Easy Build Clock with a 32 Day Runtime

[@MrBizteck]

I don't need a clock ........ But I do have a printer and time off work ......... now I need a clock !

[@michaelk4741]

Very great work ! My son printet the parts for my birthday and i had to build the clock, it needs 5 hours and gave me so much fun ! But the most fun i have ist to see the clock running !!

[@kellywilson2858]

Steve, I just came across your page. I'm very impressed by your engineering skills. Your clocks are very well thought out and I can't wait to print and build my first one.

[@stevesclocks]

Thanks. I recommend www.myminifactory.com/object/3d-print-small-8-day-clock-337387 as a great first clock to build.

[@jacquesfavre999]

Very impressive Well done You certainly have the longest running 3D printed clock

[@joetoeforsho]

Can’t wait to start printing this. I’ll keep you posted. Thank for sharing and amazing work. Keep it up 👍👍

[@cncwoodworxroc6881]

Excellent design and execution. Great design simplifies the complex to seem effortless, this is top notch. Truly impressed with the run time. Well done!

[@justalabratmr.6858]

Beautiful clock! 32 days of run time! Wow!😍

[@woodwaker1]

Steve you do some brilliant work. I have all of the parts printed and my nano programmed for the silent desk clock, just waiting on the board I ordered from you. When that one is done, I'm going to try one of your mechanical clocks. Thanks for doing these and pricing them very fairly.

[@stevesclocks]

Thanks. I am close to releasing the battery powered electromagnetic pendulum clock. I think it is much more impressive than the stepper motor clock.

[@T0phen95]

Thanks for this. This one is so much more simple to build than some of the other ones out there. I'm going to get your design and make a nice clock for myself now.

[@stevesclocks]

Thanks. I tried to make it as simple as possible without giving up functionality. The runtime improvements were a nice surprise.

[@janetmorris6792]

Nearly finished your previous model // Will be going for the plans of this in the next couple of days so I can start the print run // NICE CLOCKS

[@3dprintingfun329]

Really nice. I have been looking for 3d printed clock to make. You hit the nail on the head by using fewer store bought parts. It can be a pain to find parts on a BOM. You seem to be the best designer in this area that i have seen.

[@jmcarp0]

Yes I love the consolidation of part types.

[@Buglbiker]

I see a lot of improvements! Run already on my printer. Thanks for developing this wonderful watch. Slowly, I'm running out of space on my walls 🤪😄👍 thx Steve

[@jordyv.703]

Just bought the design! It's an amazing clock and will look fantastic on my wall!

[@jmcarp0]

Good job! Nice experience based advancement going on there.

[@marcosmoran6896]

Very nice! I was so impressed that went ahead and bought the download. I believe I figured out which parts I need to print for a given run time duration. The accompanying pdf is pretty helpful; what is missing is a step-by-step orderly assembly.

[@stevesclocks]

Hi Marcos. The pdf and the assembly videos both show the steps in the proper order used to build the clock. Feel free to message me if you have any questions along the way.

[@marcosmoran6896]

@@stevesclocks - Yes, you’re right - I made that comment too early; I should have taken more time to read through the entire PDF. Please accept my most sincere apology. 😊

[@stevesclocks]

@@marcosmoran6896 No worries. Enjoy your clock build.

[@bencromwell4453]

bought your design for the big one. cant wait for the bearings to come in. btw this design is gorgeous!

[@zabihtoosky658]

Excellent work, thanks for posting

[@domsau2]

Congratulations! You should be proud!

[@glaysonmestre]

is possible make one sistm work for years?

[@nthupmemonkey]

Amazing !! Great and beautiful design...

[@frederickwood9116]

Nice clocks. Thanks for sharing. Can a second weight be added to the system so that when one string runs out the second weight will start to run as the drive? It would allow the clocks to run longer and with a less sensitive arrangement of gears. Some sort of transition system would be needed to transfer drive over from one cog to another as the length of one string runs out. Possibly based on an additional very low gear mechanism that counts days. In this way after 15 days perhaps, the drive is transferred over to the second weight and extending the time to an additional 15 days. Thanks again for sharing.

[@stevesclocks]

The trivial solution is to double the weight and reduce the diameter of the winding drum by 50%. This achieves the desired result without needing a complex mechanism to switch between multiple weights.

