1
2
Hi Bose, I have done a little looking into MOMO Racing and have some insights for you to consider as you go forward with your project:
First, MOMO Racing uses a "geared down" optical encoding system (same unit is used in Driving Force Pro). Basically, the system uses thin slots in a rotating disk to create "counts" and to determine direction. These "counts" provide the firmward with the information it needs to know where it is between the mechanical stops. In the case of MOMO Racing, the wheel rotates less than one full revolution, so, upon startup, the wheel rotates one way until it hits a mechanical stop. The firmware resets the counter to zero, then rotates the wheel the other direction until it hits the other stop. It then divides this number (full range) by two and rotates the other direction until it reaches this new number which centers the wheel. After this calibration, when the wheel is turned the firmware keeps track of these numbers and always knows where center is. Now, the processor that stores these numbers is designed to keep numbers of only a maximum expected size. If the encoding system feeds more counts above this number then it looses counts/stops working. I do not know what this number is, but I am fairly sure that if you feed 900 degrees worth of counts to the device it will not keep track.
So, this is what this means to your project: For it to function you will need to further gear down the optical encoder to create roughly the same number of counts over the larger wheel rotation angle. MOMO Racing is a 240 degree rotational device, so to make it a 900 degree device you would need to reduce the rotation of the optical encoder disk by 3.75 times. NOTE: this greatly reduces the accuracy of the steering wheel as a one degree rotation of the stock wheel produces about 10 counts, divide that by 3.75 yields only 2.8 counts. To accomplish this gear down would require a new optical encoder frame with a double gear between the stock 14 tooth pinion and the optical disk's 35 tooth pinion. You will have to calculate the ratio based on the two gears you assemble to make this double gear.
As you can see, this is not an easy or a good way to make a 900 degree wheel out of a 240 degree wheel. So, I have a few suggustions:
First, you could make the changes to the FFB gearing I reconmmended (39:1) then, move your stops in until they produce 240 degrees. Once the system is working you could then move the stops out a little at a time to see where the "saturation" point of the processor is. This would probabily be less than one full rotation.
Or, your cound leave your current FFB gearing, which is 6:1, yielding a per degree of wheel rotation counts of 1.6, then move your stops to provide the total expected counts which would be a 1560 degrees of rotation! But, you would have no/very weak FFB.
Or, you could change the FFB gearing to yield a total number of counts equal to a 900 degree device which is a ratio of 10.4:1. This would give you some level of FFB while allowing the device to boot up. Your accracy would be less, but might still be OK. Note: if you keep your stock MOMO Racing motor pinion (20 teeth) and make a single reduction device the main gear would need to have 208 teeth. Of course, you could reduce the number of teeth on the pinion allowing a smaller main gear to be used.
Or, you could buy a DFP, use all of it's internals stock, and get true 900 degrees as it's processor is looking for a bigger number based on 2.5 turns of the wheel. One issue with this approach is the 200 degree mechanical stop in DFP. The firmware expects to "see" and operate this sliding stop during start up. Fooling the firmware could be tricky.
Or, you could buy a DFGT, use all of it's internals, get true high accuracy FFB without the problems associated with the sliding 200 degree stop as DFGT does not have this stop. But, DFGT uses a direct drive/anti backlash optical encoder (same module as G25) which you would have to deal with. You could attach your ballscrew directly to the FFB transmission main shaft (where the DFGT wheel is currently mounted) to get around this issue. Still, DFGT is a PS3 device, with all of the associated buttons. You would have to rewire the MOMO wheel with the expected wires routing the MOMO paddle shifters to the DFGT paddle shifter buttons and the other six buttons as required for your game play on a PC. This is doable, just keep the DFGT wheel panel off to the side and run wires up to the MOMO wheel.
Finally, you could buy a G25, get smooth/strong FFB, solid construction, and a clutch/shifter. I use a G25 in LFS and GTPrologue mounted on a Playseat that I customized to raise the shifter and move the pedals over to the right and am very happy with it.
Anyway, I will keep watching to see what to decide. Good Luck!
Hi Bose, at various frequencys, say, running over a bump strip, the quicky changing forces applied to the system will be better with a lighter stop.
Anyway, I don't use MSN so just let me get some more information and get back to you tomorrow.
