Nissan Juke : Juke Forums banner

261 - 280 of 317 Posts

·
Registered
Joined
·
1,176 Posts
Discussion Starter #261 (Edited)
The problem I see with replacing just the end case w/pulley and belts is that the oil pump, valve-body, and forward clutch packs are probably shot at that point as well which is why the pulleys wore out to begin with. But the diagnostic codes would probably detect if the fix worked or not. IMHO an end case replacement by itself isn't worth the effort long term.

The pulleys are thru hardened, about the hardness of the CVT push-belt. I was going to send them out for Nitriding but I didn't bother once I learned they were already thru hardened. Yep, I had some raised spalling and some grooving/pitting. I could have bought clean refurbished pulley cores, but decided mine were in great shape other than the pulley faces. I had to remove about .005" to get thru the pitting of the pulley and was able to get the pulley runout to about .0005"-.001". A 90* air die grinder was used with a 180 and 320 sanding disc to remove the heavy stuff, but it would tend to dig into the pulley and make groove marks. I then hand finished using 180 grit then 320 grit maintaining a circumferential turned finish pattern similar to the OEM pulley faces. The pulley face angle was untouched and remains @ 11*. I first tried a trick of altering the finish pattern that was at a 30* angle to gain more belt traction in the rotating direction. But then figured it might help in one direction but hurt in the other, so I went back to the stock grind pattern. Anyway, my original push-belt element surfaces were completely wiped, allowing the belt to ride on top of the oil instead of cutting thru it. The belt looked stretched but in fact it's the interior surfaces of the element that wore done creating the illusion of a stretched belt and increasing belt element gap to about .060", not good for the belts. The CVT belt is designed more like a synchro on a manual transmission, without the synchro grooves they can't shift the transmission correctly.

A lot of that wear happened in the beginning when I bought my car and Nissan underfilled the CVT fluid. Then later on I had a fluid leak thru the case and ran low on CVT fluid. The forward clutches had little to no clutch pack end gap clearance, maybe .010" if that when I dry fitted in new clutches and frictions. When the clutches heat warped then they started dragging and wiping out the clutch faces. My build increased this clutch/snap ring gap clearance to .025" to .035" so the clutches never drag on each other and wear out prematurely.

All that stuff is what I believed caused a lot of the belt damage and transmission damage. After my new valvebody swap I still had slippage but not as bad.......then the forward drum shattered after another 5k miles when I cranked the boost back upwards near 320-330 N-m. The new build replaced the forward drum, planetary, stator housing, & input shaft with 2007 Altima refurbished parts that had little to no wear indications, among all the various new soft parts and new bearings & pistons as well.

I'd be the first person to tell someone to just buy a new transmission. I exceeded $3250 on this build in parts alone (i.e. without the CVT cooler) and could probably do a new built transmission for about $3500 my cost with everything including new 901074 belt and all the rest needed.
 

·
Registered
Joined
·
1,176 Posts
Discussion Starter #262
Update:

Finally had some time to think about how to raise the CVT pulley pressure. The Jatco pulley pressure has a PID circuit that uses the Secondary Pressure Transducer as feedback to the solenoids. Changing the control piston shim/spring cap for higher preload or a heavier spring cannot work reliably.

To solve this problem, I built an electrical circuit that will intercept the Secondary Pulley pressure signal and condition it lower. This will trick the TCM into feeding even more pressure than stock. There is a comparator chip, NAND gate, (2) Transistors, & various resistors. The comparator simply takes the raw analog voltages and feeds either the transistors or the NAND gate.

The circuit was built to include a "cut-in" voltage (VDDref) such that I'm not running increased pressures all the time, just when the engine is coming into mid-range airflow. That "cut-in" voltage of about 1.5-2.0 VDC can be trimmed with a "variable resistor". The main pressure increase, or voltage decrease ABOVE this "cut-in" voltage also needs some trimming capability to add or subtract voltage, representing increased pressure in this case. You can see there is VCC = system voltage, VSSin = stock voltage coming off the Secondary pulley solenoid, VDDref = cut-in voltage where higher pulley pressure is requested.

