Transformers and DC

Thread Starter

yesplease

Joined Mar 4, 2020
41
Hi,
I'm doing a bit of searching on how to build a ring modulator as an audio effect. It looks like I'm gonna need a transformer (2 actually) and I've never used one before.
I read that transformers can only be used with AC. DC would damage the transformer. So what does that mean in case I want to feed it a square wave? as the period increases, the more this would look like DC right? If that is an issue, Is there something in the transformer specs that I could look for to determine the maximum period I should limit my input signal?
Also, does using bipolar vs unipolar signal make any difference with a transformer?

Thank you.
 

shortbus

Joined Sep 30, 2009
10,045
Your question seems to say it all, you have little or no understanding of electronics. How do you expect to design and build something that you seem to want and need right away? I doubt that you cando it for the price of this kit, I know I couldn't. Building from a kit like this will both get you what you want and help you to learn at the same time.

https://www.etsy.com/listing/125124...nwiBMhC72ulxFTxhewy2Mvvs1J96N30BoCC9cQAvD_BwE
 

crutschow

Joined Mar 14, 2008
34,428
Yes, any DC current into the transformer can cause the core to saturate.
It's the average value of the waveform that determines its DC content.
Thus a unipolar signal has an average DC value, so cann0t be applied to a transformer.

To block any DC, you can use a large series capacitor which will pass the AC but not the DC.

The transformer needs to be designed to handle the lowest frequency content (from its Fourier values) with the highest applied voltage you will be using.
That determines the maximum magnetizing current that will not saturate the transformer core.
 

MrChips

Joined Oct 2, 2009
30,806
All you have told us is about transformers, AC, DC, square wave.

There are many different types of transformers.
You need to specify the transformer, voltage, current, power, frequency ratings etc. and what voltage, current, frequency signal you want to apply to the transformer.
 

Thread Starter

yesplease

Joined Mar 4, 2020
41
Your question seems to say it all, you have little or no understanding of electronics. How do you expect to design and build something that you seem to want and need right away
I don't want it right away, as I said, I'm just researching the subject right now and trying to understand how transformers work. I found a lot of useful information but there were still ambiguities and I am asking here to clarify them.

You need to specify the transformer, voltage, current, power, frequency ratings etc. and what voltage, current, frequency signal you want to apply to the transformer.
I'm not sure why you need this information. I don't have it anyway, I'm not in the design phase. I'm in the learning phase. I just want to know if square waves can be a problem with transformers and what information I need to look for in a transformer datasheet to figure it out. If I give you all those details, then my answer will be locked to these values only. I want to be able to figure it out from the datasheet using a general principle.

@crutschow: Thank you, this answers my question perfectly. When you say "The transformer needs to be designed to handle the lowest frequency content (from its Fourier values) with the highest applied voltage you will be using", I believe you are refering the the "frequency response" parameter in some of the datasheets I see right? For example, if I see "+ 2db from 200 to 15,000 Hz" then that lowest frequency would be 200hz. I'll read up to understand what that +2db means.

As for the clarification about unipolar, that makes sense. Thanks again.
 

Thread Starter

yesplease

Joined Mar 4, 2020
41
@shortbus . I looked at the link you've provided but I still can't say if it's ok to feed a square wave audio input in that circuit. I would assume that yes, since it's specifically made for audio purposes and square waves from 20hz to 20khz are common. But they don't mention anything about needing a bipolar signal. I guess that, with what I know now, it would be mandatory. So for example, connecting this to mic would work but not on a uC's PWM output. I looked up the transformer they were using and the datasheet says Frequency Response: ±3dB, 300Hz~3.4KHz @ 1KHz 0dB. So does that mean that it will saturate if the signal is outside of that range? That doesn't seem very audio-friendly if that's the case
 

crutschow

Joined Mar 14, 2008
34,428
'll read up to understand what that +2db means.
dB (deciBel) is a relative power ratio given by 10*log(P1/P2).
For a constant impedance, this gives a voltage dB ratio of 20*log(V1/V2).

For small, interstage audio transformers, the level is often given as dBm, which is the power ratio compared to 1 milliwatt with a nominal 600 ohm transformer source and load impedance.
The +2dBm would be a power of 1.58mW into 600 ohms.
This gives a voltage of near 1Vrms.
can't say if it's ok to feed a square wave audio input in that circuit. I would assume that yes, since it's specifically made for audio purposes and square waves from 20hz to 20khz are common
The frequency response spec is for sinewaves.
A square-wave of 20Hz has lower Fourier frequency content than 20Hz, so the voltage level of a 20Hz square-wave would have to be reduced as compared to a sinewave to avoid transformer saturation.
Practically this means the the peak square-wave amplitude (1/2 the Vpp) should be no more than about 85% of the sinewave RMS maximum voltage at 20Hz.

