Replies to This Discussion

Permalink Reply by Rick Eis on May 23, 2012 at 10:42pm

Hey Andrew,

Thanks for answering my questions.  

Very cool product and I appreciate your input.  

Does your product use linear voltage regulators? :)

Rick

• ▶ Reply

Permalink Reply by Andrew Dunlop on May 23, 2012 at 11:08pm

Thank you!

That product uses a switching regulator to generate a 6V rail from high voltage, where a linear reg would not work due to the heat dissipation.  It uses a linear regulator to generate a 5V rail from the 6V rail, where the voltage drop is quite small (1V), and so the power wasted is correspondingly small.

Horses for courses - switching and linear each have their merits.

Cheers,
Andrew.

• ▶ Reply

Permalink Reply by Rick Eis on May 23, 2012 at 11:26pm

Andrew,

Sounds like a awesome idea !!!  :)

It appears our APM2 board needs to operate between 4.5 and 5.5 with an optimal 5.2 volts.

And I was looking at some adjustable linear regulators which I am sure you could incorporate.  This would enable you to match the preferred 5.2 volts for those that want to be really precise !!.....very cool concept. 

Sounds like most folks might like your unit :)

Your servo connection locking device also sounds and looks interesting :)

Thanks for sharing,

Rick

• ▶ Reply

Permalink Reply by Andrew Dunlop on May 23, 2012 at 11:40pm

Thank you again!

Yes, servo connectors can fall out if subjected to enough vibration, and so a locking mechanism was developed to overcome that embarrassing point of failure.

An adjustable linear regulator sounds like a reasonable approach, or of course a fixed 5V linear reg might be near enough.  I shall defer to the APM experts on that one.

DeveloperPermalink Reply by R_Lefebvre on May 24, 2012 at 7:09am

Andrew, that's pretty neat.  I built something similar to that myself, but much less professional.  Just a prototype I guess, but same general idea.  I wanted to keep the servo load off the APM rails.  Also I wanted to reduce wire-count going to-from the APM.  So what this does is bring in the servo signal wires only from the APM, and then passes them through to the servo output pins, along with power and ground.  Right now it supplies them with power via a 6V BEC, but it is also configured with a plug for 2S direct input whenever I get the time to install my 2S rated servos.

The 2803 chip is a driver for LED lighting, so disregard that part.

I'm not completely happy with this yet, and will make another.  I would like to build the 5V linear regulator to feed the APM into this board.

• ▶ Reply

Permalink Reply by Andrew Dunlop on May 24, 2012 at 11:15am

I think you are being too modest - it is very nice - much neater than I can ever get my hacks on veroboard.  I do support the concept of keeping large noisy servo currents away from sensitive electronics, as far as it is possible anyway.

• ▶ Reply

Permalink Reply by R. D. Starwalt on May 24, 2012 at 4:50am

Linear regs unreliable? Twas not I my lord!

I side with Andrew in that they are one of the most reliable components in any market.

They are probably also the most misapplied and abused in any market.

I mentioned witnessing an oscillation behavior when the regulator is used at or just above maximum load current. This oscillation was really a on-off-on-off-on... event. This was the device doing what it was designed to do, protect itself. As the regulator heats up (bleeding off the 'extra' energy above the drop out level) the specifications of the device change - they de-rate.

That is why there is so much literature on thermal management of regulators. When they heat up, the performance shifts. The same thing happens in bearings and mechanical things. Bridges change length when they heat up. So do roads, engine blocks, etc. (That is why you took the thermodynamics class in college).

What is a BEC afterall? A form of regulator circuit, that is all. 'BEC' is a hobby name. We EE types call them by function, a voltage regulator circuit.

It takes a power source, a battery in our case, and reduces the input value to a designed output value that will operate up to certain limits. The limits are input voltage, output current, and heat management in a simple case. Component designers then try to think up myriad ways to make the regulator work in extreme conditions. Those are more of the charts and graphs in the component literature.

Looping this back around to the APM... any form of well designed regulation that keeps the APM running within its operational specification is a good thing.

Having worked in electronics for over 35 years (and getting the B.S. later in life), now is a wonderful time to learn electronics. It has never been cheaper to experiment simply because we are buried in so much surplus/salvage/junk that has been tossed out.

