DIY discussion - Increasing power to the G25 & G27 wheels

[mrsh 0]

note that: stronger motor and the FET also. stronger motor means more current drain, so change the FET to match with the changed motor.

will it be a solution?    i understand the torque is mainly come from current. 

[TheJoker 0]

not really, by removing turns from the motors you lower the actual resistor of the motor which means if he gets the same amount of volts as before he is going faster and stronger, but using more amperes. Thats is why you should make sure you improve the cooling of the FET.

[simul8r 114]

At least you know what a re you talking about . Size is 540, i think they work together but they are working far under their limits, so stronger motors still wont change anything. @Simul8tor the only thing you culd try is remove some turns from the motors, but it is risky and you should cool the controller. But this is the only simple thing you can do to make the ffb stronger

 

Hey, at least we are all trying to achieve the same goal so, please, let's try to be a little polite about it.  Back to topic, removing turns to cool the controller board sounds odd but very interesting but as you mentioned also sounds far to complicated, well for DIY at least.  I understood what you meant about upping the amperage will not affect much but I was shooting for just a little more wattage with hopes for at least the slightest difference and without frying the board.  I wanted to at least stay within the wheel's voltage limits to be on the safe side.  As far as I recall, the ARC Team overclocked the motors using some sort highly secret booster circuit, replaced some things on the controller board and pumped it with a larger power source somewhere up to 48vdc. I thought of someway mimicking this for any possible change minus the booster while working, of course, with the boards parameters.

 

BTW, hope to hear of what you find with your G25.

[simul8r 114]

not really, by removing turns from the motors you lower the actual resistor of the motor which means if he gets the same amount of volts as before he is going faster and stronger, but using more amperes. Thats is why you should make sure you improve the cooling of the FET.

 

A bit confused here now, so more amps is good?

[TheJoker 0]

Sry mate that came out wrong, i didnt mean to insult you. But i am pretty sure there is a little missunderstanding here. By removing turns i meant we could try to...well remove turns . This means you open a motor and there you see the copper wires around the three anchors. By lowering the length of the copper wires you can make the motor faster and stronger (works with any electric motor). BUT that means more power is needed and this means the controller in the g25 could get hotter hence cooling might be required. But as i said this is the only "simple" solution that comes to my mind.

[TheJoker 0]

A bit confused here now, so more amps is good?

Sure, but you need more amps going through the motor and not theorethical amps your PSU can do. When i bought my g25 from ebay, the guy forgot to send the PSU, so i used my old one with only 0.8 amps, and still i had full ffb strength i just had the problem that it was getting a bit hot after time

[simul8r 114]

Ah, I see what you meant...motor windings (pronounced 'whine-dings') at least from experience.  Well, if your 0.8amp worked but heated up thereafter shouldn't that lead to the wheel loosing strength in time?  It is obvious my wheel feels at it's peak from a cold start then gradually looses strength during play.  I suspect this got quicker since I spliced the fans onto the supply line within the wheel's case.  I figured the extra amperage should help in this regard.

 

Again, I haven't yet put the new psu through the works as of yet.  I'm kind of revamping my rig currently and this just added to the mix.  BTW, I also purchased that other psu rated @ 25v, 2.5a; it was cheap and figured I'd get it for chits and giggles.

[mrsh 0]

what is the rated voltage of the motor? (as u said " they are working far under their limits")

if so, change the correct rated high torque motor will work. (DC24V w/ thicker wire)

[mrsh 0]

my plan,

1. change with a same size (550 OK?) & correct rated motor & FET.

2. the worst case: put a rated DC24v 775 size motor into it. and to get a bigger headroom for the giant size motor, take away the pcb separately.

[mrsh 0]

http://www.flickr.com/photos/[email protected]/6918111254/

wires are connected as common pos & neg, the 2 motors is working together !!

i think i know what can be done with upgrading the FET

[simul8r 114]

http://www.flickr.com/photos/[email protected]/6918111254/

wires are connected as common pos & neg, the 2 motors is working together !!

i think i know what can be done with upgrading the FET

 

Forgive me for not being to familiar with electronics but what is a FET?  And thanks for diving into this venture project mrsh, it's good to see progress.

