Increasing power to the logitech G920

10
United Kingdom
United Kingdom
So what i want to do is increase the power on my G920, i have looked at the G27 ARC mod that increases the voltage with a booster board and a bigger power supply. I have found the installation guide and parts list for the ARC mod but some of the instructions are incomplete, if anyone has access to the original ones i would highly appreciate it. My other idea is replacing the motors inside the wheel for ones that draw more AMPS and also have higher rpm speeds and more torque, i would also get a 24v powersupply with this that can delivery a higher amperage. Any technical help is highly appreciated
 
Hey guys i am starting a new project that will hopefully increase the response time and torque of my G920. To start off with i will be replacing the PSU with a 28v 5amp one and i will be adding some heatsinks to the motors and fets, below you can find a full list of the parts i have coming throught the mail this week.

PSU:
https://www.ebay.co.uk/itm/DC-3V-5V...var=661545335769&_trksid=p2057872.m2749.l2649

FAN CONTROLLER:
https://www.ebay.co.uk/itm/3V-35V-1...e=STRK:MEBIDX:IT&_trksid=p2057872.m2749.l2649

MOTOR HEATSINKS:
https://www.ebay.co.uk/itm/Motor-He...e=STRK:MEBIDX:IT&_trksid=p2057872.m2749.l2649

FET HEATSINKS:
https://www.ebay.co.uk/itm/Aluminiu...var=553531830300&_trksid=p2057872.m2749.l2649

DC JACK:
https://www.ebay.co.uk/itm/5-5-x-2-...e=STRK:MEBIDX:IT&_trksid=p2057872.m2749.l2649

Below i have a picture of the board that shows you in red where i will be placing the FET heatsinks, i will also be cooling the ARC chip on the board.
j59tdnX.jpg


I will be using a 120mm fan from a CPU cooler to cool the internals and this is what i have done so far to the top cover of the wheel.
SpU4SPF.jpg

z1m3MuF.jpg

It is a bit rough at the moment but i will be sanding it all down and making it look pretty, below you can find a sexy pic of the board :gtpflag:
rIjAFV9.jpg


The next step that i will take in upgrading this wheel is changing the motors and the FET's, i am still currently looking for suitable motors but i have the FET's i will be using nailed down. The G920 has 3 FET's just like the previous logitech wheels but they are different types then what you get on a G25/G27. It uses 2 x 4951GM and 1 x 9977GM, i will be replacing these with 2 x 4957GM and 1 x 9975GM. These upgrades will have to wait though until next payday but once i get them i will post an update here :D. Feel free to ask any questions.
 
Quick update, I think I found good replacement motors, logitech uses 2 x rs555sh-15260 motors and the ones I will be getting are rs555-18150. They provide 3 times the rpm of the original motor and 210gf.cm torque vs the 150 that the original motors provide.
 
Great work trying to pioneer this one! It would be a hell of a thing if you come up with an affordable solution for Logitech wheels. The closet thing I've been able find in filling in the missing dots of the ARC manual is this blog post: https://www.hsw.hu/?p=506 which can be translated and gives some indication toward what the schematic might have looked like.
 
Great work trying to pioneer this one! It would be a hell of a thing if you come up with an affordable solution for Logitech wheels. The closet thing I've been able find in filling in the missing dots of the ARC manual is this blog post: https://www.hsw.hu/?p=506 which can be translated and gives some indication toward what the schematic might have looked like.
thats a great find. Still i would need to find how their board works and im not a electrician so yeah :))) I will soon be getting the new motors and mosfets and if everything goes to plan then i should see a nice increase.
 
Quick update on what i have done so far, so all the electronics are connected inside for the fan to work and i also mounted all the heat sinks onto the board and motors.
u4WbiwH.jpg


And below is my SIM setup for now until i get the new motors and some more filament to print a handbrake mount and a new gear stick.

fngK6FZ.jpg


I can definitely say there is an improvement and from what i can tell with 28V going into the wheel it feels like it did with the original wheel mounted, so the extra 4V cancel out the negative effects of mounting a heavier wheel however i can crank the PSU all the way to 35V, unfortunately the wheel only works up to 32V for me so atm i run it at 30V and i can feel a slight increase in FFB as well as faster wheel rotation which helps in drifting.
Next update will happen when i get the rest of the parts needed. New mosfets, motors and all that good stuff. Might be a while though because of the corona virus, deliveries kinda late here in the UK :)
 
