/diy/ - Do It Yourself

Engineering technican here, I used to do these calculations all day long in school. Also not native English speaker.
Yes, that shit is fairly complicated, so I'm gonna simplify it a bit.

Generally, stiffness rises linearly with the width of a square tube. Double the width, get double the stiffness.

It also rises to the 3rd power with heigth. Make it double the heigth, get 2^3=8 times the stiffness.

It also rises linearly with the wall thickness. Double the wall thickness, get double the stiffness.

So it might seem like a good idea to make the tube as narrow and high as possible. But here, dent resistance will come and bite you in the ass. As long as the tube will not keep its structural integrity, theoretical stiffness will not help you. You need some minimum wall thickness to get the desired dent resistance. And 1mm thin metal is awful to weld, too.

So, 30 x 70 x 4 sounds perfectly alright for bending forces. For axial loads, I'd prefer a round tube, or at least a symmetrical square tube.

If you want to calculate stuff:

http://www.calculatoredge.com/civil%20engg%20calculator/beam.htm#tube

"Calculate Deflection for Hollow Rectangular Beams"

More sophisticated:

Calculate "Wy", 3rd last formula:

http://de.wikipedia.org/wiki/Widerstandsmoment

Or use this:

http://www.mobile-soft.at/widerstandsmoment-berechnung.html

Calculate bending momentum "Mb"

http://www.cnc-lehrgang.de/biegebeanspruchung/

Use drawings in the middle of the 3x2 chart.

Calculate stress:

sigma b = Mb / Wy

Multiply stress with safety factor, like 2.
Is that stress less than your steel is rated for? Then you're good. Around 100N/mm2 is a conservative rating, especially if welding is involved. If something expensive or dangerous is involved, make your safety factor 5-7.

For your axial load, buckling resistance is important, but I'm too lazy to explain that now.

http://en.wikipedia.org/wiki/Buckling

Tool and die maker here.
Bores:
Mark positions.
Center punch .
Pre-drill bores with the diameter being the chisel edge's width of the final drill bit, but do not go below 1.5mm. Drill out holes to final diameter. Use alcohol for lubrication. Deburr.

Cornered openings:
Drill many small holes in a row inside the outline (like 2mm). Open up these holes with larger drill, like 3mm, to get rid of the material between the holes. Remove center piece, clean the edges using key files.

For all steps: Practice on some scrap metal, in order not to fuck up your main project's material.

You did use a ring wrench, yes? These only exert 1/3rd of the force per edge for any given torque compared to an open-end wrench.

Therefore, open-end wrenches should only be used when there is no access for a ring wrench. But seriously, if you want to continue drilling, cut the head off flush first, with a tool I explain how to make here:

https://www.youtube.com/watch?v=z2YHmsiMmN0

Carbide drills made for stone can actually drill through hardened steel. You just have to grind the edge properly. Done this myself.

You are right. My Electronic shop's product description was wrong. HLG-185H-C1050A will work. Thanks.

@ color offset: Generally correct, but mostly only noticable at really low currents, like 5% rating.

Correction:
Use HLG-120H-C1050A
(output 74-148VDC)

The aforementioned HLG-185H-C1050A has an output of 143 - 286 V DC, which is too high.

CV is bullshit for LEDs. Unless your LEDs are tiny and you use a resistor.
The HLG-185H-C1050A mentioned by OP is absolutely the correct device to power his LEDs within specs.
The only thing I'd recommend, if someone plans a LED setup from scratch, is to run the LEDs at 50% power, to increase life expectancy and reduce heat. Like if you want 10W, buy 20W instead, and drive with half the rated current.

There we are. You need a derivate of DIN 906.
Variable cheating bolt to fake a reliable mechanical connection. Bolt head exchangable.

Naah...

No luck...

Let me look at my inventory.
No, that's not it...

Be careful, sometimes they are impregnated with poisonous stuff to keep them from rotting. This type you don't want in rooms whee you live.

