Sim Brake Thoughts

[mrbasher]

Thanks for the part number, it's ordered

As for the pedal, it comes with the master (3/4" bore) and slave cylinders (7/8" bore) and I have it in stock! I'll start onto the fitting to a G27 pedal base in the morning! There's even a skate shop close to work to fetch elastomers from (unless you have other good sources ).

The bushings used by most on the slave cylinder are urethane sway bar bushings. http://energysuspension.com has quite a few different types.

 

[mrbasher]

Again and I'm starting to feel like I need to put this in my signature lol...

current pot based pedal + something to make it stiffer (is not equal to) current load cell offerings

 

[derekspeare]

Again and I'm starting to feel like I need to put this in my signature lol...

current pot based pedal + something to make it stiffer (is not equal to) current load cell offerings

This is part of the method Emery uses - he was the first to devise it, and it's become the standard way hydraulic sim brakes are built up. Emery was the one to take the leap from what some were trying to accomplish with a load cell mashed between a brake caliper to using the slave-cum-bushing method. It's brilliant. Everyone has been copying his efforts since.

 

[mrbasher]

Emery is actually the one that told me what the "squishy part" being used on the slave cylinder was.

 

[Cote Dazur]

But one point of this thread was, that in theory you can achieve the same effect with a pot-based construction. .

Agree, but I would add that the point of this thread, the real goal, is that it can be done with more simplicity ( better reliability ) and for less money. As acheiving the same result as a load cell with a pot for the same money ( or more ) and same reliability is not advancing anything. It would just prove that the theory works.

 

[kikie]

*snip*

current pot based pedal + something to make it stiffer (is not equal to) current load cell offerings

If both systems cost the same, what would you prefer?

Pot based brake with a realistic and progressive feel?

or

An hydraulic brake with a realistic and progressive feel?

 

[mrbasher]

My design is not using a pot. In fact using a pot makes it overly complex, even more so than the load cell...

The costs involved often come down to, how simple can the design be made. The simpler it is, the less machining and molding is necessary, the less parts there are, etc. My personal hope is not to prove anything to anyone, but to offer what so many people would like at a reasonable cost, whether through me or by some other means.

It just seems plausible based upon the data and experiences I currently have. Maybe it will come to be, maybe it will not.

The beauty of stuff like this is that people still have options. Maybe it's better, maybe it's not, but it can't hurt to have another option.

 

[Fanatec...]

We need to make a difference between the linearity of signals and the linearity of force applied.
If you just make the brake stiffer by adding a foam then you get cloaser to the ideal thing but the difference is that the signals are still linear.

You want to have the last 50% to 60% of signals where you have the most accuracy. This means in the area where the force is highest and you work with muscle tension.

Our load cell brakes work exactly that way. The bigest portion of travel is in the area where the signals are still very low and it get stiff before you come to the critical area.

So both, the force and the signals are not linear and that makes the difference.

Thomas

 

[mrbasher]

Thomas,

I agree. Also, thank you for stating who is writing.

 

[kikie]

Too bad I can't order Derek's board.

If I could, I definitely would!

 

[derekspeare]

Too bad I can't order Derek's board.

If I could, I definitely would!

Everyone can order from my site

The core issue is threshold braking - it's that state when the car is achieving the most braking force in the shortest period of time (without locking up the tyres). Each road racer wants to brake a deeply as possible into the turn and apply the greatest controllable pressure on the pedal at that instant prior to wheel lock.

No one has demonstrated a reliable way to do this with pot braking. I'm surprised it's even being debated. One can easily modulate pedal pressure with a pressure brake. Few can do it successfully with a pot brake.

d

 

[kikie]

Everyone can order from my site
*snip*
d

PM sent why I can't.

 

[mrbasher]

Everyone can order from my site

The core issue is threshold braking - it's that state when the car is achieving the most braking force in the shortest period of time (without locking up the tyres). Each road racer wants to brake a deeply as possible into the turn and apply the greatest controllable pressure on the pedal at that instant prior to wheel lock.

No one has demonstrated a reliable way to do this with pot braking. I'm surprised it's even being debated. One can easily modulate pedal pressure with a pressure brake. Few can do it successfully with a pot brake.

d

Derek,

This isn't being debated... I believe most of us already subscribe to the same thought process.

