Wednesday, November 8, 2017

While Searching for my Next Engine to Build

I came across an interesting post on Pinterest: a Stirling engine-driven battery charger. 
If you haven't already noticed, that's a rendering of a 3D model, not a photo of an engine, and when I went to the original source, Interesting Engineering, it became apparent it's vaporware.  The original rendering and short little article was by a three Mechanical Engineering students who graduated before building this. So no plans and no numbers to understand the problem.

Stirling engines are nothing new, Scottish Clergyman Robert Stirling invented the concept in 1816, and they're very popular among home machinists.  Stirling engines are heat engines that operate by cyclically compressing and expanding air - a good explanation here and a good animation of how they work here.  A well built Stirling engine can turn a good sized fan blade, or spin some sort of flywheel.  There are models which will work on the warmth from your hand

I can find that a well built Stirling should achieve efficiency around 66%, but what I can't find is how to design it or how big it needs to be (physically) to deliver a desired amount of power.  In this application, you know that USB chargers are typically either 5W or 10W.  In my mind, that tells me if I want the engine to charge things at the 10W level, it needs to be sized to put out more than that.  How much more?  Since there will probably be some inefficiencies in the circuitry, lets say the generating system is 75% efficient, that means the input to the generator needs to be 10/.75 =  13.3W and the input to the engine would need to be 13.3/.66 = 20.2W.  To me, that ought to be how the size of the flywheel and the piston are calculated.

There's a handful of videos that pop up on YouTube of folks using Stirling engines as phone chargers. I don't know that they're honestly 20W charge rates, but some folks are claiming nearly that. Time for a little more research. 


  1. When these students produce an actual masterpiece with report of designed and measured performance numbers, then they deserve an actual degree. Until then, they're just copying out of an Encyclopedia.

    Google "phillips stirling mp 1002 ca" for a real working unit, the manual is and has numbers.

  2. The tea light is an interesting energy source.
    I'd favor painting the cylinder black with high heat paint, and focusing a small magnifying glass on it using the sun as your candle.
    Because I'm a notorious cheapskate.

    Just to see what you could get with that set-up.

    1. The tea light will fairly quickly paint the appropriate side of the cylinder black. No sense painting the other sides as well, since that black would then radiate heat better than the silvery metal.

      And why settle for a small magnifying glass? Why not take it here:
      and put it at the focal point? And while you're there, you can collect some of the ready-cooked birds that make the mistake of flying through the area. Or part way through, anyway...

    2. Mark - I wouldn't be surprised if the concentrated sunlight at the focal point of that array would melt the aluminum engine I'd make.

    3. That's why I'd go with the smaller array.
      The object is to heat the tube, not melt the tube. We're trying to replace the candle's heat, not pour flaming lava on it, right?

      Now, if you wanted to scale it up to power the entire domicile, we can talk, but I suspect the aluminum cylinder scaled up appropriately would require construction cranes to emplace. NTTAWWT.

    4. I suspect that it might not only melt the aluminum engine, but would probably vaporize it as well!

      Now if you only had made it with unobtanium instead...

  3. Hello SIG;

    This is off topic; I have questions about your choice of milling machine (since I am in the market for one myself) and I cannot find your email address, hence I am posting here. I believe you have the Grizzly G0704 model and I would like to know how you feel about your purchase? I am was considering that model but I became aware of the plastic "safety gear incorporated in it and do not like that feature. Since it is designed to break if the mill is overloaded, how deep a cut in steel and iron can you take with this machine? Also, in retrospect, if you had it to do over again would you consider either of these two models: the Grizzly G0795 and the Grizzly G0619. What would you consider their weaknesses? I like the option of turning the headstock on the G0619 90 degrees and I like the oil filled headstock on the G0795. The table on the G0619 is small however. I value your input as an engineer and would very much like your opinion on these machines and their merits. Please forgive the off-topic nature of my inquiry as I was not able to find an email address to PM you. Thank you for your patience with my questions.

    1. George, this might be best taken to email, but I'll start here and see what happens. If folks want to see this conversation, we can keep going here. (The email is sigraybeard at gmail dot com with the usual substitutions for the symbols. Look in the right side bar, under the list of books and then under the About Me block. You'll see a "Contact Me" paragraph.)

      All in all, I'm happy with the G0704. The machine was good as a manually operated mill from the factory. It comes with a sheet of specifications for different things and what they measured (allegedly). I think all of them were well better than spec. I can say I haven't broken the gear but I'm just one user, after all.