The circuit seems to work well on the simulator. This'll be my late summer project.

 

Attachments

·
Premium Member
Joined
·
2,562 Posts
Have you figured out how to prevent P-codes related to line pressure from popping up or is it a trial-and-error type thing?
 

·
Registered
Joined
·
1,176 Posts
Discussion Starter #264 (Edited)
For this it would be a trial and error. The ROM pack and the TCM aren't expecting this much drift in the soleniods so it will have to relearn. Sonnax recommends in this type of case to also run a higher rated spring in combination with electronics. I've looked at the Secondary Pressure control valve and/or the Secondary Regulator valve as likely candidates for spring upgrades or pre-loading with the bore endcaps. This might then be a net wash to the solenoid duty cycles, keeping the TCM happy and within it's control range, while pulley pressure increased.

I'll be bread-boarding a prototype and having our Electrical group see if there can be made any improvements. Then a case to hold the components and mount somewhere, possibly a small fuse box.
 

·
Premium Member
Joined
·
2,562 Posts
Another thing to consider is pressure vibration due to solenoid fluctuations when fluid moves past the solenoid/spools. That is main issue with the current crop of CVT valve bodies for 13+ Altima/Rogue/Sentra/Pathfinder and 15+ Maxima/Murano. When the fluid would rush past the spools and/or solenoids they would oscillate in their bores due to incorrect spring pressure or incorrect internal solenoid construction and in return the oscillations would amplify into pulsating line pressures, which in turn leads to rapid belt slip/grip and premature belt and pulley wear. This translates into a judder or vibration under light throttle acceleration similar to when a lock-up torque converter malfunctions during lock-up actuation.
 

·
Registered
Joined
·
1,176 Posts
Discussion Starter #266
That's pretty interesting, haven't experienced that on mine. Maybe the newer CVT run such light pressure on the belt that any fluctuation results in judder or slippage. One option is to run a double nested spring, they typically snub any resonance that might build up.
 

·
Registered
Joined
·
1,176 Posts
Discussion Starter #267
Been hanging out and recovering, it'll be 10 days post-op and getting healthier with re-hab.

Ordered the 2J FMIC with crashbar, 250LPH fuel pump, and Injen Upper IC pipes that seem to fit better with the FMIC. A bit pricey but hoping fitment will be spot on. I have to send my GTM 2.5" downpipe, FR testpipe,and new exhaust manifold out to Swain tech for thermal coating. I still have to order up the Mambatek 9 blade turbine wheel and rebuild kit and I should be good to go on power mods. Should easily put down 260 w.h.p. for a nice daily driver. I'll post up pics when the parts start coming in.
 

·
Premium Member
Joined
·
2,197 Posts
I will say out of the whole FMIC/piping/crash bar setup I did from 2J, I was the most impressed with the crash bar. It was so light and nicely contoured compared to the OEM.
 

·
Registered
Joined
·
1,176 Posts
Discussion Starter #269 (Edited)
Yep, the crash bar is what makes the kit nice. Saves me custom fabrication which I simply don't have time for and it looks nicely engineered. I've always had bad luck with FMIC kits requiring custom welding of pipes so I'd like to avoid it for this kit as I like everything fitting on the money.

Just got the pipes, they are already red wrinkle coated with black 2.75" and 2.5" couplers. They are very nice, the emblems will get stripped off though. Waiting on the FMIC kit, probably going to ship next week. I'll be switching to the Synapse relief valve. The Forge unit sucked badly, my EVOX synapse kit worked out nicely, so the little Juke must have one too.

I might weld in my 4" diameter Super High flow metal matrix cat into my FR testpipe. Thinking about it or just making a 2nd pipe myself with the 2.5" flex and highflow cat. These cats have like maybe 5 h.p. loss at 350-375 h.p. so they are going on my EVO X, but the Juke might also get one. Cops don't like the flameballs coming out the rear.
 

