Hobby project - how do you choose the solution?

[Borek]

While the question arose around a particular project, it is much more general. Now and then I see there are many ways of building a device that will do what I want it to do (hardly surprising). But as a beginner I often feel overwhelmed with the plethora of solutions. Cost of the elements and their availability is often similar (in a $10 hobby project I don't care about saving or loosing a buck or two), it doesn't matter much to me if I need to solder 10 or 15 elements, all approaches (at least on paper and when reading elements specifications) look sufficient for my needs - so the lack of experience makes the decision quite difficult.

Before I roll a dice, I wonder what are your practical approaches to that?

(In particular it is about making a new switcher (12 V DC to 200 V DC) for my Nixie clock. The one I have is OK but I bought it ready and it is on a separate PCB, I want to combine everything on one board. NE555, MC34063, MAX1771, LM5022 - I guess the list is much longer, I stopped searching here. From my point of view the only difference between these ICs is I have a spare 555, but I need to buy the coil, MOSFET and fast diode, which will be necessary for every other chip as well - so I am not going to spare myself shopping by choosing 555.)

 

[Guineafowl]

While the question arose around a particular project, it is much more general. Now and then I see there are many ways of building a device that will do what I want it to do (hardly surprising). But as a beginner I often feel overwhelmed with the plethora of solutions. Cost of the elements and their availability is often similar (in a $10 hobby project I don't care about saving or loosing a buck or two), it doesn't matter much to me if I need to solder 10 or 15 elements, all approaches (at least on paper and when reading elements specifications) look sufficient for my needs - so the lack of experience makes the decision quite difficult.

Before I roll a dice, I wonder what are your practical approaches to that?

(In particular it is about making a new switcher (12 V DC to 200 V DC) for my Nixie clock. The one I have is OK but I bought it ready and it is on a separate PCB, I want to combine everything on one board. NE555, MC34063, MAX1771, LM5022 - I guess the list is much longer, I stopped searching here. From my point of view the only difference between these ICs is I have a spare 555, but I need to buy the coil, MOSFET and fast diode, which will be necessary for every other chip as well - so I am not going to spare myself shopping by choosing 555.)

Usually, if a decision is difficult, you don’t have enough data. Specify performance requirements from power supply and signal into final signal out. At all stages, component choice is a balance of simplicity and cheapness with quality - as hobbyists, we have the advantage of not having bean counters breathing down our necks. And remember, there are often many solutions to a problem.

Don’t let a plurality of solutions confuse you or be a problem - it can only ever be an advantage. One of the best aspects of electronics as a hobby is the cheapness and ready availability of components.

 

[.Scott]

I'd be thinking adapting a photoflash circuit:
http://www.linear.com/product/LT3468

 

[Borek]

Specify performance requirements from power supply and signal into final signal out.

Yes, that was one of the things I checked, but looks like all the solutions I found can deliver what I need (200 V/10 mA being a generous upper limit) using the power supply I intend to use (12 V/1 A) without problems, so it doesn't help. I also thought about choosing based on efficiency, to not let things get hot, but it looks like the efficiency depends more on the properties of the MOSFET and diode (mostly their speeds) than on the IC. Also I am under impression, perhaps an incorrect one, that around 80% which is what these chips are capable of delivering the efficiency starts to depend more on things like the PCB design than the controller choice.

Basically the problem as I see it is that there are many equivalent solutions, and I wonder if you (those with a lot a practice) have any strategies of making choice - like, sticking to a brand, avoiding a brand, avoiding solutions that require high power resistors - I am just making these up, but in my experience each trade/each engineer has such rules of thumb.

 

[dlgoff]

Basically the problem as I see it is that there are many equivalent solutions, and I wonder if you (those with a lot a practice) have any strategies of making choice - like, sticking to a brand, avoiding a brand, avoiding solutions that require high power resistors - I am just making these up, but in my experience each trade/each engineer has such rules of thumb.

