Volts-per-turn vs Frequency, Inductance, and no load / full load current

[P E Schoen]

I have a toroid core that I previously wound based on my idea for using 600
Hz or 1kHz or so to get higher power. The core is about 1.6" deep and 3.4"
diameter with a 2" diameter hole. This is probably the 80 VA core from
http://www.toroid.com/standard_transformers/transformer_kits/transformer_kits.htm,
which is rated at 0.12V/turn. I wound it with about 100 turns of #18 AWG as
a secondary, and two windings of 8 turns each of #10 AWG as a primary. If
volts/turn is proportional to frequency, this should be 2 volts/turn at 1000
Hz, so I should be able to apply 16 VAC to the primary in a push-pull
configuration, using a 12 volt supply. And I should be able to get 150 VAC
out.

I measured the inductance of each primary coil as 180 uH, and the secondary
is 3.2 mH. I did a simulation with a FWB rectifier and 680 uF filter
capacitor, and I got 19 to 27 amps in each MOSFET during the ON cycle, and
1.57A at 157V into the 100 ohm load, for 246 watts. The input power is 250
watts, for 98.6% efficiency (probably unrealistic). So far, so good.

Now I disconnect the secondary, and my input is 120 watts. This is 40 watts
in each of the primary coils which I have designated as 10 uOhms resistance.
The MOSFETs (IRF2903ZS) dissipate only 165 mW. The 12V source is supplying
11 A RMS and it has a 2 mOhm resistance, so that's not where the extra 40
watts is coming from. The current through the source varies between +20
and -8 amps on one cycle and 6.2 to -24A on the other.

My question is if this is a normal magnetizing current for such a
transformer. When I change the parameters to what the transformer probably
would have at its original rating of 120 VAC, 60 Hz, with 0.12 V/turn, I get
3.2 H, and the magnetization current is about 166 mA RMS. The expected
current at its 80 VA design is 80/120 or 667 mA, about 4 times the no load
value.

And my modified transformer should be 1.33 kVA with a primary current of 111
amps, which is ten times the magnetizing current.

Magnetic theory is not my strong point. Are these figures about right? Is
this a reasonable design for a 1 kVA DC-DC converter? I like the fact that
the MOSFETs run so cool. Size and weight are not a huge factor. The iron
core toroid should be much more rugged than a ferrite design, and may be
similar in cost, especially for small quantities. And the electronics should
be very simple. A 500 watt DC-DC converter costs about $250, and two of them
would be 2.4 x 1 x 4.6 inches, and weigh about 0.5 kG or 1 .1 pound. But at
best it's 90% efficient so it would need to dissipate 100 watts of power.
Mine would be about 10 times larger and heavier, but cost about 1/5 that of
the Lambda converter: http://us.tdk-lambda.com/lp/ftp/Specs/paf500f.pdf.

If and when I finish a practical design and build and test this beast, there
are unknown factors that may come into play, such as losses at the higher
frequency. But AFAIK toroids like this can be used up to 2 kHz. According to
the following engineering bulletin
http://www.stacoenergy.com/pdf/updated/designengine.pdf, Staco variable
transformers can be used up to 2000 Hz with little or no derating. But of
course that does not imply a higher power output, so that may be misleading
or not directly applicable. However, I have seen data elsewhere that says
the usual (or somewhat thinner) nickel-steel laminations can be used at
least to 400 Hz and even over 1000 Hz.

Following is my LTSpice circuit:

