Little 1:10 audio transformer as HV transformer

Sloman

effic=

pr=

= =

--
At your mercy?

Sloman

effic=

pr=

seems =

decreases =

--
As if taste and insight were within your bailiwick.

Bill Sloman

effic=

circuit pr=

seems =

decreases =

--
Geez, Jim often posts schematics and sims and I just got kudos for a
self-heated thermistor design/schematic/sim and also posted three
versions of a pulse generator, while all you've been doing, on-topic,
is trying to one-up Jan.
---   

>What's your opinion about using audio transformers to make high
>voltages?

Sloman

effic=

pr=

seems =

decreases =

--
You think boogers are tasty?

Bill Sloman

effic=

circuit pr=

seems =

decreases =

They are designed for passing low frequencies, and usually not much DC, so the cores tend to saturate at low currents, and there tends to be a lot of winding capacitance. OK for low power stuff.

John

Bill Sloman

effic=

circuit pr=

seems =

decreases =

Not often. He rarely contributes anything these days but snide snips and tales about great things he did decades ago.

It's easy enough to check. He's running roughly 5% on topic lately. Sad.

John

ll Sloman

t effic=3D

cuit pr=3D

seems =3D

reases =3D

.

There isn't a lot of evidence about how the mass of the proton might change with time, and even less for the existence of other universes, so these points are wide open to discussion. There's lots of evidence supporting the idea that we are experiencing anthropogenic global warming. Ivar Giaever doesn't seem to appreicate it, any more than you do, but insisting on discussing the subject as if there wasn't any serious evidence supporting the hypothesis is anything but sensible.

This isn't a sensible statement either - anthropogenic global warming is about the undeniable accumulation of greenhouse gases in the atmosphere and the unavoidable consequences of this for the global temperature. Concentrating one's attention on the temperature measurements as if they were central to the case is anything but sensible - demented is the word that come to my mind.

It won't go away just because there are fewer stories about it in the popular press. The sad fact is that we aren't doing anything like enough to slow it down, and our kids are probably going to be decidedly cross with us for this neglect in a couple of decades, when the consequences are going to be a bit more obvious.

-- Bill Sloman, Nijmegen

On a sunny day (Sat, 17 Sep 2011 12:17:09 -0700) it happened John Larkin wrote in :

efficiency,

prescaler

be about 100kHz,

efficiency (core loss at 100 kHz?)

%...

caps,

transformers.

frequencies.

no

core losses at 100kHz.

it sees about 220pF MOSFETcap.

then.

other PMT:

Yes, I could use 2 diodes and a 2x multiplier to get to 1200V from 5 with my circuit. Some remarks, I see your 'generator' is an opamp. Do you remember the UJT (uni junction transistor), I am sure you do. I did a TV (vidicon) camera timebase with HV generator and drove it from a UJT, with the right capacitor and resistors you can just get the right pulse ratio. If the supply is stable, then the UJT is awfully precise on frequency, no drift. It can also easily be synced. A 555 could be used here too. But I think your design uses too many caps and diodes in the HV, this is exactly what I was trying to avoid. Mine uses only about 11 mA with a 10 MOhm load, and that is a lot less then all the LEDs that are also in the design for user interface etc. And mine is stabilised, works really well. Usually when you have many multiplier stages the Zi increases, and the stabilisation becomes more of an issue.

In all of this actually you can do it with just an inductor, and a HV (say 700V or so Vce capable) transistor, like these guys:

Now that is fun. note how he senses Imax too in a clever way. Cycle by cycle current limiter.

