Saturday, June 21, 2008

Mushrooms and transformers

I hadn't been to Woolies for a while because Nicole has been doing all the shopping while I work. But because we both had a reasonably busy week we went together again for the first time in 7 weeks I think. Anyway, Nicole likes mushrooms.

The apples were displayed nicely.

After doing our shopping I bought a new pair of pants and after that I drove us to Henley. We wanted 2 things:
1) See the beach and smell the ocean
2) Get coffee and cake

This is the Henley beach jetty. I don't think I've ever photographed it from this angle before... Exciting!

There was also bit or sunset going on. Gets a bit boring after a while ;-)

Imagine living in one of these houses. You have to see the sunset every day! What a life...

This is the place we had coffee and cake. A quite nice little discovery.

Remember the power supply I started building? I ran into a small issue: HEAT. The end transistor can handle 200°C max. but not without cooling. I was using one 2N3055 which would have had to handle 34 volts x 3 amps = 102 Watt, in pure heat. That's a small heater... To cool this I would need a very BIG heat sink (cooling element) and also forced cooling using a fan. I no like.

The problem is this: The power supply's end stage is basically the external load in series with the internal transistor. The total voltage over these two is always the same in my old design: 34 Volts. If the voltage over the external load is only 1 Volt(this can be regulated), there is 33 Volts over the transistor. When the external load is also sucking 3 Amps through that transistor, the transistor is consuming 3 x 33 = 99 Watt of power.

So I bought the above toroidal (ring) transformer which turns 220V into 2x 12V. Rectified this is 2 x 17 Volts. So I can make the total voltage over transistor plus external load 17 Volts or 34 Volts. I'm going to make a switch that sets the output to two seperate ranges: 1-15 Volt and 14-30 Volt. This way the voltage over the transistor can never be more than 16 Volts. I'm also reducing the max. output current to 2 Amps AND I'm going to use 2 transistors. So now the max. power dissipation of one transistor will be 16 Volts x 1 Amp = 16 Watts! Easy to cool down :-)

Oh, by the way, if you have read all the above, you're a hero! And if you understand it as well, you're also a geek! :-)

6 comments:

Sneezer said...

i passed right over the technical jargon because i was waiting to see the toy transformer you bought, sigh.

Vincent said...

@ Sn33z0r: I'm sorry to disappoint you, Optimus Prime.

Gledwood said...

everywhere by the sea loooks the same!!

take care i'm kind of ok

Anonymous said...

The switch.. you engage either one or two coils. Looks to me like you then could switch between 0-30 at 1A max and 0-15 at 2A max. Depends on what you want to use the thing for i guess.. breaking your range with a switch looks like a not so good idea to me. Imagine you're playing with something in the 10-20V range..

Also i vaguely remember putting two transistors in parallel without them having a separate regulating currency is not such a good idea.. but my memory is getting rusty here..

Alternatively, why don't you put in some more elements (double what you have) to make it possible to individually regulate both outputs? From zero, you then regulate one from 0-15 and the other from 15-30. Two power supplies in series so to speak.. make 2x2 outputs and you're totally flexible.

Or you could stick with your original 99 watt design. If the tor can handle it, why not let it steam? or chuck in 2 to half the problem. How often will you need 3A at 1v?

Looks like a fun project :-)

kees said...

@Mic "How often will you need 3A at 1v?" When you're welding train wheels to the tracks...

Vincent said...

@ Gled: Yay! You're alive :-) Yeah, there's usually sea and beach hehe.

@ Mic: You're a hero-geek! But that's no news to me anyway lol.

Putting 2 coils in series is harmless as long as you do it in phase with each other. Each coil can actually handle up to 4 Amps. I believe something similar was even done in the 25V/2A supply we build at the same time no?

Using transistors in a parallel way is not uncommon either as long as you give each it's own output resistor.

For the original plan you still need to cool and for that you need a VERY big heat sink. I simply don't have the room for that.

I take your point about being less flexible but I don't see a need for having to vary the voltage from 10-20 in one application. Well, not fluently anyway.

Anyhow, I have to go back to Jaycar today because I need more wire :-/

@ Kees: That's going to need a bit more power I think. Interesting scenario though :-)