Paul

===================== 12V-160V.asc =====================

Version 4
SHEET 1 880 680
WIRE 304 128 224 128
WIRE 320 128 304 128
WIRE 432 128 384 128
WIRE 496 128 432 128
WIRE 560 128 496 128
WIRE 592 128 560 128
WIRE 96 144 -160 144
WIRE 224 192 224 128
WIRE 320 208 272 208
WIRE 432 208 432 128
WIRE 432 208 384 208
WIRE 592 208 592 128
WIRE 496 224 496 128
WIRE -496 240 -576 240
WIRE 96 240 96 224
WIRE 96 240 -496 240
WIRE 96 256 96 240
WIRE 272 272 272 208
WIRE 272 272 224 272
WIRE 320 272 272 272
WIRE 416 272 384 272
WIRE -320 288 -384 288
WIRE -80 288 -256 288
WIRE -384 304 -384 288
WIRE -256 336 -256 288
WIRE 96 336 0 336
WIRE -496 368 -496 240
WIRE 304 368 304 128
WIRE 320 368 304 368
WIRE 416 368 416 272
WIRE 416 368 384 368
WIRE 496 368 496 288
WIRE 496 368 416 368
WIRE 592 368 592 288
WIRE 592 368 496 368
WIRE -160 384 -160 144
WIRE 0 384 0 336
WIRE -576 416 -576 240
WIRE 592 432 592 368
WIRE -320 464 -320 288
WIRE -208 464 -320 464
WIRE -80 464 -80 288
WIRE -48 464 -80 464
WIRE -576 528 -576 480
WIRE -496 528 -496 448
WIRE -496 528 -576 528
WIRE -384 528 -384 384
WIRE -384 528 -496 528
WIRE -256 528 -256 416
WIRE -256 528 -384 528
WIRE -160 528 -160 480
WIRE -160 528 -256 528
WIRE -80 528 -160 528
WIRE 0 528 0 480
WIRE 0 528 -80 528
WIRE -80 608 -80 528
FLAG -80 608 0
FLAG 592 432 0
FLAG 560 128 Vout
SYMBOL ind2 80 128 R0
SYMATTR InstName L1
SYMATTR Value 180µ
SYMATTR Type ind
SYMATTR SpiceLine Rser=10u
SYMBOL ind2 80 240 R0
SYMATTR InstName L2
SYMATTR Value 180µ
SYMATTR Type ind
SYMATTR SpiceLine Rser=10u
SYMBOL ind2 240 176 M0
SYMATTR InstName L3
SYMATTR Value 32m
SYMATTR Type ind
SYMATTR SpiceLine Rser=200u
SYMBOL nmos -208 384 R0
WINDOW 3 56 102 Left 2
SYMATTR InstName M1
SYMATTR Value IRF2903ZS
SYMBOL nmos -48 384 R0
WINDOW 3 56 102 Left 2
SYMATTR InstName M2
SYMATTR Value IRF2903ZS
SYMBOL voltage -496 352 R0
WINDOW 123 0 0 Left 2
WINDOW 39 24 132 Left 2
SYMATTR SpiceLine Rser=2m
SYMATTR InstName V1
SYMATTR Value 12
SYMBOL diode 384 288 M270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D2
SYMATTR Value MUR460
SYMBOL diode 320 224 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D3
SYMATTR Value MUR460
SYMBOL polcap 480 224 R0
WINDOW 3 24 64 Left 2
SYMATTR Value 560µ
SYMATTR InstName C1
SYMATTR Description Capacitor
SYMATTR Type cap
SYMATTR SpiceLine V=600 Irms=2.9 Rser=0.018 Lser=0
SYMBOL res 576 192 R0
SYMATTR InstName R1
SYMATTR Value 100
SYMBOL voltage -384 288 R0
WINDOW 123 0 0 Left 2
WINDOW 39 -43 57 Left 2
WINDOW 3 -192 268 Left 2
SYMATTR SpiceLine Rser=50m
SYMATTR Value PULSE(0 10 250u 10n 10n 495u 1000u 100)
SYMATTR InstName V2
SYMBOL voltage -256 320 R0
WINDOW 123 0 0 Left 2
WINDOW 39 -43 57 Left 2
WINDOW 3 -322 268 Left 2
SYMATTR SpiceLine Rser=50m
SYMATTR Value PULSE(0 10 750.5u 10n 10n 495u 1000u 100)
SYMATTR InstName V3
SYMBOL diode 320 144 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D1
SYMATTR Value MUR460
SYMBOL diode 384 384 M270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D4
SYMATTR Value MUR460
SYMBOL polcap -592 416 R0
WINDOW 3 24 64 Left 2
SYMATTR Value 2200µ
SYMATTR InstName C2
SYMATTR Description Capacitor
SYMATTR Type cap
SYMATTR SpiceLine V=25 Irms=2.9 Rser=10m Lser=0
TEXT 72 384 Left 2 !K1 L1 L2 L3 1
TEXT -584 624 Left 2 !.tran 0 200m 0 1u startup
TEXT 160 440 Left 2 ;Primary 2x8 turns 2V/turn at 1000 Hz
TEXT 160 472 Left 2 ;25 A peak magnetizing current for 180u 1kHz
TEXT 160 496 Left 2 ;27 A peak current at 100 ohm load

 

[John S]

I have a toroid core that I previously wound based on my idea for using
600 Hz or 1kHz or so to get higher power. The core is about 1.6" deep
and 3.4" diameter with a 2" diameter hole. This is probably the 80 VA
core from
http://www.toroid.com/standard_transformers/transformer_kits/transformer_kits.htm,
which is rated at 0.12V/turn. I wound it with about 100 turns of #18 AWG
as a secondary, and two windings of 8 turns each of #10 AWG as a
primary. If volts/turn is proportional to frequency, this should be 2
volts/turn at 1000 Hz, so I should be able to apply 16 VAC to the
primary in a push-pull configuration, using a 12 volt supply. And I
should be able to get 150 VAC out.

I measured the inductance of each primary coil as 180 uH, and the
secondary is 3.2 mH. I did a simulation with a FWB rectifier and 680 uF
filter capacitor, and I got 19 to 27 amps in each MOSFET during the ON
cycle, and 1.57A at 157V into the 100 ohm load, for 246 watts. The input
power is 250 watts, for 98.6% efficiency (probably unrealistic). So far,
so good.

Now I disconnect the secondary, and my input is 120 watts.

I hope you mean 120VA! From your LTSpice example I get only a couple of
watts at no load.

Cheers,
John

 

[Tim Williams]

I have a toroid core that I previously wound based on my idea for using
600 Hz or 1kHz or so to get higher power. The core is about 1.6" deep
and 3.4" diameter with a 2" diameter hole.

Ok, so it has a 1.6 x 0.7" cross section, or 723 mm^2. At Bmax = 1.2T,
that's 867uWb/t maximum flux.

I wound it with about 100 turns of #18 AWG as a secondary, and two
windings of 8 turns each of #10 AWG as a primary. If volts/turn is
proportional to frequency, this should be 2 volts/turn at 1000 Hz, so I
should be able to apply 16 VAC to the primary in a push-pull
configuration, using a 12 volt supply. And I should be able to get 150
VAC out.

8 turns gets you 6937uWb flux, which is 30.8V RMS sine at 1kHz. If you're
driving it with a square wave, that becomes 27.7V peak. If you're driving
it in push-pull, the total across 16 turns is 55.5V, etc.

