This joystick has a “gear”, so the potentiometer makes a full turn.
Interesting! Thanks for the clarification. Where did you find this circuit? It might be interesting to see the original.
Interesting. I have seen specifications of 120Hz and 180Hz for proportional valves from a couple manufacturers, but perhaps the moving mass / inductance in the valves your circuit controls, is higher, necessitating the lower frequency.
Do you mind telling what brand / type the valves are?
While this is not likely the way I would design a circuit, the RV2 resistance to ground is:
Full CCW = ground
Full CW = ground
Mid point = 2.5k to ground.
For the last 20 years or so, using HALL sensors has become more common for such industrial applications. Hall sensors have “inifinite” life, don’t need gears that can wear out or get misaligned) If safety is a real issue, such are for example with gas pedals in cars, big cranes, hydraulic stuff, etc, then even two hall sensors are used. I think there are special hall sensor chips for cars (gas pedals) which have two independent sensor circuits in the same IC.
And also of course, hall sensors are cheap 
Yes, we’ve discussed that, and why it’s like that. Did you read the thread?
This idea, with the potentiometer in center at zero speed and the trim potentiometers aside, comes from my own mind. I had to figure it out when i buld my own all electric mini loader, that i accelerate and steer by a joystick. The loader must stand still as i release the joystick. And produce 5 to 0 kohm (for the motor controller) in both directions (forvard and back). Plus that i had to build in a microswitch for sensing forward/back.
For the PWM-part of the joystick in this thread i used the main parts from a standard PWM-regulator, and modified it with my ideas. Only paperwork, that not looks so great (tried to upload it, but i am only allowed to post one image). Therefore i want to explore Kicad.
I understand that people question my mid pot solution, with both ”ends” to ground, as my electronic-guru-friend shouted the first second he saw it ”Are you really stupid? That wont work!”. Silence five seconds, then, with a nicer voice ”Oh yes, it does……”
2 Likes
I dont know the brand right now, but i will look for it when i visit he customer next time.
You are right, but this skylift (e.g. boom lift) is from about 1980 so i choosed to work this way. I guess i need a microcontroller to work with Hall:s?
You are right about safety. Especially on a skylift. My PWM-unit generates no output if the lever is released. If that fails, the output is stopped by the microswitches, that are open when the lever is released. If those both fails, at the same time, there is a ”Dead mans grip” (a foot pedal that the operator must press) that stops the engine if released. Maybe there should be a fourth safety system, if all three fails simultainiusly, but i dont know what that could be.
1 Like
Heart rate sensor + bluetooth receiver? No pulse, and the machine stops.
Nice - it’s a good design which I probably wouldn’t have thought of (although I’m only a hobbyist). Just one point: the resistance goes from 0 to 2.5k, not 5k. (I made the same mistake myself.)
Where do you stop? A fifth safety system? A sixth?
It’s a rhetorical question - there is no right answer for everyone. At this point in your design you would normally employ a Reliability Engineer (I was one) who will model the various failure modes and calculate the probability of all three protection devices failing simultaneously. You would then have to decide whether that is an acceptable risk or not. If you want to reduce the risk further, the Reliability Engineer would advise you on the best way to reduce that risk. It might involve another layer of protection, or changes to the existing layers.
Reliability Engineering is a fascinating job. One lesson you should take is that it is not possible to reduce the risk to zero. Another lesson is that Reliability Engineers are brilliant at spotting “gotchas” that might undermine your efforts to make something safe (or reliable).
1 Like
This FAQ explains trust levels for posting images etc.
To self promote yourself to “Basic level”, all you need to do is open and read one more different thread for just a couple of minutes and then “New User” restrictions will be lifted.
A very nice looking machine, you’ve built 
2 Likes
Regarding, “I renamed the references from R1, C1 to something more understandable, like 4k7, 0.1uF and so on, which might have been a mistake……”
Yes that is a mistake. It goes against assumptions all experience gives us.
Let me anticipate another. It is a mistake to put any component value into the PCB silk screen or copper. Component values will change.
Thank you, i have now completely rebuilt my project. Started from skratch, leaving the references as they are. All worked very well.
I encourage you to read this design review guide.
While it is written by people using a similar professional software, Altium, the requirements it identifies are VERY much the same because they are requirements for a good schematic and PCB which is largely not specific to the software in which you do the design.
I am going to guess this will introduce you to a lot of requirements which are all worth while learning.
Hope this helps.
Regarding the original question, " trying to get the two Op-amps, and its power pins, into the same LM393 footprint. But all the time Kicad places two LM393, instead of one. "
In this very voluminous thread, I did not find any reference instructions or a tutorial on how to use the library part for the LM393 with the two op-amps and power pins.
While I too have fought this in the past it is some time ago and I realized I do not understand it enough to explain it or use the feature.
Better yet a tutorial with an example project (files we can upload) which uses such a part or parts.
Anyone?
Indeed. For that reason I have mostly tuned out of this thread
For me this works:
- In the Schematic Editor press a to add a symbol.
- Type: lm393 and press enter to accept it.
- Do three left clicks with your mouse to place all three units of the LM393 symbol
And then, the result is:
The problem here is that OP has changed some settings in KiCad. RefDes is hidden for example. There are also a lot of other issues with the first screenshots (misused labels, shorted capacitor) and those detract from the question. This is also not the place for a generic tutorial. For those it’s better to find a tutorial, or just read the manual on basic KiCad use. What does work well for a forum is for things not in the manuals, or still unclear after reading the manuals, but you do have to be more precise in asking the question.
The Symbol Library Browser is opened.
The Symbol required is found. ( in this case, just for something different, a LM339 )
A description of the symbol is found at the bottom of the sheet: Cyan arrow.
The box up top (Red arrow) will show “Unit A” if it is a multi-unit symbol. If the symbol is a single unit, the box will be empty of text.
If the symbol is multi-unit; clicking on the triangle (Magenta arrow) will display a list of the total number of units available for the symbol. (in this case, 5 units).
Five clicks will result in each of the five units being placed on the schematic.
That’s about all there is to this function. 