      As for how deep a cut to take, I've been using the Speeds and Feeds calculator from and I cut aluminum at speeds that amaze me. I'll run the spindle RPMs at max, near 2400 RPM, feed at 25 IPM, and the program still tells me my combination of RPMs and feed is too slow so the tool life will go down. I haven't cut hundreds of feet of aluminum or steel with it yet, so maybe the gear is a "just waiting to happen" thing, but I'm considering upgrading the motor so that I can get more RPMs out of it. Then I can feed faster.

      I think that obeying the speeds and feeds calculator helps ensure I don't do things that will break that gear. I think that if you try to take too aggressive a cut for the spindle, that's when trouble hits. Mills are a system, designed to be used in certain ways. When you exceed what they're designed for, of course things can break.

      Of the two you mention, I like the G0795 specifically because of the bigger table than the 619. I like that both of them emphasize power tapping, although in that area, the 619 seems to be the winner, with motor reverse buttons on the end of every arm of the quill.

      The headstock on the G0704 can be turned horizontally, although it's a bit of a pain. It's fairly well balanced, but it's clearly not a design feature. It seems like someone said, "for a few more Yuan we can make the head rotate" so the managers said to do it. If it had been designed to change a lot, I'd think they'd come up with detents - say like a woodworking miter saw - to stop it at 0 and 90 at a minimum.

      Because it's a popular machine, the G0704 has a lot of support content available for it and that was a factor. If you've poked around my pages on documenting the conversion, you know about Hoss's DVD on converting it, and if you're read my conclusion to that series, you'll know you can buy one that's already been converted to CNC from Automation Technologies. I've seen other people selling their versions of the CNC conversion parts like Hoss' design. There are other people who do things for the 704, and CNCZone has an active forum for Benchtop machines

    2. I for one would be interested in hearing as much as you would like to say about the matter as one of these bench top type mills is VERY high on my wish list.
      I am a member of a Yahoo group for the mini lathe and quite a bit of the mill stuff seems to migrate over.
      IIRC, The Little Machine Shop had a special going on recently on one of their models that came with DRO's and a Bluetooth Tablet already installed to make CNC an easy upgrade.
      Not cheap of course but I have dealt with them and they are way up there on my list of people I would not hesitate to recommend.

      If you do take this to an Email conversation then please feel free to CC me. I still use my old Gmail address for some things. Bustednuckles AT Gmail DOT COM


    3. SIG:
      Thank you so much for your in depth reply to my numerous questions about the 704! It is encouraging to learn that the plastic safety gear is not as fragile as I had thought it to be. There is much useful information in your reply that I value such as the reference to the CNC Zone Benchtop machines forum. Coincidentally, I was looking at LMS's 5500 Hi Torque bench mill that Phil mentioned as an alternative, and though it looks like a very good option, I don't trust shipping all the way from CA to PA where I live and Grizzly has an outlet right here in PA. Again, thanks for the reply and since I have more questions I will be in touch via email.

    4. Phil;
      Thanks for your offer to email you with questions on this topic. I have been looking at the LMS 5500 as I noted to SIG above and it looks like a very nice machine. My only concern, as I mentioned to SIG is damage to the machine in transit to PA. Were that to happen, trying to get parts, repairs and/or especially having to return the mill to CA might be a troublesome undertaking. Even so, I have questions for you, too on your thoughts about LMS's offerings and appreciate your offer to talk about this via email, which I will do.

    5. Part of my original series of posts on choosing a mill is here.

      The LMS 5500 is a slightly bigger version of the LMS mill I was considering, the 3990. The 5500 might be an X3 or a new class mill from Sieg, but the work envelope is only a bit smaller than the G0704 and quite a bit bigger than the 3990. The features and general look of the 5500 seem very similar to the G0704 (the generic model is BF20), and the LMS is about $300 more than the 704.

      Like Phil says, I've ordered from LMS many times. My lathe is their 3540, and I have all sorts of tools and accessories from them. I wouldn't hesitate to order from them. On the other hand, having a Grizzly store would be like the Candy Factory to a kid, so I'd definitely be there.

  4. SiG, while you're reasearching.....add scalability to the list. Aesop has a point: Under potential future circumstances there won't be much need for a 20 watt phone charger if there's no cellular system to access (the major fault, among many smaller ones, of the BioLite thing), but an inexpensive and highly efficient quiet way to produce a few hundred watts would have value.

