*most recent entry is at the top of the page.  to begin at the beginning, you must go to the end.

Jump to: 1st Entry  2nd entry  schematics  3rd entry 4th  entry 5th entry

6/17/05

well, its official.  Its done.  It's not perfect, but it will do.  If I get the bug to do it, I will probably add current limiting and sensing, more features and such, but right now, I have a functioning digitally controlled dual output power supply with a range from +- 0 to +- 16.3 volts and I am happy.  Pictures of the final steps of completion are posted below.  Enjoy.

6/14/05

    The birth of a new programmer.  for details, go to my electronics infomarmation page on how to make a pic programmer from a Digilent JTAG3 parallel cable.  It saved my butt.  I replaced the burnt out PIC, and located a socket for the new one to go into, sooooo, I shant have to do that again.  I loaded the firmware on, and lo and behold, IT WORKS!!!  here is a picture of in its "ugly duckling" stage.

6/13/05

    THE DEATH OF MY PROGRAMMER!!!  oh it hurts.

so I took my own advice and decided to finish this darn thing.  well, I really got into it.  in the above picture, you can see that everything is installed (well, you will just have to take my word for it...)  I took the time to socket EVERYTHING that might need to be replaced, EXCEPT for the pic.  yes, thats right, in my infinite wisdom and haste ( I had no more sockets left... ) I decided to just solder the pic in place.  real smart.  "it wont get hurt" I thought.  "it will be protected by the 7805 should anything go wrong"  "its not doing anything other than reading the A to D and running an LCD, checking some buttons, what could go wrong?" well, I could accidentally plug the 7805 into its socket BACKWARDS !!!  but wait, I'm getting ahead of myself here.  you want to hear the whole story.

    So I get everything all soldered up... boy it looks pretty.  I got 3 buttons, 1 knob with a built in power switch, 3 green LEDs for power good indicators, I got my LCD all hooked up (socketed of course), I got the ad7303 wired up and socekted, I put in pins to make it easy to program the chip in circuit, I even socketed the ceramic resonator.  I was proud of this work, and eager to see that LCD light up with some cool info.  So I load up my test program (the one in the picture with the LCD display) which would be minimally functional power supply software, and try to program it.  well, all goes well till the verify.  then it konks out saying that the memory did not verify.  BS! I read the memory and it looks fine.  I try again.  no dice.  I check the oscillator, and its not even runnning!  what the heck!  did I grab a bad pic?  no, this one was brand new...  so I try and retry and try again.  no dice (oh btw, this is the MbasicPro software I was telling you about earlier, with their programmer.  that the programmer highlighted in red in my picture) I do everything I can think of with it, but it still wont program properly.  so I decide to switch over to their hex loader software (its a bare bones assembler, but it was the most recently updated software from them, and the included firmware update for my programmer causes it to cease functioning when used from MbasicPro, remember, the one that hasn't been updated in 2 years)  so I update the firmware for the programmer, erase the device, read the device to ensure that it actually had been erased (MbasicPro said it was erasing it, but after 4 erases, the program was STILL in memory), the hex file of the read was all FF's (hallelujah) so I loaded the dual power supply hex file and downloaded it.  HOLY MACKERAL!  THE OSCILLATOR IS OSCILLATING!!!  oh yeah!  I am riding high! but the LCD is not working, so I decide to reset the pic by pulling out the 5 volt regulator.  yup.  this is where it happens.  I put the regulator back in (backwards) and my LCD screen goes black.  "hmm, thats odd" I think to myself and *POP*  the magic smoke comes out of my programmer.  you see, I powered my programmer from my circuit, and when I plugged the regulator in backwards, I can only assume it did not regulate to 5 volts.  it must have gone pretty darn close to 20 volts.  so needless to say, my programmer is dead.  my pic is no longer oscillating too if you were curious.  my LCD is ok, and I had unplugged the AD7303 from the circuit because I thought It might be interfering with the programming.  and I have no way of reprogramming the pic, because my only programmer is dead.  so, there is a good possibility that I fried my pic too, and that will mean unsoldering 28 pins.  ugh.  I think I am going to buy a socket.