People need to get out of their head the concept of conventional pot brake pedals. That is not what we are speaking of here.

 

[RacerXX]

Everyone can order from my site

The core issue is threshold braking - it's that state when the car is achieving the most braking force in the shortest period of time (without locking up the tyres). Each road racer wants to brake a deeply as possible into the turn and apply the greatest controllable pressure on the pedal at that instant prior to wheel lock.

No one has demonstrated a reliable way to do this with pot braking. I'm surprised it's even being debated. One can easily modulate pedal pressure with a pressure brake. Few can do it successfully with a pot brake.

d

I think it's more than that…speaking from the Engineer / Racer perspective.

IRL (in real life) we trail brake into a turn trying to ride the edge of the friction circle and get the most important part of the turn (first &#8531 correct considering what preceded and follows the present turn you are facing. That circle represents the full grip the tires can produce when faced with any combination of lateral and longitudinal forces. Tires can pretty much grip about the same in any direction hence the friction circle description. It's more of an oval though, biased towards the direction in which the tire patch is longest.

IRL when you trail brake into a turn you are feeling the g-forces, YAW, steering angle, rate of change of steering position, resistance to steering, and pedal pressures along with what you see. In simulation land we are missing a few of those. If you trail brake hard IRL you can feel the car yaw besides the g-forces. Even the motion devices do little with yaw. So when you are fully braking hard and optimally into a turn it is not just lockup you are concerned with, it's the car rate of change in heading too. In sim land without feeling yaw we need to go by the car's heading aided by what is felt in the wheel and pedals.

Some of the above noted is why when you set brake bias IRL you don't just do straight line stopping tests. You also do "J Turns" to help evaluate stability when cornering and braking. SAE even has a standard procedure for J Turns.

 

[derekspeare]

It's the ability to produce a brake pedal input sufficient to bring one's virtual car around the corner in the quickest and most efficient manner - simple. And as we all agree, the difference between the racing world and the sim racing world is dramatic and substantial.

d

 

[kikie]

I guess if I use a brake caliper, I don't need a slave cylinder?

 

[mrbasher]

I guess if I use a brake caliper, I don't need a slave cylinder?

You need something to "squish" to allow pedal movement. That is why you put the bushings on the slave cylinder.

 

[kikie]

I have finally, after a few "almost" sleepless night, made up my mind. I'm not going to buy a new wheel and a new graphics card any time soon. And if/when I do buy a new wheel, I know how I'm going to make myself a nice DIY hydraulic system.

I'm going to wait for the tech demo of Assetto Corsa to be released and then I'll decide if I start building my own DIY hydraulic pedals. Probable, not sure yet, I'll be the T500RS of which I'm going to try to upgrade the pedals to an hydraulic system.

 

[pilmat]

OK, here we go!

Hydraulic Brake Pedal Components
​

And before I get accused of being a hack NO, I did not do that to those poor G27 pedals!!!!! I'm actually quite upset by it as I now have to craft a heel rest...

The components so far:

• G27 pedals, completely stock apart from the lobotomy.
• CNC 340SBU clutch kit (10.25" pedal @ 5.2:1 ratio, 3/4" bore master cylinder and 7/8" bore pull type slave cylinder).
• mV pressure transducer.

As this is drifting away from the original topic, I'll start a new thread with this build and come back here with results

Here is the Hydraulic Brake Build thread.

 

[JogoAsobi]

As this is drifting away from the original topic, I'll start a new thread with this build and come back here with results

Please keep us posted, its not OT, its just an extension of your Sim Brake thoughts being realized!

 

[kikie]

Reading through the entire thread can be confusing.

The OP speaks of the signal/input (voltage) read by the game (sim) (reads a position) is the same from all three systems :

* potmeter
* load cell
* hysraulic

So, which ever system you use, the game will always see a "voltage" input signal.

So, why build a load cell or an hydraulic system when the game (sim) gets the same signal from a pot? Besides the realistic brake feeling and the advantage you get from muscle memory, there is no differenc between a pot based brake and a load cell/hydraulic brake. Correct?