[@frederickwood9116]

@ I’m talking out of my hat here as I have not built let alone designed a clock. However I thought there was a constraint around loading of the mechanism and precision ease and also sensitivity to small environmental fluctuations like knocks. Thanks for the reply. One day I’ll get to this. Preferably with the kids too :-)

[@michaelbegley5487]

There is an interesting property of pendulum clocks that when two are out of sync and attached to the same object they will sync up over time. This is because they cause the object to vibrate a small amount. You can see the two clocks sync after they are both on the wall. At first, they are almost moving in the opposite direction of each other, but after he cuts to the next clip they are almost perfectly in sync.

[@stevesclocks]

Synchronized clock beats seems likely if both clocks have similar pendulums with one second beats. It probably takes a few hours for the pendulums to synchronize. I use shorter pendulum lengths that are appropriately sized for each clock. The small clock beats every 0.615 seconds and the large clock beats every 0.659 seconds. They might look to be slightly synchronized one minute, but would be completely out of sync a minute later.

[@fortun8diamond]

In order to get the full month of runtime, you need to hang it really high (such as along a stairwell or in a room with a really high ceiling). The weight drop height is 12 feet or 3.5 m. I did an estimation and the winding drum rotates about once every 17 hours. This is estimated 4-5 inches of drop per day. To get the full month runtime with around 7-8 feet of drop height, the weight needs to be heavier and the gear ratios must be changed again. With standard drop height, this clock needs only one winding per 2 weeks, which is good enough.

[@stevesclocks]

There are many ways to increase the runtime. This clock uses different gear ratios between the winding drum and the minute hand gear that rotates once per hour. All runtime options use a reasonable drop of about 52" or 1.3m. A pulley in the weight shell doubles the runtime. The longer runtime options need more drive weight, but the frame can only handle around 11lb or 5kg. The result is the 32 day runtime option running with 5kg is functional, but it has very little power reserve. The pendulum amplitude is barely strong enough to keep the clock running. The beat must be perfectly balanced for the clock to keep running. This clock runs better with runtimes in the 8-15 day range using drive weights of 3-4kg. I agree with what you say. 2 weeks, or even 1 week, is good enough and it makes the clock much more reliable.

[@kicksnj2]

First time for me needing a 3D printer. What would I need to buy without breaking the bank, but not sacrificing quality and have ease of use and setup. Never done this before. I am a old timer that has built all his clocks out of wood and now want to try this way. Just bought the files for the clock above.

[@stevesclocks]

Every printer in this list all3dp.com/1/best-3d-printer-reviews-top-3d-printers-home-3-d-printer-3d/ except the resin printers and the one in the "toy" category is capable of printing this clock. Pretty much any modern printer with a print area of at least 160mm by 160mm can print it. Some of my larger clocks require a print area of 210mm by 210mm. If you think you may want to try the others, go with at least a medium size printer, or any printer without "mini" in the name.

[@dennisgignac5103]

Fantastic project. Love it !!!

[@WarlikeWifee]

Just bought the STLs for this clock since my husband is an apprentice horologist at a clock shop. We might end up just using some leftover weights from some cuckoo clocks 😅 we're wanting to get it printed so he can build the clock with our boy

[@stevesclocks]

Any weight with a pulley will work. Most people don't have extra weight shells sitting around so they need to print them.

[@veluprabhakaran5980]

Congratulations very nice.. can i get blueprint for this clock?

[@stevesclocks]

www.myminifactory.com/object/3d-print-32-day-clock-easy-build-156759

[@richardmcpherson8322]

Mine has been running for a little over a week now. I love it!

[@stevesclocks]

Glad you like it. What runtime option did you use?

[@richardmcpherson8322]

I used the 10 day to start with. I don’t mind winding once a week. Lol

[@stevesclocks]

@@richardmcpherson8322 Good choice. It should run very well. I run mine in 32 day mode, but the one I am sending to my mom is 7.5 day mode since it has a really large swing.

[@a330turbinex7]

Superb!!! Thanks Steve!

[@andyjhawk007]

Looks superb Steve. Will give it a go soon.

[@JoeWayne84]

Excellent work Sir, God bless!

[@raphi72]

How much slower is the clock running in summer vs in winter? I guess the plastic pendulum is expanding resp shrinking.

