Anyone here an EE? Kinda need help..

mrCide

Diamond Member
Nov 27, 1999
6,187
0
76
I'm not too smart in this field (I know the basic).

I'm trying to build a PSU to power a bulb (don't ask). Bulb requires a starting voltage of 1200v and a constant supply of 550v +-10% AC. Amperage required is 5mA +- .05. And lastly it requires frequency of 40KHz..

Here's what I know (little) :): [edit] create oscillating circuit with caps/transistors/inductor and send through transformer to bump up voltage.. [/edit]

I'm not necessarily asking for someone to design it (actually I am), but what I'd like is a point in the right direction.. TIA..

..bumping this..
 

Demon-Xanth

Lifer
Feb 15, 2000
20,551
2
81
Sounds like a high frequency florescent light. Best advice I can give is do a search on what goes into a florescent light ballast.
 

blahblah99

Platinum Member
Oct 10, 2000
2,689
0
0
Yikes, with that kind of voltage and current, I'd stay away from trying to build my own ballast for that light. Just do what demon-xanth said and go buy a pre-made ballast. It'll save you some time and trouble.
 

thraxes

Golden Member
Nov 4, 2000
1,974
0
0
Agreed, those numbers are most certainly not safe to play around with yourself.

Even if you do manage to build and test the PSU without killing yourself it would be a very large safety hazard. Try to get premade one!
 

Superdoopercooper

Golden Member
Jan 15, 2001
1,252
0
0
I am an EE, and I would hightly recommend against trying to build something. Since you can buy a new flourecsent for between $10-50, depending on the size/type, it's not worth hours of pain, and potential $$ you lose from electrocuting yourself.

I could do the project... maybe. Converting the 60Hz up to 40Khz makes the problem magnitudes mroe complicated... but I wouldn't want to try.
 

mrCide

Diamond Member
Nov 27, 1999
6,187
0
76
I suppose I'll tell you what my project is so it makes a little more sense.

I have 20 cold cathode flourescent lights (you know, the case modder type ones).. I want to power them in parallel from an AC source, i want to build a small box with 20 ballasts wired in parallel series or something or other. and have wires come out from that to the lights.

The inverter supplied converts 12v DC to the specs I listed below.. picture of what i'm talking about. Anyhow, Obviously I can buy 20 more and buy a seperate PC PSU to power them but that's NOT an option I want to take. :)

It doesn't seem to difficult minus the frequency stuff. Only problem that I'm not an EE.. :) And uh.. if it makes you guys feel better I shocked myself a few times both on the wires that were bare on the light (i split one side to measure) and on the actual inverter.. and it really wasn't all that bad.. maybe cause it's only 5mA has somethin to do with it? Or I'm superman.. ;)

Anyhow.. that's what i'm tryin to do.. thanks for the replies too
 

mrCide

Diamond Member
Nov 27, 1999
6,187
0
76


<< That board looks pretty simple to me. If you can get me pictures of the component and trace sides, I'll see what kind of schematic I can come up with. I recently took apart my klipsch pre-amp board and came up with this schematic modification here.. >>



cool.. i figure it should be easier to build than even this because there's no DC=>AC going on :) here you go though.... backside, tia.
 

Sahakiel

Golden Member
Oct 19, 2001
1,746
0
86
I don't know anything about cathode lights, but from what I'm reading from your posts, you're really heading off the deep end.
My advice: if you're gonna do something with 1200v, best way is to get someone who knows what he's doing to do it. You screw up and accidently touch the wrong wires while testing or something, and we're gonna have bbq night. Your best bet, if you're really serious about this, is to either modify an existing PSU or to build your own from scratch. Both are basically the same difficulty and process.
Also, to get 1200V to output only 5mA requires 240,000 ohms. I think radio shacks sells up to 100k 1/2 watts, but not sure.
Your inductors would be used only to keep constant current. Drop em in series with the resistors and if you don't have too high a voltage across them, it'll take less than a day to power them up and you won't risk an explosion. Either that, or I'm thinking about ramping capacitors.
Oh, and about shocking. Running 1200V across the fingers of your hand won't kill you. It may fry your nerves and you'll lose motor control in your hand, but that's it. Running 1200V from your right hand across your heart to your other hand or your feet is another matter.
 

blahblah99

Platinum Member
Oct 10, 2000
2,689
0
0
supercooper,

Yup, the hiss is gone and the midrange is a bit clearer. I replaced the generic op-amp with an OPA4227 and replaced the 330 resistors to 520, which reduced the gain by about 5dB.
 

blahblah99

Platinum Member
Oct 10, 2000
2,689
0
0
alrite, i took a quick look at those pictures of the board you had and it looks like the oscillator is a colpitts or pierce design with an inverter as an amplifier. The output is fed into a step-up transformer.

Look up colpitts, hartley, and pierce oscillators on google.
 

Superdoopercooper

Golden Member
Jan 15, 2001
1,252
0
0
BlahBlah... how the heck did you know that? Where did you go to school? Are you "old"? Those are terms I've NEVER heard before. :D
 

sharkeeper

Lifer
Jan 13, 2001
10,886
2
0
With that many lamps you may want to consider RF excitation. Of course, that isn't within the reach of the novice experimenter!

Cheers!
 

blahblah99

Platinum Member
Oct 10, 2000
2,689
0
0
Im only 23 and studying EE at UCLA as an undergraduate... I think those designs were used as earlier forms of oscillator before crystals were introduced. But nevertheless, its in the book "Microelectronics" by Sedra & Smith, which is an excellent book for any circuit designers.

The designed shown in the picture above resembles something similar to those oscillators. I dont think flourescent lighting is too picky on frequency, so as long as its within operating specs.

I'm not sure if this would work or not, but you might be able to build a 40khz sine wave generator using op-amps and, amplify that with transistors, and feed the results into a transformer.