Segway build.

[MattF]

Correct, in that both springs will be at rest when the shaft is in its current position. However, the simple fact that they would be secured on either side of the shaft rather than in the single location they are now would cause what to happen when you rotate the shaft...? The pin, (or collar), would rotate with the shaft, pulling both springs in the process. One would travel under and one over. You could also use them as a centring mechanism, if you make the chassis fixing points adjustable, to counter for any spring tension change or mismatch and the like.

 

[Wendelspanswick]

I think Matt means:

l-w-O-w-l

With l being the chassis ends and O is the shaft.

 

[MattF]

Yup, that's it Wendel. Glad to see I'm not the only one who doesn't do graphics/CAD software btw.

 

[ale36]

quick paint drawing to show it better

 

[gerrymoore]

Thanks for the diagram...I have tried it this way but I need to get different springs for it to work. In the meantime, I have re-jigged the original and it seems fine for now....don't actaully need much movement left and right.
Moment of truth next Monday when I hope to have the completed coding and be able to try it.....will have the video camera on for it's maiden voyage!!

 

[ale36]

Any progress on this?

 

[gerrymoore]

Hi..no good news, unfortunately. We had a go last week...it's difficult as my 'programmer' lives a long way away so we don't get much time to 'play'. We have two problems...the chassis is flexing when stood on and this is causing the chains to 'bind' - solution is to firm up that part of the machine.

The firmware is 'under development' so we still have to get it sorted....we went 'backwards' on the day and ended up worse than when we started. The steering is proving fun to code and at one time the machine was spinning like a top...luckily with no-one on it!!

Be a few weeks now before we can have another go...meanwhile I hope to post some video of our efforts to youtube.

Thanks for your interest..I'll post any new developments.

 

[DrNik]

with your steering wouldnt it be easier to have it like this

 

[gerrymoore]

Hi Nik...the steering 'problem' is with the code rather than the mechanicals..although your idea looks good.
We only need a very small amount of movement as it just moves a little pot. The differences in resistance are sent via the AD converter to the motor drives and turn one faster than the other...or you can turn on the spot by having one forward and one reversed...as we found out when the batteries flew out under centrifugal force!!

It turned out to be very sensitive to get right but didn't help that it seemed OK when we were testing 'off ground' but then when we tried for real it went a bit 'bananas' (not too technical am I!!!)

It's gonna be a steep learning curve...it took the original Segway guy many years and millions of dollars to get right...I'm reading his story at the moment...a really clever guy!!

Patience, patience!!!

 

[ale36]

Hi Nik...the steering 'problem' is with the code rather than the mechanicals..although your idea looks good.
We only need a very small amount of movement as it just moves a little pot. The differences in resistance are sent via the AD converter to the motor drives and turn one faster than the other...or you can turn on the spot by having one forward and one reversed...as we found out when the batteries flew out under centrifugal force!!

It turned out to be very sensitive to get right but didn't help that it seemed OK when we were testing 'off ground' but then when we tried for real it went a bit 'bananas' (not too technical am I!!!)

It's gonna be a steep learning curve...it took the original Segway guy many years and millions of dollars to get right...I'm reading his story at the moment...a really clever guy!!

Patience, patience!!!

Didn't he die falling from a Cliff testing on out?

 

[gerrymoore]

No...I think that guy had bought the UK or European rights...he was out on an all-terrain version, ran over a tree root which sent the Segway off the cliff and he was killed.
I'm only just reading about the origins of the Segway. They projected sales of 22 million or more in the first 10 years...not sure where they all are?? Mind you...I can't even build one!!!

 

[ovaltineo]

I'm new to this forum coz I don't own any welder . But I do know a bit about building Segways, so I'd like to jump in. My Segway clone is very similar to Gerrymoore's. The main difference is that I'm using a plywood instead of a CNC-cut metal. The only tools I have are an electric cordless drill and a hand saw, you see. So it really is a poor man's Segway. My first version uses my own custom-built motor controllers, but I am now swapping this with the low-cost BTN7960 controllers. For some reason, I can only attach one image at a time, so I'll have to split my post.

 

[ovaltineo]

Here it is with the lid on.

 

[ovaltineo]

Here's the new electronics with the BTN7960 controllers.

 

[ovaltineo]

And here is how I implemented the steering sensor - it is an MPU600 attached to the rotating steering bolt. The steering is not exactly heavy duty but it works !

 

[DrNik]

nice work ovaltino, quite envious i want one,

Gerry get the batteries charged on the quattro (bbc: life on mars pun)

 

[gerrymoore]

Hi Ovaltino...and well done..you've beaten me to a working segway!! I think plywood may have been easier for me!!

You are using the same motors, wheels, motor controllers and I presume Arduino Mega?

What giro/accelerometer are you using..we have the MPU6050. The steering is a 5k linear pot and the central position is sent to the a/d converter to give left and right...in theory but not working too well as yet

Unfortuantely we can't seem to get the overall coding to work...the machine either 'bogged down' or went manic!! I'm waiting for my 'programmer' friend to come up with a better solution.....do you have anything on the coding side you can suggest to help?

 

[Wendelspanswick]

We want on a Segway Safari through Mount Edgecombe Park near Plymouth on Sunday. Nine of us on off-road Segways hurtling round the estate (well 12mph seemed like hurtling!). Really enjoyable day out but I would hate to try and write the software for one, its seems incredibly complex, they can sense whether you are ascending or descending an incline and adjust the sensitivity accordingly and the governor for the speed control angles the Segway backwards as you hit maximum speed to stop you leaning any further forward.

 

[Llama]

Hi Ovaltino...and well done..you've beaten me to a working segway!! I think plywood may have been easier for me!!

You are using the same motors, wheels, motor controllers and I presume Arduino Mega?

What giro/accelerometer are you using..we have the MPU6050. The steering is a 5k linear pot and the central position is sent to the a/d converter to give left and right...in theory but not working too well as yet

Unfortuantely we can't seem to get the overall coding to work...the machine either 'bogged down' or went manic!! I'm waiting for my 'programmer' friend to come up with a better solution.....do you have anything on the coding side you can suggest to help?

Assuming all the code is logical, how well is your processor keeping up with the tasks? It's a reasonably complex task so if you are not careful with your timing you'll end up with all sorts of weird things happening if calculations are taking too long, or if you are interrupting before it's finished.

 

[gerrymoore]

Hi Llama...don't think that's the problem..the Arduino Mega runs at 16mhz so has plenty of oommph to run our code.....I think any interupts are at around 100hz.

The code we are using is based on one successfully used before with an analogue MPU.....these are now unobtainable as everything has gone digital.....so it's a case of getting it to work with the 6050 libraries.

Unfortunately, the coding is beyond my very limited knowledge...I used to be good with BASIC!!!....so I'm waiting on my colleague to do a recode but we live a long way from each other which makes it more difficult.