3C85- 3C90 - What's next?

[Yzordderex]

My back is against the wall here.

They designed a new control board using dsp instead of micro and I am
left to mop up after them once again. They've doubled power from
supply.

I'm looking for a better material to run flyback. Needs to saturate a
little higher and run a little cooler. They're almost running core
into saturation when the thing gets warmed up. Unfortunatly its a 555
running it, so nothing in the way of protection for fet except a small
source resistor (degenerative feedback) to allow the thing to survive
power up.

I suppose there are some materials I should look at, but haven't been
in that loop for a while.

Regards,
Bob
N9NEO

P.S. Paul Mathews - did you get my email I had sent you directly?

 

[Fred Bartoli]

"Yzordderex" <yzordderrex@verizon.net> a écrit dans le message news:
a69e22f6.0403301206.11d6e79f@posting.google.com...

My back is against the wall here.

They designed a new control board using dsp instead of micro and I am
left to mop up after them once again. They've doubled power from
supply.

I'm looking for a better material to run flyback. Needs to saturate a
little higher and run a little cooler. They're almost running core
into saturation when the thing gets warmed up. Unfortunatly its a 555
running it, so nothing in the way of protection for fet except a small
source resistor (degenerative feedback) to allow the thing to survive
power up.

Use what's made for : an UC3842 is as cheap as a 555.

As for your ferrite, you're not looking for a better material but for a
better design :

Flux = L I = n B S
For a flyback in discontinuous mode (I take it's not wise to use continuous
mode if you're to asking such question) you have an output power :
P = (F L Ipk^2)/2 = F n Bpk S Ipk / 2

To lower the peak induction, you can increase either Ipk, n or F or several
of those.
There's also a relation between Ipk, n (or L) and F when you want (and you
want) to stay in discontinuous mode, but I'll let you do some maths too ;-)

BTW what are your constrains ? Why a flyback ?

Fred.

 

[legg]

On 30 Mar 2004 12:06:02 -0800, yzordderrex@verizon.net (Yzordderex)
wrote:

My back is against the wall here.

They designed a new control board using dsp instead of micro and I am
left to mop up after them once again. They've doubled power from
supply.

I'm looking for a better material to run flyback. Needs to saturate a
little higher and run a little cooler. They're almost running core
into saturation when the thing gets warmed up. Unfortunatly its a 555
running it, so nothing in the way of protection for fet except a small
source resistor (degenerative feedback) to allow the thing to survive
power up.

I suppose there are some materials I should look at, but haven't been
in that loop for a while.

Core size in present layout?

Target power level?

Is there any air movement?

RL

 

[Genome]

"Yzordderex" <yzordderrex@verizon.net> wrote in message
news:a69e22f6.0403301206.11d6e79f@posting.google.com...
| My back is against the wall here.
|
| They designed a new control board using dsp instead of micro and I am
| left to mop up after them once again. They've doubled power from
| supply.
|
| I'm looking for a better material to run flyback. Needs to saturate a
| little higher and run a little cooler. They're almost running core
| into saturation when the thing gets warmed up. Unfortunatly its a 555
| running it, so nothing in the way of protection for fet except a small
| source resistor (degenerative feedback) to allow the thing to survive
| power up.
|
| I suppose there are some materials I should look at, but haven't been
| in that loop for a while.
|
| Regards,
| Bob
| N9NEO
|
| P.S. Paul Mathews - did you get my email I had sent you directly?

YEAH!!!!!

WOT ABOUT SOME PROPER SPECS!!!!!

dna

 

[Yzordderex]

Ok, thanks for all the very nice comments. Hope I can respond in
kind.

First, for Fred.

The edict from above was "NO CONTROL CHIP". Design was implemented
almost 8 years ago, and at the time the control chip was quite a bit
more expensive than the 555. The 555, with the standard feedback loop
that everybody has been doing, was a step above the self-oscillating
supply it replaced, with all of it's peccadillos I might add. There's
probably a half million of these supplies out in field and to my
knowledge the failure rates for 240, 480, and 600v controls have been
very very low.

I'm fully aware of the 3842 controller. I was one of the first users
of the controller in the mid 1980s - back when there were problems
with the UV lockout circuit in the engineering samples. Anyone else
remember that far back?

And yes, I will admit that the best approach would be a redesign. If
this magic material I'm looking for doesn't appear soon I will be
forced to open the gap a tad, and crank up the frequency. I prefer
not to run continuous.