 

EDIT: Disregard, looked it up instead.  FET - field-effect transistor and wild guess is you are referring to the MOSFETS - metal–oxide–semiconductor field-effect transistors....okaayy, got it.

[TheJoker 0]

@mrsh you need to get a motor where the axle is looking out the back of the motor

also how exactly do you plan to swap out the FETs?

[TheJoker 0]

does anybody know someone selling his g25/27 but only the wheel? Then i could experiment a bit, on YT i see lots off ppl using the pedals and shifter with other wheels

[simul8r 114]

does anybody know someone selling his g25/27 but only the wheel? Then i could experiment a bit, on YT i see lots off ppl using the pedals and shifter with other wheels

 

Posted that question earlier, no response.  Was willing to take busted ones because usually they end up being a encoder problem.

 

BTW Joker, I even thought of the idea of adding a additional pair of the same motors totaling 4 to one wheel. What would be your insight to this?

[TheJoker 0]

you would need another g25 and use 2 controllers, just adding 2 motors is not going to work

[mrbasherman 41]

The FETS used in the wheel are good to 50 volts and 3 amps. They should not need to be upgraded.

 

However, some of the other components such as clamp diodes and caps may need to be looked at. These may not all be rated high enough. Many caps are rated at 50 volts though, so they probably aren't an issue.

 

The motors themselves while rated at 24 volts, can have a lot more volts run though them. The stepper motors on my CNC machine are rated at something like 6 volts if I remember correctly, but I run 48 through them. It's standard practice. However, you want to avoid exceeding the amperage rating on the motors if possible. It's a bit of a juggling act...

 

Anyhow, all of this is about how well you can keep the motors and components cool. If you could reach an ideal state in which all heat is drawn away from the motors and FETS, they'd last forever even at 48 volts. That however is just not possible...

 

There is some math to be done here before getting to carried away hooking up different PSU's, motors, etc. Unfortunately, I don't have the time to do it right now but I can help when time permits.

 

The first thing I would do is find some heat sinks that will fit the 550 motors in the wheel. Then you need a fan pulling air from outside the wheel, blowing directly on those. You need an exhaust hole as well. For the FETS, get some ram heat sinks.

 

Don't even mess with the windings on the motors, you'll just mess them up.

 

Lastly, this has been sort of explained a few times but I wanted to clear it up anyhow:

 

A watts rating or amperage rating on a PSU is a rating of it's capacity to provide a rated value. It's not the same thing as voltage. So, lets say the PSU is rated at 2 amps, but the wheel only uses 1 amp. In this case, the PSU is only going to provide 1 amp to the wheel. Until the wheel or whatever is connected to the PSU exceeds the rating on the PSU, you will see no difference in performance. If that rating is exceeded, you will see a degradation in performance.

[simul8r 114]

Once again basherman, really appreciate you offering your knowledge to the cause.  All the help we can get is always beneficial.

 

So, just to recap what we know or don't know so far....

 

G25/27

 

MOSFETS           :  50vdc @ 3amps

CLAMP DIODES : ?

CAPACITORS     : possibly 50vdc max

MOTORS             : 24~48vdc (and since original Logitech PSU is rated 1.75amps we assume at least 1amp for each motor)

HEAT SINKS       : 1 Ram sink across the top of the MOSFETS, 2 cylindrical motor sinks for 540 or 550 rc motors

FAN/S                  : direct cooling over motors

 

I don't suppose you know what this secret booster is all about that ARC incorporated into the mix?

 

BTW, that is about the most progress I've seen since the beginning of this discussion.

[mrsh 0]

mrbasherman,  what is the model # of the FET? search the data sheet will know the exact value.

btw, 50V @3A almost 150W (theoretically) and it has 2 connected as parallel (300W?). 