You should see my 'stock pile'. I will never live long enough to employ all the parts I have. Resonable offers to come get some will be accepted. ;)

• ▶ Reply

Permalink Reply by Brad Hughey on May 24, 2012 at 7:40am

Hey, R.D., we must have "come of age" about the same time.  I originally wanted to be an astronaut, but my eyes were too lousy, even after the then pioneering treatment of orthokeratology.  I settled for a civilian course of study in aviation, taking my PPSEL check ride 4 days after my 17th birthday.  Quickly realizing there was no way I could buy credentials at any price to compare with the surplus military talent after the end of the Vietnam "conflict", I changed my major to EE.  And the rest, as they say, is history.

Just to insert my opinion on the topic of this thread, linear regulators certainly DO still have their sweet-spot applications, particularly where the voltage drop and current load is low.  Speaking of history, I do indeed remember the very first "consumer" product to have a switching power supply was an HP calculator of 1970-ish vintage (back when HP was an engineering company rather than a vendor of high-margin printing supplies).  The Apple II+ had the first mains-powered switcher I ever saw (learning 6502 assembler and playing Sublogic flight simulator - perhaps a bit more of the latter).  Really, it was the need for 5V logic supplies at high current without having to deliver them with a forklift that drove switchers.

I've always said (platitude alert) that "a cool transistor is a happy transistor".  That is the primary reason (aside from transient overvoltages) linear transistors (biased active) fail.  I'll be the first to admit that thermal management will be a challenge if you wrap a 7805 with black tape and stuff it inside a polystyrene fuselage.  Those transistors will not be happy.  :-(

• ▶ Reply

DeveloperPermalink Reply by R_Lefebvre on May 24, 2012 at 7:43am

Yes, I agree with the 7805 cooling.

That is why mine is bolted to a hunk of aluminum, and sitting in the rotor wash.

• ▶ Reply

Permalink Reply by Brad Hughey on May 24, 2012 at 8:07am

I assume the folks in Olathe chose a switching topology because the "black tape treatment" is exactly what happens to most R/C electronics.  The first time I ever saw a hobby ESC I wanted to immediately grab an Xacto knife and hack away the shrink wrap covering the heat sink.

• ▶ Reply

DeveloperPermalink Reply by R_Lefebvre on May 24, 2012 at 8:10am

Yes, it's pretty nonsensical the way they always cover the heat sink in shrinkwrap.  What really is the idea here?  If you are going to do that, might as well use a hunk of solid aluminum rather than the extruded aluminum heat sink!

I remember back in the day when RC Car ESC's came in a nice plastic case with a big heat sink on top!

• ▶ Reply

Permalink Reply by R. D. Starwalt on May 24, 2012 at 6:51pm

Funny, I wanted to be an astronaut too. Then again I grew up in Florida. Embry-Riddle was a serious want-to but like you, Vietnam had Aeron Eng grads pumping gas after the spin down. Jacksonville Univ was too rich for my blood and my interest in math and physics lead me to an AS degree in XET (X-ray Eng Tech) and if I had stayed another quarter, I could have doubled in EET.

Fame and fortune called and off into the medical xray imaging business I went. In 92 came the AS EET and I continued on with Univ Transfer. The closest I got to being an astronaut was an undergrad program in 94 - The Space Life Science Training Program(competed 3 years to be selected). The 6 weeks at KSC/Canaveral was great and essentially a NASA Boot Camp. That is a subject for another conversation but I could have died a happy man then (walked under one Orbitor in the OPF and got to watch Columbia launch next door to some Air Force folk tracking her with a Star Wars project on her vertical fin - check out assorted launch pics from the 90's and see if you can spot the package).

The BSEET was finally done in 2010 from South Carolina State. Unlike electrons, I never take the path of least resistance and built a spectrum analyzer for my Senior Project at SCSU. The power management on that project was fun but at least with mains power of AC, the regulation concerns were much simpler. Being a ham operator gave me a serious edge also. (There is another story I could tell about a class mate's first attempt at soldering a uP fuel injection project - MegaSquirt. Short version - 'Brandon, you been soldering very long?')

34 years in the medical imaging business has shown me many advances in Elec Eng technology. The best thing is that time to market of useful new stuff has been serious shorted.

I'm buying the next round at the next meeting of The Old Guys of Electronics.

• ▶ Reply

• ‹ Previous
• 1
• …
• 4
• 5
• 6
• 7
• 8
• Next ›
• Page