I measured the inductance of each primary coil as 180 uH, and the
secondary is 3.2 mH.

The magnetic path length is roughly 215mm. At a typical average
permeability of 10k, you'll have an inductivity of 42uH/t^2, or 2.7mH
primary inductance.

Steel is very nonlinear, so you might expect the average value to be as high
as four times that, and the initial value to be perhaps only a tenth -- that
you measured 180uH suggests your initial permeability is around 670, which
is reasonable.

When it comes to transformers, all you need to check are:
1. If the average permeability is over 1000 or so, magnetizing current will
simply be too small to care about. Don't even measure or calculate
inductance, because as you can see, it varies like crazy. A transformer may
bear superficial resemblance to an inductor, but it serves a very different
purpose -- power conversion, not energy storage.
2. Make sure that you have enough turns and core area, so that the flux
density remains below saturation at the highest voltage and lowest frequency
you'll be running at. This is calculated:
N = Vrms / (4.44 * F * Bmax * Ae)
where N is the number of turns, Vrms is the RMS sine wave voltage applied to
those turns, F is frequency, Bmax is saturation flux density (usually 1.2T
for steel), and Ae is the cross sectional area of the core.

You can calculate inductance as an exercise, but like I said, it's not
representative of actual results:
L = N^2 * mu_0 * mu * Ae / l_e
where mu_0 is the permeability of free space (~= 0.001256 uH/mm), mu is the
permeability of the core material, and l_e is the magnetic path length.
Without the N^2, this is A_L, the inductivity of the core.

If an air gap is present (typical of E-I choke designs; obviously, difficult
to arrange for stripwound toroids), the path length changes. Specifically,
compared to air, the core has an equivalent path length of l_e / mu, which
clearly becomes negligible when mu is thousands. So, for an air gap of just
a few tenths of a milimeter, the airgap dominates the inductance, inductance
drops, and energy storage goes up -- because the saturation flux remains
approximately constant, while the inductance drops linearly (for small
gaps), and current at saturation rises linearly. Energy goes as current
squared, so energy rises linearly also, which makes sense because air is a
good way to store magnetic energy.

I did a simulation with a FWB rectifier and 680 uF filter capacitor,
and I got 19 to 27 amps in each MOSFET during the ON cycle, and 1.57A
at 157V into the 100 ohm load, for 246 watts. The input power is 250
watts, for 98.6% efficiency (probably unrealistic). So far, so good.

If you wish to perform a simulation to determine what voltage and frequency
you can run, you need a nonlinear core model. The simplest way is a voltage
integrator feeding diodes, feeding back the diode current as terminal
current. Because the diode voltage is phase shifted by the integration, a
resistance in parallel with the diodes looks inductive at the terminals; a
capacitor in parallel with the diodes looks resistive at the terminals. In
this way, eddy currents can also be modeled.

If you just want to get the required parameters (volts/turn or whatever),
you're better off doing it analytically with the simple ratio I provided
above. SPICE doesn't solve your problems for you, it just tells you how
wrong your guess is...

My question is if this is a normal magnetizing current for such a
transformer.

In short, no. Initial permeability is way off, that's all.

this a reasonable design for a 1 kVA DC-DC converter? I like the fact
that the MOSFETs run so cool. Size and weight are not a huge factor.
The iron core toroid should be much more rugged than a ferrite design,
and may be similar in cost, especially for small quantities. And the
electronics should be very simple.

Not really.

The MOSFETs run just as cool in a properly designed ~100kHz circuit, and the
ferrite core will be cheaper, smaller and run cooler. (Typical ETD25ish
core will do well over 250W at ~100kHz, and might cost $10 from most
suppliers, not $50.)

The ferrite may be brittle, but if you dropped that big toroid on a concrete
floor, the sharp corners of the core would bite right through the windings.
The smaller package means it's easier to put inside a chassis which deflects
blows, or it could be potted without excessive expense (still, potting
compound isn't cheap!).

The electronics will be exactly as simple either way if you don't mind a
design without any ruggedization. I haven't looked at your LTSpice circuit
but I'm guessing it's little more than a chopper. What happens if someone
shorts the output? Excessive input voltage? Reverse input voltage? How
about EMI/RFI performance -- snubbing, filtering, common mode noise?

A current mode controller isn't too difficult to design and build (or get an
OTS chip to do it), but all the protection and filtering gets tedious.
Pretty soon you have a bucket of $1 parts that, for whatever reason, costs
about,

A 500 watt DC-DC converter costs about $250, and two of them would be
2.4 x 1 x 4.6 inches, and weigh about 0.5 kG or 1 .1 pound.

....About that!

But at best it's 90% efficient so it would need to dissipate 100 watts
of power.

The efficiency could be higher, but that's not bad as conversion goes.

Tim

 

[P E Schoen]

"John S" wrote in message

On 2/25/2012 5:35 PM, P E Schoen wrote:

I hope you mean 120VA! From your LTSpice example I get only a
couple of watts at no load.

Duh

When I deleted the connection from the secondary to the diodes, the node
numbers shifted. I changed the schematic by adding labels to the important
nodes. Now it makes a lot more sense!

Thanks,

Paul

 

[Fred Abse]

I have a toroid core that I previously wound based on my idea for using 600
Hz or 1kHz or so to get higher power. The core is about 1.6" deep and 3.4"
diameter with a 2" diameter hole. This is probably the 80 VA core from
http://www.toroid.com/standard_transformers/transformer_kits/transformer_kits.htm,
which is rated at 0.12V/turn.