6/12/05

    well, I haven't written much, and I haven't done much either, except... Solder up the rectifier, 5 volt regulator and dual tracking regulator.  its all ready to go to accept its "brains"  the only problem is, I have developed a serious dislike for PICs.  More specifically, I have developed a serious dislike for my compiler;  the BasicMicro MbasicPro.  When I bought it about 3 years ago, it was great.  they were coming out with updates for it alot, and there was terrific customer support.  But when people started wanting 18F support, no one answered.  before long, the forum no longer received the frequent visitations by the creators of Mbasic.  it has now been over 2 years since the last update, and frankly I am pissed.  I used to think the product was great.  I have since learned how to program in C, and now LOATHE coding in basic.  the tokenized code was huge and slow, and well, don't get me started.  anyway, I have come to dislike the pics ( I am now using the AVR atMega series.  NO DIVIDE BY FOUR CLOCK SPEED!!!!  in your face microchip.) and MbasicPro.  Also, I wanted to use a bootloader in my project so I could easily update the firmware without the use of the programmer (which I also bought from Mbasic, and dislike immensely.  IT HAS NO PROTECTIVE CASE!  for God sakes, prototypers like me have all sorts of metal and wires laying around on my bench!  sooner or later, somethings going to short out and POOF!  no more programmer!  man.  plus, I purchased the "intelligent ZIF adapter" to make my life easier... Mbasic came out with an update for their software about 2 years ago, and the ZIF adapter STOPPED WORKING ALL TOGETHER! ARRGGHH!!! anyway, back to the bootloader) but their bootloader code is broken (it broke when the update came out.  yeah, the one two years ago, and still no fix) so I am forced to look elsewhere for a bootloader and... so I think you might get the picture.  so what I think I am going to do is just put in a pic 16F876, an LCD display, a knob, a button or two, sew the whole thing together and get on with my life.  maybe USE the power supply instead of dreaming about all the features it could have...  so here is a picture of my current progress:

The 7805 is socketed, the LM324 is socketed, and the pass transistors are socketed (yeah, those little dinky guys :)  what isn't soldered yet is the 120 ac, pic and related hardware, DA converter (its output will be connected to the blue wire going nowhere), the LCD, POTS, buttons, LEDS, etc.  so I figure, I might jsut do that tonight and get this project done.  so, the specs will be:

1. knob controllable

2. LCD readout of voltage

3. 2 to 3 buttons that wont do anything right now

4. no serial port

5. 0 to 17 volts

6. current totally dependent upon my pass transistors, but possible to 2 amps

7. no current sense

so, yeah, compare that to my previous specs and get a hearty laugh.  but hey, I will have built it. it will be done... or will it.  bwah ha ha ha ha!

6/1/05

    I finally got the box I am going to try and shoe-horn all this stuff into.  Here are some pictures of beauty and the box:

The nifty red circuit board in the background is the Digilent Cerebot.  its an AVR Atmega64l with 128 kB of external SRAM, power supply, 6 pin headers all the way around, onboard LED's, multiple programming options, its pretty cool.  Anyway, that's the box I hope to put it all into.

 

5-21-05:

    I have begun developing the microprocessor control of the power supply.  I wrote a program which samples the output of a potentiometer and then directs the DAC to output that exact value.  This means that the PIC has complete control over the output of the power supply.  That means that the output can be controlled by any number of sources such as the on board potentiometer, an application communicating via the serial port or USB, an external EEPROM with a specific output recorded on it ... the possibilities are endless.  I also began interfacing the LCD display to the project.  Currently the LCD display is displaying the potentiometer voltage (program voltage) and the DAC output voltage.  P indicates "Potentiometer" and C indicates "Converter" for the DAC.

basically, I am reading the pot through one of the channels of the pic's onboard ADC, converting that 10 bit number into an 8 bit number (shift left by 2), displaying that voltage on the LCD, instructing the DAC to output that 8 bit number, reading the DAC's output on another channel of the pic's ADC and then displaying that voltage out on the LCD.  so I have a side by side comparison of "program" voltage and "output" voltage.  lots of room for conversion error though isn't there?

I also began putting together the schematics of the project:

here is the raw, unregulated positive and negative supplies.  this part of the circuit is capable of delivering +- 20v at 2 amps.

Here is the 5 volt regulator and the dual tracking regulator.  Please note that if you intend to build this circuit, choose appropriate pass transistors and heat sink them properly.  Under heavy loads, these transistors will be dissipating ALOT of power.  Also note the reference voltage.  This is where the DAC will be controlling the dual tracking regulator.

5-17-05:

My current project is a dual polarity power supply.  The specs are still in flux, but the voltage range is determined.  I like using op amps, but hate the dual supply aspect, so I thought I would build my own dual polarity power supply.  I am using a dual tracking regulator, so the two outputs are not independently adjustable.

Specs:  (lol, some of these are the result of rampant featuritis, some are actual)

Voltage range:  0 to +- 17 volts (done)

Current range:  0 to 2 amps, adjustable (the power is definitely there, I just need a way of monitoring and regulating it.)

LCD readout of voltage, current and “other stuff”

Can be used as a single supply with floating ground, or earth ground 

RS232 operation by computer (not sure what I would need this for, but it would be sooo cool) maybe even USB!

Firmware updateable through RS232 or USB

Fused for protection /with indicator (duh, but not done =)

Main power switched via voltage knob (almost done)

This is a picture of the current breadboard circuit.  What you see here is a PIC 16F876 driving a AD7303 (dual 8 bit DAC) which acts as a programmable voltage reference for the dual tracking regulator.  Currently the PIC is running through a test program, which is ramping the DAC from 0 to 255.  You can see the output from the dual supply on the scope screen.

What this means at this stage of development:

1. I have dual polarity output from 0 to +- 17 volts
2. I have a “computer controlled” output with 8 bit accuracy, yielding an output voltage of 0.067 volts per bit.