Hi Bose, I think I understand now: the "square mass" I refered to is your "moving stop". It moves up and down the ballscrew and stops at the ends. Is this correct?
If so, as I mentioned, your "stop" must be accelerated and deaccelerated at what ever frequency that the FFB system is sending the wheel. This "absorbs" some of the forces and degrades your experience.
As far as what you can do. If you use the same motor I would increase the gear reduction to near stock or higher using stronger gears or perhaps a belt drive. If you use a belt drive tension on the belts will be critical to the feel and losses in the system. Please note that the electronics are design with power constraints based on the stock motor/power supply and can be burned out with a higher amperrage motor. There is also "over current/back EMF" protection on the board which can affect your driving experience if the configuration is not correct.
Next, I would make the "stop" weigh less (lower the mass).
I need to find out a few things about the optical encoder on MOMO Racing which will help you determine what range you can obtain, as MOMO Racing, as you know, is a less than one full turn device (240 deg's I think) as so it does not need to calculate what turn it is on. This is very different from DFP/DFGT/G25 which are multi turn devices which BTW would be more straight forward for you to modify.
Please tell me if my assumptions about the "stop" are correct and let me think about this and get back to you later.
Hi Bose, I can tell you a bit about the MOMO Racing FFB transmission which might make your lack of power at the wheel clearer. MOMO Racing uses a 39:1 overall gear reduction. The motor produces approximatly 42 mNm of torque. If you multiply 39 times 42 you get 1638 mNm at the SHAFT. But, at the wheel rim you only get about 12.8 (wheel dia is about 256 mm). Your set up appears to use the first reduction gear from MOMO Racing attaching it directly to the screw shaft to which the wheel is attached. The motor pinion has 20 teeth, the first reduction gear has 120 teeth yielding a 6:1 reduction. This means that at your SHAFT you are getting only 6X42 mNm=252, and therefore at the wheel rim you are seeing less than 2 mNm. You can see why a double reduction gear set is required: the second reduction pinion on the double gear has 20 teeth, the main gear has 130 teeth yielding a 6.5:1 reduction, if you multiply 6.5X6=39. It is this power of mutiplucation that produces the torque that is needed from a cost effective motor/power supply.
Another aspect of FFB response is inertia in the system. Everything that has mass has inertia. Interia is what makes someting that is not moving hard to get moving, and something that is moving hard to stop. Increased inertia in the FFB system absorbs the force producing an overall smaller "feel" at the wheel. By adding large metal parts the inertia of the system increases decreasing the available force. Another serious aspect of this added inertia is tooth strength. The teeth in MOMO Force's transmission are desgned to accept a given bending stress without failing. The highest stress happens at the mechanical "stops". Inertia makes the system (wheel, shaft, gears, motor rotor) want to keep rotating. The stop makes the whole thing come to a very quick stop. Force equals Mass X Acceleration (F=Ma). In this case very large forces are created because the (de) acceleration is very high. And, due to the nature of FFB gear drives, the same teeth see this force over and over again until that fail due to fatigue of the plastic. Your system has alot more rotating mass which will destroy the plastic teeth of the original MOMO Racing first reduction gear.
Another aspect of FFB transmission design/function is refered to as "frequency response". This is the ability of the transmission to reproduce the frequency (back and forth forces) at the wheel rim. Gear based transmission have small gaps between the teeth of the gears called backlash. This backlash limits the highest frequency that can be felt at the wheel. Think of it this way: the motor turns back and forth at a given frequency. As it does it must "take up" the space between the teeth before any force can be transfered to the next gear tooth. IF it is changing direction so quickly that it is turning around BEFORE the gap is removed, then no force is transfered to the next tooth and the maximum frequency response has be reached. Control of the system backlash is of very high importance in FFB transmission design.
As you stated, these dc motors have a "shocky" feel unpowered. This is called "cogging" force and is a function of the motor's design. Logitech FFB motors are custom designed to minimize the cogging effect as much as possible. But, cogging and motor torque are directly related: less cogging means less torque. A higher transmission reduction ratio minimizes the cogging feel: there is less feel at the wheel's rim as small wheel rotation angles result in large motor rotation involving many cogging steps. BTW: G25 uses a 16:1 reduction which caused the need for very special, very low cogging, custom motors to minimize the cogging feel at the wheel.