About 8 turns per (RMS) volt, which is rule of thumb for "cooking"
transformer iron, with 60Hz sinusoidal excitation. Probably won't be any
good at 600, let alone 1000Hz.

Square wave makes things worse. The iron won't be suitable for your
application. You need thin (.003") mumetal or ferrite.

I wound it with about 100 turns of
#18 AWG as
a secondary, and two windings of 8 turns each of #10 AWG as a primary.
If volts/turn is proportional to frequency, this should be 2 volts/turn
at 1000 Hz, so I should be able to apply 16 VAC to the primary in a
push-pull configuration, using a 12 volt supply. And I should be able to
get 150 VAC out.

Running your simulation, I see 1600 amps peak in the FETs, and the primary,
under load.
That won't reflect real life.

The expected
current at its 80 VA design is 80/120 or 667 mA, about 4 times the no load
value.
And my modified transformer should be 1.33 kVA with a primary current of 111
amps, which is ten times the magnetizing current.

VA in one direction equals VA in the other direction. TANSTAAFL.

Magnetic theory is not my strong point.

I won't argue with that ;-)

You need a better transformer model, including the characteristics of the
iron. There's an ideal transformer model that comes with LTspice, that you
can adapt by shunting the primary with a hysteretic model inductor (RTFM).

In a nutshell:

Your model is not adequate.

Your iron isn't up to the job.

Use ferrite. Something like Epcos E cores in N97 material, or a toroid in
N49.

 

[John S]

in message


Duh

When I deleted the connection from the secondary to the diodes, the node
numbers shifted. I changed the schematic by adding labels to the
important nodes. Now it makes a lot more sense!

Thanks,

Paul

Yes, BTDT. I now label the important nodes that I want to track. Once
labelled, they don't change.

 

[Fred Abse]

Running your simulation, I see 1600 amps peak in the FETs, and the primary,
under load.

That was down to the goddamn Greek mu. There was one I hadn't spotted
(C1). They don't travel well on Usenet. Apart from inrush, current
now looks sensible.

You really need to check the "convert Greek mu" in Netlist options for
stuff posted to Usenet.

I still don't think it will work as expected at 600/1000 with that iron.

 

[P E Schoen]

"Fred Abse" wrote in message

That was down to the goddamn Greek mu. There was one I hadn't
spotted (C1). They don't travel well on Usenet. Apart from inrush,
current now looks sensible.

You really need to check the "convert Greek mu" in Netlist options for
stuff posted to Usenet.

I didn't know about that. It is now checked.

I still don't think it will work as expected at 600/1000 with that iron.

I agree that this may not be a very practical design. I made a new circuit
which limits the MOSFET current to about 90 amps, but now instead of having
1000 amps inrush, I have about 1000 watts. I used a 30 ohm load to get 750
watts output.

So I added an inductor (470 mH) to the output, which reduces the power to
acceptable levels during startup, but it also adds weight, size, and cost.
The current limit transistors might not be needed in this case.

I also tried it with an extreme overload of one ohm. The power in each
MOSFET becomes 416 watts. Not good. And the output current is 6.7 amps
rather than the normal 4.5. So I really need better overcurrent and short
circuit protection.

The high start-up current is due to the output capacitor. So I added a
current limiting resistor and discharge diode, which seems to work
acceptably.

Maybe I will still build the basic design and try it using the toroid, just
to see how well (or not) it works. The current limit transistors are not
really needed if I limit the capacitor charge current. So that will simplify
the circuit and still keep things in a safe range. I can try a light load
first, then go for the maximum. I have some 2000 watt heaters 220V which are
about 26 ohms, so that will give me close to the 750 watt target. I think
I'll enclose it in a steel box in case anything explodes.

Then, I change the transformer to one using ferrite. I bought some big E
cores and bobbins some time ago. They are something like 3" square, and
probably good for several kW. I'll probably design it for 160 / 320 VDC and
12 VDC input so I can use a single battery to get enough voltage for the DC
bus in my 2 HP 240V 3 phase motor drive. I just want to see how well the
combination runs a 1.5 HP motor. And it may be handy to have a portable
three phase source.