I have just been looking to upgrade my G27, the ARC booster is just just a board with replacement higher powered Fet's on it, if you look at one of the pics in the link posted you can see the 2 motor fet's have been removed from the stock board and the ARC booster board has been wired in to where the fet's used to be, I assume the ARC kit has a couple resistors and diodes to lower the voltage to whatever component senses the input voltage and which would cause the switch off you had at 35volts
 
I'm gonna be attempting to do my own arc mod with my DFGT. Problem is, I can't really figure out exactly where they're tapping power on the PCB. based on the G27 it looks like they're tapping a diode on the hot side of the power input circuit, and using the other diodes leg as a method of tapping into the ground plane of the PCB? besides that, it looks like the way it's working is the power side circuit is bypassing all the source pads of the fets, and only using the gate signaling, and using the power tapped from the PCB to tie into the 4 gates of the fets, then the drains are wired into the motors as usual on their own pcb?

It seems like the limiting factor of the PCB is the source circuitry limiting it from running at higher voltages, so bypassing the PCB's circuitry is what allows for the higher voltages.

I think the mod itself is VERY basic but the reason why it commanded such a high price was because they epoxied in all the mosfets into their heatsink block so no one could find out that what they were doing was very basic and only uses like 5 dollars worth of parts.
 
thats really strange. I've had my dfgt for 10 years and have been running it with a bigger motor that runs higher rpm and has more power than the stock, without any wear on that gear what so ever.
 
Bumping this. Any progress? I want to just add in the upgraded motors. Do I need to upgrade anything else if that's all I want to do?
If you want to change out the motors then the mosfets need changed too, otherwise you will burn the original mosfets and fry your wheel
 
I have done the voltage mod to a few wheels, G27, G920, and 2 G29s. This included a 30v PSU, heatsinks on the FETS and the motors, 40mm fans in the "grills", 60mm fan above the board, adding exhaust holes, and wiring the fans and FFB to a switch.

I have been very happy with the improvements from these mods but I'm an engineer and like to tinker. However, I am in no way an electronic engineer, so I have a couple questions.

1. The motors you specified were a 18-24v motor, the stock motors are 12-24v, wouldn't these make for a much larger "dead zone" in the wheel?

2. Instead of replacing the FETS, would it be possible to use some sort of relay in conjunction with a separate power supply to deliver the increased volts and amps?
 
I actually didn't change out the mosfets and the wheel ran fine. Until I dropped a screw driver in the open case while it was on.

The motors definitely got hot but I did 1.5- hour play sessions without issue. The only problem I had with this mod was adapting the hall sensor reluctor wheel thing.
 
What do you have in place to dissipate heat?
Did the Mosfets get hot?
How did you adapt the sensor?

I was thinking about going with the RS555SH-2670 instead of the 18150. The 18150 have only 26% more torque than stock, while the 2670 has 89% more torque. Another pro is the operating voltage, 9-30v for the 2670 as opposed to 18-30v for the 18150 and 12-28v stock. This lower voltage operation should even decrease the center "dead zone". The downside is the 2670 uses 40% more amps than even the 18150. There is also a post on FB about using the 2670, they had near perfect linearity and 7nM of force, thats higher than the CSL 1.1. While there isn't a ton of info and it looks as if this is a completely custom setup it does make me want to try those out.

I have two G29 wheels, after your report, my plan now is to play it safe with one and install the 18150 motors and the other install 2670's. I'm hoping the worst thing that happens with the bigger motors is the FETS blow at which point I have to replace them anyway.




bg1.png
 
Last edited:
so wait, basically so long as i can deal with the extra heat on the fets and the motors i can just keep uping the voltage straight to the wheel? (i understand it will go pop at a certain point) so i dont need to actually **** with the board in order to actually deliver the extra voltage to the motors?

I have done the voltage mod to a few wheels, G27, G920, and 2 G29s. This included a 30v PSU, heatsinks on the FETS and the motors, 40mm fans in the "grills", 60mm fan above the board, adding exhaust holes, and wiring the fans and FFB to a switch.

I have been very happy with the improvements from these mods but I'm an engineer and like to tinker. However, I am in no way an electronic engineer, so I have a couple questions.

1. The motors you specified were a 18-24v motor, the stock motors are 12-24v, wouldn't these make for a much larger "dead zone" in the wheel?

