Full-Wave Bridge Rectifier Filter + Regulator DC Power supply

millwood

Joined Dec 31, 1969
0
The issue of whether the transformer parasitics would be material was first mentioned in connection with this thread. For a 60Hz transformer at the power level under discussion in this thread and in a capacitor input filter, the transformer parasitics will always be material.
*always*? really?

I have not encountered one single such case, but I guess I missed the memo, :).

Even though you may not have encountered them, they exist.
I don't dispute the fact that they do. I dispute the notion that they are *always* material.

In the circuit in this thread, run the simulation without the series 37Ω, but with the 5.6 mH series inductance.
I don't dispute the methodology. I just have doubts about the measurements. I have a Signal transformer, dpc-12-2000, 24vac, 12vct@2amp. a small (tiny) transformer. It measures less than 1ohm dc resistance on the 2ndary (or 0.5ohm dc resistance per 2ndary winding), and about 60ohm dc resistance on the primary.

I put the primary in serial with a 25w 110v light bulb and shorted the 2ndary. I got about 10vac across the primary.

If the transformer were a 60ohm resistor, the reading would have been 11 - 12v.

doesn't that suggest that the transformer is fairly close to be ideal? and the thing is just the size of a small chicken egg.

if you have seen a 1mh or a 2mh choke, you will absolutely question the reading of 5.6mh leakage inductance from a "tiny" transformer too.

those tube guys would be extremely delighted to get that much inductance out of a tiny transformer / choke, leakage or otherwise.
 

The Electrician

Joined Oct 9, 2007
2,971
*always*? really?

I have not encountered one single such case, but I guess I missed the memo, :).
What I said was "For a 60Hz transformer at the power level under discussion in this thread and with a capacitor input filter, the transformer parasitics will always be material."

Since the load current in the circuit under discussion is about 4 mA, and the diode currents are about 16 mA RMS, 5 mA avg, even the transformer whose measurements I first gave is overpowered.

What I'm saying, and I repeat to be absolutely clear, is that a 60 Hz transformer of power rating commensurate with the load in the circuit under discussion in this thread, will always have parasitics that will have a substantial effect on the circuit performance. Transformers that small always have poor regulation, and that means disproportionately large parasitics.



I don't dispute the fact that they do. I dispute the notion that they are *always* material.
I didn't say that they are "always" material, without qualification. I was very clear that I was referring to the circuit in this thread. I said that a transformer suitable (of commensurate power rating) for this thread's circuit will always have parasitics that have a material effect.



I don't dispute the methodology. I just have doubts about the measurements. I have a Signal transformer, dpc-12-2000, 24vac, 12vct@2amp. a small (tiny) transformer. It measures less than 1ohm dc resistance on the 2ndary (or 0.5ohm dc resistance per 2ndary winding), and about 60ohm dc resistance on the primary.
The spec sheet shows this as a 24 VA rated transformer. That is WAY overpowered for this circuit, which is only supplying about 1/10 watt to the load.

I put the primary in serial with a 25w 110v light bulb and shorted the 2ndary. I got about 10vac across the primary.

If the transformer were a 60ohm resistor, the reading would have been 11 - 12v.
You've probably calculated the resistance of the bulb from its 25W rating, but that's with 120 VAC applied. Incandescent bulb resistance is a strong non-linear function of the applied voltage. You're applying 120-10 = 110 volts. Rather than infer the current from the bulb rating, why don't you measure the actual current in the transformer primary?

But the transformer shouldn't be a 60Ω resistor, anyway. If this is a 24 volt output transformer, 120 VAC in, then the turns ratio should be about 5:1. Therefore, the 1Ω secondary resistance should be reflected to the primary as 25Ω. Adding that to the 60Ω primary, the primary should look like 85Ω, not 60Ω. And this neglects the effect of whatever leakage reactance there is.

doesn't that suggest that the transformer is fairly close to be ideal? and the thing is just the size of a small chicken egg.
No, because the impedance seen at the primary with the secondary shorted should be more than 85Ω, not 60Ω.

if you have seen a 1mh or a 2mh choke, you will absolutely question the reading of 5.6mh leakage inductance from a "tiny" transformer too.

those tube guys would be extremely delighted to get that much inductance out of a tiny transformer / choke, leakage or otherwise.
Yes, I've seen such chokes and having seen them doesn't lead me to question the transformer measurement; I accept what the meter says. I also know from experience and from examples in texts that such a leakage is entirely typical.

You're probably referring to power inductors.

The fact that a physically small component has a large inductance is not particularly significant. I have a 7 henry inductor that is in a molded cylinder, 30 mm in diameter and 18 mm tall. It's the current rating that determines how big an inductor will be (along with the inductance, of course).

The primary inductance of the very small transformer I've been referring to is 13 henries, with the secondary not shorted. That's quite typical, but it couldn't be used as a power inductor because there's no air gap in the magnetic path.

I have a Stancor transformer rated at 12 VA, P/N LB-1224. With the primaries jumpered for 120VAC operation and the secondaries jumpered for 24 VAC operation, the primary DC resistance measures 50.53Ω and the secondary measures .92Ω. This is the closest thing I have to the transformer you measured. The winding resistances are very similar, though its rating is only half yours, and the construction is different. Specs are here:
http://www.stancor.com/wrdstc/pdfs/Catalog_2006/Pg_025_26.pdf


I also have a Calrad transformer rated at 24 VAC @ 1 amp with EI construction, which is what your Signal transformer appears to use.

It has a primary DC resistance of 67.4Ω and a secondary resistance of 3.19Ω.

None of these transformers, your Signal, my Stancor and Calrad, are commensurate with the power level of the circuit in this thread; they're much too large. Therefore they don't provide counterexamples to my statement that "a transformer suitable (of commensurate power rating) for this thread's circuit will always have parasitics that have a material effect."

Signal's home page shows that they make transformers down to the .1 VA range. A commensurate transformer for the circuit in this thread would be something like .5 VA.

To help you believe that small transformers like these can have leakage inductances in the millihenry range, I've attached 3 photos of the LCR meter's measurement at the secondaries of the three transformers with the primary shorted.

The first is the very small transformer I first mentioned in this thread.

The second is the Stancor transformer mentioned above.

The third is the Calrad transformer.
 

Attachments

millwood

Joined Dec 31, 1969
0
What I'm saying, and I repeat to be absolutely clear, is that a 60 Hz transformer of power rating commensurate with the load in the circuit under discussion in this thread, will always have parasitics that will have a substantial effect on the circuit performance.
I just find that very hard to believe, and it hasn't been my experience

And this neglects the effect of whatever leakage reactance there is.
I guess the point of my exercise is to say that since the voltage measured off the transformer isn't that far off the one calculated assuming it being a pure resistor, the leakage inductance cannot be that big of a deal.
 

millwood

Joined Dec 31, 1969
0
I did some quick googling on this leakage inductance thing. here is a summary of what I found:

1) leakage inductance is obviously dependent on a lot of factors, including configuration, size, core materials, etc. in general, the larger the transformer is, the higher the leakage inductance is.

2) typical transformers are designed so that the leakage inductance is no more than 10% of the primary inductance.

3) typical values of leakage inductance are hard to come by, but for those torroidal transformers used in smps, their values are from a few nh to a few uh. large power transformers (a few kw to hundreds of kw) may have leakage inductance in the mh range, but then their primary inductance is huge too.

now I think about it, is it possible that the "m" in the "mh" reading means "micro" henry (uh), not milli-henry (mu)? that makes more sense number-wise.
 

The Electrician

Joined Oct 9, 2007
2,971
I just find that very hard to believe, and it hasn't been my experience
I guess you just haven't designed a power supply to supply .1 watt @ 24 volts.

Neither have I, but I know that very small 60Hz transformers such as the one I have, and even smaller ones, have very poor regulation and the transformer parasitics are proportionately much larger than in bigger transformers.

You should tear open a small wall wart and remove the little transformer to play with. Some of the newer wall warts have switchers in them. You can tell them because they aren't very heavy. Get a heavy one and it should have a 60 Hz transformer inside.

I guess the point of my exercise is to say that since the voltage measured off the transformer isn't that far off the one calculated assuming it being a pure resistor, the leakage inductance cannot be that big of a deal.
I assume you're talking about your Signal transformer. For an EI style core, and at the power level you have, the leakage reactance is probably substantially less than the resistance.

For example, the Calrad transformer I measured has a secondary referred leakage reactance of .79Ω, but a resistance of 6.5Ω. The resistance is more than 10 times the leakage reactance.

But the Stancor transformer I measured has a two bobbin flat core construction. The secondary referred leakage reactance is 1.38Ω and the resistance is 1.69Ω. They're almost the same, and the leakage reactance should be included in a model to get good results. The leakage reactance with that kind of construction is always higher than with purely concentric wound construction.

In very large transformers such as the 100kva one I cited, the leakage reactance was 1.51Ω and the resistance was only .58Ω. Here the leakage reactance is 3 times the resistance.

The relative proportions of leakage reactance to resistance vary widely with size and construction of a transformer. Some transformers are deliberately made with large leakage inductance to serve as a current limiting mechanism, such as neon sign transformers.

Back to the very small transformer I've been discussing, here are 2 more images from the LCR meter.

The first shows the primary inductance with the secondary open.

The second shows the secondary inductance with the primary open.

Notice that the secondary magnetizing inductance is 1.8 H, and the leakage inductance measured at the secondary is .0056 H. The leakage inductance is much less than 1% of the magnetizing inductance.

In both of these images, the resistance is due to core loss and has nothing to do with copper loss.
 

Attachments

The Electrician

Joined Oct 9, 2007
2,971
now I think about it, is it possible that the "m" in the "mh" reading means "micro" henry (uh), not milli-henry (mu)? that makes more sense number-wise.
Are you referring to the LCR meter images I posted? If so, then, no, it's not possible that "mH" means microhenries. When the meter measures microhenries, it shows "μH".

See the attached image showing the measurement of a 33 uH inductor.
 

Attachments

millwood

Joined Dec 31, 1969
0
In very large transformers such as the 100kva one I cited, the leakage reactance was 1.51Ω and the resistance was only .58Ω. Here the leakage reactance is 3 times the resistance.
now, 1.5ohm at 50hz means 4.8mh leakage inductance for a 100kw transformer.

that's a number that I can believe in.

5mh for a tinny weeny transformer?

no way.
 

millwood

Joined Dec 31, 1969
0
here is something to think about:

http://www.plitron.com/audio_3050.asp

it is a high quality audio transformer from Plitron. 40h primary inductance, 10mh of leakage inductance, 6" in diameter, and weighs a good 12 pounds.

here is another: http://www.plitron.com/audio_4002.asp: 535h primary inductance, 7mh leakage inductance, weighing a cool 5 pounds.

with this kind of numbers, I really question if you can have 5mh of leakage inductance out of a tiny transformer. or maybe the lrc meter needs a calibration?
 

The Electrician

Joined Oct 9, 2007
2,971
5mh for a tinny weeny transformer?

no way.
What are you saying? That you don't believe the measurements I showed you?

Take your Signal transformer and short the primary.

Put a 10 uF high quality (non-electrolytic) capacitor in series with the secondary. Excite that series combination with an audio frequency signal and tune until you find the resonance (should be around 1 kHz, plus or minus a factor of ten). Then calculate the apparent inductance of the secondary winding with the primary shorted.
 

millwood

Joined Dec 31, 1969
0
What are you saying? That you don't believe the measurements I showed you?
I am saying that there is something not right about those measurements. the numbers don't make sense.

Take your Signal transformer and short the primary.
a little bit too complicated. Measure the serial resistance of the winding, Rseq, and put a resistor, Rs, in serial with the winding. apply a ac signal, Vin of f hz.

measure the voltage on Rs as Vs. those are the parameters you would need to know to work out the inductance.
 
I am saying that there is something not right about those measurements. the numbers don't make sense.
Explain why they don't make sense.

a little bit too complicated. Measure the serial resistance of the winding, Rseq, and put a resistor, Rs, in serial with the winding. apply a ac signal, Vin of f hz.

measure the voltage on Rs as Vs. those are the parameters you would need to know to work out the inductance.
So go ahead and make the measurement and show your calculation for the inductance.
 
we have been talking about that since Page 4 of this thread. :)
We've been talking about a lot of things. Remind me which ones that don't make sense to you.

One thing I will discuss is this. You think that it's reasonable for a 100kva transformer to have leakage inductances in the mH range, but not for a small transformer.

Consider that a large transformer has a large core cross sectional area, so that the number of turns on a winding to support a given voltage is lower than for a small core. Further consider that inductance varies as the square of the number of turns.

Very small transformers have a large number of turns of small wire on their windings. That's why they have winding inductances in the henries to tens of henries range, and can easily have leakage inductances in the millihenry range.
 
Top