[@stevesclocks]

If the clock is placed in a barn with huge temperature variations, then it will have large rate changes. If the clock is in a house with heating and air conditioning, then it will usually be accurate within a minute or two when you wind it. A small turn on the adjustment nut twice a year will bring it back to spec.

[@raphi72]

@@stevesclocks Sure. Well, I guess it doesn't have the precision of a clock like this one: kzbin.info/www/bejne/hF7anmZvlpullc0

[@stevesclocks]

@raphi72 Only two changes would be needed to remove almost all of the variation in a deadbeat escapement clock. The pendulum length needs to be constant. A carbon fiber rod or wood dowel are big improvements, although they are more effort than simple printed parts and barely needed in a temperature controlled house. The other improvement is to make the pendulum swing isochronous. This would require a suspension spring with curved cheeks. A weight driven clock has a constant drive power, so it usually keeps a stable pendulum swing. Spring driven clocks benefit a lot more from this change. One of the wooden gear clock web sites has a quote similar to: "If you want to know the time, look at your watch (or phone). If you want to see art, look at the clock you built". I am perfectly happy designing simple clocks that are accurate to about a minute per week.

[@richardshelby-dunn5826]

I'm looking for a clock that will tell time, date (month/day), moon phase, zodiac sign. I have found a lot of clocks that will do 1 or 2 parts but never one that does all of them at the same time. Is that even possible or is that too thick to be a functional wall clock? That is one beautiful clock though!

[@stevesclocks]

It is always possible to add new features. Some can be squeezed into the existing frame. Others force the clock to grow in size or thickness. Eventually it starts looking like the astro skeleton clock.

[@marksholcomb]

If you increased (doubled) the diameter of the winding drum, wouldn't you double the moment imposed on the gear 8 axis? Then half the effective weights? I may give it a try. Beautiful Design and the more I dig into it, appreciate the thought and time that went into every aspect of the design.! I have been looking for a doable 3d printed clock for a long time. Thanks for your hard work.

[@harry3life]

I'm not experienced with building clocks. What I would like to do is make one out of cut aluminium. Would it be possible to fabricate this from laser cut aluminium and it would still work?

[@stevesclocks]

I am not experienced with laser cut aluminum, so I don't really know. The gear spokes and rim should be strong enough in aluminum but would need to be made stronger if you wanted to laser cut them from plywood. You would need to cut the large gears and pinions separately and merge them together along with the shafts. The frame would need to be redesigned from components. Traditional clocks use brass gears with steel pinions. Brass on brass or steel on steel would wear out much faster. It might work much better if you could cut the pinions from steel or brass. You might have to experiment with the drive weight and runtime options. I have no idea if your clock would require more or less drive weight compared to 3D printing.

[@fluxwood]

Great design!

[@Noold]

So cool! Would love to see a video on your design process!

[@stevesclocks]

The process starts with a basic layout on scratch paper. Then a 2D CAD sketch from the front of the clock, moving gears around until things look good. Various configurations are tried, like changing a 40 tooth gear with an 8 tooth pinion to a 50 tooth gear with a 10 tooth pinion to balance a gear elsewhere in the clock. I move things around to make sure gears don't interfere with other arbors or support columns. I play around with various gear ratios so the pendulum length is proportional to the clock. theindex.nawcc.org/CalcPendulumLength.php is a great simple gear train calculator. The next step is a 2D side profile to determine gear and pinion thicknesses. The weight shell gears are thicker than the others so they can support the heavy static loads. Finally, I can start creating 3D models of the parts. Most of the effort consists of the 2D outlines of the gear teeth and spokes. I start with tooth profiles using Gearotic, but add my own touches to make the gears easier to print. Generating the 3D models is relatively easy once all the 2D work is finished. Each clock goes through a few iterations before it is completed. Sometimes parts collide and it wasn't noticed until the actual parts can be assembled. Other times, I realize a slightly different layout would look better. There is usually a box full of experimental parts for each finished clock. Preparing the design to release on MyMiniFactory is the final step. Each part needs to be saved as a separate STL file and quality checks get run on it. Generating the assembly manual also takes a while.

[@Noold]

@@stevesclocks Sounds wildly complicated! Fascinating though.

[@MikeyMystery45]

Awesome design from what I can tell. I want to try this design with what I have available to me.

[@stevesclocks]

This clock uses the fewest non-printed components of any of my clocks. And it should fit on most 3D printers with a 160x160mm print area.