And square-waves are not common in natural audio, even though you can have a square-wave at an audio frequency, which may be used for some types of audio testing.
Frequency Response: ±3dB, 300Hz~3.4KHz @ 1KHz 0dB. So does that mean
It means that the deviation in the signal level ±3dB over that frequency range at some nominal power level, as determined by the transformer size.
It has nothing to do with the maximum signal the transformer can carry without saturating.
 
Last edited:

Thread Starter

yesplease

Joined Mar 4, 2020
41
dB (deciBel) is a relative power ratio given by 10*log(P1/P2)
I knew that part:)

Square waves are common in electronic music (you said "natural audio" so you were probably thinking this too) and analog synths. And many people are adding ring modulators in their modular synths. So I'll try to figure out what makes it possible and what makes their transformers suitable for such an application. The information you gave me today is a very good start. Thank you very much.
 

MrChips

Joined Oct 2, 2009
30,806
There are huge differences between a power transformer and a signal or pulse transformer.
Hence we cannot make any recommendations until we know your application.
 

shortbus

Joined Sep 30, 2009
10,045

kubeek

Joined Sep 20, 2005
5,795
I knew that part:)

Square waves are common in electronic music (you said "natural audio" so you were probably thinking this too) and analog synths. And many people are adding ring modulators in their modular synths. So I'll try to figure out what makes it possible and what makes their transformers suitable for such an application. The information you gave me today is a very good start. Thank you very much.
Transformers are rarely used in the audio path, except for line level signals such as microphones and DI boxes. The main use is in the output stage of a tube amplifier. So, can you tell us what you actually have in mind and why do you think you need a transformer for audio/music kind of waveform, i.e. 20Hz to 20kHz?
 

Thread Starter

yesplease

Joined Mar 4, 2020
41
@shortbus the thing is that this very design is what led me to start wondering about my question in the initial post. I want to buy the kit (actually, I wanna buy the parts separately since it will be a lot cheaper). But there is nothing to be learned from the kit since they don't explain what kind of signal can be used with it.

Wikipedia was, of course, one of the first thing I read. After reading many resources, my conclusion was that you cannot feed DC to a transformer. So it made me think that a square wave (i.e 0-5v) at 1hz really looks like I'm applying a dc source of 5v for .5s and then I turn it off. My OP was about validating if that reasoning was right or not in the context of using a transformer. This is not specifically answered in any of the litterature I read. Or perhaps it is, but in a way that would require me to deduce it if I had more technical knowledge about electronics in general. So I figured I could ask a question on a forum at that point.

@kubeek. Thank you for offering your help. I think I got the answer I wanted from @crutschow. But in case you want to elaborate more: I was looking at circuits such as the one that @shortbus posted in post #2. People use 2 transformers and 4 diodes to create an effect that can be interesting in a guitar pedal or synth module. Looking at those kind of circuits made me go down the rabbit hole of trying to understand how transformers work in general.

Since starting this thread, I started looking into a TY-145P. That transformer seems to be used by many hobbyist for making ring modulators. But I dont know what type of signal they are driving into it (other than it's audio freq range). Not sure if it's good for 1v, 5v, 15v. Bipolar? Unipolar? So I'm trying to deduce what kind of signal would be "supported" by reading the datasheet. I know it's all in there, I just need to figure it out. Or get help from people on a forum :)
 

MrChips

Joined Oct 2, 2009
30,806
A 0-5V square wave at 1 Hz…
You are way off the mark!

What is a transformer?
A transformer is two coupled inductors. Look up mutual inductance.
So inductance is important. Look up inductance.
How does inductance change with frequency?
Why wouldn’t a transformer work at 1Hz?
Or does a transformer work at 1Hz?
What difference does the shape of the signal make?
Unipolar or bipolar? Does a transformer care?
 

Thread Starter

yesplease

Joined Mar 4, 2020
41
Thank you @MrChips . I understand how inductance works. Or at a minimum, I know that a magnetic field gets generated by a changing electric field and the reverse happens too. A transformer is just 2 inductors where once creates a magnetic field and the second one converts that magnetic field back to an electrical current.

