Sim Brake Thoughts

  • Thread starter pilmat
  • 231 comments
  • 36,820 views
kikie
But you'll lose resolution when you replicate a hydraulic system on a pot based system (when recalibrating the brake pedal in windows)?

So as Jef badger said on TIW, a hydraulic system is overrated!

Buts still a hydraulic system looks so cool, doesn't it? :cool:

What is it: a hydraulic system or an hydraulic system?
Click to expand...
Measuring small amounts of travel with a pot would definitely raise that issue. Ofcourse insane gear ratios in the linkage would help but at the same time would potentialy introduce large amounts of play and imprecision. My choice would be hall sensors. You can get magnets in various sizes, an RE magnet such as from a CD-ROM laser (5x5mm) could get the job done and if rigged properly, it would be a winner.

Yep, in my opinion hydraulic sim brakes are just not worth it, ofcourse if you can justify it to yourself, it's definitely a fun project.

Edit:
Well, the only way computers can read input signals is in an electronic(voltage/resistance) way. You cant get a computer to compute things like pressure without converting it into an electronic signal ;) Though a feature that would allow to alter the brake responsse curve would be nice and there's probably some programs or plugins that allow that, I don't know for sure as I'm not familiar with PC gaming but there probably is some possibilities :)
 
Last edited:
Introducing large amounts of play and imprecision = losing resolution when not using an insane gear ratio.

So for the moment, as far as I know, a(n) hydraulic brake pedal with pressure transducer is a way to overcome the resolution (linear pot with a non linear force, travel of the pedal ) problems when you use a pot based brake pedal which simulates a(n) hydraulic brake.
Or
A hydraulic brake with caliper an a load cell connected to the caliper instead of the slave cylinder and the transducer.
 
They're definitely legit ways of having a good brake, but if you are intending to use a loadcell, then I don't see the point in using hydraulics to transmit the force to it.

I didn't say that a realistic pot based brake can't be precise, a well designed and well built system would work maad. I'm really interested to see what mrbasher had came up with :)
 
What Jet Badger has stated is the basis for the brake I have been working on :) In essence...
 
Hydraulic isn´t that just 100 % about feel and 0,0 % about true performance. What is the benefit of delayed pedal return compared to the immediate springy type? If anything that should be a negative as you have no control over the delay?
 
A slave cylinder without the bushings or a floating caliper, will affect the stiffness and throw of the brake pedal. correct?

Does a hall sensor have to move in a curved manner for it to correclty measure the movement of the pedal or can a hall sensor also be used in a horizontal/vertical way to measure, let's say, a piece of rubber?

Is it also possible to replace the potmeter of the e.g. G27 or the T500 brake pedal with a hall sensor and still connect the pedals to the wheel?

How does a hall sensor actually work? You probably need a hall sensor and something that measure it's movement.

In essence, you need to construct a brake pedal is such a way that the pot or hall sensor doesn't measure the travel of the pedal but precisely measure the forces applied to, let's say a large piece of rubber (in case of the hall sensor)?
 
This thread is facinating ( to me ) lots of very interesting facts and idea, I have learned a lot with it and also looking at the other thread from Pilmat with the mod to the G27 pedal set.
I now understand that it might possible to emulate a load cell brake with a potentiometer based design.
What I am still puzzled with is, what is the advantage of using the potentiometer over the load cell? or is more like a dare, trying to prove a theory?
 
Cote Dazur
This thread is facinating ( to me ) lots of very interesting facts and idea, I have learned a lot with it and also looking at the other thread from Pilmat with the mod to the G27 pedal set.
I now understand that it might possible to emulate a load cell brake with a potentiometer based design.
What I am still puzzled with is, what is the advantage of using the potentiometer over the load cell? or is more like a dare, trying to prove a theory?
Click to expand...

One can accomplish the benefits of pressure based braking with a fixed pedal and a load cell. It is not possible to do that with a potentiometric system.