My question is;

if you can stiffen the pot based brake pedal with whatever you use (rubber, foam, skate board truck bushings, hydraulic system without the pressure transducer) and you recalibrate the brake pedal in windows (gamecontrol) or in the game profiler or in the sim itself, aren't going to get the same result as a signal coming from a Load cell/hydraulic brake pedal?
Stiffen the pot based brake pedal will reduce the travel of the pedal, thus reducing the travel of the potmeter. Right?
But if you recalibrate the brake pedal, doesn't the game (sim) sees the smaller amount of travel from the pot as full braking when you push the pedal to the max?

 

[pilmat]

Please keep us posted, its not OT, its just an extension of your Sim Brake thoughts being realized!

Don't worry, this thread will not be forgotten I don't want to clutter this great discussion with a build, but as soon as I get into the testing of even the brake feel, I'll be posting here.

Reading through the entire thread can be confusing.

The OP speaks of the signal/input (voltage) read by the game (sim) (reads a position) is the same from all three systems :

* potmeter
* load cell
* hysraulic

So, which ever system you use, the game will always see a "voltage" input signal.

So, why build a load cell or an hydraulic system when the game (sim) gets the same signal from a pot? Besides the realistic brake feeling and the advantage you get from muscle memory, there is no differenc between a pot based brake and a load cell/hydraulic brake. Correct?

My question is;

if you can stiffen the pot based brake pedal with whatever you use (rubber, foam, skate board truck bushings, hydraulic system without the pressure transducer) and you recalibrate the brake pedal in windows (gamecontrol) or in the game profiler or in the sim itself, aren't going to get the same result as a signal coming from a Load cell/hydraulic brake pedal?
Stiffen the pot based brake pedal will reduce the travel of the pedal, thus reducing the travel of the potmeter. Right?
But if you recalibrate the brake pedal, doesn't the game (sim) sees the smaller amount of travel from the pot as full braking when you push the pedal to the max?

Well said, thank you! The thing extra to this I have come to see is that when you make a pot based system "short enough and stiff enough" in travel to get the desired feel, you will end up losing resolution. That is unless the pot input is non-linea,r as in a cam based pot installation.

 

[mrbasher]

Well said, thank you! The thing extra to this I have come to see is that when you make a pot based system "short enough and stiff enough" in travel to get the desired feel, you will end up losing resolution. That is unless the pot input is non-linea,r as in a cam based pot installation.

This is very simple to overcome and is done in a similar manner to that employed when using a load cell or millivolt pressure sensor. Edit: and, it requires no extra stuff like non-linear pots, gearing, etc.

However, nothing currently exists that allows this. If you feel like doing some digging, I explained this situation and how to "fix" it over at InsideSimRacing

 

[kikie]

*snip*

Well said, thank you! The thing extra to this I have come to see is that when you make a pot based system "short enough and stiff enough" in travel to get the desired feel, you will end up losing resolution. That is unless the pot input is non-linea,r as in a cam based pot installation.

Losing resolution, what does that actually mean for controlling the brakes of a car ingame? Less detailed control?
What does the "cam" part of a cam based pot installation mean?

This is very simple to overcome and is done in a similar manner to that employed when using a load cell or millivolt pressure sensor. Edit: and, it requires no extra stuff like non-linear pots, gearing, etc.

However, nothing currently exists that allows this. If you feel like doing some digging, I explained this situation and how to "fix" it over at InsideSimRacing

I suppose you have the same username on InsideSimRacing?
I'm going to dig until the every page on the forum is "disorganised".

EDIT: I just rememberd something the ECCI guy told me once in our email conversation. People don't "feel" the difference between a 10-bit or a 12- bit resolution (I was asking for information about their potmeters). It's all about modulating the brake pedal properly.

I'm starting to understand!

 

[Zathra5_]

So, why build a load cell or an hydraulic system when the game (sim) gets the same signal from a pot? Besides the realistic brake feeling and the advantage you get from muscle memory, there is no differenc between a pot based brake and a load cell/hydraulic brake. Correct?

The load cells that I know of are very stiff and would contribute almost nothing to brake feel. The brake feel is from the mechanical system that transmits the forces to the load cell (or pot) from your foot. The advantage you get from "muscle memory" is also due to the mechanical system and not the measurement technique.