[mrsh 0]

can the big gear be separate from the wheel shaft ? i want to replace it to a belt drive sys.

anyone knows the gear ratio? i guess its about 1:11-12. with a standard 540 size motor, max torque generated on the wheel: ~3-4.5kg/cm. how about T500?

wonder to put a gear box before the gear since the speed isnt a big issue, the torque is

[mrsh 0]

Posted that question earlier, no response.  Was willing to take busted ones because usually they end up being a encoder problem.

 

BTW Joker, I even thought of the idea of adding a additional pair of the same motors totaling 4 to one wheel. What would be your insight to this?

i think there is not enough space......

[mrbasherman 41]

mrbasherman,  what is the model # of the FET? search the data sheet will know the exact value.

btw, 50V @3A almost 150W (theoretically) and it has 2 connected as parallel (300W?). 

 

There are actually 4 FETS, but only 2 chips. One chip is a dual N channel FET and the other is a dual P channel FET. The models are: IRF7342 and IRF7103.

 

They are undoubtedly configured as an H-bridge. However, I do not know if the motors are wired in series or parallel without taking the wheel I have apart to look.

 

The motors are apparently around 22 ohms resistance each, in case you'd like to do some more math.

 

That said, the stock FETs burn out sometimes as is, which is why I said the needed additional cooling. The thermal plane they are on is not all that large and asking more of them without additional cooling would not be a good idea. They also do not have a thermal pad that is soldered directly to that plane. (kind of dumb really considering) Edit: I wanted to add here... Generally when a FET dies in one of these, it is because one of the motors has. Usually this is due to the motor resistance decreasing due to wear. Over time this is bound to happen unless they are cleaned. When the motor resistance starts to fall, the current running through the FETs begins to rise.

 

On that note. I'm not sure what the "secret booster" is that ARC added to the mix but it's likely the opposite. It's possible that it's a chopper circuit which is designed to maintain the voltage going to the FETS, but reduce actual current. This is done through very fast PWM. The result would be highly increased detail due to the increased voltage and less highly increased overall power. (like holding the wheel against the FFB)

 

I should add... I don't mind participating in the discussion, but I take absolutely no responsibility for anything you guys do to your wheels. 

 

OH, and one more thing... 50 volts AT 3 amps may theoretically be 150w, however, that's only if you are sending 50 volts at 3 amps. The chip can't handle both at once, not for long anyhow.

[simul8r 114]

I should add... I don't mind participating in the discussion, but I take absolutely no responsibility for anything you guys do to your wheels. 

 

Understood, we definitely hold ourselves responsible for anything we do to our wheels.  In the meantime I will try to keep an ear out for a tester wheel throughout the discussion.

 

Thanks basher.

[simul8r 114]

They are undoubtedly configured as an H-bridge. However, I do not know if the motors are wired in series or parallel without taking the wheel I have apart to look.

 

Is there anything that we can do to help you with this.  I can dismantle mine but I would need some direction towards what I'm looking for.  We can use those pictures from that link you posted earlier.  Just circle and make points using Microsoft paint for example.

[simul8r 114]

... Generally when a FET dies in one of these, it is because one of the motors has. Usually this is due to the motor resistance decreasing due to wear. Over time this is bound to happen unless they are cleaned. When the motor resistance starts to fall, the current running through the FETs begins to rise.

 

On that note. I'm not sure what the "secret booster" is that ARC added to the mix but it's likely the opposite. It's possible that it's a chopper circuit which is designed to maintain the voltage going to the FETS, but reduce actual current. This is done through very fast PWM. The result would be highly increased detail due to the increased voltage and less highly increased overall power. (like holding the wheel against the FFB)

 

How then can we maintain the motors without having to pry them open.

 

A 'Chopper circuit', huh, very cunning of them calling it a 'booster'.

[TheJoker 0]

@basher they are wired parrallel, i got mine open.