Pretend to you have load cell pedal with 1mm of pedal movement, that movement being essentially the flex any force would create on it (which it shouldn't, but for an argument, let's pretend). Let's also say that our load cell is a 100kg element connected to a 12bit controller board (preferably one of mine :)). That system gives us 4096 steps of resolution over 1mm of pedal travel. The output of that pedal (braking force) is controlled by how hard you press on the pedal with your foot.

Now let's say we have a potentiometric pedal with 1mm of pedal travel connected to a 12bit controller, and that pedal has a gorilla die spring in it to make sure it only travels that 1mm. By my calculations the pedal may rotate three degrees at best. That means the pot will use about 34 of 4096 steps of resolution. And to move the pedal 1mm you have to exert 100kg of force on it.

Would you rather have 41 steps of resolution per kg of force or 0.34 steps of resolution per kg. of force? Now let's make our load cell pedal arm fixed, the pedal travel being zero. You still get 4096 steps of resolution from 0 to 100 kg. of force. How much does the pot brake do? Zero. See the difference?

These debates have been exhausted ad nasuem long ago. It's well settled practice that pot braking in any form is woefully inferior to pressure braking. One measures position and the other force. The two are not the same and cannot be made to be. Us humans can remember muscle exertion and strain much better than limb or digit position. Using force in a sim racing brake is intuitive and natural - and just like what's in a real car.

Use a load cell or a hydraulic system if you have the money. Don't reinvent the wheel.
 
The only problem I have with a load cell or hydraulic system is that you can't connect your pedals to a wheel. If you make a load cell brake out of the T500 or the G27 brake pedal, you'll have to get a joystick controller and a LC amplifier which makes these pedals stand alone. Unless you use the Boding LC mod or the G27 LC mod.

If a game doesn't recognize more controllers, you don't have pedals.

From what I understand what Jet said, you can definitely make a pot based brake pedal that will have the same result as a load cell/hydraulic pedal.
If you can transfer horizontal movement into a circular movement, a pot based brake can be designed in such a way that it does the same thing as a load cell/hydraulic brake. If what Jet sais is true and I understand what he's saying.
 
derekspeare
One can accomplish the benefits of pressure based braking with a fixed pedal and a load cell. It is not possible to do that with a potentiometric system.

Pretend to you have load cell pedal with 1mm of pedal movement, that movement being essentially the flex any force would create on it (which it shouldn't, but for an argument, let's pretend). Let's also say that our load cell is a 100kg element connected to a 12bit controller board (preferably one of mine :)). That system gives us 4096 steps of resolution over 1mm of pedal travel. The output of that pedal (braking force) is controlled by how hard you press on the pedal with your foot.

Now let's say we have a potentiometric pedal with 1mm of pedal travel connected to a 12bit controller, and that pedal has a gorilla die spring in it to make sure it only travels that 1mm. By my calculations the pedal may rotate three degrees at best. That means the pot will use about 34 of 4096 steps of resolution. And to move the pedal 1mm you have to exert 100kg of force on it.

Would you rather have 41 steps of resolution per kg of force or 0.34 steps of resolution per kg. of force? Now let's make our load cell pedal arm fixed, the pedal travel being zero. You still get 4096 steps of resolution from 0 to 100 kg. of force. How much does the pot brake do? Zero. See the difference?

These debates have been exhausted ad nasuem long ago. It's well settled practice that pot braking in any form is woefully inferior to pressure braking. One measures position and the other force. The two are not the same and cannot be made to be. Us humans can remember muscle exertion and strain much better than limb or digit position. Using force in a sim racing brake is intuitive and natural - and just like what's in a real car.

Use a load cell or a hydraulic system if you have the money. Don't reinvent the wheel.
Click to expand...

Derek, I have a great respect for your products and your opinions. You contribute openly in these forums and you make great points that a lot of us don't always see.

But this one I will have to argue with you about. The basic assumption of your pot inferiority in this argument is that you are measuring the pedal's rotation directly. A simple four bar link would easily amplify the pot's rotation to achieve a much greater resolution. Is it as good a load cell or pressure transducer? Here you have a good argument, especially when you consider the convenient packaging of a load cell. But couldn't the milli-volt output from the pot also be amplified just like we are doing with the pressure transducer to provide a greater resolution?