Finally, I have question about your design: what purpose does the "screw shaft" and associated "square mass" serve? It appears that the original MOMO Racing wheel is attached directly to the shaft. If so, then the screw and added system mass do nothing for you that a straight shaft would do, while adding cost and interia to the system.
Hope this helps. As you can see, a great deal of thought and engineering go into making FFB wheels that work well and are affordable.
Good luck with your project!
G25 was designed from the ground up as a two motor single reduction transmission with reduced inertia being only one of many reasons.
There is a "lever switch" that must be in the "open" position for the sequential mode to work. You must have damaged/jammed or altered this switch when you modified your shifter. The Sequential LED comes on just as you start turing the Sequential/H shifter Knob CCW because the switch is "made" when the Knob is in the "H shift" mode.
What you are seeing is the base (copper alloy) plate used before the nickel plate is applied. Although looking at the boot area it does look like rust. How long have you had/used your G25/Shifter for? Also, before you sanded it did any other abrasive material get on the shifter? Just really curious, thanks!
The Shifter Shaft is made from cold rolled carbon steel, and plated to prevent rusting (The Wheel frame, Pedal Caps are Stainless). The only "ring" on the shaft is made from plastic and allows the boot to slide up and down the shaft and since you noted that it is rusting also I am confused as to just what is going on. It should not be rusting. Could you please post some pictures? Thanks!
Note: if the plating wears down, the copper base plate shows through (kinda brown looking) and this might be what you are seeing. You might rub your finger on it to see if it comes off or stays put.
The first FFB retail PC wheel came out in the spring or summer before MS, and Logitech's first FFB wheels (1998). I tested one (it had a very toy like appearence with a "cut out top" style rim. The face had a decal with gages printed on it. It was belt drive and work fairly well. For the life of me I cannot remember who built it, except that Immersion did the engineering.
Once booted on your PC press "Select" "R3" (at the same time) then the right paddle shifter to cycle the mechanical sliding stop between 200 and 900 degrees.
The DFP defaults to 200 degrees when plugged into a PC. If you plug it into a PS2 running GT4 it will boot up and reconfigure to 900 degrees. There is a electro-mechanical sliding stop that does this. Once booted on your PC, you can press - "select" and "R3" (at the same time) then, while holding them down, press the right paddle shifter. This cycles the sliding stop either in or out. If it's in (200 deg mode) only the left LED is on, if it's out (900 deg) both LED's are on.
The ball detent holds the gimbal (shifter) in all gears as well as neutral. As you found if you remove it it will not stay in gear. The clicking sound was "designed in" to produce an audible response for gear shifts.
The click is caused by the diameter of the spring/ball hole in the gimbal being larger in diameter than it needs to be just to align the ball. A 4 mm deep X 6.30 mm dia counter hole allows the ball to "click" againist the side of the gimbal hole when the ball snaps into the metal hole in the latch plate. If a 4 mm long "shim cylinder" with a ID of 5.9 mm and a OD of 6.3 mm (wall thickness 0.20) were pressed into the gimbal spring hole it would greatly decrease the volume of the click. If it causes you so much discomfort it might be worth a try. (shim stock wrapped in a cylinder might work, just make sure that the ball can move in and out freely and that you do not press it in more than 4 mm. Also, make sure that it cannot come free and jam the shifter gimbal)
There is nothing that can "slip" in the gear train in DFP. The best way to determine if it's the wheel or the game, is to plug it into a PS2 and see if it centers up. Please note: DFP calibrates by rotating the wheel "stop to stop", counting "counts" from its optical encoder, then dividing by two to find the center. If the wheel is held during calibration it will not calibrate properly (if at all). Now, if something gets in the way of the "stops" causing the counts to be off, then the center will be off. This could be some debris that get lodged between the rotating stop and the sliding stop surfaces.
DFP is a nearly completly different design top to bottom. The wheels are not compatable.
They do share the main gear, double reduction gear, motor pinion, clamps and optical encoding parts.
Last edited by mcman, .