[@garrettnelson2293]

What size print bed will I need for the largest parts?

[@stevesclocks]

The large clock needs a 210x210mm bed and the smaller version needs a 160x160mm bed.

[@MACISUS]

Amazing! I can’t wait to print this! I have spent lead from my air rifle range so free weight! I love this as I’ve always loved a grandfather clock and this is so much better option and takes less space! Thanks! Would you consider providing a one piece pendulum as I have an ender 5 plus upgraded to direct drive that lets me print up to 45cm tall prints! Can’t wait to start printing the gears and clock face! I’m even considering glow in the dark for the hands and time numbers. 👍🏻

[@anthonyvallot1752]

Steve, I see other folks have built your clock designs. How do I get one of your 32-day clock kits, Tony Vallot

[@stevesclocks]

You can buy the plans at www.myminifactory.com/users/StevePeterson and print your own clock.

[@GoogleUser-vv9wy]

This is amazing. Good job

[@SunilPatil-eg9cr]

I want to build this clock , how can i proceed

[@stevesclocks]

It is a 3D printed design intended for each builder to print their own version. Links to the STL files are in the description. Any modern 3D printer with a large enough print surface is capable.

[@Tb64]

Thinking about buying this print but I wanted to know if I can print this on my Delta printer, the build area is only 200mm diameter x 200mm tall. Is there parts larger than this?

[@stevesclocks]

The only part that doesn't fit a 200mm round plate is the lower segment of the pendulum shaft at 206mm long. It fits diagonally on a 160x160mm square plate. You could shorten the center straight portion or cut off the bottom tip that is purely cosmetic. All other parts will fit. Height is never an issue.

[@humbertonieto2868]

Hola soy de colombia tienes unos diseños increíble mi pregunta es si todo los componentes del reloj se imprimen en 3d o solo una parte y como puedo hacer para los diseños

[@stevesclocks]

All the visible parts are 3D printed. The few non-printed parts are listed in the description above. Google Translate: Todas las piezas visibles están impresas en 3D. Las pocas piezas no impresas se enumeran en la descripción anterior.

[@solarguy6043]

Super informative! What is the expected lifespan of such a clock (guessing is ok and likely required). The bearings pretty much eliminate wear on the high stress points. Any lubricant on the other moving pieces???

[@stevesclocks]

My first clock design has been running nonstop for around 3 years. This easy build clock is around 1 year so far. No wear is visible in any of them, so 5-10 years or more seems reasonable to expect.

[@solarguy6043]

@@stevesclocks Excellent! And of course, spare parts are not hard to come by.

[@rickibaron4816]

I would like to purchase your designs, but I need files that work on a CNC machine (I have 2 of these that I built myself). I do not own or even know where I would begin to try and work with a 3D printer. (I am from the older generation). I am currently make an all Acrylic/Lexan grandfather clock.

[@stevesclocks]

I believe that every clock should be designed to take advantage of the material and construction methods used to build it. This clock has been completely optimized for 3D printing. The gears have just enough material to be printed but would need to be strengthened to be suitable for CNC routing. The frame would also need to be redesigned. It could be ported with some rework. However, you may be much happier with a design from someone that specializes in wooden clock designs. I may design wooden gear clocks in the future, but for now all of my clocks are optimized for 3D printing.

[@rickibaron4816]

@@stevesclocks Could I possible get the number of teeth on each gear. I am very good at redesigning things to work. My current grandfather clock only runs of one day. I would like it to be able to work for a month on its own. Thank you for a very interesting designed clock. I may try and figure it out by closely inspecting your video. If you should ever decide to to try out any wooden Clock designs please let me know. Thank you.

[@stevesclocks]

@@rickibaron4816 You can download the assembly guide from www.stevesclocks.com/sp5 I recommend starting with the 8 day runtime option.

[@rickibaron4816]

@@stevesclocks Thank you very much. This is exactly what I needed. When I get it going I will let you know (It might take a week or two).

[@Mrelectric423]

I’m interested!!

[@isicohen2350]

Hello, I downloaded your interesting "Pendulum clock", and printed almost all parts. But... I cannot understand (in assembly, pages 17-18 ) what is , and why "music wire". Where can i find it? I ordered (and received) 1.5 mm stainless steel road. Can i use it instead this "music wire"? Thanks for your answer. Isidor.

