Inspired by the recent receipt of my hot end I decided that it really was time I strung up the spectra line and got the printer a little more finished. Well it's finally finished and up and running. 

After running some test prints there' few issues I need to address...

The first thing I need to 'fix' was my hobbed bolt. There were definite repeating patterns in the layers that seemed to correspond with the rotation of the extruder. I marked the extruder wheel and then watched the pattern repeat as the wheel rotated a full revolution. The pattern was little more then visible difference in the 'transparency' of the filament as it printed but I wanted to be sure to eliminate all issues, after all print quality is the name of the game.

My previous attempt at making the hobbed bolt was a very Heath Robinnson affair. I did not use a 3D printed guide but instead used a couple of 608 bearings in the vice along with my hand held battery drill. Whilst it did look quite good I will accept the possibility that it could be to blame for my poor print quality. So wanting to get things right this time around I decided to download a suitable 3D printed tool to assist.

Most of the hobbed bolt tools that I had seen simply consisted of a square block with two holes through it at right angles - one for the bolt and one for the tap. (the serrated groove is made with a 6mm tap). The blocks have recesses for bearings - 608s in the case of the bolt and a 624 for the tap. There are also some that use a dremel with a cut off wheel and a mini dividing head type of arrangement to cut gives around the circumference of the bolt. Neither really appealed to me. I had no 624 bearings and the slots made with the dremel seemed very course.

In the end I found a tool to use with a lathe. It looks a bit like a knurling tool except that it holds the bolt in two 608 bearings and the tap fits into the chuck. The best part was that I did not need the small bearing. I downloaded it and printed it out (good job I've got the makerbot here at the moment). If you want to download it yourself it is thing 3780


Most of the tutorials that I have seen and the info that I found recommend to use an M6 tap. I used this for the first bolt but then got to thinking about the diameter of the groove in relation to the filament size and decided to make an M4 bolt too. Making the hobbed bolt is simply a case of mounting the bolt holder in the tool post and the tap into the chuck. I then set the lathe to the slowest speed and slowly set the bolt to the tap using the cross slide. The bolt slowly spins as the tap acts like a worm drive; cutting the serrations as it turns. I let the bolt rotate a few times before finishing it. The important thing here is to make sure that you cut the groove at the right height I did this by putting the bolt into the extruder and marking it with a marker pen. You can then set the bolt correctly by shimming it with washers.

With the new bolt made I set everything up and tested the extruder. It ran and fed filament okay so I proceeded to do some more test prints. The prints were the same. 

Last night I noticed that the temperature did not appear to be accurate. I measured the hot end with my laser thermometer and it gave a wildly different result to what was being displayed on the readout. It was far too hot. The thermistor in the J-Type hotend is a 100k type but as there are several different types of 100k thermistor I wondered if I might have got the settings wrong. Reading up on RepRap and the Repetier software it is possible to create your own thermistor tables based on some simply readings.


Using the following spreadsheet - www.reprap.org/wiki/File:MeasuringThermi...a-ThermistorBeta.ods you measure the resistance at two separate temperatures to calculate the Beta values that you can then enter into the Repetier firmware which it will use to create a generic thermistor table.

I took the resistance readings at room temperature (nominally 100000 ohms @ 25 degress) and again at about the operating temperature (I measured 9K7 @ 198 degrees) this gave me a calculated Beta value of 2457. I entered this into the Repetier firmware and uploaded it.

I then set the printer on pre-heat and compared the temperature readout with the measured temperature from my temperature gun. The result was much better but not perfect. I then noted that the Rumba board has an onboard pullup resistor so I adjusted the settings to include the onboard 4K7 resistor and tried again. Better but still way off.

What I noted watching the temperature was that the onboard temperature readout rose until it got to about 8 or 90 degress but then slowed whilst the actual temperature kept rising. Eventually the onboard temperature appeared to level out at 180 degrees (the preset setting for PLA) which it appeared to maintain with no issues excpet that the actual measured temperature was somewhere nearer to 290 degrees.

I tried various configurations but was unable to get the thermistor to accurately control the hot end. In the end I decided to hook up the spare hot end that I have to try that instead.

Initially I reverted to the baseline settings for a 100k thermistor figuring that I could monitor that temperature with the laser thermometer and kill the power should it get too high. I chose the PLA pre-heat settings and watched the temperature rise. The temperature readout on the printer was about 10 degrees off of the measured temperature but it seemed to be fine. Looks like there was a duff thermistor in the old hot end.

I set up the same test that I had previously been having issues with and watched as it printed out with no issues.

From what I can see the thermistor was faulty allowing the hot end to get too hot. This had the effect of overheating the PLA which caused the initial layers to print very poorly. Once the hot PLA from the hot end chamber was consumed and PLA was flowing through the hot end the print stabilised a bit. With the thermistor working correctly and the hot end at the correct temperature the print was much better.

I managed to print the test print with no dramas and also printed a 25mm cube. There is still a bit of tweaking to do but the printer is finally working. Now that it prints I can concentrate on improving the quality which should hopefully just be a matter of tweaking