# pulsed DC and transformers

Discussion in 'Electronic Basics' started by [email protected], Jun 21, 2008.

1. ### Guest

I'm trying to figure out if I could drive a transformer with pulsed
DC, and have it work (not saturate, overheat, etc.)

Ideally I'd like to pulse a self-wound toroidal transformer with
12VDC, on the primary, and get 48VDC out of the secondary, center-
tapped to get +/-24V on each leg. I plan to use a 555 in astable
mode, 50% duty cycle, driving a mosfet to feed the transformer
primary, at frequencies I haven't determined yet - anywhere from 100
Hz to 100 kHz.

Wikipedia has this:

The time-derivative term in Faraday's Law shows that the flux in the
core is the integral of the applied voltage.[11] Hypothetically an
ideal transformer would work with direct-current excitation, with the
core flux increasing linearly with time.[12] In practice, the flux
would rise very rapidly to the point where magnetic saturation of the
core occurred, causing a huge increase in the magnetising current and
overheating the transformer. All practical transformers must therefore
operate under alternating (or pulsed) current conditions.

so it implies a transformer can be driven by pulses. In practice,
what happens when driving a transformer with pulsed DC? Should I
stick to low frequencies? Go for higher frequencies?

Thanks,

Michael

2. ### Bob EldGuest

Keep the DC out of the transformer, especially a torroid as they have little
tolerance for DC offset. Transformers or inductors that tolerate DC offsets
have air gaps or the equivalent to limit saturation.

The simple way is to drive the transformer through a capacitor, thus
blocking all DC into it's primary. The secondary will convey the pulse, but
the zero volt position will be somewhere in the middle of the pulse so that
the area above zero equals the area below zero. DC can be restored with
appropriate diodes.

For example say the original pulse train was positive going 4 Volts for 1
msec. then zero volts for 3 msec.

If the transformer ratio is 1 : 1, then the secondary would show 3 Volts
positive for 1 msec and 1 Volt negative for 3 msec. The positive area equals
the negative area. Of course, if the pulse width changes, the zero line
would shift to equalize areas. Faraday's law.

3. ### JoergGuest

Yes, I'll second that. Else the time from core saturation to plume of
smoke can be just microseconds. Last time I did that (experimental PWM
stalled while I was doing other work in the office) it took two days to
get the stench out of the area.

[...]

4. ### Guest

Gee, power electronics seems like so much fun. Maybe I chose the
wrong major at the university... ;-)

Thanks,

Michael

5. ### JoergGuest

Chemistry can be fun, too. One of my friends ended in the hospital. "My
lab is now one hellacious black hole and the door is gone as well."

6. ### Guest

What did he try to do... make nitrogen tri-iodide or something silly
like that?

Yep, chemistry can be fun. So far I made 40 cc's of ethanol from the
kids' leftover food scraps. Just have to scale it up by a factor of
10^9 or so...

Michael

7. ### JoergGuest

I don't remember. But it made the news.

There was an ad for an ethanol machine that can make 30gal/week or so at
home. In the fine print: You had to pour in lots of leftover booze.
Yeah, right. Who'd have gallons of "leftover" Whiskey for that?

8. ### Guest

Haha.

A co-worker told me he was going to buy this:

http://www.efuel100.com/t-product.aspx

Only \$6,988 after tax credits.

"To operate simply load EFuel100 feedstock (table sugar with ethanol
yeast mix) into the fermentation tank and select the fermentation
option on the control panel to begin the process. It will take between
10lbs to 14lbs of sugar to produce 1 gallon of ethanol. The
MicroFueler is capable of producing 5 gallons of ethanol per day once
fermentation is complete."

He asked me for my comments on the process flow diagram. I gave him
about 5 areas where I would do it differently... plus I told him, with
sugar about \$2.50 for 5 lbs at Wal-Mart, it didn't seem economical.
But he's already sent in the down payment... we'll see how it goes...