I'm sure these points have all been discussed at length in the past, but many of us are new to the debate and the information is spread out around inter-web. The points in here were only to show that all the current braking technologies measure the energy stored in an elastically deformed piece of material and provide that feedback to the computer in a linear scale.

Two of the important things I have learned from this thread:
  1. The scale of measurement is of high importance. This way we can give the best modulation information to the computer. Throttles and brakes are not on-off switches :)
  2. The "feel" of the control is of higher importance than things like travel and stiffness. If we can get a pedal that can provide the highest degree of muscle memory feed back and not be so position based, we have a better pedal.

I would like to thank everybody who has contributed and is following this discussion. It is informative and collaborative without being confrontational, which is the spirit of my intention in starting it 👍
 
@derekspeare

Please enlight me on hall sensors. Using a small enough magnet they can be made to use all of the resolution even in a case of 1mm travel. Or atleast I think they can, you being more knowledgeable in this topic I'd like to hear what are your arguments about that.

In your example you proved that a pot can do the same as a loadcell. Its only drawback being the resolution.

Even if you put the loadcell directly under the pedal, it would still have a travel of atleast 1mm due to the flex of the loadcell. That is how they work, if you could stop the loadcell from bending in any way under force (the only known method to do this is witchcraft) it would output zero no matter how much force you apply to it.

A pot and a spring also measure force. As someone(pilmat?) said before, a loadcell is a very stiff spring with a position sensor.

At the end I think you missed a lot of discussion where we covered the position vs 'muscle tension' thing.


@kikie

That is a very popular misconception, you can make loadcell modded pedals connect directly to the wheel. The only thing that is needed is a loadcell amplifier. You can connect it to the original pedal wires exactly the same way you would to a joystick board.
 
Last edited:
Jet Badger
@derekspeare

Please enlight me on hall sensors. Using a small enough magnet they can be made to use all of the resolution even in a case of 1mm travel. Or atleast I think they can, you being more knowledgeable in this topic I'd like to hear what are your arguments about that.

In your example you proved that a pot can do the same as a loadcell. Its only drawback being the resolution.

Even if you put the loadcell directly under the pedal, it would still have a travel of atleast 1mm due to the flex of the loadcell. That is how they work, if you could stop the loadcell from bending in any way under force (the only known method to do this is witchcraft) it would output zero no matter how much force you apply to it.

A pot and a spring also measure force. As someone(pilmat?) said before, a loadcell is a very stiff spring with a position sensor.

At the end I think you missed a lot of discussion where we covered the position vs 'muscle tension' thing.
Click to expand...

A good load cell element doesn't flex - At all. Believing that a load cell is supposed to flex is misplaced.

Hall sensors need strong mags. I've written quite a bit about hall sensors. See the iRacing forums and search up the topics I've posted. The Allegro 1302 is the one to use. The datasheets for it are useful.
 
Last edited:
The load cell doesn't flex to the naked eye... But that is how they work...

When you have a moment, press the brake pedal on your CSP set and watch the load cell. press it nice and hard.
 
derekspeare
A good load cell element doesn't flex - At all. Believing that a load cell is supposed to flex is misplaced.

Hall sensors need strong mags. I've written quite a bit about hall sensors. See the iRacing forums and search up the topics I've posted. The Allegro 1302 is the one to use. The datasheets for it are useful.
Click to expand...

My brain does not compute. Even the 'good' loadcells (I believe the one that is used in CSTs is a good one) clearly have been designed to flex and I am pretty sure it does flex.
I have a full stash of bathroom scale loadcells and they are also clearly meant to flex, and they do.
If they don't flex, tell me how exactly do they measure The Force?

I believe rare earth magnets are strong enough? I hope they are because I can't get one off my fridge for half a year now.
 
Jet Badger
@derekspeare

Please enlight me on hall sensors. Using a small enough magnet they can be made to use all of the resolution even in a case of 1mm travel. Or atleast I think they can, you being more knowledgeable in this topic I'd like to hear what are your arguments about that.

In your example you proved that a pot can do the same as a loadcell. Its only drawback being the resolution.