[@stevesclocks]

Music wire is hardened steel wire. It is also called spring steel because of its stiffness. Stainless steel rod should also work. It is probably a bit softer than music wire, but the forces on the 1.5mm diameter wires are fairly small. It should be ok.

[@Ninjump]

Steve! Amazing work and design. I've just gotten back into 3D printing and started looking into clocks again. Do you think I could have the gears laser cut from aluminum so I could anodize them?

[@stevesclocks]

It should work. The gears are relatively straightforward. You would have to assemble pinions onto each gear at the proper locations. They are printed as a single unit. The frame also has a lot of print-in-place components.

[@Ninjump]

@@stevesclocks appreciate the reply! Off to go make my frame 👍🏾

[@johnnypearce5355]

how do i get the files to build this clock?

[@stevesclocks]

www.myminifactory.com/object/3d-print-32-day-clock-easy-build-156759

[@MJ-ee2rb]

When winding the clock or any clock does it keep time?

[@stevesclocks]

There are several ways for a weight driven clock to keep time when winding. Maintaining power or a Huygens endless chain are two common methods. They are typically only needed on precision regulators. This clock that has an accuracy of about a minute per week will lose another 20 seconds when it is wound. It doesn't really need the extra complication of providing power to the escapement when winding.

[@CDubs720]

This is Awesome!

[@mateotovar4548]

This is beautiful

[@alexglikman8629]

Looks nice. Now i print your previos clock. Please do this model to prusa mini

[@stevesclocks]

The 32 day version just released is perfectly sized for a Prusa Mini.

[@rynait]

interesting, wonder if anyone tried fabricating this clock with cnc router or cnc laser cutter machine?

[@stevesclocks]

It might need to be scaled up by around 50-100% so the gear teeth are large enough. The spokes and rims would also need to be widened to give them extra strength. The overall look of the clock would stay the same. Some day I will try to build one.

[@claustrox]

I am gonna buy your project. Is the quality of the ball bearings crucial? There are a lot of brands and prices over the internet can you suggest a proven good one?

[@stevesclocks]

I use generic bearings that cost around US$5-10 for 10. After cleaning out the factory grease, select two that appear to have the lowest friction for the pendulum support. Rubber seal bearings are common and the seals are easy to remove. High price bearings with tight tolerances might have slightly higher friction in the mode that they are being used in my clocks. I have not found a bad batch of generic bearings that would not work. The worst case is typically 1 or 2 bearings that are slightly gritty and they can still be used for the weight shell.

[@claustrox]

@@stevesclocks thank you print started :-)

[@alexpmK3]

Brilliant

[@kerimyigit5370]

Plastomechanic genius.

[@David-vm1rs]

Hello, i am printing your clock. How heavy should the Bob be?

[@stevesclocks]

The rate is determined by length, not weight. Some weight is needed to provide stability. I think the bob weighs 6 to 8 ounces when filled with pennies or washers.

[@David-vm1rs]

@@stevesclocks thanks for the fast answer. I Just fill IT With a couple of Euros und try it 👍

[@OspreyGozo]

Nice project! How do you set the time?

[@stevesclocks]

Just rotate the minute hand. There is a friction clutch that allows the hands to move without affecting power going to the escapement.

[@makewithmegma]

Amazing work😍❤🔥👌👍

[@isicohen2350]

Hi, how do you print the 2 colors

[@isicohen2350]

i downloaded all the files, but i don't understand how to print the "frame-dial-numbers" in two colors. for me it can be better to print only the numbers, and glue them to place.

[@isicohen2350]

OK, sorry for my question, i sliced the numbers in th slicer , and printed them. now I can glue them in place...Thanks

[@stevesclocks]

The color change can be done by adding a filament change in the slicer. PrusaSlicer makes it really easy. Move the slider in the preview panel and click on the "+" symbol, then re-slice. Cura seems to be a bit more effort. Here is one description www.instructables.com/How-to-Change-Filament-Mid-print-Using-Cura-48/

[@TheFluffyDuck]

How well does it keep time?