Thanks,

Paul

(Here's the modified circuit:
=================================================

Version 4
SHEET 1 896 680
WIRE -96 80 -144 80
WIRE 128 80 -96 80
WIRE 304 128 224 128
WIRE 320 128 304 128
WIRE 416 128 384 128
WIRE 560 128 496 128
WIRE 592 128 560 128
WIRE 224 160 224 128
WIRE -592 176 -672 176
WIRE -544 176 -592 176
WIRE -256 176 -544 176
WIRE 16 176 -256 176
WIRE 128 176 128 160
WIRE 128 176 16 176
WIRE 128 192 128 176
WIRE 496 192 496 128
WIRE 320 208 272 208
WIRE 416 208 416 128
WIRE 416 208 384 208
WIRE 592 208 592 128
WIRE -352 224 -368 224
WIRE -64 224 -272 224
WIRE 16 240 16 176
WIRE -432 256 -496 256
WIRE 496 256 448 256
WIRE -368 272 -368 224
WIRE -256 272 -256 176
WIRE 128 272 112 272
WIRE -144 288 -144 80
WIRE 112 288 112 272
WIRE 224 288 224 240
WIRE 272 288 272 208
WIRE 272 288 224 288
WIRE 320 288 272 288
WIRE 416 288 384 288
WIRE 496 288 496 256
WIRE -496 304 -496 256
WIRE 448 304 448 256
WIRE -64 336 -64 224
WIRE 48 336 -64 336
WIRE -592 368 -592 176
WIRE -432 368 -432 256
WIRE -192 368 -400 368
WIRE 48 368 48 336
WIRE 64 368 48 368
WIRE 304 368 304 128
WIRE 320 368 304 368
WIRE 416 368 416 288
WIRE 416 368 384 368
WIRE 448 368 416 368
WIRE 496 368 448 368
WIRE 592 368 592 288
WIRE 592 368 496 368
WIRE 704 368 592 368
WIRE 160 384 112 384
WIRE 704 400 704 368
WIRE -672 416 -672 176
WIRE -256 416 -256 352
WIRE -240 416 -256 416
WIRE -144 416 -144 384
WIRE -144 416 -160 416
WIRE 16 416 16 320
WIRE 32 416 16 416
WIRE 160 416 160 384
WIRE 160 416 112 416
WIRE -144 432 -144 416
WIRE -400 448 -400 368
WIRE -400 448 -432 448
WIRE -256 448 -256 416
WIRE 16 448 16 416
WIRE 16 448 0 448
WIRE 160 448 160 416
WIRE -672 528 -672 480
WIRE -592 528 -592 448
WIRE -592 528 -672 528
WIRE -496 528 -496 384
WIRE -496 528 -592 528
WIRE -368 528 -368 352
WIRE -368 528 -496 528
WIRE -320 528 -320 496
WIRE -320 528 -368 528
WIRE -144 528 -144 512
WIRE -144 528 -320 528
WIRE -64 528 -64 496
WIRE -64 528 -144 528
WIRE 48 528 -64 528
WIRE 160 528 48 528
WIRE 48 592 48 528
FLAG 48 592 0
FLAG 704 400 0
FLAG 560 128 Vout
FLAG -544 176 in
FLAG -96 80 dr1
FLAG 112 272 dr2
SYMBOL ind2 112 64 R0
SYMATTR InstName L1
SYMATTR Value 180u
SYMATTR Type ind
SYMATTR SpiceLine Rser=10u
SYMBOL ind2 112 176 R0
SYMATTR InstName L2
SYMATTR Value 180u
SYMATTR Type ind
SYMATTR SpiceLine Rser=10u
SYMBOL ind2 240 144 M0
WINDOW 0 -19 22 Left 2
WINDOW 3 -46 46 Left 2
SYMATTR InstName L3
SYMATTR Value 32m
SYMATTR Type ind
SYMATTR SpiceLine Rser=200u
SYMBOL nmos -192 288 R0
WINDOW 3 53 65 Left 2
SYMATTR Value IRF2903ZS
SYMATTR InstName M1
SYMBOL nmos 64 288 R0
WINDOW 3 56 67 Left 2
WINDOW 0 59 38 Left 2
SYMATTR Value IRF2903ZS
SYMATTR InstName M2
SYMBOL voltage -592 352 R0
WINDOW 123 0 0 Left 2
WINDOW 39 24 132 Left 2
SYMATTR SpiceLine Rser=2m
SYMATTR InstName V1
SYMATTR Value 12
SYMBOL diode 384 304 M270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D2
SYMATTR Value MUR460
SYMBOL diode 320 224 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D3
SYMATTR Value MUR460
SYMBOL polcap 480 192 R0
WINDOW 3 32 56 Left 2
SYMATTR Value 470u
SYMATTR InstName C1
SYMATTR Description Capacitor
SYMATTR Type cap
SYMATTR SpiceLine V=600 Irms=2.9 Rser=0.018 Lser=0
SYMBOL res 576 192 R0
SYMATTR InstName R1
SYMATTR Value 30
SYMBOL voltage -496 288 R0
WINDOW 123 0 0 Left 2
WINDOW 39 -43 57 Left 2
WINDOW 3 -194 303 Left 2
SYMATTR SpiceLine Rser=50m
SYMATTR Value PULSE(0 10 250u 10n 10n 450u 1000u 100)
SYMATTR InstName V2
SYMBOL voltage -368 256 R0
WINDOW 123 0 0 Left 2
WINDOW 39 -48 62 Left 2
WINDOW 3 -322 296 Left 2
SYMATTR SpiceLine Rser=50m
SYMATTR Value PULSE(0 10 762u 10n 10n 450u 1000u 100)
SYMATTR InstName V3
SYMBOL diode 320 144 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D1
SYMATTR Value MUR460
SYMBOL diode 384 384 M270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D4
SYMATTR Value MUR460
SYMBOL polcap -688 416 R0
WINDOW 3 24 64 Left 2
SYMATTR Value 10u
SYMATTR InstName C2
SYMATTR Description Capacitor
SYMATTR Type cap
SYMATTR SpiceLine V=25 Irms=2.9 Rser=100m Lser=0
SYMBOL res -160 416 R0
SYMATTR InstName R2
SYMATTR Value 2.5m
SYMBOL res 144 432 R0
SYMATTR InstName R3
SYMATTR Value 2.5m
SYMBOL res -256 208 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R4
SYMATTR Value 100
SYMBOL res -416 464 R180
WINDOW 0 36 76 Left 2
WINDOW 3 36 40 Left 2
SYMATTR InstName R5
SYMATTR Value 100
SYMBOL npn -256 400 M0
SYMATTR InstName Q1
SYMATTR Value 2N3904
SYMBOL npn 0 400 M0
WINDOW 3 -43 97 Left 2
SYMATTR Value 2N3904
SYMATTR InstName Q2
SYMBOL res -144 400 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R6
SYMATTR Value 100
SYMBOL res -240 368 R180
WINDOW 0 36 76 Left 2
WINDOW 3 36 40 Left 2
SYMATTR InstName R7
SYMATTR Value 2k
SYMBOL res 128 400 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R8
SYMATTR Value 100
SYMBOL res 32 336 R180
WINDOW 0 36 76 Left 2
WINDOW 3 36 40 Left 2
SYMATTR InstName R9
SYMATTR Value 2k
SYMBOL ind 400 144 R270
WINDOW 0 32 56 VTop 2
WINDOW 3 5 56 VBottom 2
SYMATTR InstName L4
SYMATTR Value 1n
SYMBOL res 480 272 R0
SYMATTR InstName R10
SYMATTR Value 30
SYMBOL diode 432 368 M180
WINDOW 0 24 64 Left 2
WINDOW 3 7 -24 Left 2
SYMATTR InstName D5
SYMATTR Value MUR460
TEXT 144 304 Left 2 !K1 L1 L2 L3 1
TEXT -584 624 Left 2 !.tran 0 200m 0 1u startup
TEXT 320 456 Left 2 ;Primary 2x8 turns 2V/turn at 1000 Hz
TEXT 320 528 Left 2 ;25 A peak magnetizing current for 180u 1kHz
TEXT 336 560 Left 2 ;77 A peak current at 30 ohm load
TEXT 320 496 Left 2 ;Secondary 100 turns 2V/turn at 1000 Hz