Setting up the G25 in LFS can be very daunting, and often seems like something is wrong with the wheel. The best way to tell if your wheel is working correctly is to plug it into a PS2 and run GT4. If all works here then your setup in LFS (or whatever PC game you use) is wrong. Don't assume that the wheel is at fault as often a simple mis-direction in the "options" section of LFS can make it feel like it's defective.
Also, understand that G25 has no mechanical 200 deg stop (like DFP), only "soft stops" generated by the FFB motors, and the mechanical stop at 900 deg's. These stop points must be defined by the user.
The simplest way to determine if your G25 is working correctly (sequential shift only, no clutch) is to plug it into a PS2 (GT4 works well). Often the configuration of the PC affects the function of the device (any device). PS2 does not have these issues and therefore provides a quick check on function.
The mounting holes are M6X1, not M5. Very cool install!
DFP/MOMO Racing/G25 all have "thin desk extenders" that snap onto the clamp to allow use with "thin desks". They are shipped installed. To get the full 52 mm grip you need to remove these extenders by pushing them on the side and "snapping them off". Once off they can be snapped back on if needed.
Note: the middle "clamp" on MOMO Racing is not intended to be a clamp. It is a "anti tip" device that should not be tightened too tight. If this middle clamp arm in in the way it can be removed with a screw driver as this part uses "PEM" inserts which are also reusable.
Are you pushing down in the neutral position and holding it down as you put it is sixth? Reverse "makes a switch" at the base of the shifter shaft and then the latch pin slips into a reverse slot under the sixth gear slot.
OK, got it! Very clever! So, correct me if I'm wrong (again
): you could use a corded USB game pad and do the same thing. Just mount the game pad PCB inside the G25 case, and then run wires from the game pad through the shaft up to the wheel to buttons on the wheel.BTW; I think you got lucky with your set up; DFP uses a 40:1 reduction between the motor and the wheel (double reduction transmission). Then it uses a "reducer" at the optical encoder (the little gear on the motor shaft) to drive the encoding wheel. This was done to reduce the speed of the encoder wheel to prevent dropped counts. The total ratio between wheel and encoder is about 16:1. Since G25 (a single reduction transmission) uses a direct drive encoding wheel (at 16:1) no reducer was required. By attaching the DFP encoder to the second motor on G25 the optical encoder rotates at a 6.4:1 ratio (instead of 16:1), so it does not "see" the total number of counts it expects (it "expects 2400 lock to lock, but "sees" only 960), but that's where the centering opto saves you as it is a "default" center if the counts are "wrong". A standard DFP will boot up and calibrate without this centering opto. One other interesting thing about your set up is the fact that DFP goes through a defined 200 deg, then 900 deg, then 200 again calibration sequence. The firmware "expects" to see the 200 deg sliding stop switches switch back and forth, and in your set up they don't. I find this very curious and the only thing I can think of is the centering flag is saving you.
Hi SatCP, correct me if I'm wrong: you now cannot use G25's Shifter or Pedals (clutch), and the maximum force you feel is lower than before.
G25 must feed two motors, therefore the PWBA is designed to take more current from the bigger power supply. DFP's PWBA is designed for about half that power.
BTW, you might try disconnecting the wires to the "centering opto" that you mounted on the 900 deg sliding stop to see what happens.
You can always use the included M6X1 threaded bolt holes (hard points) in the bottom of the pedals (that Logitech included at added cost and engineering time) to attach your pedals to a piece of plywood (use countersunk holes). This piece of ply could be shaped to set under you chair, or be long enough to reach the wall to prevent sliding. Of course, on carpet you can use the included carpet lock, another added feature unique to Logitech pedals (patented) which is also an added cost/complexity that helps the pedals stay put. Finally, you could buy a simple race sim set and use all of the "hard points" that Logitech thought to include so all three componets can be bolted down. I know of no other vendor that provides this many solutions at this price point.
Note: if you do use the hard points make sure you review the included "bolt pattern templates" to make sure you do not use too long of machine screws that could damage the top case of your device.
The reverse gear on G25's shifter is activated by two inputs that the firmware is looking for: "sixth gear" output from the two potentiometers plus the reverse switch. This will be tough to get around since it is not a simple button press, but multiplexed inputs.
1
2FGED GREDG RDFGDR GSFDG