2. Instead of replacing the FETS, would it be possible to use some sort of relay in conjunction with a separate power supply to deliver the increased volts and amps?

could you please confirm for me this, "basically so long as i can deal with the extra heat on the fets and the motors i can just keep uping the voltage straight to the wheel? (i understand it will go pop at a certain point) so i dont need to actually do anything with the board in order to actually deliver the extra voltage to the motors?"

this is purely in relation to getting a stronger feeling from the wheel, i understand i will have to make other modifications in order to get any real positive development out of it
 
Last edited by a moderator:
Yes you can safely up the voltage on the wheel to 30v without doing anything else, however when things get hot you will lose FFB.
 
awesome cheers mate, have just cut and patched the original psu plug to an old 28v lappy charger,

now to hunt around for some fans haha
 
Last edited by a moderator:
I literally left the case off and had a small desk fan on it the whole time. for the G29, I just hot glued the wheel on. Some versions use a magnetic reluctor and that worked fine. For the G27, I designed a 3d printed holder for it (search on thingivserse for it). I still had to glue on the encoder wheel itself though. I did about 2 months on it before I accidentally fried it.
 
I was looking at the specs on the stock motor FETS, they are rated at 60v and 3.4a, which should be absolutely fine for the 18150 motors as they max out at 2.06a each. However, the single FET that looks like it controls the entire power supply is only rated at 3.5a is a slight concern, as the two bigger motors will be above that limit at peak amp draw. The bigger, 2670 motors that max out at 7.7a, will absolutely blow up all of the stock FETS, light your house on fire and probably open up a black hole, distorting space and time, resulting in humans never evolving to the dominant species or just earths destruction.

@protomor Did you notice any difference in either the center dead zone or lack of minor details since the 18150 motors operation range is 18-30v rather than 12-24v for stock? Did you notice the single FET getting hotter or a burning smell from anything?

The voltage of the new motors may not be a huge issue for me as I mostly play iRacing and can set the minimum force higher, after I test it with the "wheel check" utility. For instance, right now, according to the software, my minimum force on my G29 is 1-2%, my G27 is 16-20%. I'd be interested to see what these motors will test out at.

For anyone who is better with electronics than me and to avoid the destruction of mankind: What would be the appropriate MOSFETs to use for the 2670 motors ? I understand how MOSFETs operate but that is about it. I would think I would need something that can handle at least 40v, 20a for the single FET, and 10a for the motor FETs.
 

Attachments

  • WheelCheck_1_82.zip
    124.2 KB · Views: 246
Last edited:
I saw some on eBay but if I remember correctly they were expensive, like 50 USD and came from China so probably will take a while.
 
I was looking at the specs on the stock motor FETS, they are rated at 60v and 3.4a, which should be absolutely fine for the 18150 motors as they max out at 2.06a each. However, the single FET that looks like it controls the entire power supply is only rated at 3.5a is a slight concern, as the two bigger motors will be above that limit at peak amp draw. The bigger, 2670 motors that max out at 7.7a, will absolutely blow up all of the stock FETS, light your house on fire and probably open up a black hole, distorting space and time, resulting in humans never evolving to the dominant species or just earths destruction.

@protomor Did you notice any difference in either the center dead zone or lack of minor details since the 18150 motors operation range is 18-30v rather than 12-24v for stock? Did you notice the single FET getting hotter or a burning smell from anything?

The voltage of the new motors may not be a huge issue for me as I mostly play iRacing and can set the minimum force higher, after I test it with the "wheel check" utility. For instance, right now, according to the software, my minimum force on my G29 is 1-2%, my G27 is 16-20%. I'd be interested to see what these motors will test out at.

For anyone who is better with electronics than me and to avoid the destruction of mankind: What would be the appropriate MOSFETs to use for the 2670 motors ? I understand how MOSFETs operate but that is about it. I would think I would need something that can handle at least 40v, 20a for the single FET, and 10a for the motor FETs.
Remember you throwing in 7nm motors in there might strip the plastic gears
 
I was thinking that as well but seeing that project I linked above it seems as if it is working... Now as far as the reliability, only time will tell.
 
I having minor problem with my g920 the wheel is slight crooked and I have researched and learned putting washer behind the motherboard works so I did it, but it’s still slightly off center but I think it is the motor itself that is the issue and was wondering if it’s possible to replace
 
Back