 

[P E Schoen]

"Robert Macy" wrote in message

What is the 12Vdc supply? The best vehicle battery I've seen
barely does 300 to 500A into a dead short. That calculates to
more like 20m, not 2m Rseries. Especially at start up inrush
currents.

Yes, that is more reasonable. And for my prototype I plan to use a booster
battery and connection through a cigarette lighter jack so the contact
resistance and the wires will probably have enough resistance (and
inductance) to limit the surges to under 100A. Also, the output filter
capacitor will probably have higher ESR than what is in the simulation
model. With real components, I would expect less current surges.

My current limiting scheme is fatally flawed because it puts the MOSFETs
into a linear mode which burns a lot of power. As a simple improvement, I
wonder if a 556 dual timer could be set up with a 200 mV threshold and 50A
100mV shunts (2 milliohm) so that the MOSFET would be turned off completely
when 100 amps is reached. But it would need to be latched on until the drive
reverses. This could probably be done by connecting the trigger to the gate
drive signal. But then maybe I should do it all with a PIC, or a PWM
controller like an SG3526.

Mostly I just want to see just what the performance of the toroid will be at
1kHz. So I just need a 12V square wave good for about 50A. A 600 watt
amplifier? Or maybe it would be better to reverse drive the transformer. I
could use my VF motor drive to get 400 Hz, but that's PWM, with a 15 kHz
carrier.

What I may want to look at are spot welding transformers. From what I can
tell, they use a toroid and 1000 Hz, and a 32 kVA (20% duty cycle) unit is
19kG, so a 1kVA continuous version would probably be less than 5kG:
http://www.portablewelders.com/welding-products/welding-transformers-technical-data
http://www.portablewelders.com/welding-products/medium-frequency-welding-transformers

I do not think these are ferrite. They might be powdered iron, but I think
they are tape wound toroids. Here are some that are rated at 50-2500 Hz, and
from 5 to 5000 kVA!
http://www.romanmfg.com/products_weldingtransformers.html

Here is some information on some types of steel laminations used for various
frequency ranges:
http://www.vias.org/eltransformers/lee_electronic_transformers_03_03.html
Even the C-97 power transformer laminations (14 mil) are shown as good up to
about 800 Hz. And the C-95 (5 mil) are good up to 7 kHz for power
applications.

Here is some information on typical toroid transformers
http://sound.westhost.com/xfmr2.htm, where it shows core loss to be about 1%
of VA rating at 60 Hz. I might expect it to be perhaps no more than 20% at
1kHz then. The E-I transformer was about 20% core loss, so clearly not
usable at even 300 Hz.

This seems useful:
http://www.magmet.com/pdf/TransformDesignConsiderat.pdf

For the toroid transformer I have, the no-load core loss at 60 Hz is about
0.5%, so I think it will be good at 1000 Hz:
http://www.toroid.com/custom_transformers/custom_toroidal_transformer.htm

And here are some steel core materials good up to 50 kHz:
http://www.jfe-steel.co.jp/en/products/electrical/supercore/index.html

Paul

 

[P E Schoen]

"P E Schoen" wrote in message

My current limiting scheme is fatally flawed because it puts the
MOSFETs into a linear mode which burns a lot of power. As a
simple improvement, I wonder if a 556 dual timer could be set
up with a 200 mV threshold and 50A 100mV shunts (2 milliohm)
so that the MOSFET would be turned off completely when 100
amps is reached. But it would need to be latched on until the
drive reverses. This could probably be done by connecting the
trigger to the gate drive signal. But then maybe I should do it
all with a PIC, or a PWM controller like an SG3526.