Hi! You do not need a charge controller, as the charging current will be very low when you use solar panels. Unless they are huge. But if you stay below 0.1C charge ratio, charge termination is not critical.
Example: You use eneloop 2000mAh cells (all other cells are shit).
Max charge current without termination would be 200mA. All you need is a diode between the cells and the solar panels, so that no reverse current flows.
3 Eneloop cells in series will be enough.
So you should buy solar cells with at least 6V rating, and max 200mA rating.
Buy LEDs instead of mini glow bulbs. You need resistors to limit the current of each LED to around 15mA. For a white LED, 40 Ohm is pretty spot on. Google LED resistor calculator.

Source: I started RC in '93 and know what I'm doing.

But seriously, unless you are doing it just because, hook the light to the power pack you drive your boat with.

https://www.youtube.com/watch?v=FdthSQDH8qk

Like that, very similar system. The tapered end of the sealing ring, or olive, always points away from the nut.

Yeah sorry english is not my natural language. So let's replace "creating hydrogen" with "separating hydrogen from oxygen" to make everybody happy :-)

I know how to make one.

First, you need a shitload of cells. I slaughtered some LaptopPacks from ebay and have now around 40 good cells. But I am an RC Geek and have all kinds of computerized chargers to test cells. If you don't have this, buy new name-brand cells, ie avoid everything with "fire" in the name.
Second you need a charge circuit, precisely this:

http://www.ebay.com.au/itm/DC-DC-Converter-CCCV-Power-Module-LM2577-LM2596-Charger-Module-/121409411294?pt=LH_DefaultDomain_0&hash=item1c449098de

You can power it with a large variety of input voltages, meaning solar is also acceptable.

Third, you need a Converter to make the USB-typical 5V from it.

http://www.ebay.com.au/itm/2PCS-PFM-Control-DC-DC-USB-0-9V-5V-to-5V-dc-Boost-Step-up-Power-Supply-Modul-M29-/291156172199?pt=LH_DefaultDomain_0&hash=item43ca429da7

Problem is, they work down to 0.9V, so they might deep-discharge your cells.
To prevent this, put a protection circuit between the cell pack and the USB thing:

http://www.ebay.com.au/itm/5pcs-1S-Li-ion-Lithium-18650-Battery-Input-Ouput-Protection-Board-PCB-DW01-/181441124951?pt=LH_DefaultDomain_0&hash=item2a3ebbda57

One of these will do.

I decided it is not worth the hassle and bought this one from Fasttech:

http://lygte-info.dk/review/Review%20USB%20battery%20box%206x18650%20Qidian%20UK.html

Search theyr power bank secion, they have an identical thing there, but matte black, not shiny. Works very well.

Damn, how refreshing it would be if some of these "inventions" were actually built by people with some scientific background, or at least the ability to test things like efficiency, energy input/output etc in a scientific manner. Why are all these websites look like the unorganized sites of conspiracy theorists?

If you burn hydrogen, oxygen and hydrogen atoms get combined, and release energy. To separate them, one needs exactly the same amount of energy. It's like magnets. They pull together with the same force one needs to pull them apart. No amount of resonant frequencys or magic is going to change that. Chemical energy is chemical energy, and for every reaction which gives you energy, there is a reaction is opposite direction which needs the exact same amount of energy. These are the laws of nature, and they are called laws for a reason. Burning hydrogen can never release more energy than creating hydrogen. If you wind up a spring drive, you can't get more energy out than you put in.

Water injection does actually work:
http://en.wikipedia.org/wiki/Water_injection_%28engine%29

But this has nothing to do with actually burning water or hydrogen.

tldr:
You can't beat science or the laws of nature.

Well, I had once tried to get my crank arms of the bearing, but failed due to not having a pulling tool. I forgot to put the bolt back in. Over time, it got worked loose, and finally fell of. So my arm experienced practically the same as OP. I put it back on, tightened the bolt properly, and rode that bike for 15 years without problems. So, an arm becoming loose and falling off does not need neccessarily that it gets destroyed in that process. Before messing around otherwise, I would put a piece of scotch tape over the nut to prevent it from falling off and getting lost. After every ride, I would briefly check if the nut is still as tight as before the ride. Also, take the wrench with you in case you feel that it becomes loose during the ride. If it did not become loose during the next 50km or so, it is safe to assume that the crank arm did not get damaged somehow, and will sit firmly in it's position in the future. I would still remove the nut and assemble it again with loctite just to be sure.

