input Maximum Voltage 10Vrms
Turns Ratio 1:1
Primary (input) Impedance 600 Ohms
Primary (input) DCR 64 Ohms
Primary (input) Inductance 3H
Secondary (output) Impedance 600 Ohms
Secondary (output) DCR 68 Ohms
Frequency Response 20~20K Hz., <1dBu
THD+Noise <0.05% @ 1K Hz.
Plot Frequency Response/THD+N
Insertion Loss 0.5 dB
Bobbin Material Glass Filled Nylon 6/6
Flamability Rating Class B 130°C
Core Material M6 29 ga. line grain oriented steel
Termination 0.187" (3/16") quick disconnects
Mounting Zinc plated steel channel/frame
Weight 0.4 lbs.
"
3 Henries ...... That's a leaky one alright.
And 64 Ohms leaked up to 600 Ohms ...

 

[Steve Eddy]

Even at DC, I would guess

Sure.

They work in Washington the same as they work anywhere else.

se

 

[jneutron]

This is not excess noise, as it is all thermal in origin. I am still waiting to hear of any sources of excess noise in a transformer.

When I test coils around high permeability laminated cores, my HP 4284A calls eddy current losses "resistance" I use the Ls-Rs model.

So if those losses are resistive by nature, then the noise generated by the resistive losses of eddies would be a modulated noise. No signal, no noise..

Cheers, John

 

[jneutron]

"
Primary (input) Inductance 3H.....

3 Henries ...... That's a leaky one alright.
...

Why is 3 henries "leaky"? I must have missed something..

What about 680 henries?

Cheers, John

 

[john curl]

JN, what does this have to do about the discussion in hand?

 

[john curl]

Sorry about the typos. want instead of what, and instead of in, it is difficult for me to fix these typos, given that I a under continuous moderation. Sometimes, like now, it gets embarrassing.

 

[hitsware]

Why is 3 henries "leaky"? I must have missed something..

What about 680 henries?

Cheers, John

That was a jab at Steves idea that ideal transformers
have no inductance except leakage.

 

[john curl]

NOTHING is ideal, that is why we need intelligent engineers to see the weaknesses and work around them.

 

[jneutron]

JN, what does this have to do about the discussion in hand?

Please try to follow the discussion John.

3 henries was "leaky". So I asked, what was that in reference to, as I missed that from earlier in the thread.

The 680 henry magnet I'm working with must be considered "open floodgates" then.

You really should have asked about the Ls-Rs model, and how an HP 4284A classes eddy losses as a series resistance within the wire, in addition to the reduction of inductance caused by the eddies..

This implies that any current within the wire not only generates noise within the wire, but the slew rate of the current will cause the eddy currents to generate noise as well. A first derivative based amplitude modulation of a noise source..

I wonder how that would show up on a spectrum analyzer..

Cheers, John

 

[hitsware]

Negating 'idealness' for a moment,
isn't it a fact that the mutual ' inductance '
of the coils of a transformer is the basis
of its operation ?

 

[Steve Eddy]

Now, what IF you made a transformer of essentially the same material in the same way? After all, a transformer is made of 2 coils, the secondary usually being very much like the 1H head. Could there be a similarity in behavior? I should think so, and I measured it, long ago.

What was it you were saying about apples and oranges earlier?

se

 

[thune]

I read the VdV paper to suggest that when the source impedance is > to >> than the primary impedance (reflected from secondary), a potential high-pass-filter is created that weakens the midbass at threshold levels. However, when the source impedance is less than 1/2 the impedance seen on the primary, these effects are minimal. So I'm not clear how this might even apply to MC transformers. I like this kind of study, as it tries to find a reasonable explanation of the increased detail percieved from triodes when compared to pentodes used as output devices (taking the preference as a *given*.)

 

[Steve Eddy]

3 Henries ...... That's a leaky one alright.
And 64 Ohms leaked up to 600 Ohms ...

No no, the 3 Henries is primary inductance, not leakage inductance.

se

 

[jneutron]

No no, the 3 Henries is primary inductance, not leakage inductance.

se

sheesh...I wish my inductance was primary..the 5 gauss line is over 20 feet away.

John

 

[Steve Eddy]

That was a jab at Steves idea that ideal transformers
have no inductance except leakage.

Ideal transformers don't even have leakage inductance.

C'mon, Mike. Surely you know what "ideal" means in this context.

se