What's the transformer winding insulation rated for? Have you considered core saturation and losses?
That little transformer probably won't like being pushed to such a high voltage.
-- Paul Hovnanian mailto:Paul@Hovnanian.com ------------------------------------------------------------------ Porsche 928: 0 to c in 2.125 years, 2.435 light-years per mile^3 of gas
Only a clueless fool would assume that unlimited energy could be stolen from an orbital encounter. Only a complete moron wouldn't notice that both body's new orbits will contain the same point (where they left each other's "influence"). IOW, there is a limit to the energy "stolen".
Hi, clueless moron.
On a sunny day (Wed, 21 Sep 2011 09:11:07 -0700) it happened "Paul Hovnanian P.E." wrote in :
As I wrote, the insulation for most enamel wire is about tested at 500 Veff for 1 minute. Should be OK for 500 V peak. The one I have wound is over 2 chambers, so that halves the voltage if done right. Look at the scope pics for saturation etc, it is not saturating. Read the threat.
pick
as
are
which is
of
from
body's
ONLY on a two body problem, the current problem is 4 or more. The asteroid, the earth, earth's moon, and the sun minimum. For accurate orbits Jupiter, Saturn, Neptune and Uranus should be considered. We wouldn't want to hit the earth or moon due to sloppy calculation now = would we.
refail: Without providing any numbers or links you still sound like a nattering nabob of doom.
asteroids
is
Moron, it takes three bodies, two in orbit around another.
No, you're a clueless moron. It's obvious to everyone here.
pick
as
are
, which is
b of
from
h body's
ld
Since krw's response includes neither numbers nor links - as usual - joshepkk does seem to be well ahead on points.
-- Bill Sloman, Nijmegen
Venus, pick
as
There are
energy, which is
nabob of
stolen from
body's
would
No. It does not. Momentum exchange occurs in the two body case, especially if one (or both) of the two bodies is spinning. The = Earth-Moon case proves that.
asteroids
which is
You're absolutely clueless. Figures.
Once again.
Here is the negative output version. I needed it because this feeds a PMT, and I want the anode close to 0V:
+5V | [ ] R1 100k PIC B | comparator + ------------------------------------| +0.1V C2IN+ p16 | C1 | === 10n | | [ ] R2 10M /// +5 | | FR1Y | | ---------|First there was an error in the diagram, one that I cannot, because of the ASCII art, get away with by stating that it was fly droppings: The polarity point of the transformer was in the wrong place, this is now corrected.
It is nice to be able to set the PMT voltage (or any other thing you use it for), and it would be even cooler if you could do it from zero to full output under software control. In a flash of light (LEDs) it came to me, simply use the PIC DAC output to set the voltage at the 'other end' of R1. The PIC DAC is a resistor divider chain with an analog demultiplexer, and it has a rather high impedance. Several kOhm, so you may need to make R1 a bit smaller for the same maximum output. You can then write soft to for example set the DAC via RS232 or buttons on a front panel. You can select any reference in the software for that DAC, from 1.024V up to 4 x
1.024V for internal reference, or Vdd if you want to use the +5V as reference, The PIC dataheet (18F14K22 I used) will tell you more. If you want manual control with a potmeter, use a 1kOhm linear one between 0 and +5V, and connect the top of R1 to it. I find with the pot I can go from zero to -500V output in this circuit, Without load the circuit will revert back to hysteretic control at lower than 250 V out, but with even a 10MOhm load (scope probe) it stays in current mode voltage control all the way to zero. Voltage stabilisation stays at about 2% or better (I described in a previous posting why), I think you can, by selecting a different value for C1, probably keep it in cycle by cycle mode with any load, because there is already a small pre-load from R2. The PIC DAC will let you set the -500V in 31 steps that makes about 16 V per step, very good control of a PMT.I hereby declare this circuit public domain of course you should mention Jan Panteltje if you re-publish it. And sci.electronics.design, as some people here contributed to the inspiration.