I would expect that a 0.0001hz AC would still generate a magnetic field of the same frequency but the reason that doesn't work is something about impedence I think. According to wikipedia: "At low frequency the reactance falls; at DC, the inductor behaves as a short circuit. As frequency increases the reactance increases and at a sufficiently high frequency the reactance approaches that of an open circuit."

But I'm not really interested in knowing if the transformer will let current pass or not (I know... current does not pass thru, but rather by inductance). I wanna know if the transformer will break or if the sourcing circuit will get damaged. What scares me is the part where they say "The inductor behaves as a short circuit". By "short circuit", do they mean that a high current will go through? The datasheet says that the primary impedence is 600Ohms. So I guess that this would rather mean that connecting a transformer on a DC source would simply make the transformer become a 600Ohm load. If that's the case, then I'm not worried anymore. Assuming that the source can provide 8ma easily and that the transformer can also handle 8ma. Because it only means that transformer would work within a certain frequency range and just act as a resistor outside of that range but it wouldn't actually damage any part of the circuit. And that would make sense, because the Ring Modulator kit that is being sold in post #2 doesn't have any protection at all on the input but it is being sold to people who will use LFOs from 0.5hz to 100hz.

EDIT: I believe that the resistance would be this: "11. DCR: Primary (1-3) 44Ω Nominal Secondary (4-6) 58Ω Nominal" instead of what I tought was "primary impedence". So a DC fed to that transform, on the primary side, would simply act as a 58Ω resistor?

Is my understanding correct?
 
Last edited:

shortbus

Joined Sep 30, 2009
10,045
Transformers are rarely used in the audio path, except for line level signals such as microphones and DI boxes. The main use is in the output stage of a tube amplifier. So, can you tell us what you actually have in mind and why do you think you need a transformer for audio/music kind of waveform, i.e. 20Hz to 20kHz?
That is true for most cases, but what he is after is to cause distortion, controlled distortion but still a distortion. Many rock songs use various "pedals" to cause audio distortion and this is one of them.


Wikipedia was, of course, one of the first thing I read. After reading many resources, my conclusion was that you cannot feed DC to a transformer. So it made me think that a square wave (i.e 0-5v) at 1hz really looks like I'm applying a dc source of 5v for .5s and then I turn it off. My OP was about validating if that reasoning was right or not in the context of using a transformer. This is not specifically answered in any of the litterature I read.
The wiki does answer that. What the ring mod does is add a DC component to the audio sine wave, that is what makes the audio sound like it does. Or it can add or subtract different frequencies.
"If one of the original signals was instead a square wave, the output would be quite different, since a square wave consists of a sum of an infinite number of odd harmonics, and each harmonic will generate its own set of sidebands" from - https://en.wikipedia.org/wiki/Ring_modulation
 

MrChips

Joined Oct 2, 2009
30,806
Thank you @MrChips . I understand how inductance works. Or at a minimum, I know that a magnetic field gets generated by a changing electric field and the reverse happens too...

Is my understanding correct?
That is not correct.
You do not need a changing electric field in order to generate a magnetic field. A constant current will generate a magnetic field, e.g. a solenoid. The reverse is not true.

When we model electrical/electronic components we have to take into account all unforeseen R, C, and L factors when they matter.

1645631508707.png


At the simplest level we model an inductor based on its reactance XL.

XL = ωL
where
XL = reactance in ohms
ω = angular velocity in radians/second = 2πf, where f is frequency in Hz
L = inductance in henries

At the next level of approximation, we include the resistance of the wire, RL.

1645631437641.png

Here is where things get complicated. We cannot just add DC resistance RL and reactance XL.
We need to add the complex reactance which results in impedance.

Thus the impedance of an inductor is:

ZL = RL + jXL
ZL = RL + jωL

j is a mathematical operator that takes into account the fact that the DC resistance and inductor's reactance are 90° out of phase.

At 0Hz, XL = 0
Hence ZL = RL at DC.

At anything above 0Hz things get complex because the current and voltage are not in phase. This is why knowing the frequency of the application is vital to understanding the behaviour of the transformer. There is no one size fits all answer.
(You get phase shifts when you put audio through an inductor.)

So if you thought that a transformer is a simple device to transform one voltage to another, think again. Things are not so simple. Knowing the math involved helps to explain the mystery.
 
Top