The Universality of this IC is amazing. It's fairly basic,Open-Collector output, no Feedforward, no integrated High-Side driver or synchronous rectifier provisions, but those features can be added without too much grief
I use this in combination with a High-Side driver IC, and inverter, and a couple of transistors to implement my Sync-Buck Converter, "RPC1"
Dual outputs are combinable, ORed together for a single-ended design, or separate for a push-pull.
No inherent limitations on SMPS topology
Or, the IC doesn't even care if you just drive an LED with its output
I like to build circuits, use standoffs to mount them to a pine or oak board
Put salvaged plexiglass over them
This one uses the schematic above to create an LED strobe light
It's also a test case for my "RPC1" regulating power converter Buck board PCB
I wanted to have a quick-grab device I could configure in different ways, but primarily for varying the effective resistance of a low resistance, ie especially for applying an electronically-variable load
One application is for solar battery-charging.
When the battery is full, you get to choose whether you disconnect the solar panel, or you can apply a proportional load.
This circuit is great for adding that load
I was planning a Bike Light with a Lithium Battery bank and therefore needed a way to
I need to use power often on the workbench to test my little transistor, IC and LED circuits. I need a rough idea of the current.
Often these are one-off circuits for some niche project like testing some special 1800K LEDs, and I want to use the actual power cube I'm going to deploy it with. It's fairly easy to re-route my 12V battery input to this circuit instead from a power cube
This circuit utilized a quasi-logarithmic scale, with a 10:1, a 2:1 and a 5:1 ratio step. A little confusing maybe, but it portrays the feeling
I had this TO-39 transistor. I wanted to see how well it could work as an RF amp
I connected this amp to an N5247B Keysight Vector Network Analyzer and measured S21, ie a ratiometric measure of "how much power out compared to how much power in"
At one point, I had some values that didn't work well - this is an attenuator. It would be crazy to design an active attenuator by designing a really really poor amplifier.
Usually flat performance is more useful. This circuit has some peaks and valleys - often you're trying to flatten out dips and peaks
With a series-connected Voltaic stack, if the elements are of differing capacity - which they will be, they can't be EXACTLY the same - it's useful or necessary to balance the cells.
The bottom Voltaic element, referenced to ground, is self-explanatory. The others, the Vgs is additive to its position in the stack..... hence the need to limit the Vgs on the upper shunting PMOS FETs, so I used Zener diodes to clamp their gate drives.
This was an early brainstorm session. Later in the Smooth Power project I put load resistors so the PMOS aren't directly shorting the Voltaic cells
For the fringe case where a DC solar pump runs directly off a solar array, the boundary conditions of sunrise and sunset are problematic for both the DC motor itself.... and the ability to utilize the energy that would otherwise be wasted either heating the motor.... or simply open-circuiting the solar.
This circuit continuously samples the main bus (solar) voltage, starts the pump if above a threshold, and then either
I tried to make this too feature-full. Would have been better off not trying to implement so many different settings.
In the process of testing, I damaged the little digital meter, and haven't gotten it up and running. But you can't win them all.... and it was good practice
While all my features would be useful for my needs and typical use cases, the amount of little details such as designing and building instrumentation amps to drive the meters bogged me down on this project.
Still planning to vitalize somehow.
Each of these 6 LEDs has each of its Red, Green and Blue leads connected to its own IO pin in the PIC18F45k42 uC.
Each one is PWMed to get the appearance of variable brightness.
While there are 3 PWM modules on this IC - that would only support 1 RGB LED. I didn't use them. Here I've implemented "bit banging" to effectively get 18 PWMs.
This scheme uses a large proportion of the available processor time. Although, the PWMing might be much faster than needed.
MainFlasher-F42 (c)
Download