Here is the software controlled negative output version. I needed it because this feeds a PMT, and I want the anode close to 0V:
PIC DAC output p19 -----------------------------------
0 to + 5V | in 32 steps | [ ] R1 100k PIC B | comparator + ------------------------------------| +0.1V C2IN+ p16 | C1 | === 10n | | [ ] R2 10M /// +5 | | FR1Y | | ---------|
There's two layers of insulation between adjacent windings, so it should be good for 1 kV; and if you have 100 turns, and separate 'em into ten sections of ten turns each, no section has any two wires over 1kV apart when the output is 10kV. A little bit of spaghetti and some wraps of polyester (mylar) tape or (oldschool) paper, are all the separation you require. One can make almost any HV with a minor customization of the transformer.
The oldest DC/DC converters used saturating audio transformer cores in a one-transisor blocking oscillator; it isn't efficient or neat, but it worked fine.
With the Source voltage raising, does this keep the Vgs above the Vgs(thr) ? I would think if you're not careful you could put the MosFet into linear region, thus causing a little heat.
Wouldn't it be better to have the current sink on the other side of the xformer primary ?
Jamie
MOSFET,
switch off.
limiter:
absent,
Cool!!
On a sunny day (Mon, 03 Oct 2011 18:19:32 -0400) it happened Jamie wrote in :
The aim is to go for 100mV peak at the shunt. That leaves plenty for the logic level switching MOSFET.
Not sure what you mean by that.
On a sunny day (Mon, 3 Oct 2011 12:36:03 -0700 (PDT)) it happened whit3rd wrote in :
Yes, I used 2 compartments for this thing, but I did not insulate and wind very carefully for the test coils. The real ones will have some insulation tubing too.
I added some test turns on top of one and grounded those, 1kV out. I smelled ozone :-)
(added ten test turns to find the exact turns per volt, did not count when winding).
Yes they were great as blocking oscillator in TV sets, both for H and V. Until I discovered the UJT.
I think inductors are great fun, I bought several of those E cores with formers, you can make any voltage you want with that.
From context, I think Jamie wonders if the current sense resistor R3 would be better placed in series with +5V and the top of the transformer primary, rather than between the source of Q1 and ground.
--Winston
Exactly.
Jamie
(...)
Jan mentioned that current through the FET measured 11 mA, so the voltage across the current sense resistor R3 cannot exceed about
11 mV. You can see from the FET data sheetExample: As shown in "Fig 1. Typical Output Characteristics" 4 V from the PIC to the gate of the FET will saturate it to the point that it will happily pass 2.2 A at a source to drain voltage of 0.1 V. If we extrapolate the 4 V line down to say 3.989 V, we see that the FET will still be quite saturated and well capable of
2 A Id at the same 0.1 V (d to s) drop. The circuit thus has ~182 times the current sinking ability that it needs, even after the 11 mV reduction in Vg.There *are* circuits (some high-side D.C. 'Electronic Speed Controllers') where low battery charge can fry the high side FETs when run at high temperatures, so the effect does happen. Just not in this circuit.
--Winston
On a sunny day (Wed, 05 Oct 2011 19:13:12 -0400) it happened Jamie wrote in :
OK, I did not see Winston's post, but if that is what you mean, then this funny answer: I could not find a comparator with 2 '-' inputs ;-)
The whole idea of this circuit is to do both the cycle by cycle current limiting, and the voltage compare, with 1 comparator.
I thought I could get away with it by doing the voltage feedback via a resistor network to the '-' of the PIC comparator, and the shunt in series with the primary from the +5V, but somehow could not get it to work. This may be my lack of persistence, or some basic thing. The voltage feedback should however always set the reference level of the current limiter. So I put the current sense in the '-', and control the reference for it with the output voltage on the '+' input of the comparator. That worked first time. If you can find a comparator with 2 '-' inputs, of get it working with the shunt in the '+', let me know.
The other reason for keeping with this, and not looking any further is of course that you can so beautifully, and fully linear, control the output from the top of R1, from 0V to maximum, with 0 to +5V.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.