Oh great, I crafted my response under the impression that the threaded pard was largely broken off with only a stub remaining. Ah well, better this way than having to tear everything apart.

Hey mate,

you got a cone bearing.

http://www.radforum.de/threads/174557-was-ist-das-fuer-ein-tretlager

The ring with the slots is a locking nut having the same thread as the frame. The barely visible thread on the cone, the part in the middle with the width flat, should indicate which direction to turn to get it loose. The part on the other side may turn in the opposite direction. You probably need to get the crank assembly off before being able to reach it. Take it to your dealer and demand a new one. But this part is very outdated and hardly used any more, and least not in Germany. Well, the new part screws in first on the chain side nice and tight. Then you screw in the cone on the other side, and adjust it until the bearing begins to move with a tiny bit of resistance. Then you lock it in place tightening the locknut, and reassemble the crankarms.

Google "fatigue failure" and have a look at the broken part, maybe it's the same.

No motor oil. It contains additives specifically needed for the situation in engines, where they are exposed to residues of gasoline, water, dirt.

Get manual transmission oil, preferrably mineral, not synthetik, as this is more agressive against paint. If you want to clean metallic parts, soak them in automatic transmission fluid. Afterwards, pour it over your enemies cars, it will fuck up the paint on the long run.

Capacitors have 3 qualities:
1) capacity
2) inner resistance
3) heat tolerance

There is no such thing as a capacitor specifically designed for audio applications. A capacitor gets charged and discharged, and it does not care if it sits in an amp, a power supply, or on a mainboard.
So, what you do is decide on the capacity, and if you need a "low ESR" (low inner resistance) capacitor, and buy it.
If it's rated 105°C and 5000h, even better. Unlike commercial server power supplies, which run in hot environments 24/7, a normal sound system gets used an hour or two a day max. You really should not worry, this is no difficult environment for caps.

Hi!
One way to get lasting computer stuff is to avoid consumer grade stuff. Get office grade stuff instead. If the IT guys in any company get complaints about a specific brand over and over again, they won't buy it any more, and the manufacturers know it.
Instead of a new Acer Laptop, buy a used Thinkpad. Mine is from 2007, and runs gread wit an SSD. HMMs (hardware maintenance manuals) are to be found for every model, and they describe every step down to the tinyest fucking screw that needs to be taken out to change any part that is in there. Spare parts can be found all over ebay. My screen was yellowish when I bought it used. €50 on ebay for a brand new one! Consumer laptops of my classmates in technical college were failing left and right, mine held up.
Next thing, printers. Avoid consumer grade ink jets like the plague. I have an office grade HP Laserjet 2300d.
€100 on ebay, refill toner €15 for two fucking 6000 page cartridges. Heavy like a tank, but also built like one. I feed it all the wrinkled leftover paper to print on the backside, hardly ever a paper jam. I am actually looking forward to every print job I have, because I just love machines that run like a swiss clockwork!
Screw those:
http://theoatmeal.com/comics/printers

Hi!

Wiha or Wera is very good already, but I buy nothing else but Walter Schröder.
http://www.walter-schroeder.de/schraubendreher/index.html
Try the X-Gripp series.
For the slotted ones, I use the VDE type ment for electrical installation work. Here, the shaft is slimmer, and after removing the insulation, it is easy to reach even deeply recessed screws that some devices have. Normal slotted screwdrivers can be quite broad in the area behind the tip.

Hi!

Wiha or Wera is very good, but I buy nothing else but Walter Schröder.
http://www.walter-schroeder.de/schraubendreher/index.html
Try the X-Gripp series.
For the slotted ones, I use the VDE type ment for electrical installation work. Here, the shaft is slimmer, and after removing the insulation, it is easy to reach even deeply recessed screws that some devices have. Normal slotted screwdrivers can be quite broad in the area behind the tip.