If you are just making one, then none of this is relevant IMO.

 

[jim hardy]

I say for a home project - always steal a good idea.. .That's how one learns , use what's already out there for an idea launch platform..
@.Scott 's suggestion seems to me an excellent one.

Since film cameras are virtually gone they're dirt cheap in junkshops. Most point & shoots have a built in Xenon flash circuit capable of 400 volts, so do disposables . Around here $2 is top price.

Take the magnetics out of a junk camera and add a regulator of your own design ?
If it's a decent camera salvage the lens too, for homemade telescope eyepiece...

What a time we live in -
"I will fear no boredom. My junkbox runneth over."

old jim

 

[Averagesupernova]

As a hobby, is the goal mostly learning aside from the obvious need of a power supply? @jim hardy has a very good approach. Learn from what someone else has done. If this is mostly about learning, why not try several different approaches?
The way I see it in design is this: Ready made more integrated components such as a power supply modules, display modules, etc. raise the risk of redesign or re-layout due to obsolescence. Company that builds the module goes out of business and you need to find another module which may or may not be a drop in replacement. More discreet components that you would build a switcher out of will likely not disappear from the market. The flip side to this is increased design time, potentially more risk in reliability due to a new and potentially flawed design. Probably others as well. After all, a great design is the best set of compromises.

 

[Svein]

With several decades of experience, I have landed on the following practice:

1. Start by writing down at least 5 different solutions to your problem. At least one of them should be "far out".
2. Add some pro & contras to each solution
3. Start out by implementing the "best" solution, but do not throw away the others. You may be coming back to them later.

 

[Tom.G]

With several decades of experience, I have landed on the following practice:

1. Start by writing down at least 5 different solutions to your problem. At least one of them should be "far out".
2. Add some pro & contras to each solution
3. Start out by implementing the "best" solution, but do not throw away the others. You may be coming back to them later.

Especially number 3!

 

[Borek]

Thank you guys.

Ordered a bunch of elements, will try to make more than one switcher just to experiment

And I have "invented" my own approach, LTSpice simulation suggests the idea is right, will see how it performs in reality.

 

[Averagesupernova]

Wouldn't mind seeing some pix of your endeavors as time goes.

 

[Borek]

Wouldn't mind seeing some pix of your endeavors as time goes.

No problem. I know it is worth discussing things with you all, I have learned a lot from a 555 thread. Not that I am surprised, after all - PF is PF.

 

[Borek]

The prototype of the solution I selected as the most promising works quite nicely

https://www.physicsforums.com/threads/555-based-switcher.927668/

 

[berkeman]

While the question arose around a particular project, it is much more general. Now and then I see there are many ways of building a device that will do what I want it to do (hardly surprising).

As you gain more experience in EE/electronics, you may find yourself using your hobby projects to learn more advanced skills. For example, a roommate of mine back in undergrad in the 1980s built a nice stereo from discrete transistors, and built a simple 8-bit microcontroller from logic gates. He could have used one of the new Intel uC/uPs at the time to build up his processor, but he chose to build it up from discrete gates to challenge himself and use it as a learning lesson.

So on your switching power supply project, if I were building it as a hobby project, I'd probably build it up using a PIC uC as the loop controller. That involves programming the DSP code and the feedback loop code and overcurrent detect and shutdown code, etc. I've had the need to do something like this in my work a few times over the past few years, but they never became projects (they weren't simple switcher projects, but complex power supplies with unusual requirements, so using a PIC for the central controller may have made sense).

In any case, enjoy the fun of your projects!

 

[Borek]

So on your switching power supply project, if I were building it as a hobby project, I'd probably build it up using a PIC uC as the loop controller.

My adventure with electronics started last year with an Arduino, so I already consider myself an Atmel guy The idea of building the switcher around one of the Atmel uPs crossed my mind, especially as my previous project was based on ATtiny13.