In a system without viscoelastic elements, if you know the forces applied, you can calculate the distance traveled and if you know the distance travelled you can calculate the forces applied. Given an accurate enough pot or accurate enough load cell, there would be no difference. However, in a system with increasing stiffness, at full braking small changes in distance will correspond to larger changes in force (as shown in the picture):

Therefore your pot would have to be relatively more accurate and have higher resolution. This might be why people notice a difference with bodner cables and why load cells are preferred over pots.

 

[Jet Badger]

This is a very interesting thread right there. Even by sticking a piece of rubber under your pot-based pedal you are effectively creating a loadcell, a contraption that translates applied force to distance(position), just in a traditional loadcell the elastic element is steel. If your piece of rubber has linear force vs distance properties then you have a precise force sensor under your foot. All force/pressure sensors work in the same way - they translate force into deformation of material and then measure that deformation. In retrospect a pot is a loadcell without the bendy part. Now the reason why pots are bashed for 'being not realistic' in brake applications is that they are usually attached to the pedal arm and measure its travel which is not properly related to the force needed to press it. For a brake to be realistic its first 85 or so % of travel should result in something like 10% of braking, and the latter 15% of travel should pack the rest 90%. That curve would look like something like this __/ , so not linear at all and judging the braking force from the pedal travel with a linear pot would obviously be wrong. To do that precisely we need to measure the force that passes through the system, that can be seen well in CST pedals - the pedal rod moves indepently from the force sensor and is linked to it only through a spring and a rubber damper, Id like to point out that that spring with damper is only to give pedal its feel and to return it into place. If we'd replace the loadcell with a rubber cylinder and measure its deformation using a pot or a hall sensor, we would have the same result.

I'm saying that its not the pots that are the main flaw in pot-based brakes, it's the way they are linked to the pedal.

 

[kikie]

This is a very interesting thread right there. Even by sticking a piece of rubber under your pot-based pedal you are effectively creating a loadcell, a contraption that translates applied force to distance(position), just in a traditional loadcell the elastic element is steel. If your piece of rubber has linear force vs distance properties then you have a precise force sensor under your foot. All force/pressure sensors work in the same way - they translate force into deformation of material and then measure that deformation. In retrospect a pot is a loadcell without the bendy part. Now the reason why pots are bashed for 'being not realistic' in brake applications is that they are usually attached to the pedal arm and measure its travel which is not properly related to the force needed to press it. For a brake to be realistic its first 85 or so % of travel should result in something like 10% of braking, and the latter 15% of travel should pack the rest 90%. That curve would look like something like this __/ , so not linear at all and judging the braking force from the pedal travel with a linear pot would obviously be wrong. To do that precisely we need to measure the force that passes through the system, that can be seen well in CST pedals - the pedal rod moves indepently from the force sensor and is linked to it only through a spring and a rubber damper, Id like to point out that that spring with damper is only to give pedal its feel and to return it into place. If we'd replace the loadcell with a rubber cylinder and measure its deformation using a pot or a hall sensor, we would have the same result.

I'm saying that its not the pots that are the main flaw in pot-based brakes, it's the way they are linked to the pedal.

I see you followed the link I gave you over at thewayIplay.

 

[pilmat]

...

...

Thank you You both seem to see this the way I intended the conversation: that the pot is not the problem and that a load cell is only one possible solution!

When I say a pot and spring are equivalent to a load cell, I mean that both deform an elastic element and have some way of measuring it. Yes, a load cell is a convenient packaging of this and it naturally provides a means to measure a wide range of forces over a short stroke.

You both also see that by trying to measure linear pedal movement to capture a non-linear force application is an issue that is not well handled with the lower priced commercial pedals. Mr. Basher has a solution, and I think I can guess what it is. We'll look into that one in a bit. Right now I'm building a hydraulic brake to do some evaluation. That thread is here: Hydraulic Brake Build (G27 Based)

 

[Jet Badger]

Kikie, I certainly did

 

[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?

The only way all this can change is when sim developers program their sim is such a way that it reads a pressure instead of a voltage (signal). In that case a (n) hydraulic system is true to the real thing.

What is it: a hydraulic system or an hydraulic system?