·
Super Moderator
Joined
·
1,573 Posts
hey pboglio, can I snag your stock crash bar?
 

·
Registered
Joined
·
1,176 Posts
Discussion Starter #271 (Edited)
-Squirtbrnr,

I'll let you know, might need it in the future.

Update:


While I'm waiting on my FMIC to come in, did a little more work on the CVT electrical pressure control circuit. I had lots of trouble with the NAND gate not firing correctly. The 5-volt NAND gate was getting confused with the comparator outputting above and below 2.5V, causing confusion on logic high/low states. I switched to a 2-volt version and it solved the problem. A 3k ohm resistor was installed between the NAND gate and transistor to better control the original voltage signal. The circuit now has a clean signal from 0.5V (0 psi) thru 4.5V (870 psi). I'm getting some minor non-linearity in the conditioned signal from 4.0V thru 4.5V but I can live with it. The circuit simulator looks good, breadboard is next.

The potentiometer R10 controls the "cut-in" voltage, here set at 1.25V trigger point. This sets the "trigger point" to start the decreased voltage. Why is this advantageous? Well, the factory go to great lengths to run the bare minimum pressure to extend belt life, and the increased pressure is only needed at higher torques. In the circuit diagram you can see the pressure transducer signal Vss=0.5V. At this value, the unmodified signal runs to the NAND gate and the transistor U3, here showing 502mV or 0.502V. This is close to the original signal of 0.5V, basically 0 psi pulley pressure condition.

The potentiometer R11 controls the "increased pulley pressure signal". I've set it for 92% of the original Vss = 2.5V signal off the pressure sensor, varying from 0.5V-4.5V. Basically a mid/high torque condition. Transistor U3 "stock circuit" is now deactivated and the "modified circuit" U2 is now active with a trimmed voltage of 2.28V (Vss = 2.5V).

The TCM will now see this reduced sensor voltage and attempt to trim the "Secondary Pulley Solenoid" by 8-10% to compensate, retaining the trim memory on the ROM pack mounted to the valvebody as well as the TCM. This "adjusted" solenoid pressure then gets sent to the "Secondary Pressure Control Valve", which in turn controls the "Secondary Regulator Valve" that directly controls the Output pulley. This will now push the Solenoid trim adjustment +8% or -8% and potentially against it's upper range, which could throw a code. In this event a "stiffer" spring of say +10% in (1) or both of the control spool pistons might bring the trim adjustment to around +/-0. Similar to MAF scaling to bring fuel trims to 0, same concept.

Why haven't I talked about torque converter or clutch pack pressure increases? Well, because they are a little more complex and I have modified clutch pack with (4) forward clutches good for 330 lb-ft, the stock converter being good to about 300-312 lb-ft by my estimates. The AMSOIL CVT fluid boosting those numbers even higher, more than plenty. This would then limit any interactions within the CVT to the Secondary/Output pulley only, greatly simplifying the tuning/troubleshooting process.

Challenges are several. The cut-in voltage would ideally change with each manual gear ratio, since pressures are also different. For now, setting the trigger point at 1.25V will work well for gears 1 thru 6. It'll be triggering a bit low for 1st gear and a bit high for 6th gear but it'll work. A simple voltage divider circuit also works, but once the pressures start getting higher for those high boost runs, it's really a nice feature to ease off the pressure at lower torque conditions.

Finally, I'm trying to find a way to deactivate the automatic upshift in manual mode. My car has run to 7000 rpms in sport mode(automatic) and normal mode (CVT) in the past but shifts at 5950 rpms in manual mode. At this point I'm looking at some options but I'd prefer it be done in the ECM/TCM. It would seriously increase my high rpm horsepower without requiring alot of torque.

More pics of the additional parts (fmic, turbine, strut bar, CAI) as they come in.

 

Attachments

·
Registered
Joined
·
1,176 Posts
Discussion Starter #272
I've been recouping from my hip surgery so haven't played much with the transmission. My engine rebuild is now starting to take full swing so I'll be phasing into that and winding this thread down a bit.