Reason for flyback is I need +15, +5 (the regulated winding), -9, and
a gate drive winding common to the negative bus to supply gate drives
for 3 lower igbt's in inverter. Supply also provides bus voltage
information to micro via the -9v winding when mosfet is on. Also -10v
is derived from gate drive winding when mosfet is on. FYI, the mosfet
sits on top of the transformer with it's drain connected to the +bus.
Allows us to meet, or almost meet some would say, the agency spacing
requirements for safety. Optocoupler can barely keep up with all the
dancing going on around it.

Now for Mr Legg. Core is EI-19. And power level is about 8w-9w. I'm
running same core set in a larger drive near 12w The problem is that
this transformer has a 27v feedback winding, as I play a different
game for driving the big modules. Can't just swap out for existing
transformer. I have 2 designs, one for 240v mains, and another which
works on 480-600v mains. Forget air movement, its in an enclosure
within an enclosure.

Mister DNA, I ran across your site yesterday. I hope to return again
soon. As far as specs go, I don't think I need to get any more
involved than I already am.

And lastly Mr. Bloggs,

They are what I call the people down the hall who failed in their
simple challenge to redesign one control board to interface with the
rather large and well established family of power boards - already in
production. The simple fact they failed is not so much a problem with
me. It is their repeated attempts to blame my designs for their
failure which really irks me. Read on. They are the people who
called me one afternoon and asked, "where is the differential
amplifier on the power board that tells the micro what the bus voltage
is?" I had to remind them that that amplifier was never on this
product line and had not been used in over three years. We use a
transformer winding to do that now sonny. Where ya been? They tried
to jam it up my ass the following Monday morning at the engineering
meeting. Said I had a problem with voltage sensor on all my products.
I basically tore them a new asshole in front of whole department.
Ended up being one resistor change on they's control board - a
prediction I hade already made to boss. He has little use for the
handwaving, fingerpointing, federation of dunces himself.

They are the people who decided the new board needed a 40 cent reset
chip - parked next to an unused comparator. (Actually comparator is
supposed to monitor some other power supply, but they don't know how
to bias the thing, so in effect does nothing.) The reset chip they
tried to jam up my ass because the +5v supply has a dip in it (always
had) when the lower transistors turn on to charge the upper bootstrap
caps. They said they weren't really concerned with some dynamic
control issues that I was having with the power supply. I got their
magic chip taken off the board asap.
They're the same crowd that asked me why I sent a +15v trip signal up
to the microboard. The answer was because that's what 'they' asked
for when first power board designed 3 years ago.
The same crowd that just told me (at the eleventh hour) the new micro
pulls twice the current that the old board did. They are a marginally
incompetant bunch, and that's the nicest thing I can find to say about
them. The list goes on and on.
Well now they are having their own set of problems with this new
design they got going on. Seems they are beginning to care about some
dynamic control issues of their own now. Their motor just about
stalls when they put a little step load on it They told me they
needed to measure the motor leakage inductance. Were going to put a
step voltage on the stator and watch the current rise. YUK YUK and
good luck to them. They gotta make a show next month.

Hope this answers your question and gives you a good idea of who they
are. I for one would have at least looked at a schematic of the power
board I was trying to interface to. Wouldn't you????

Sorry for the tone, but if you had to deal with they/them on a daily
basis, I'm sure your sanity, as my own is, would be in question.

Now about that material? Anything better than the 3C90? that might
run a little cooler and saturate a tad higher? Should I stop wasting
my time and just regap part and blame it on them? heheh. I know you
guys really want to redesign my supply, but I'm really all set. Just
need a little shove in the right direction here.

if you guys really want to know what I'm up to go here

http://mysite.verizon.net/vze6qsvc/coolradiostuff22/

What's all this "they" stuff? How many personalities do you have anyway?
And why don't you try communicating with yourselves more?

regards,
Bob

 

[legg]

On 31 Mar 2004 14:04:32 -0800, yzordderrex@verizon.net (Yzordderex)
wrote:

Reason for flyback is I need +15, +5
(the regulated winding), -9, and a gate
drive winding common to the negative
bus to supply gate drives for 3 lower
igbt's in inverter. Supply also provides
bus voltage information to micro via the
-9v winding when mosfet is on. Also -10v
is derived from gate drive winding when
mosfet is on. FYI, the mosfet sits on
top of the transformer with it's drain
connected to the +bus. Allows us to meet,
or almost meet some would say, the agency
spacing requirements for safety. Optocoupler
can barely keep up with all the dancing
going on around it.

