understanding transformer distortion

[joaquins]

I have always found magnetics rather arcane with its terminology that I find hard to visualise. Ohms law is no problem. I can imagine current (electrons) flowing through a resistor and how much gets through depends on the constriction of the resistance and the force you push it with (voltage). What are the magnetic equivalents?

Cheers

Ian

I guess you know, but the magnetic flux could be seen as a current, and the permeability as the conductivity, so is pretty much the same thing, in the middle you change from one type to the other, you change from electric to magnetic and then electric again, but you could think as a third electric path, non linear, but we are used to non linearities as well with semiconductors and the absorption in capacitors, just thinking in the magnetic flux as a third electric path could help. The only difference is that you don't have a particle moving around caring the energy (usually the electrons), but the movement of those electrons make the magnetic flux. The easiest maybe is think as an abstract way of electric flux with different rules, taking into account the absorption and non linearities of the core material, the other losses by joule effect are part of the usual electric path (as the winding resistance) which should be no problem to you, used to work with those all the time, as you said, ohms law.

JS

 

[CJ]

5500 : 2800 ohms
68 : 23.5 DCR

used to build pulse transformers, huge double C core the size of two bricks,
had about ten turns, lots of inductance from the huge core, hi-freq response and low DCR from small copper loss, hi Z because they ran at a high frequency, crape wrapped and tanked in silica oil to keep the 100 KV pulse from jumping ship,

"A pulse transformer is a transformer that is optimised for transmitting rectangular electrical pulses (that is, pulses with fast rise and fall times and a relatively constant amplitude). Small versions called signal types are used in digital logic and telecommunications circuits, often for matching logic drivers to transmission lines. Medium-sized power versions are used in power-control circuits such as camera flash controllers. Larger power versions are used in the electrical power distribution industry to interface low-voltage control circuitry to the high-voltage gates of power semiconductors. Special high voltage pulse transformers are also used to generate high power pulses for radar, particle accelerators, or other high energy pulsed power applications.

To minimize distortion of the pulse shape, a pulse transformer needs to have low values of leakage inductance and distributed capacitance, and a high open-circuit inductance. In power-type pulse transformers, a low coupling capacitance (between the primary and secondary) is important to protect the circuitry on the primary side from high-powered transients created by the load. For the same reason, high insulation resistance and high breakdown voltage are required. A good transient response is necessary to maintain the rectangular pulse shape at the secondary, because a pulse with slow edges would create switching losses in the power semiconductors.

The product of the peak pulse voltage and the duration of the pulse (or more accurately, the voltage-time integral) is often used to characterise pulse transformers. Generally speaking, the larger this product, the larger and more expensive the transformer.

Pulse transformers by definition have a duty cycle of less than 0.5; whatever energy stored in the coil during the pulse must be "dumped" out before the pulse is fired again.

 

[CJ]

let's have a look at air gaps vs distortion, sound exciting? 

 

[abbey road d enfer]

Tangent: Considerations regarding DC resistance, AC impedance, and level.  I noticed an audio transmitter modulation transformer with the following specs, which are quite far from what we expect for small signal line level audio.

5500 : 2800 ohms
68 : 23.5 DCR

With an unknown small (er than a brick) transformer, we'd assume it to be 600 : 250 roughly.

Clearly, a modulation xfmr is not a low level part. It's similar to an output xfmr, in fact it IS an OT with a high Z secondary.
These are indeed very odd figures.
The Z/R ratio is very different than what you get with usual OT's. Z/R ratio is directly dependant on the core size. The higher the rated power, the higher the Z/R ratio.
But here it looks like a factor 10 has been lost somewhere.
A typical 10W transformer for EL84 would show Z=5000 ohms, DCR=100-200 ohms for the primary

 

[CJ]

last of the distortion pages,  warning:images may be distorted,

 

[emrr]

5500 : 2800 ohms
68 : 23.5 DCR

It does look like factor of 10 has been lost but indeed, these are the specs factory printed on the transformer case. 

 

[CJ]

nothing odd about the DCR,Marshall OPT for EL34's has a DCR of about 17 ohms in each leg for about 34 total, Fender is  higher with around 100 ohms total Pri DCR but it has a smaller stack than the Marshall so  more turns are needed to get the same Henries,

you will find higher DCR on the smaller pwr tubes like EL84 and 6V6, like 300 ohms for a Princeton Reverb OPT, but less plate current means smaller wire so resistance goes up, even on a smaller core,

wire size vs DCR is not very linear,
0 .080" #32 has 163 ohms/1000 ft, .064 A/cir mil
0.040" (half the dia.)  #36  has 670 ohms/1000 ft, .025 A/cir mil

modulation xfmr will have skin effect at hi freqs, so bigger wire is needed,

 

[12afael]

the manufacturer say that the distortion is higher with low impedance because the you have more signal in the primary. with a 600ohms it form a divider with the input winding of 50% and with 50ohms you have a ~92%.  I found this information useless. they should measure the distortion at the same level or at the same distortion.  it seem that they are reading just the oscilator output before the source impedance (a serie resistor).

i'm right or i'm missing something?

 

[PRR]

> all transformer ratios are 0 dB.

+1 uh, -1. Agree that transformer is ~~ -1dB more-or-less.

___________________________________

> different hookups and what they do to max level
> 500 Ohms 16 dBm
> 200 Ohms 12 dBm
>  50 Ohms  6 dBm
>  17 Ohms  1 dBm

This is mis-use of "dBm".

It should at least be "dBu".

Very preferably it should be Volts. The THD is a function of Volts pretty-much no-matter what the impedances are (as long as they are "nominal" or lower).

Taking so-called "dBm" as "dB Volts above 0.775V" we get these numbers:

500 Ohms  4.9V
200 Ohms  3.1V
50 Ohms  1.55V
17 Ohms  0.87V

Note that 500 to 50 Ohms is 3.16:1 of voltage, and 4.9V/3.16 is 1.55, exactly what "6dBm" would mean if it were a Voltage and not a Power.

Conversely you can see this as a "Volts per turn" spec (same as power transformer ratings). You may not know the actual turns, but you sure better know how to figure turns-ratios for various connections, and the max clean Volts will go in the same ratio.

Now compute the Power of say 0.87V in 17 Ohms. 0.044,5 Watts, 44.5 milliWatts. Take the Power (*10) dB of 44.5mW above 1mW. My abacus is showing +16.5dBm, essentially (+/-10% round-off?) the *same* Power as the +16dBm on the 500 Ohm tap.

(My math may be slipping due to 500/600 Ohm difference, but half-dB is generally loose-change in transformer specs.)

 

[PRR]

> an audio transmitter modulation transformer with the following specs, which are quite far from what we expect for small signal line level audio.
5500 : 2800 ohms
68 : 23.5 DCR

What frequency range?

What power level?

If you accept a 300Hz cut-off you can wind less turns of fatter wire and get a lot less DCR.

If sized "big", perhaps >100 Watts, the geometric proportions lead to larger H and lower DCR. Since I used to live near a 25,000 Watt audio modulator (next to a 50KW clear-channel AM broadcaster), I know you can find HUGE modulation transformers (too big to drag home).

Put-together: >1KW AM rigs will usually be full-bass Broadcast, and AM <1KW would almost always be Ship or Police bass-less voice-only.

 

[emrr]

Part # would suggest pre-war RCA AM broadcast transmitter.