[@stevesclocks]

It usually stays within a minute or two per week. The accuracy is mostly dependent on pendulum length and the escapement type. The pendulum length is easily adjustable. It is influenced by temperature changes, but I find this clock to be just as accurate as my other clocks with carbon fiber pendulum shafts. The deadbeat escapement in this clock is very tolerant of changes in drive weight and slight changes in the beat. Changing the drive weight from 5 to 8 pounds will increase the pendulum amplitude slightly, but it seems to keep similar time. An anchor escapement would drift all over the place.

[@maelleamand]

I looooove it!!!

[@thelastofthelemmings6279]

Hello, Very excellent job! I'm not entirely certain of what I'm looking for. I am trying to setup a weight and pulley system to water my greenhouse (Just a slow drip, not much pressure) And I stumbled upon your brilliant creation. I was wondering if something like this is even possible? And figured you're probably a good person to ask. 👍

[@stevesclocks]

I have no idea how to adapt a mechanical clock into a greenhouse watering system. It seem like there should be ready made battery powered solutions that would be much more reliable.

[@voldy3565]

Nice clock! I wish it was symmetrical, though.

[@stevesclocks]

Thanks. The primary design constraint was to make the assembly as simple as possible. Moving the pendulum to the side helped with this goal. It also improved the efficiency so the runtime was increased. I have other more symmetrical clocks posted at MyMiniFactory. Assembly is slightly more complex, but still not too bad for a completely functional clock. Runtimes range from 4 to 8 days.

[@kimberleyander]

Very beautiful clock I would love to print this Please post the files on thingiverse if you can

[@stevesclocks]

All my designs are posted to www.myminifactory.com/users/StevePeterson

[@alexglikman8629]

I start print your amazing clock. What weight you recomend for 10 day clock ? 2-3-4 kg ?

[@stevesclocks]

2kg should be a good starting value. 2.5kg would give a stronger pendulum swing.

[@alexglikman8629]

@@stevesclocks Between 2 to 2.5 kg. Thank you. And what weight of pendulum bob you recomend ?

[@stevesclocks]

@@alexglikman8629 The pendulum Bob weight is not critical. Anything between 4 to 8 ounces is fine.

[@umnavegador9998]

Downlod do projeto?

[@stevesclocks]

www.myminifactory.com/object/3d-print-32-day-clock-easy-build-156759

[@AlfredoCronistaViajes]

¡Espectacular!

[@qreaigh]

Is the weight shell completely full of BBs

[@stevesclocks]

Yes, and sometimes it even needs an extension to add more. It all depends on the runtime option chosen and the overall gear train friction.

[@AhmadKhadiri-Haz]

No way it's the most quiet clock ever

[@stevesclocks]

It ticks like any mechanical clock with an escapement.

[@AhmadKhadiri-Haz]

@@stevesclocks every time I see an 3d mechanical clock it make sound sometime

[@ПашаПитецкий]

Клас 😊❤

[@vmdirt]

I've made 2 of these now although I've only been able to get them working reliably with the 7 day gearing. kzbin.info/www/bejne/fardgaeGr6ekZtk

[@stevesclocks]

I also converted my 32 day clock to a 7 day clock. 32 day mode turned out to be very aggressive. The pendulum amplitude was barely enough to keep the clock running and the beat always need to be reset after winding. 7 day mode is much more reliable. Another issue seems to be that some brands of PLA are stickier than others. And they might develop stickiness over time. I have started adding a light coat of grease to the pinions. It doesn't seem to affect PLA and the clocks run much better. The pendulum amplitude nearly doubled. So far, the clock has been running about 3 months with no side effects from the grease.

[@vmdirt]

@@stevesclocks I used a small file to dress the gear teeth which helped and also used a little white lithium grease on the gears. I am very happy running the 7 day gears. Thanks for the great design!

[@stevesclocks]

@@vmdirt My oldest clock has a section in the manual about cleaning up the elephant's foot. I used to manually clean up every gear with a file. Improvements in the slicer, or maybe just finding the right settings, and I no longer do any gear cleanup except drilling out the center holes.

[@vmdirt]

@@stevesclocks I'm pretty new at printing and unless I print the gears with a brim there is a 50% chance one side of the gear will curl up because it doesn't stick to the bed. It's a lot of extra work to clean up the gears after removing the brim but it's 100% effective. I would love to be able to print gears without having to do any cleanup. I have spent a lot of time trying to get the bed height right and I use hair spray with seems to help with most things but those gears still give me trouble, usually the thin ones.