Well, it's still a good wheel, but I just don't like the noise it's making. My 8 year old MS sidewinder ffb is like 10 times more quiet.
I also dont really know what to do, do you guys think I should return it?
Same here. It's a wheel with flaws, but the goods make it up. Yes, of course I could buy way better pedals from BRD, or an excellent shifter from Frex, or... But nothing can beat the price tag of the G25.
I'm using it in 540-720° mode and I really love it. It's somewhat noisier to the left, but still it's way more silent than my old DFP. Maybe I'm just lucky to have one of the good G25s? On the other hand... A product's overall impression can be completely wrong on forums because very often the people who don't have any issues with it don't post, whereas the ones with problems don't leave any chance to point others at it . I mean, I know quite a few people who bought a G25 too, and they all love the wheel.
I just tried this and I can confirm it. The noise is completely gone if you take out one motor (doesn't matter which one). But that's of course not an option
But then, what's causing it and why don't have all wheels the same problem?
If one motor is removed, do you think there'll be a loss in the FF felt? Not the strength as thats obvious, but I mean does the wheel software instruct each motor to do different things and so you'd loose out on some feel with just one motor.
Well, I'm happy you guys found the cause of the noise. I think i just gonna return this g25 and hope I get a better one which doesn;t make a lot of noise, and if that one still makes noise i'll just wait for a new revision.
Hmmm, so if it is the motors, then I suppose there is a problem with the whole "twin-motor" theory? It reminds me of when I would build something with my friends Lego robot toys, or Erector set... If you had two motors working in unison with gears, if one was not in sync with the other, you would get grinding and friction. Interesting how obvious the problem can be... this has to be it.
maybe the anti backlash actually makes it worse though all the ideas i had on why it would do that didnt really work out
anyway seems rather obvious that if you load a cog from 2 sides everything needs to be alinged and synched perfectly to work the way its supposed to
so maybe its a design flaw after all potentially made worse by chinese production quality that didnt come up during the beta stage when the wheels were built by swiss mechanics ... or something along those lines
Both motors are placed in parallel, so they do exactly the same. So no, you would not loose out on feel but the strength will obviously be a lot lower.
I wish I could take apart the anti backlash cogwheel, but I didn't dare to take off the bearing from the axle. The bearing is pressed quite hard on the axle, and with all the plastic I didn't want to break things. It would be interesting though, to see what happens if you take the anti backlash cog out and only use the main cogwheel.
I hope Logitech drops the double cogwheel design in a successor and goes for the professional approach by using slant tooth gears. That effectively elimininates both backlash and especially noise when made correctly.
I'm pretty sure that they will never use slant tooth gears because slant tooth gears require that the bearings (+the chassis) support the cog axles axially. The tolerances to not allow the axles to move axially would need to be more "tighter" to eliminate the backlash. It would also be more expensive as you would probably need bearings on the 2nd axle as well. If you used 2 axle design with just one pair of cogs, you would need to support the motor axle axially as well to protect the motor and eliminate backlash. In short that means imho that slant tooth gears are not applicalbe solutino in FF transmissions.
The way to go is to use belts. However the plastic cogs are so cheap to manufacture that in price there isn't really anything to compete with them. And they are quite reliable as well.
I am starting a project to modify the DFP to use some kind of belt driven system. The problem is that the total ratio of the gears in DFP is close to 19.5:1 between the wheel and motor so I need to decide whether I'm sacrificing some torque and use just one belt or to use two belts and have the increased inertia...
full agreement on the bit about belts
except for one: they could use double slant tooth gear which from my understanding have all the benefits of slant teeth without axial forces
about the dfp
have you though about using a stronger motor ? that might require some reverse engineering to figure out exactly how the electronics work and exachange the motor driver to something that can driver a larger current
ive thought about to maybe try and mod the dfp with a belt and a brushless motor if i find the time
It would be super expensive to use double slant tooth gear because you would need to specifically order those from some plastic stuff maker, ehm, corporation. And because of the wear there would eventualy be play, no matter how excellent and precise your tolerances were. I can't think of anything that uses double slanttooth gears, made from plastic so I can't imagine it as a good alternative for a cost-efficient solution
About the dfp mod, I'm in the very beginning of the process. I have thought about using bigger motor but it would require some heavy or intelligent modifications to the DFP's electronics. And the electronics is not really the strong side of me :o.
The main thing what I'm aiming for is to make the transmission between the motor and the wheel as light as possible plus to eliminate _all_ play. That may include using plastic belt wheels and plastic axles to save weight everywhere. After all it seems that I'm using 10:1 ratio, and it reduces the amount of available torque = less power to turn the wheel for the motor but also less resistance for me to turn the wheel. There are still few things to find out and solve before I can actually start making anything. First I need the go and buy a steering wheel, some bearings and axles plus the belt and the belt wheels... And I haven't really explored my junkyard stuff, yet
Anyhoo, gotta figure out some more stuff and on monday go find some nice hc hardware to get this started
main problem i see with this (excpet for the oney youve stated) is that the sensor is mounted at the motor so unless you use some different seonsor the result will be a wheel with half a dfps precission
I've been thinking that I use cogs there to get close to the original 19.5:1 ratio for the optical sensor. I just need to find out how DFP gets the information about how many degrees it has turned, and more importantly, how it recognises the "end points" of its rotation...