OK, I made a circuit as described with two 555 timers. It seems to work
well, and the current limit is such that it folds back to only a few watts
with an overload (I used 1 ohm). I may need to provide a more stable
reference voltage than the 12V supply, however. The CV for the timer is
about 200 mV, which gives a threshold of 200 mV or 100 amps on the 100mV 50A
shunts. But I don't know if a real 555 will work at such a low voltage. The
schematic for the threshold comparator shows two diode drops to the common
resistor. The CMOS version may have a comparator that includes ground. Some
schematics show both PNP and NPN transistors.

Paul

===========================================================

Version 4
SHEET 1 896 680
WIRE -432 0 -1088 0
WIRE -672 32 -1056 32
WIRE -96 80 -144 80
WIRE 128 80 -96 80
WIRE -992 112 -1024 112
WIRE -672 112 -672 32
WIRE -672 112 -768 112
WIRE 304 128 224 128
WIRE 320 128 304 128
WIRE 416 128 384 128
WIRE 496 128 416 128
WIRE 560 128 496 128
WIRE 592 128 560 128
WIRE -400 144 -720 144
WIRE 224 160 224 128
WIRE -1088 176 -1088 0
WIRE -992 176 -1088 176
WIRE -720 176 -720 144
WIRE -720 176 -768 176
WIRE -672 176 -672 112
WIRE -592 176 -672 176
WIRE -544 176 -592 176
WIRE 128 176 128 160
WIRE 128 176 -544 176
WIRE 128 192 128 176
WIRE 496 192 496 128
WIRE 320 208 272 208
WIRE 416 208 416 128
WIRE 416 208 384 208
WIRE 592 208 592 128
WIRE -112 224 -320 224
WIRE -16 224 -32 224
WIRE -256 240 -768 240
WIRE -432 256 -432 0
WIRE -432 256 -496 256
WIRE 496 256 464 256
WIRE -320 272 -320 224
WIRE 128 272 112 272
WIRE 464 272 464 256
WIRE -144 288 -144 80
WIRE 112 288 112 272
WIRE 224 288 224 240
WIRE 272 288 272 208
WIRE 272 288 224 288
WIRE 320 288 272 288
WIRE 416 288 384 288
WIRE -1056 304 -1056 32
WIRE -992 304 -1056 304
WIRE -704 304 -768 304
WIRE -496 304 -496 256
WIRE -16 336 -16 224
WIRE 48 336 -16 336
WIRE -1216 352 -1264 352
WIRE -1168 352 -1216 352
WIRE -704 352 -704 304
WIRE -704 352 -1168 352
WIRE -592 368 -592 176
WIRE -432 368 -432 256
WIRE -400 368 -400 144
WIRE -192 368 -400 368
WIRE 48 368 48 336
WIRE 64 368 48 368
WIRE 304 368 304 128
WIRE 320 368 304 368
WIRE 416 368 416 288
WIRE 416 368 384 368
WIRE 464 368 464 336
WIRE 464 368 416 368
WIRE 496 368 496 336
WIRE 496 368 464 368
WIRE 592 368 592 288
WIRE 592 368 496 368
WIRE 704 368 592 368
WIRE 160 384 112 384
WIRE -1024 400 -1024 112
WIRE -992 400 -1024 400
WIRE -672 400 -672 176
WIRE -672 400 -768 400
WIRE 704 400 704 368
WIRE -672 416 -672 400
WIRE -256 416 -256 240
WIRE -240 416 -256 416
WIRE -144 416 -144 384
WIRE -144 416 -160 416
WIRE 32 416 16 416
WIRE 160 416 160 384
WIRE 160 416 112 416
WIRE -144 432 -144 416
WIRE -400 448 -400 368
WIRE -400 448 -432 448
WIRE 160 448 160 416
WIRE -1264 464 -1264 352
WIRE -1168 464 -1168 352
WIRE -992 464 -1088 464
WIRE -688 464 -768 464
WIRE -720 528 -768 528
WIRE -672 528 -672 480
WIRE -592 528 -592 448
WIRE -592 528 -672 528
WIRE -496 528 -496 384
WIRE -496 528 -592 528
WIRE -320 528 -320 352
WIRE -320 528 -496 528
WIRE -144 528 -144 512
WIRE -144 528 -320 528
WIRE 48 528 -144 528
WIRE 160 528 48 528
WIRE -688 544 -688 464
WIRE -16 544 -16 336
WIRE -16 544 -688 544
WIRE -720 560 -720 528
WIRE 16 560 16 416
WIRE 16 560 -720 560
WIRE -1056 592 -1056 304
WIRE -992 592 -1056 592
WIRE -704 592 -704 352
WIRE -704 592 -768 592
WIRE 48 592 48 528
WIRE -1264 640 -1264 528
WIRE -1168 640 -1168 544
WIRE -1168 640 -1264 640
WIRE -1024 640 -1024 400
WIRE -1024 640 -1168 640
WIRE -672 640 -672 528
WIRE -672 640 -1024 640
WIRE -1088 672 -1088 464
WIRE -112 672 -112 224
WIRE -112 672 -1088 672
FLAG 48 592 0
FLAG 704 400 0
FLAG 560 128 Vout
FLAG -544 176 in
FLAG -96 80 dr1
FLAG 112 272 dr2
FLAG -1216 352 CV
SYMBOL ind2 112 64 R0
SYMATTR InstName L1
SYMATTR Value 180u
SYMATTR Type ind
SYMATTR SpiceLine Rser=10u
SYMBOL ind2 112 176 R0
SYMATTR InstName L2