Your claim is that the problem is simply one of
overheating and the resulting lower saturation levels.

How is this assertion justified?

What temperatures are currently achieved? What
temperatures were previously considered acceptible?
The close proximity of the mosfet suggests that there
could be other heat sources affecting thermal rises.
What are the recent symptoms of circuit overload?

3C85 (which I assume the circuit started off with
x-odd years ago) will not exibit thermal run-away
below , if it is the dominant source of heat.
3C93 will self-regulate around 135degC -
generally an unacceptible situation in commercial-
grade isolation components.


Does your circuit depend on core saturation
to function as intended? Saturation loss-limited
designs are typically characterized by operating
frequencies lower than 50KHz. What frequency is
presently attempted? Is this frequency stable over
the load range?

Now for Mr Legg. Core is EI-19. And power
level is about 8w-9w. I'm running same core
set in a larger drive near 12w The problem
is that this transformer has a 27v feedback
winding, as I play a different game for driving
the big modules. Can't just swap out for
existing transformer. I have 2 designs, one
for 240v mains, and another which
works on 480-600v mains. Forget air movement,
its in an enclosure within an enclosure.

I don't fully understand the limitations that you
are imposing here. Are you saying that turns alterations
are not permitted to solve this problem? Just changing
the core material will not solve problems that surround
saturation limits.

Increasing the gap would allow operation at an increased
frequency, at reduced peak flux. Are you allowed to do so?

The core size is not immediately recognizable to me.
An E19/8/5 (US lamination E187) is permitted between
400 and 800mW total dissipation for self-rise of
between 30 and 60degC, so that 12W of through-put
power is not an unrealistic expectation.

To maintain operation in Complete Energy Transfer
at full load (with a 50% duty limit) at 100KHz
will require the largest gap available in this core
size and a full turn per output volt on the secondary
windings to keep flux density below 240mT (~400mW core
loss in 3F3 material).

As these can be stacked windings, the output complexity
doesn't neccessarily invoke a copper-loss penalty.

RL

 

[Ken Smith]

In article <a69e22f6.0403301206.11d6e79f@posting.google.com>,
Yzordderex <yzordderrex@verizon.net> wrote:
[....]

I'm looking for a better material to run flyback. Needs to saturate a
little higher and run a little cooler. They're almost running core
into saturation when the thing gets warmed up.

I don't think the saturation flux is much higher on any of the newer
materials. Can you raise the frequency slightly? If you go up in
frequency, you can go down in inductance by increasing the gap. A bigger
gap will allow slightly more energy in the inductor before it saturates.
The combination of slightly more cycles per second and slightly more
energy per cycle means the max power before saturation will be higher and
hence the inductor can be run at a lower fraction of max.

My books are all packed but I remember a number like 3F4 as a newer
material that has better high frequency characteristics. Switching to
something with all the other losses lower may allow you to get the
temperature down enough.

What are you using as the output side rectifier? Perhaps a change to a
Schottky can bail you out by saving some secondary side losses.

Have you considered a Berquist "gap-pad" to suck the heat out of the
inductor?

How much extra IR loss do you have at the running frequency? If it is
enough to matter you could try winding with wire rope or Litz.

Unfortunatly its a 555
running it, so nothing in the way of protection for fet except a small
source resistor (degenerative feedback) to allow the thing to survive
power up.

These turkeys didn't steal my design for running 3 supply dynamic RAMs on
a ZX80 did they? I didn't patent the idea. I guess I should have.

--
--
kensmith@rahul.net forging knowledge

 

[Yzordderex]

Sorry Mr. Legg. Core is the EE style you mentioned. Got my vowels
all mixed up.

After reading my own post - I know, I know, I should have done that in
first place - There exists another front for confusion. First ps was
designed about 7 or 8 years ago. Then about 3 years ago we designed
another drive platform and used old supply from 8 yrs ago. We're
talking of 2 designs. One with 800v/1a smd part and another 1500v
TO-220 plastic part. And yes, today I would use a control chip. Back
then I was willing to let somebody who knows nothing about power
supplies micromanage my development efforts. Today it's basically
f**k off I'm doing it my way. I'm no flyback expert, but I do ok at
it.