    I doubt anyone would be interested in a 350-500 watt fossil fuel generator, even one with Honda's legendary reliability, but a compact, portable $2-per-watt gizmo that will operate 24X7 at 35 dB with low grade heat input (perhaps even stored hot water) would have a market.

  5. I ran across this little beast a few minutes ago,

    some of the parts look ungodly difficult to make and I couldn't do it in two lifetimes but the little joker is Bad Ass and just looks cool too.
    At 26 CC's it would be plenty big enough to run a small generator.

  6. I am guessing that your home's central AC rejects enough heat to charge a lot of phones.
    I am also guessing that the amounts of heat rejected by the freezer and the refrigerator are large enough to run a Stirling Cycle engine.
    Although maybe a thermocouple farm placed on the condenser side of the AC, freezer, and refrigerator would be more than enough to charge phones.
    On the other hand, the Stirling Engine would be way, way cool.

  7. Buy a (or a couple) cheap walk light from Walmart. The type with an on/off switch so that the charge from the solar cell will not be used to power the LED. Set it out in the sun and it provides a light (I use this in my motorhome as a reading light or just a general light to play cards at night). Or you can replace the cheap battery that comes with it with a larger capacity (2300 mWH) battery and use four light units then put the four batteries into a portable battery pack charger. This provides a USB outlet that will charge many different devices.

  8. Lots of interesting feedback here. If you haven't go look at that video Phil posted at 0512.

    One of the things I found while researching is that there seems to be commercially available kW class Stirling engines. These guys claim to have 10kW Stirling engines as products specifically for power back up. A search for kW Stirling engines brings up lots of answers, but I'm not sure we could just go somewhere and order one, like you can a Honda gasoline or Generac propane or LNG generator

    The Stirling as a class has the advantage of being quieter than an Internal Combustion engine. No explosions going on, no valves slapping open and closed or other things. After all, they report numbers like 60 to 70% efficiency and a car engine might hit 30%. Noise represents inefficiency.

    1. "Noise represents inefficiency."

      Scientifically precise, but only to someone who's never accelerated a Harley, a stock car, a dragster, or a radial piston aircraft engine.

    2. I didn't say it wasn't pleasant, I said you can tell the more efficient engine by which one is quieter.

      Do you remember some congress critter saying the automakers need to make the hybrids noisier because people don't hear them coming?

  9. Stirling engines brings up lots of answers, but I'm not sure we could just go somewhere and order one, like you can a Honda gasoline or Generac propane or LNG generator

    There's a fix for that - a small machine shop and a web presence. I'd bet a lot of the parts could be farmed out to Ruger - they seem to have the high precision casting thing down pat, and IIRC Ruger does perform some design and build for non-firearm outfits. Economies of scale and all that if enough are made....

    No reason why a 5-10KW (or bigger....) Stirling wouldn't work, but my thought (above) RE: 350-500 watts was more along the lines of a small, quiet, efficient system to keep a battery bank charged as a supplement to solar panels. I can acres of panels, but they don't work at night and output declines on cloudy days. Being able to easily put 350-500-750 watts back into the cells while I slept, using a small propane burner fed from a 250-500 gallon tank, could be useful (example: A Sundanzer chest fridge (or freezer, they're the same except for thermostat) uses <200 watts/day at 12V).

    1. Better question: Can you outperform the conversion efficiency of solar cells with solar heating on a notional Stirling?
      And at what price point for buy-in?

      This could be dangerously close to a business proposition, and last I looked, ground glass lenses and steam aren't very susceptible to CME or EMP effects, now are they...?

      A 10KW power supply (or several) that would charge batteries to the end of their useful life-cycle with few/no moving parts, minimal investment, no noise footprint, and zero EMP vulnerability?

      My interest is officially piqued.

      We won't even talk about using the power to pump water, then separate it with electrolysis, and use the hydrogen for a power cell.

      It ain't perpetual motion, but it's damned close, and gives you instant redundancy.

      Adios, power grids.

    2. Or, a square yard flat, molded polycarbonate Fresnel lenses sitting on top of an 8 inch square grid for mechanical support. Lens off the shelf from Edmund Scientific, or a cheaper clone. Made from polycarbonate (Lexan) for hail and UV resistance. If you're custom molding, make it a shallow dome shape so it's stronger and doesn't need supports in the middle. In another approach, someone invented a mechanism for individual mirrors in a square dish shape to track the sun in two axes to focus onto a fixed spot on a tower, using only linkages and two motors.

  10. Picture of square dish concentrator is at right column of: