ampage Tube Amps / Music Electronics	For current discussions, please visit Music Electronics Forum.

ampage archive

Vintage threads from the first ten years

Re:trannies, hold up!

1/6/2000 2:37 PM
keith	Re:trannies, hold up! Thanks for the info! Keith

1/6/2000 6:44 PM
Scott Swartz	Are you referring to the fact that the 1st trans needs to be rated for the combined VA of both transformers? This I knew, probably should have noted in first post. Or, is this a magnetic core related issue? The winding relationship equation gives no insight of a problem, and unfortunately that is about the extent of my transformer knowledge. Anyone have a technical explanation? I hate to post incorrect info.

1/6/2000 6:44 PM

Scott Swartz

Are you referring to the fact that the 1st trans needs to be rated for the combined VA of both transformers? This I knew, probably should have noted in first post.

Or, is this a magnetic core related issue? The winding relationship equation gives no insight of a problem, and unfortunately that is about the extent of my transformer knowledge.

Anyone have a technical explanation? I hate to post incorrect info.

1/6/2000 7:30 PM
R.G.	Lemme try. The magnetic field inside the core of a transformer has a maximum value, beyond which it saturates. Since any winding on a magnetic core looks like an inductor, if you put an AC voltage on the winding, the inductance limits the AC current. As long as the saturation value of field inside the core is not reached, this inductance will keep the AC current within calculable bounds. The things that affect the AC current are peak AC voltage and frequency. If you think of the positive half cycle of AC, as long as the voltage is positive, the current in the winding ramps upwards. When the signal goes into the negative half cycle, the reversed voltage starts ramping the current down again in the direction of negative saturation. The current is the time integral of the instantaneous voltage divided by the inductance (a result of V=L di/dt). If either the voltage is too high or the time is too long, the current reaches the place where the core saturates, and then the inductance drops precipitously and the current can then skyrocket. This is the point where you begin to smell that familiar I-squared-R smoke because now only the wire resistance limit current. So - for AC in an inductor, the AC voltage across it has to be limited to some maximum, and the frequency must be some minimum as lower frequencies give the same voltage longer to saturate the core. You can trade these off. If you half the voltage, you can go to 1/2 the frequency, all other things being equal. However, the AC power line is a fixed frequency, 50 or 60Hz depending on which country you're in. Power transformers are designed for that frequency, and as a result, the maximum voltage any winding can support without saturating the core has some fixed voltage limit. If you put significantly more voltage on a power transformer winding than it was designed to handle, it will saturate the core. Since all the windings on a transformer have a fixed voltage per turn of winding, it doesn't matter which winding you drive. A 120:12VAC transformer usually steps down to get 12VAC from a 120VAC source. It can also produce 6VAC if you feed it 60VAC in the primary, or 3VAC if you feed it 30VAC in the primary. It also works as a step up. Feed it 12VAC in the 12V winding, and you get 120VAC (neglecting losses...) out the 120VAC winding. Feed it 6VAC in the 12VAC winding and you get 60VAC on the 120VAC winding. However, this freedom stops at the max design voltage. If you feed 24VAC into the 12VAC winding, you get a little more than 120VAC out the 120VAC winding (this is all the core will support) and a lot of smoke and potentially flames if you leave it like this. If you take two 120:12VAC transformers and hook the 12V windings together, you can put 120VAC on either 120VAC winding and get 12VAC on the 12VAC windings and 120VAC on the other 120VAC winding. If you do a 120:12vac driving a 6V:120V transformer this way, though, you're putting a 2 times overvoltage on the 6V winding, the second transformer's core saturates and it starts drawing huge currents, limited only by the wiring resistance in the first transformer and the resistance of the 6V winding. BOTH transformers overheat, the second from saturation and the first by overloading. The limitation on not saturating a transformer by overvoltage is different from the limit of the amount of rated power. In a well designed transformer run at proper voltage and frequency, the AC current under no load is usually under 1-2% of the maximum power it can deliver to its secondary - the losses are low. Any secondary loading pulls power (in effect) directly from the primary through the magnetic field in the core without changing the value of the magnetic field conditions in the core as the energy in from the primary is just balanced by the energy out to the secondary (this is one reason you can't saturate a transformer by loading a secondary, by the way; it just can't happen). If you put too much load on the secondary, again the transformer begins to overheat, but it's because the secondary load is causing too much I squared R heating in the primary and secondary windings from the increased current. I hope that helps. Post again or send email if you have questions.

1/6/2000 7:30 PM

R.G.

Lemme try.

The magnetic field inside the core of a transformer has a maximum value, beyond which it saturates. Since any winding on a magnetic core looks like an inductor, if you put an AC voltage on the winding, the inductance limits the AC current. As long as the saturation value of field inside the core is not reached, this inductance will keep the AC current within calculable bounds.

The things that affect the AC current are peak AC voltage and frequency. If you think of the positive half cycle of AC, as long as the voltage is positive, the current in the winding ramps upwards. When the signal goes into the negative half cycle, the reversed voltage starts ramping the current down again in the direction of negative saturation.

The current is the time integral of the instantaneous voltage divided by the inductance (a result of V=L di/dt). If either the voltage is too high or the time is too long, the current reaches the place where the core saturates, and then the inductance drops precipitously and the current can then skyrocket. This is the point where you begin to smell that familiar I-squared-R smoke because now only the wire resistance limit current.

So - for AC in an inductor, the AC voltage across it has to be limited to some maximum, and the frequency must be some *minimum* as lower frequencies give the same voltage longer to saturate the core. You can trade these off. If you half the voltage, you can go to 1/2 the frequency, all other things being equal.

However, the AC power line is a fixed frequency, 50 or 60Hz depending on which country you're in. Power transformers are designed for that frequency, and as a result, the maximum voltage any winding can support without saturating the core has some fixed voltage limit. If you put significantly more voltage on a power transformer winding than it was designed to handle, it will saturate the core.

Since all the windings on a transformer have a fixed voltage per turn of winding, it doesn't matter which winding you drive. A 120:12VAC transformer usually steps down to get 12VAC from a 120VAC source. It can also produce 6VAC if you feed it 60VAC in the primary, or 3VAC if you feed it 30VAC in the primary. It also works as a step up. Feed it 12VAC in the 12V winding, and you get 120VAC (neglecting losses...) out the 120VAC winding. Feed it 6VAC in the 12VAC winding and you get 60VAC on the 120VAC winding.

However, this freedom stops at the max design voltage. If you feed 24VAC into the 12VAC winding, you get a little more than 120VAC out the 120VAC winding (this is all the core will support) and a lot of smoke and potentially flames if you leave it like this.

If you take two 120:12VAC transformers and hook the 12V windings together, you can put 120VAC on either 120VAC winding and get 12VAC on the 12VAC windings and 120VAC on the other 120VAC winding. If you do a 120:12vac driving a 6V:120V transformer this way, though, you're putting a 2 times overvoltage on the 6V winding, the second transformer's core saturates and it starts drawing huge currents, limited only by the wiring resistance in the first transformer and the resistance of the 6V winding. BOTH transformers overheat, the second from saturation and the first by overloading.

The limitation on not saturating a transformer by overvoltage is different from the limit of the amount of rated power.

In a well designed transformer run at proper voltage and frequency, the AC current under no load is usually under 1-2% of the maximum power it can deliver to its secondary - the losses are low. Any secondary loading pulls power (in effect) directly from the primary through the magnetic field in the core without changing the value of the magnetic field conditions in the core as the energy in from the primary is just balanced by the energy out to the secondary (this is one reason you can't saturate a transformer by loading a secondary, by the way; it just can't happen).

If you put too much load on the secondary, again the transformer begins to overheat, but it's because the secondary load is causing too much I squared R heating in the primary and secondary windings from the increased current.

I hope that helps. Post again or send email if you have questions.

1/6/2000 8:20 PM
Dave James	...excellent. Thank you. DJ

1/6/2000 9:11 PM
Scott Swartz	Thank you for the excellent explanation. I understand exactly what the problem is. The only question I still have is exactly how much safety factor is designed into a typical core. I'll try a few experiments with a filament trans, a variac, an ammeter, and most importantly fuses tonight and post the data of when it saturates. I have a 6.3VAC, 3A Magnetek trans I bought recently from Mouser laying around, I'm pretty sure. My apologies for the incorrect post.