I didn't want anything more than a "Try this material out" kinda
response, but you guys being made of the same stuff as me - the stuff
they make good engineers outta, I understand your curiousity and your
willingness to do a total redesign for me. Perhaps a competition is
in order - I write the spec and you guys design the supply. Would
leave me more time to play with my radios.

After my unnecessary and scathing description of they yesterday (a
well thought out and not for one second regretted description it was!)
I figure I owe you all an equal treatment of the problem at hand.

Ok, my concern is mostly, but not all, with fet. The transformer core
(under full load at operating ambient temperature) measures about 93c,
Fet 106 with a small heatsink. The supply runs about 70kHz. At this
point xfmr saturates at about 140ma, 20ma over where the peak current
is running - 120ma. Mosfet is (I wanna say, but don't quote me) a
200ma 1500v Sanyo part. Not too much margin here, but haven't had one
take off yet. When a start command is given the three lower Igbt's
take turns turning on and charging the upper bootstrap capacitors.
The transformer saturates some while caps are charging and fet takes a
little hit. Probably not a problem as three charging events over in
short time. 5v supply dips down to about 4.6v and then back up with a
little overshoot. The UV chip the dunce down the hall put in was set
to trip at 4.75v.

Yes, I could easily redesign the transformer and get what I need out
of it. BUT same transformer(s) are used from fractional HP no fans,
no charge relays, to 10HP, two relays and a good sized fan.
Unfortunatly the 555 being what it is does not have a lot of range.
Can't just throw in another transformer and off to the races. Might
have to do a little tweaking (or at minimum take a look at) each
supply to make sure full load at bottom end(say 350vdc) is ok, not
continuous, and can supply the power, make sure will stay in
regulation at 850vdc with light load too. So would rather not have to
write big ecn across the board and make a lot of noise with
production. Would rather pay another nickle for core set and quietly
slip in when nobody looking. If a nickle makes an across the board
reliability change then I am happy. My group is interrogated for
every penny we spend, while idiots down the hall are throwing money at
the control board like it's on fire and money will put it out. I
think we all know what happens when projects are micromanaged and
every last penny is taken out of it. You find in a short time that
it's just the reliability you've beaten out of it.

Hope I've covered all your questions. If not, no prob, just ask away.
Thanks for helping out.

And I hope you've taken a look at my website before I change the
marquee. I like to give it a makeover every once in a while, and
getting ready to do so soon.

Regards,
Bob
N9NEO


legg <legg@nospam.magma.ca> wrote in message news:<6b9n609ud2hs9ek25uhlurs51nigocl5a9@4ax.com>...

On 31 Mar 2004 14:04:32 -0800, yzordderrex@verizon.net (Yzordderex)
wrote:

Reason for flyback is I need +15, +5
(the regulated winding), -9, and a gate
drive winding common to the negative
bus to supply gate drives for 3 lower
igbt's in inverter. Supply also provides
bus voltage information to micro via the
-9v winding when mosfet is on. Also -10v
is derived from gate drive winding when
mosfet is on. FYI, the mosfet sits on
top of the transformer with it's drain
connected to the +bus. Allows us to meet,
or almost meet some would say, the agency
spacing requirements for safety. Optocoupler
can barely keep up with all the dancing
going on around it.

Your claim is that the problem is simply one of
overheating and the resulting lower saturation levels.

How is this assertion justified?

What temperatures are currently achieved? What
temperatures were previously considered acceptible?
The close proximity of the mosfet suggests that there
could be other heat sources affecting thermal rises.
What are the recent symptoms of circuit overload?

3C85 (which I assume the circuit started off with
x-odd years ago) will not exibit thermal run-away
below , if it is the dominant source of heat.
3C93 will self-regulate around 135degC -
generally an unacceptible situation in commercial-
grade isolation components.


Does your circuit depend on core saturation
to function as intended? Saturation loss-limited
designs are typically characterized by operating
frequencies lower than 50KHz. What frequency is
presently attempted? Is this frequency stable over
the load range?

Now for Mr Legg. Core is EI-19. And power
level is about 8w-9w. I'm running same core
set in a larger drive near 12w The problem
is that this transformer has a 27v feedback
winding, as I play a different game for driving
the big modules. Can't just swap out for
existing transformer. I have 2 designs, one
for 240v mains, and another which
works on 480-600v mains. Forget air movement,
its in an enclosure within an enclosure.

I don't fully understand the limitations that you
are imposing here. Are you saying that turns alterations
are not permitted to solve this problem? Just changing
the core material will not solve problems that surround
saturation limits.