About the precision, the original ratio was around the 19,5:1 and the rotation of the wheel is 900 degrees so basically the precision is 19,5 * 900 = 17550 (precision units . Actually the optical sensor just "rotates" 17550 degrees when the wheel rotates 900. You would get the absolutical precision by knowing how many "samples" or signals the sensor gets during a full 360 degree rotation and then dividing the 17550 by it). If I use 450 degrees of rotation and install the sensor on the motor (ratio ~= 10:1, with the belt drive system), I have precision of: 10 * 450 = 4500, which is roughly a quarter of the original. With some kind of transmission for the sensor I can get to 9000 (double the ratio from 10 to 20) but it still quite far from the original value. Or is it? Actually you get the maximum precision only when you use the wheel in 900 mode and use all of it. I have no idea about the precision of the DFP's sensor, but there are ways to find it out. And the sensor doesn't even "read the rotation" directly from the motor axle, there are small cogs related. But the bottom line is, the more the sensor rotatesm, the more accurate the wheel is.
If it really turns only 720 degrees then I can give you the answer. There is a small chip near the big cog which has an inductive sensor on it. There is also a strap of metal attached on the big cog. The inductive sensor notices this metal strap while the motor turns the wheel, giving a signal to the DFP telling that the wheel is turn 360 degrees to left or right, or is centered. Actually the sensor just gives signal to the DFP's electronics at 360 angles.
Does the DFP calibrate itself if you put into extrmeme left/right position before switching the power on? I can't test this now...
I asked tabout this mysterious chip at RSC and, even if it is mostly speculating, it seems to be the only reason why the chip is there. Check this link for (some) info and pics of the chip
And it is going to be a tough job finding a steering wheel that doesn't cost too much. Almost all 260-280mm diameter wheels are above 100€. Some I found were around 80€ but especially the "racy" wheels seem very nice but cost over 200€ :nol2:. Building one by myself doesn't too interesting either . Using the DFP wheel is out of question, it is just too small imho.
EDIT: and maybe it's better to continue this discussion at the RSC, because it's getting very OT on this thread already
so it probably turns left and right through 720 (triggered by the inductive sensor) and works out the centre and clicks per degree of the "normal" sensor from that
it does
iirc if you have it at one of its extremes it turns to one of the 720 positions, calibrates a full 720 cycle from there and then goes on to calibrate itself in 200
id much rather discuss it in a forum where i dont have to wait 5 minutes for my post to show up
But how does the DFP know when it is on the extreme end? It can't be the optical sensor because the optical sensor works differently and thus can't give such information. Without this information it is quite hard to convert the DFP to use 450 degrees (as I have planned to do). One option would be that the DFP has some sensors on the PCB that "notice" if the motor draws excessive amperage to power to turn the wheel. Dunno, all guessing
There is a good thread about the subject already, and the chances are that the geeks who know this stuff will spot this endeavour more likely at RSC than here
either some way to sense its hit the hard stop (button or something similar) or no turn sensor response i guess
i know but i just refuse to use that forum for anything lengthy
i experimented with the claibration prcedure a little
seems like it always does a quick short jab to the right to check if it can turn that way and then tries to turn through 720
My wheel makes the same noise in both directions or similar to same. Like in movie when turned to right.
But its still left abit noiser then to right if i turn it really fast. Now I dunno how stong was the recording on that movie but doesnt sound anything like that when turning to right. Abit more tinny bit different but not like that.
But this is for sure the motor direction sounding like that.
I can see a problem if you go for 900° turns during a night even without FF. I had a momo Racing before which was much lauder anyway then I got a MOMO Force which is just a tinny bit noiser. Or maybe not it just make a more electric noise. Well it sound healthy and since I dont use more then 275 degress shouldnt be a problem at night.
Mine was also noisier in one direction (left). It is the nosiest when I don't connect it to the PC or the walloutlet. I opened it to see how it was build up and I must say that it looks strong and solid. I checked the screws, but they all looked good. I also don't think its the screws, it is probably the motors or the gearing in between..
The funny thing is when you disconnect one or both the motors and you turn the wheel.. (when it isn't connected to anything) Then the wheel is very light to turn.. This means that when the G25 isn't connected, the motors work like generators (logical) and are 'braked' by the (passive) electronics..
So when you turn the (disconnected) wheel, the resistance that you feel are mostly the motors that can't turn freely and that will probably cause resistance in the cog's and might resonate.. I think that is what I hear with my wheel..
. I mean, I know quite a few people who bought a G25 too, and they all love the wheel.
. Actually the optical sensor just "rotates" 17550 degrees when the wheel rotates 900. You would get the absolutical precision by knowing how many "samples" or signals the sensor gets during a full 360 degree rotation and then dividing the 17550 by it). If I use 450 degrees of rotation and install the sensor on the motor (ratio ~= 10:1, with the belt drive system), I have precision of: 10 * 450 = 4500, which is roughly a quarter of the original. With some kind of transmission for the sensor I can get to 9000 (double the ratio from 10 to 20) but it still quite far from the original value. Or is it? Actually you get the maximum precision only when you use the wheel in 900 mode and use all of it. I have no idea about the precision of the DFP's sensor, but there are ways to find it out. And the sensor doesn't even "read the rotation" directly from the motor axle, there are small cogs related. But the bottom line is, the more the sensor rotatesm, the more accurate the wheel is.