Looking at the transmission design I decided to see what I can do to further reinforce it before I wrap up my epic transmission rebuild. The clamping force problem has been addressed with the electronics I mentioned earlier. However, there remains some strength concerns on the forward clutch drum. Having gone thru extensive computer analysis, I've decided to attempt to reinforce the clutch hub with a high strength collar insert. The hub collar will be constructed of 4340 steel/nickel alloy. It'll be strong enough to get the job done without requiring an exotic alloy like 300M or costing a fortune. It's not as thick as I'd like, but it's as big as the mating parts will allow since everything is a close fitment. My machine shop is waiting on prints for this week, hopefully in 3 weeks he can knock out 5-6 pieces. I'm building a few extras in case I build more transmissions in the future.

The idea is to have this additional collar machined and then I'll hone this part myself to match fit perfectly onto the existing foward clutch drum. This slip fitment is kinda important, since to prevent cracking of the clutch drum spline hub the fitment needs to be close. Permatex bearing retainer compound is the method I'll be applying to make this work as it's strong and will fill the minute gaps. This'll transfer the load forces thru each part so they are equally stressed as one piece. I'd have liked to TIG or Braze the parts together, but they couldn't take the thermal stresses without cracking due to the Nitride coating.....not worth the risk.

The FEA analysis I ran showed an improvement of about 28-30% strength, allowing 400 N-m (294 lb-ft) at over 100,000 full power cycles. Normally the stocker would be done by 8,000 full power cycles based on my analysis. That's a huge improvement. There is no real way for me to test this unless I do it on the vehicle and turn up the boost. The goal will be 370 N-m or 272 lb-ft of torque. I'll be pushing the transmission at this torque level.......but I've upgraded the transmission in all of the critical areas that failed on my original transmission.......so the weak points have been identified. With the extra clamping force I have planned, it shouldn't slip. Belt wear will be high, but that's how it goes.

At this point I'm leaving the stock Bosch 901066 in place, or in this case the updated Bosch 901083 belt. I've reached the physical limit of what I can achieve in this particular build over the course of 9-10 months. If I do a "Stage 2" 400 N-m transmission in the future, it'll include the Bosch 901074 pushbelt with the necessary modifications to make it fit/function.



Forward Clutch Drum Asm3 - Reinforced.JPG Forward Clutch Drum Asm3 - Reinforced2.JPG
 

·
Super Moderator
Joined
·
9,823 Posts
Interesting mod for that failure point. It will hold of course. ;)
 

·
Registered
Joined
·
1,176 Posts
Discussion Starter #274
It should hold 300 lb-ft of torque with good durability. I found 300M Vanadium steel alloy for the raw material, pretty cheap too. My machinist will cut it in the annealed state which is 110,000 psi. It's as strong as 4130 with heat treatment.
 

·
Registered
Joined
·
1,176 Posts
Discussion Starter #275
Figured I'd pop in. Nothing new at this point on the transmission, it's sitting on the bench waiting to be installed. Got my cylinder head under development and engine block waiting to be machined for the upgraded pistons/rods.

The master plan was to build the engine to handle higher rpms so the transmission can survive without going crazy on torque levels. This is how my EVO X is tuned right now, because of the weak factory connecting rods. It's limited to about 325 lb-ft but I'm revving it out occasionally to 7600 rpms to make the approx. 400 h.p. The midrange is mild, but the high rpm rush is pretty sweet. Big mid-range torque is fun......but it breaks rods and transmissions. The Juke will be more fun revving the motor out.......and safer too.

Anyway, the Juke cylinder head has been race ported at this point, upgraded/heavier valve springs and higher duration camshafts on order. The block will be upgraded with forged rods, forged pistons, higher strength head gasket, RS oil pump, RS rod bearings. I'll be switching over to UpRev Tuner so I can tune it myself and bump the rev limiter to 7500 rpms. The engine will be tuned to make about 240-250 lb-ft torque at the 7500 rpm band, so horsepower potential will be around 350 h.p. For now, the Mamba 19T will get me by until I can upgrade the turbo in the future.