On a sunny day (Sat, 17 Sep 2011 12:24:21 -0700) it happened John Larkin wrote in :

There was recently a study that showed simple BW overhead sheets work much better than powerpoint presentations in conveying information.

efficiency,

prescaler

Then use a different MOSFET. In switching applications a MOSFET with a higher RDSon may actually be more efficient because the gate charge will be significantly lower and therefore it will switch faster.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
nico@nctdevpuntnl (punt=.)
--------------------------------------------------------------

On a sunny day (Sun, 18 Sep 2011 10:42:02 GMT) it happened snipped-for-privacy@puntnl.niks (Nico Coesel) wrote in :

efficiency,

prescaler

frequencies.

no

Well I wound my own transformer today, the second one for this, used a real nice setup: Dremel end fits exactly in coil former, Dremel supplied by current limited lab supply. Dremel in vice on desk, bobbin with wire on vice tightening rod... Hold the wire with a paper tissue so I do not burn my fingers, and fill up the coil former. I did one layer primary X turns, and then filled up both chambers of the coil former. tested it, perfectly in resonance, same low current comsumption. Measured the turns ratio: 1:33. Yippee! More voltage for the same current... The original conrad 1:10 audio transformer:

Its 100kHz resonance on top of the about 4 kHz drive:

MOSFET needs to be on for a large percentage of time here to get enough flyback voltage.

My transformer made with the above method:

Better Vds waveform, only on for a very short time:

Gate drive directly from PIC:

This one is interesting, measuring Id with a 10 Ohm resistor from the +5, 4 kHz drive:

The fast spike at the start is likely the Cds (220pF) charging. The vertical straight jump down is caused by the voltage drop over the sense resistor, The linear ramp is the current increasing in the primary, no saturation to be seen here. Sort of all that :-)

While I don't much doubt this, I'd like some explanation. Is it (useless) information overload or something inherent in the media? IOW, what if both have exactly the same information content... Cite?

BTW, I hate PP, too.

How can you argue with this PP lucid depiction of US strategy in Afghanistan?

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

On a sunny day (Sun, 18 Sep 2011 10:05:08 -0500) it happened " snipped-for-privacy@att.bizzzzzzzzzzzz" wrote in :

Well, I never watch that sort of stuff, you have google, I cannot remember where I read it, I read many sources. But I think it had something to do with brain activity wasted on perceiving layout rather than content.

On a sunny day (Sun, 18 Sep 2011 11:20:47 -0400) it happened Spehro Pefhany wrote in :

Well, looks like a picture of Obamas brain. Just a big mess, he turns the world into.

On a sunny day (Sun, 18 Sep 2011 12:13:52 -0400) it happened Jamie wrote in :

efficiency,

prescaler

frequencies.

can no

flyback voltage.

kHz drive:

resistor,

seen here.

No, does not seem needed here. I measure the HV and feed it to a comparator of the PIC to do hysteretic on /off of teh PWM. In the unlikely case of overload the 5V supply will current limit.

efficiency,

prescaler

frequencies.

no

flyback voltage.

kHz drive:

resistor,

seen here.

Are you using current sensing feed back?

Jamie

efficiency,

prescaler

frequencies.

can no

flyback voltage.

kHz drive:

resistor,

seen here.

/off of teh PWM.

The reason I asked is, we did a circuit not too long ago where I used current sensing on the primary for both forward current and FLyback current. At the head of the circuit was the bias set that formed the correction to maintain constant output. The net results of this gives you a variable carrier frequency due to no time being wasted on the flyback cycle. As soon as the circuit detects no more energy and demand was still there, the next on duty cycle would start and wait for max current or a partial of that to build in the primary..

This allows me to not need a uC or an PWM oscillator, since it's basically doing it's own oscillation. I just had the comparator circuit regulating the current point with the on cycle and monitoring the flyback cycle for no energy so the circuit could get on the next cycle. This also makes it easy to work with a transformer of many different induction sizes on the primary and variable loads.

Jamie

"The cognitive style of powerpoint" by Edward Tufte, There's a scan online at

Tufte also wrote "The visual display of quantitative information" and "Envisioning information", both classics and themselves both excellent illustrations of what they teach. Highly recommended.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510
845-480-2058

hobbs at electrooptical dot net
http://electrooptical.net