Even if you put the loadcell directly under the pedal, it would still have a travel of atleast 1mm due to the flex of the loadcell. That is how they work, if you could stop the loadcell from bending in any way under force (the only known method to do this is witchcraft) it would output zero no matter how much force you apply to it.

A pot and a spring also measure force. As someone(pilmat?) said before, a loadcell is a very stiff spring with a position sensor.

At the end I think you missed a lot of discussion where we covered the position vs 'muscle tension' thing.


@kikie

That is a very popular misconception, you can make loadcell modded pedals connect directly to the wheel. The only thing that is needed is a loadcell amplifier. You can connect it to the original pedal wires exactly the same way you would to a joystick board.
Click to expand...
Really? That is the best news I've heard today. I didn't know that you could connect the LC amplifier directly to original wires. Thinking of it, it sounds logical.
Luckily there're smart people to teach stupid people like me! Thanks Jet!

So you could also connect a pressure tranducer to a LC amplifier and connect this directly to the original wires and the wheel?
 
My utilitarian principles override anything else. If each man finds and solution for his own needs and realizes a material benefit as a result, I applaud him.

Have fun with your efforts!

d

:gtpflag:
 
@kikie
I'm not sure how exactly a pressure transducer outputs its signal but if it's the same as a loadcell, it should work with a lc amp. At the end of the day how do people with hydro setups interface the transducer with the control board? You would need to connect it the same way to the place where the pedal pot was connected.
 
Kikie,

You can certainly make your own pedals with a load cell or transducer and connect them to your wheel directly. However, in the case of a wheel like a T500, the pedal voltage is 3.3v and this is generally speaking, below the excitation voltage necessary for most load cells. Thus, it may not provide the intended results when connected to a T500, not without providing 5v to the load cell.
 
mrbasher
Kikie,

You can certainly make your own pedals with a load cell or transducer and connect them to your wheel directly. However, in the case of a wheel like a T500, the pedal voltage is 3.3v and this is generally speaking, below the excitation voltage necessary for most load cells. Thus, it may not provide the intended results when connected to a T500, not without providing 5v to the load cell.
Click to expand...
In that case you will come up with a solution so I can buy your "solution". :P


Thanks for answering, PM of mine was not necessary. :guilty:
 
kikie
In that case you will come up with a solution so I can buy your "solution". :P


Thanks for answering, PM of mine was not necessary. :guilty:
Click to expand...

Hmm... I apologize but my plate is actually not just full, but overflowing already... Maybe some day though. :)
 
Weeks? LOL is right ;)

I can put whatever info you guys need up here and someone else can build it... It's not that hard. The only issue is that it would end up on a bread board or similar and sorta look thrown together. It would however work...
 
Ok but I have no electronics background!





Question for all to ponder on. :sly:

LC amp with variable gain control like Derek's or Leo's LC amp without this feature?
 
My guess is a voltage regulator and an external power supply. Amirite?

Kikie, that is just a matter of replacing a fixed resistor with a trimpot on the board if I understood your question right. On most amps the gain is fixed and for that a resistor is responsible, if you replace that resistor with a variable resistor(pot/trimpot), you would have variable gain control. Though it wouldnt be a direct replacement because a pcb that was meant for a resistor wont have the right pins. You would need a new pcb.


There is something I want to show

The little thing on the bracket is a magnet, as the pedal moves it flexes the bracket a bit, so using a hall sensor by that magnet should result in the cheapest pressure sensitive brake ever. There is still testing needed to do and probably it will not work as well as expected but its still something.

cam00158.jpg
 
Jet Badger
My guess is a voltage regulator and an external power supply. Amirite?
Click to expand...

That is certainly an option... I'd suggest the following setup:

5 volt cell phone charger and 3.3 volt low dropout voltage regulator. Then power the load cell via the 5 volts coming from the phone charger and power the load cell amp via the output of the 3.3 volt regulator. This will ensure that the output of the load cell amp does not exceed the 3.3v expected by the T500.