Stay tuned.
 

·
Super Moderator
Joined
·
1,573 Posts
Stay tuned.
Haha, I see what you did there!

On a side note, talking with Anthony from Drunkmann Tuning, he said there’s no reason to bump the rev limiter on the engine.

“Also the limiter is around 6900rpm. There is no good rreason to be above 6700rpm the car stop producing power at 6600rpm or so and starts to lose power past that. RPM and power are non-linear are smaller turbocharged cars as the turbo tapers.
6500rpm is more than enough and where the powerband is happy.”
That being said, you have an extremely modified CVT and engine combined with a turbo upgrade so you may see a benefit to increasing the rev limiter.
 

·
Super Moderator
Joined
·
9,823 Posts
Yes he does not know about the springs etc and prolly remembers my valves floating around 7500rpms. I could not hear it but he did.
 

·
Registered
Joined
·
1,176 Posts
Discussion Starter #278
Haha, I see what you did there!

On a side note, talking with Anthony from Drunkmann Tuning, he said there’s no reason to bump the rev limiter on the engine.



That being said, you have an extremely modified CVT and engine combined with a turbo upgrade so you may see a benefit to increasing the rev limiter.
Lol.....stay tuned.....literally.

Well, here's an example. On my EVO X, with the highflow cat, stock compressor inlet pipe, and enclosed airbox (AEM) the car would struggle revving past 6200-6300 rpms. With the 3" testpipe, 3" compressor inlet pipe, and open airbox the engine will rip to the 7600 rev cutout pulling hard if I don't pay attention to the upshift point. The EVO X heads flow much more than a Juke modified head, but the analogy holds I think. On the Juke, the engine upgrades are going to easily let the engine rip to 7500 rpms. Actually what's going to limit me is the rpm limit of the CVT belt. The engine could easily spin to 8000 rpms, the CVT belt would struggle. I'm not too worried, I'll adjust the rev limiter as high as I can reliably go, but it'll be higher than 6900-7000 rpms.

Mac,

Valve float at 7500 rpms.......that's interesting. Thanks for the data point, I was wondering how good the stock springs were. Joe @ 2J basically told me the same thing, spring upgrades were a waste of time for the most part.
 

·
Super Moderator
Joined
·
9,823 Posts
On OEM springs. The stiffer springs won't.

Sent from my LM-G710 using Tapatalk
 

·
Registered
Joined
·
4 Posts
@pboglio

*Correcting the sec pressure with a fixed linear offset will be incorrect as the solenoid pressure vs current profile is not linear
*The maximum clamping is indeed still limited by the system pressure, a stiffer spring will not bring anything as the spring force can be neglected compared to the hydraulic clamping force, however the system might still have some margin.
*Overwriting the sec pressure will potentially decrease ratio stability as the primary pressure should follow the higher secondary pressure, but the system does not know this hence it will need to rely on the ratio PID to compensate for this
ratio as well but only very limited.
*Increasing the input torque and compensating the sec pulley pressure for this might maybe work, but will you do the same for the drive/reverse clutch and TC? Probably the TC control both controls the switching from TC mode to lockup mode and hence simply increase or decreasing the current is unlikely to work. I also assume the adaptive control on the TC only works for the lockup clutch biting point and does not impact the TC switching current.
*fatigue of the pushbelt and variator shafts might become a problem due to the higher clamping.
*Increasing the rev limit will lead to higher centrifugal forces on the prim cylinder parts, this in combination with higher clamping pressure as a result from the reaction (through PID) on the increased sec pressure might overload the pulley pressure cilinder parts.

I came accros this thread by accident, nice project! Would you happen to have a re-build manual which you constructed yourself or a mapping of system pressures vs solenoid currents? Basically any information would be of interest to me. What oil do you use? Where do you source parts, website?

Regards,
Xander
 
261 - 280 of 317 Posts
Top