1/6/2000 9:11 PM

Scott Swartz

Thank you for the excellent explanation. I understand exactly what the problem is. The only question I still have is exactly how much safety factor is designed into a typical core. I'll try a few experiments with a filament trans, a variac, an ammeter, and most importantly fuses tonight and post the data of when it saturates. I have a 6.3VAC, 3A Magnetek trans I bought recently from Mouser laying around, I'm pretty sure.

My apologies for the incorrect post.

1/6/2000 10:28 PM

R.G.

quote:
"The only question I still have is exactly how much safety factor is designed into a typical core."

Hah! It's good to find a sane, inquisitive mind.

The unfortunate answer to your question is - it depends. Although I've glibly talked about saturation as though it's a thin black line that you're OK on one side of and dead on the other side of, it's not that black and white.

Magnetic saturation in most linear irons and ferrites is somewhat gradual. There isn't any line, only a gradually decreasing inductance and increasing magnetizing current. Generally this happens over about 15-20% of the intended voltage range, so a nominally 120VAC winding will just get sloppier up to maybe 144Vac. This is nice if you're designing transformers that get occasional abuse. They don't just start smoking, but rather just run a bit hotter as the voltage goes up. This gives them some tolerance for surges and modest abuse.

But put 50-100% overvoltage on them and they'll start overheating.

In general. What actually happens is that the transformer designer is juggling the amount of iron, number of turns and wire size against his boss's competing urges to have things cheap and not have things burn up during the warranty period. The designer makes a guess based on experience, budget, and other sloppy factors about just how much waste heat he can afford, how little iron he can use, and how few turns of how skinny a wire gauge he can get away with. To top it off, his boss will (at rare intervals!) give him more money if his designs are consistently cheaper and last long enough than other designers. So the urge to push just enough into saturation on power transformers is strong.

That's another way of saying that cheap transformers will have thicker iron, skinnier wires, use more magnetizing current (that the customer pays for) and run hotter than expensive transformers that use thinner laminations, thicker wire and less magnetizing current - for exactly the same VA rating.

As far as I know there isn't any standard of practice for just how much overvoltage a winding will take, and for how long before something burns.

1/7/2000 5:01 PM
Scott Swartz	I tested a Stancor 120V:12.6V, 1 amp filament transformer as a stepup transformer using a variac on the 12.6V winding and monitoring the current on the 12.6V and the no load voltage induced into the 120V winding. Data is as follows: Induced----------Current (A) Voltage 60-------------------0.05 90-------------------.16 120------------------.48 130------------------.78 140------------------1.16 150------------------1.72 160------------------2.34 170------------------2.98 180------------------3.67 190------------------4.2 200------------------5.17 Several observations: The transformer started to hum at 130V and got louder as the core saturated more. This is a fairly old trans and it appears that the core losses are fairly high. I charted the data in Excel and the data fits a curve defined by a y=A + Bx^4.07. It wouldn't let me set the intercept as zero, but of course it should be, and maybe the 4.07 power is supposed to be theoretically 4. I don't have any references on trans design to check this out. I could email the chart to anyone interested. The Magnatek trans I thought I had is already wired into an amp and I didn't want to unsolder and remove it. Also, the trans I tested I got on sale for $1 from Angela Instruments (I bought 10!), so I wasn't risking much.

1/7/2000 5:01 PM

Scott Swartz

I tested a Stancor 120V:12.6V, 1 amp filament transformer as a stepup transformer using a variac on the 12.6V winding and monitoring the current on the 12.6V and the no load voltage induced into the 120V winding. Data is as follows:

Induced----------Current (A)
Voltage

60-------------------0.05
90-------------------.16
120------------------.48
130------------------.78
140------------------1.16
150------------------1.72
160------------------2.34
170------------------2.98
180------------------3.67
190------------------4.2
200------------------5.17

Several observations:

The transformer started to hum at 130V and got louder as the core saturated more.

This is a fairly old trans and it appears that the core losses are fairly high.

I charted the data in Excel and the data fits a curve defined by a y=A + Bx^4.07. It wouldn't let me set the intercept as zero, but of course it should be, and maybe the 4.07 power is supposed to be theoretically 4. I don't have any references on trans design to check this out. I could email the chart to anyone interested.

The Magnatek trans I thought I had is already wired into an amp and I didn't want to unsolder and remove it. Also, the trans I tested I got on sale for $1 from Angela Instruments (I bought 10!), so I wasn't risking much.

Page 2 of 3