Another option is to use what is called a DC/DC step-up converter. These can be purchased specifically to step-up 3.3v to 5v. You would then do basically the reverse of what I said above. 3.3v goes into the load cell amp and the output of the step-up converter powers the load cell. You can also build these, but honestly, I wouldn't bother.

For a 3.3v regulator, I would suggest this MCP1700-3302E/TO (look up the part on digikey or something) You also need a cap with it or it won't work right. Reading the datasheet is a good idea ;) They have example circuits in there.

Jet Badger
Kikie, that is just a matter of replacing a fixed resistor with a trimpot on the board if I understood your question right. On most amps the gain is fixed and for that a resistor is responsible, if you replace that resistor with a variable resistor(pot/trimpot), you would have variable gain control. Though it wouldnt be a direct replacement because a pcb that was meant for a resistor wont have the right pins. You would need a new pcb.
Click to expand...

You can replace the resistor by doing the following:

On the pot, connect one of the outer legs to the middle leg, then run a wire from the middle leg and the other leg into where the resistor was and solder the leads where the resistor was. :)

Jet Badger
There is something I want to show

The little thing on the bracket is a magnet, as the pedal moves it flexes the bracket a bit, so using a hall sensor by that magnet should result in the cheapest pressure sensitive brake ever. There is still testing needed to do and probably it will not work as well as expected but its still something.

cam00158.jpg
Click to expand...

Badger did you build those? I was curious who did that...

On the subject of hall sensors. The Allegro 1302, if powered by 5v will output 2.5 volts when no magnetic field is present. (this is the quiescent output if looking at the datasheet) If you move a "north" field toward the device, the voltage will go down towards 0 volts. If you move a "south" field toward the device, the voltage will go up towards 5 volts. I hope that makes sense... This can be overcome with an opamp circuit. The device is meant to sense magnetism on one face, not both. (face meaning one side of the sensor)

These are cheap and fun to play with, and don't take any special circuitry. You should buy one! (if you haven't) $1.95 at jameco.com :) http://www.jameco.com/webapp/wcs/stores/servlet/Product_10001_10001_2135881_-1

Anyhow, I wanted to mention that because it's the first thing people notice and get stumped on. :)
 
Yep, those are mine :)

Thanks! I didn't know that. I'm still yet to make my rotary hall sensor work, the voltage drops to 0v when the sensor is grounded no matter what. It probably isn't a real short circuit as the psu doesn't shut off but I guess I'll need to buy more sensors :)

The sensors I can get locally are these: http://lemona.lt/index.php?page=item&i_id=55462 TO-92UA they're 2 bucks, I don't really know how good they are but I as I said above I couldn't get one to work :D
 
Oh...well... that explains it. :| Dammit. I have checked the datasheet in the past but obviously got lost between pretty illustrations :D

Oh, at the time I bought that sensor it was the only hall sensor of any kind that store had. I see now there's much more! Thanks, I'll definitely check that one out 👍

Anyway, during my little research I found out that people acheive the best results with hall effect when the magnet rotates infront of the sensor. My ghetto hall sensor looked like this:

cam00152g.jpg
 
Looks good to me!

Yes rotational is the simplest method to get a full scale result from the sensor. As long as it's analog output ;)

There's another example I saw recently that was well done using believe it or not, a Bic pen... It was rotational. I'll see if I can dig that one up. The design was a drop in pot replacement, sort of.

Edit: That was easy.. First result on google lol http://www.mycockpit.org/forums/i-o...-dead-simple-hall-effect-pot-replacement.html

Edit again.. That whole thing is a good read. (meaning I recommend reading it) ;)
 

Similar threads

I
Fanatec CSL V2 brake modification question....
Replies
3
Views
2K
Ian JB
I
sideswiper
True brake
Replies
0
Views
774
sideswiper
sideswiper
TheNormsk
Started to get knee pain on left leg with hard brake pedal
Replies
6
Views
4K
IfAndOr
IfAndOr
super_gt
Fanatec CSL Elite Pedals V2
Replies
4
Views
2K
Abyss1
Abyss1
T
Truebrake spring's
Replies
1
Views
2K
IfAndOr
IfAndOr

Latest Posts

Back