SYMATTR Value 180u
SYMATTR Type ind
SYMATTR SpiceLine Rser=10u
SYMBOL ind2 240 144 M0
WINDOW 0 -19 22 Left 2
WINDOW 3 -46 46 Left 2
SYMATTR InstName L3
SYMATTR Value 32m
SYMATTR Type ind
SYMATTR SpiceLine Rser=200u
SYMBOL nmos -192 288 R0
WINDOW 3 53 65 Left 2
SYMATTR Value IRF2903ZS
SYMATTR InstName M1
SYMBOL nmos 64 288 R0
WINDOW 3 56 67 Left 2
WINDOW 0 59 38 Left 2
SYMATTR Value IRF2903ZS
SYMATTR InstName M2
SYMBOL voltage -592 352 R0
WINDOW 123 0 0 Left 2
WINDOW 39 24 132 Left 2
SYMATTR SpiceLine Rser=20m
SYMATTR InstName V1
SYMATTR Value 12
SYMBOL diode 384 304 M270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D2
SYMATTR Value MUR460
SYMBOL diode 320 224 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D3
SYMATTR Value MUR460
SYMBOL polcap 480 192 R0
WINDOW 3 32 56 Left 2
SYMATTR Value 470u
SYMATTR InstName C1
SYMATTR Description Capacitor
SYMATTR Type cap
SYMATTR SpiceLine V=200 Irms=50 Rser=0.05 Lser=0
SYMBOL res 576 192 R0
SYMATTR InstName R1
SYMATTR Value 30
SYMBOL voltage -496 288 R0
WINDOW 123 0 0 Left 2
WINDOW 39 -43 57 Left 2
WINDOW 3 -194 303 Left 2
SYMATTR Value PULSE(0 10 250u 10n 10n 450u 1000u 100)
SYMATTR InstName V2
SYMBOL voltage -320 256 R0
WINDOW 123 0 0 Left 2
WINDOW 39 -48 62 Left 2
WINDOW 3 -322 296 Left 2
SYMATTR Value PULSE(0 10 762u 10n 10n 450u 1000u 100)
SYMATTR InstName V3
SYMBOL diode 320 144 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D1
SYMATTR Value MUR460
SYMBOL diode 384 384 M270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D4
SYMATTR Value MUR460
SYMBOL polcap -688 416 R0
WINDOW 3 24 64 Left 2
SYMATTR Value 10u
SYMATTR InstName C2
SYMATTR Description Capacitor
SYMATTR Type cap
SYMATTR SpiceLine V=25 Irms=2.9 Rser=100m Lser=0
SYMBOL res -160 416 R0
SYMATTR InstName R2
SYMATTR Value 2m
SYMBOL res 144 432 R0
SYMATTR InstName R3
SYMATTR Value 2m
SYMBOL res -16 208 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R4
SYMATTR Value 100
SYMBOL res -416 464 R180
WINDOW 0 36 76 Left 2
WINDOW 3 36 40 Left 2
SYMATTR InstName R5
SYMATTR Value 100
SYMBOL res -144 400 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R6
SYMATTR Value 100
SYMBOL res 128 400 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R8
SYMATTR Value 100
SYMBOL Misc\\NE555 -880 208 R0
SYMATTR InstName U1
SYMBOL Misc\\NE555 -880 496 R0
SYMATTR InstName U2
SYMBOL res -1152 560 R180
WINDOW 0 36 76 Left 2
WINDOW 3 36 40 Left 2
SYMATTR InstName R7
SYMATTR Value 50
SYMBOL polcap -1280 464 R0
WINDOW 3 24 64 Left 2
SYMATTR Value 10u
SYMATTR InstName C3
SYMATTR Description Capacitor
SYMATTR Type cap
SYMATTR SpiceLine V=25 Irms=2.9 Rser=100m Lser=0
SYMBOL res 480 240 R0
SYMATTR InstName R9
SYMATTR Value 30
SYMBOL diode 448 336 M180
WINDOW 0 24 64 Left 2
WINDOW 3 24 0 Left 2
SYMATTR InstName D5
SYMATTR Value MUR460
TEXT 144 304 Left 2 !K1 L1 L2 L3 1
TEXT -584 624 Left 2 !.tran 0 200m 0 1u startup
TEXT 320 456 Left 2 ;Primary 2x8 turns 2V/turn at 1000 Hz
TEXT 320 528 Left 2 ;25 A peak magnetizing current for 180u 1kHz
TEXT 336 560 Left 2 ;110 A peak current limit
TEXT 320 496 Left 2 ;Secondary 100 turns 2V/turn at 1000 Hz

 

[Jasen Betts]

What is the 12Vdc supply? The best vehicle battery I've seen barely
does 300 to 500A into a dead short. That calculates to more like 20m,
not 2m Rseries. Especially at start up inrush currents.

Stop bying used batteries from the wreckers.


700 CCA (that's into a 7.2V load, not a short) is not uncommon,
the smallest I've seen was rated 300CCA

For the 300CCA battery, If the internal resistance is ohmic,
that would be 600A into a short.

 

[Jamie]

Stop bying used batteries from the wreckers.

Sounds like you speak with experience

700 CCA (that's into a 7.2V load, not a short) is not uncommon,
the smallest I've seen was rated 300CCA

For the 300CCA battery, If the internal resistance is ohmic,
that would be 600A into a short.

Jamie