Increasing the gap would allow operation at an increased
frequency, at reduced peak flux. Are you allowed to do so?

The core size is not immediately recognizable to me.
An E19/8/5 (US lamination E187) is permitted between
400 and 800mW total dissipation for self-rise of
between 30 and 60degC, so that 12W of through-put
power is not an unrealistic expectation.

To maintain operation in Complete Energy Transfer
at full load (with a 50% duty limit) at 100KHz
will require the largest gap available in this core
size and a full turn per output volt on the secondary
windings to keep flux density below 240mT (~400mW core
loss in 3F3 material).

As these can be stacked windings, the output complexity
doesn't neccessarily invoke a copper-loss penalty.

RL

 

[John Woodgate]

I read in sci.electronics.design that Yzordderex
<yzordderrex@verizon.net> wrote (in <a69e22f6.0404010957.5351c6ca@postin
g.google.com&gt about '3C85- 3C90 - What's next?', on Thu, 1 Apr 2004:

Today it's basically
f**k off I'm doing it my way. I'm no flyback expert, but I do ok at
it.

I didn't want anything more than a "Try this material out" kinda
response, but you guys being made of the same stuff as me - the stuff
they make good engineers outta, I understand your curiousity and your
willingness to do a total redesign for me. Perhaps a competition is
in order - I write the spec and you guys design the supply. Would
leave me more time to play with my radios.

You want to hire Genome, keep him away from the alcohol and perhaps
interesting herbs. I feel sure he could sort out your problems in so
short a time that you would be astounded at the modesty of his fee. (;-)
--
Regards, John Woodgate, OOO - Own Opinions Only.
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

 

[legg]

I didn't want anything more than a "Try this material out" kinda
response, but you guys being made of the same stuff as me - the stuff
they make good engineers outta, I understand your curiousity and your
willingness to do a total redesign for me.

Try 3F3 (not 3F4 or 3C9x) and increase the gap slightly, if only by
adding a fishpaper spacer.

If I knew what gap or turns in currently used, I could evaluate the
likelihood of success more accurately. A frequency shift may be
unavoidable.

Sticking a small bead in series with forward peak-rectified outputs
could reduce fet stress, without seriously compromising the output
voltage.

After my unnecessary and scathing description of they yesterday (a
well thought out and not for one second regretted description it was!)
I figure I owe you all an equal treatment of the problem at hand.

Ok, my concern is mostly, but not all, with fet. The transformer core
(under full load at operating ambient temperature) measures about 93c,
Fet 106 with a small heatsink. The supply runs about 70kHz. At this
point xfmr saturates at about 140ma, 20ma over where the peak current
is running - 120ma. Mosfet is (I wanna say, but don't quote me) a
200ma 1500v Sanyo part. Not too much margin here, but haven't had one
take off yet. When a start command is given the three lower Igbt's
take turns turning on and charging the upper bootstrap capacitors.
The transformer saturates some while caps are charging and fet takes a
little hit. Probably not a problem as three charging events over in
short time. 5v supply dips down to about 4.6v and then back up with a
little overshoot. The UV chip the dunce down the hall put in was set
to trip at 4.75v.

Yes, I could easily redesign the transformer and get what I need out
of it. BUT same transformer(s) are used from fractional HP no fans,
no charge relays, to 10HP, two relays and a good sized fan.
Unfortunatly the 555 being what it is does not have a lot of range.

Whatever change is made, you will have to differentiate the new
transformer from the old anyways, if only to apply it in positions
where it is required first, while old inventory is absorbed in
locations where the old part is still useful.

Use of the new transformer could also be tied to a single component
change that increases the controller's frequency.

If you have two different and identifiable parts in the system, you
could choose to keep it that way, removing any requirement to
re-engineer locations where the old part still functions.

For inventory compatability, testing the new core part in the old
transformer should produce no noticeable performance changes, save a
modest reduction in temperature. This would maintain core purchasing
scale and obsolete the older material, when it runs out.

RL

 

[Yzordderex]

I was trying to make the change as painless as possible. If I could
find a 'newer' material that would just drop in - and away we go -
that would please me. I'll for sure take a look at the 3F4 material.
I'm aware of the tradeoffs between permiability, losses, and
saturation flux densities, so my optimism is of course guarded.