Michael

9. ### JoergGuest

It's disturbing. I see many well-educated people fall prey to hype
technologies. Predominantly the overly eco- bio- whatever leaning folks.
The millisecond they get criticized they often react with "Ah,
baloney" but without being able to furnish any data to support their own
theory. Kind of like warmingists ...

10. ### Guest

On Jun 23, 11:51 am, Joerg <>
wrote:

~snip~

Yes, truly.

The site I posted above used to mention some sort of tax-free, cheap
inedible sugar from Mexico. I tried seeking more info via the web,
but no luck. (What kind of sucrose is inedible ??? And why?) The
website's FAQ has since removed the reference to the cheap Mexican
sugar.

One of the ways I would have done it differently, by the way, would
have involved just going to restaurants and collecting starchy food
scraps (rice, bread, potatoes)... toss in a little bit of alpha-
amylase enzyme (0.1% by mass) to convert starch to glucose, that the
yeast can eat.

Global warming... well, if it's true, I can start a coconut farm in

Michael

11. ### JoergGuest

Very shortly thereafter you'll read in the papers that some crooks
lifted a vat of starchy food scraps from the so-and-so diner. There
ain't enough of that stuff. Look at the theft cases of "discarded"
french fries oil. Some people are becoming desperate.

A while ago I was passed by a Mercedes 300 TurboDiesel in the Bay Area.
He pulled into the right lane in front of me but I had to back off. This
car let off such a strong burger shop smell that my mouth began to water.

Not likely. We used to be able to grow citrus around here. No more, that
stuff dies in freezes every winter.

12. ### Guest

Ok, thanks.

If I'm reading this right, I want RC to be <= the time period T.

I'm not sure exactly what the primary resistance of my self-wound
toroidal transformer will be, but let's just say it's one ohm.

So let's say I'm operating my 555 driver at 7 kHz. So, C must be <=
(0.000143 s) / (1 ohm); C <= 143 uF, correct?

So a 100 uF capacitor should (just barely) satisfy the requirement.
(Or a 10 uF, or a 1 uF... right?)

Can it be an electrolytic? If so, the (+) side should be closest to
the 12VDC input, right?

Thanks,

Michael

13. ### JoergGuest

More like 22uF. Calculate it so it's less than 1/10th of the transformed

Theoretically yes. But be careful, electrolytics can only take a certain
amount of ripple. Found that out the hard way as a kid. Of course I had
to build the biggest honking amp a 230V/16A circuit could stomach. A
470uF/400V cap decided it had enough of this. Lift-off. Left a crater in
the plaster ceiling and a hole in the carpet.

IOW ceramic would be better.

14. ### Guest

On Jun 23, 2:11 pm, Joerg <>
wrote:

~zap~

Ah, thanks! A good thing I asked the experts. "But be careful,
electrolytics can only take a certain amount of ripple." didn't appear
in MY circuits book.

Just in time for Fourth-of-July fireworks...

(I'd better test my circuits with small loads first.)

Michael

15. ### JoergGuest

Small loads, see if anything gets hot.

BTW the carpet turned out to be a pain. Needless to say my parents were
not enthused and I vowed to fix that. Found a carpet remnant on sale at
a store 5 miles away, 15ft by 7ft, barely large enough. Paid, rolled it
up, strapped it to the frame of my bicycle and pedaled home. Got a cramp
from that unhealthy position and had to push it the last 4.9 miles.
Stopped at the library, picked up a book about floor covering
techniques. At least I learned how to lay carpet and how to replace
broken spokes on bicycle wheels. All because that one dreaded capacitor
couldn't hang on.

16. ### Guest

On Jun 23, 2:52 pm, Joerg <>
wrote:

....

And the ceiling?

17. ### JoergGuest

That was just a little crater, some of the plaster and concrete had
fallen out. Pa gave me a gypsum bag, I mixed some up, went on a ladder
and spackled it. Of course, then I had to re-paint the whole ceiling and
that is not one of my favorite tasks.

18. ### theamber

6
0
Jun 24, 2008
Pulsating DC to transformer.

Use a "Pulse transformer" that is a transformer that is optimised for transmitting rectangular electrical pulses.
The lower the freq. the better with 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.