PIC controlled 12v waveform generator

The '3055 is a relatively low beta transistor, hfe can be as low as 5 and and the PIC may not be able to source enough current. Use a PIC to drive a small switching transistor that controls the gate of a power MOSFET. A switching supply MOSFET would be a good choice...in fact, take a look at the IRF and ONSEMI websites for the gate-drive methods.

Hi,

The lamps are two per set, with each set being driven by a high-frequency inverter that runs on 12vdc, and they have a nice growing / shrinking effect as you vary the voltage to the inverter. You can check them out at

Ryan

Which lamps in particular are you dealing with, and how many lamps per "set"?

Thought I'd look up PNP signal transistors, like the plastic equivalent of the 2N2907, maybe the 390x or 440x (I can never remember which of those is which), and decided to check Radio Shack - I've seen 10-packs there before, for pretty reasonable, given the convenience. So, I go to their website, and do a search on 'pnp transistor', and the first hit is MPS2907, at $0.69 ea, with a picture of a TO-220.

Idiots.

Thanks, Rich

drive a

look at the

equivalent

those

their

MPS2907,

I was noting the 2n3906 as being _similar_ to 2n3904 according to them. That TIP42 with 65W dissipation and an Ic max of 6.0mA is interesting too. I didn't realize it was such a high voltage transistor. ;-)

ummmm, ok.....

So is there anything similar to the 3055, with a slightly higher beta that could be run from a PIC? I only need to switch about 500mA at 12v (~6W at peak) per transistor. The MOSFET idea is fine, but then I have to balance two transistors in their transition region, which seems harder... Is it really so complicated to get a simple 12v waveform?

Ryan

I'm trying to get the lights to phase on and off, not just jump from the on/off states. So like a sine wave, but above 0v. This way they have a cool growing/shrinking effect, as they turn on from one end to the other. So I need to operate the transistors in the transition region... I'll check out the IR* parts you recommended... thanks!

Exactly what kind of "waveform" are you looking for? If you just want a squarish wave, then a MOSFET is a piece of cake. You don't have to really worry about bias. MOSFETs are nice because the gate (base equiv) is a voltage controlled junction, not current controlled. In fact the input impedance of the MOSFET is so high that you can assume that it is basically infinite. The "on resistance" Rds(on) can be very low. So low, in fact, that you can switch very large amounts of current without needing a heatsink.

When I'm tinkering, I like to use NTE2985 parts. These are way overkill for switching 500mA, but for tinkering around they are just fine. These transistors can switch 30A with an Rds(on) of .05 ohms with 5V drive applied to the gate. Do the math, the dissipation is very, very small for even large currents. I love these things. ;-) The other important thing about these parts is that they are known as "logic level" MOSFETs. This means that the gate is turned fully on with less than 5V. Google on "logic level mosfet" and have fun.

For smaller currents, others should be able to recommend cheaper, smaller parts. I believe the IRF510, IRF520 and IRF530 might be interesting to you. Digikey sells the IRF510 for about a $1 in single quantities and can handle something like 5A, but the Rds(on) is about 10 times as high as the NTE2985 (~.5 ohms). I only used the NTE parts because that was what a vendor had in stock and I got a decent price on them. I was actually looking for something else at the time, but NTE cross reffed the part to the NTE298x series, so that's what I bought. ;-)

You use PWM, and ramp up and down the PW control value. Especially with a PIC. and Especially especially if you're talking about a 2N3055 to drive an LED!

Have Fun! Rich

wrote

Aha, well then I suppose that you _can_ do it that way, but how about if we just turn the bulb full on and full off really fast instead (hundreds of times per second)? Then by varying the ratio of on time to off time over seconds of time, you get a varying brightness in the bulb. Google up PWM (pulse width modulation), it's what you really want to do. ;-)

Hi,

I am interested in capturing the slow "growing" effect that you see when the voltage is ramped up, where the light actually fills the tubes from one end to the other. This is an effect that can only be seen at approx. 1 Hz or slower, with a sine wave type input. I'm hoping to also get the higher frequency thing working also, so I'm looking for a frequency range of 0-1000 or 0-10000 Hz, with a fine control of the frequency at slower than 1 Hz... I'm going to try playing with the PIC PWM control value and writing some BASIC or whatever to control the frequency modulation and control method. I'm still nervous the PIC won't source enough current for the 3055's, but I'll check out the IRF510, IRF520 and IRF530 transistors when I get time. I'm not too experienced with PIC design, does anyone have recommendations for a (simple or cheap) PIC / programmer package to use for this project?

that should work ok, though not with a sine input of course, unless your circuit has a bridge rectified input.

for what? The reservoir will turn 1kHz into dc, and if you removed that to modulate the bulb at 1kHz I dont think you'd see anything but a lit bulb.

It wont, use a darlington if you dont want to buy the fets.

NT