My mantra is "if it ain't broke.....
There is a delicate balance I've had to maintain with this power
supply in order to keep it cheap and reliable. Read my previous
whinings, if they ever show up here. I don't want to upset that
balance. Since the transformer(s) are used across most of the product
line, it's not a matter of redesigning and testing one control. I
have to test a whole family of units. I'm not one to shotgun a
change, and say "good enough". I'll probably take the safe (or lazy
if you must) approach and create a new part for the particular
controls that are affected. I hate pulling out part numbers because
of the costs to maintain a new part.

And No, I was the Turkey who designed the 555 in. Didn't steal design
from anyone. It's one of my own cluster ***ks Ken.
Thanks again and best regards,
Bob
N9NEO



kensmith@violet.rahul.net (Ken Smith) wrote in message news:<c4h9g1$uts$2@blue.rahul.net>...

In article <a69e22f6.0403301206.11d6e79f@posting.google.com>,
Yzordderex <yzordderrex@verizon.net> wrote:
[....]
I'm looking for a better material to run flyback. Needs to saturate a
little higher and run a little cooler. They're almost running core
into saturation when the thing gets warmed up.

I don't think the saturation flux is much higher on any of the newer
materials. Can you raise the frequency slightly? If you go up in
frequency, you can go down in inductance by increasing the gap. A bigger
gap will allow slightly more energy in the inductor before it saturates.
The combination of slightly more cycles per second and slightly more
energy per cycle means the max power before saturation will be higher and
hence the inductor can be run at a lower fraction of max.

My books are all packed but I remember a number like 3F4 as a newer
material that has better high frequency characteristics. Switching to
something with all the other losses lower may allow you to get the
temperature down enough.

What are you using as the output side rectifier? Perhaps a change to a
Schottky can bail you out by saving some secondary side losses.

Have you considered a Berquist "gap-pad" to suck the heat out of the
inductor?

How much extra IR loss do you have at the running frequency? If it is
enough to matter you could try winding with wire rope or Litz.

Unfortunatly its a 555
running it, so nothing in the way of protection for fet except a small
source resistor (degenerative feedback) to allow the thing to survive
power up.

These turkeys didn't steal my design for running 3 supply dynamic RAMs on
a ZX80 did they? I didn't patent the idea. I guess I should have.

--

 

[Nico Coesel]

yzordderrex@verizon.net (Yzordderex) wrote:

My back is against the wall here.

They designed a new control board using dsp instead of micro and I am
left to mop up after them once again. They've doubled power from
supply.

I'm looking for a better material to run flyback. Needs to saturate a
little higher and run a little cooler. They're almost running core
into saturation when the thing gets warmed up. Unfortunatly its a 555
running it, so nothing in the way of protection for fet except a small
source resistor (degenerative feedback) to allow the thing to survive
power up.

You could add a current sense resistor and pull the reset low by an
extra transistor whenever the current goed beyond a certain point.
Works very nice as a cycle-by-cycle over current protection.

I suppose there are some materials I should look at, but haven't been
in that loop for a while.

There is also 3Fxx material which is slightly better than 3Cxx

--
Reply to nico@nctdevpuntnl (punt=.)
Bedrijven en winkels vindt U op www.adresboekje.nl

 

[Ken Smith]

Have you looked at MPP core materials? They have about twice the
saturations flux but higher AC losses.

I think MetGlas is even higher but you can only get a few shapes.

--
--
kensmith@rahul.net forging knowledge

 

[Yzordderex]

Hi Nico,

Looks like your post is a little lagged, as mine usually are. I'm
going to have to load in a news client one day.

Ferroxcube guy says try 3C92 or 3C94 material. He's gonna send some
over. Perm looks to be a little lower, but the gaps gonna set the
inductance anyway, so I'll give it a shot. Funny how they change the
scaling on those material curves all around to try and trick ya into
thinking they got some really great thing going on.

Yea, I could do all kinds of things. As I said (in posts after yours)
I want to keep change rather quiet. Don't need to be giving the
spin-masters down the hall any ammo.

regards,
Bob





nico@puntnl.niks (Nico Coesel) wrote in message news:<406d9af5.150336722@news.planet.nl>...

yzordderrex@verizon.net (Yzordderex) wrote:


You could add a current sense resistor and pull the reset low by an
extra transistor whenever the current goed beyond a certain point.
Works very nice as a cycle-by-cycle over current protection.

I suppose there are some materials I should look at, but haven't been
in that loop for a while.

There is also 3Fxx material which is slightly better than 3Cxx