2004 to 2020 Mazda 3 Forum and Mazdaspeed 3 Forums banner
1 - 17 of 17 Posts

· Registered
'21 BMW 3-Series
Joined
·
311 Posts
Discussion Starter · #1 ·
I've noticed in the past couple of months a slight jerk between 35-45mph when slowly pressing the accelerator.
It happens both on cruise control and without and seems that cylinder deactivation is not on (just by looking at the animation on the infotainment).

It seems to me like the issue is around the torque converter lock-up as it happens when I slightly accelerate after being on engine brake.

Does anyone experienced something similar to this?
 

· Registered
Joined
·
1,413 Posts
The on/off throttle transition is hard for the engine mgt system to manage; to execute smoothly. The results, in a Mazda, seem to be pretty good. I have other cars, Toyota's, with port injection, with a more ragged transition. One Toyota is an automatic; one is a std... Both are '99's. My Mazda is a '17, Mazda6 2.5, manual transmission. Mine does not have CD. Those automatic Mazda's with CD have some sort of "pendulum" type mechanism associated with the torque converter to supposedly ease the roughness as CD comes-in or goes-out... Could this be malfunctioning, possibly? Then again it may be performing in a normal manner.
 

· Registered
Joined
·
144 Posts
Mine does something similar, but it doesn't seem to be speed-dependent. Minor throttle inputs sometimes cause the torque converter to unlock instead of forcing a downshift (which is what it SHOULD do most of the time). I often watch the tachometer fluttering up-and-down by a few hundred RPMs when driving in gear. To me, Mazda's unique torque converter system with the lockup clutch seems compromised. There's a lot of harshness, sluggishness, and surging in the power delivery.
 

· Registered
2021 Mazda3 Turbo Sedan
Joined
·
46 Posts
It's the converter locking/unlocking/slipping quickly. The non cylinder deactivation and turbo cars also have the same operation but to a lesser degree due to the differneces in the torque converter itself. Pictures are from a non cylinder deactivation car and have a stack up of clutches that operate much smoother. The cylinder deactivation cars do not have room for a set of clutches, just one single clutch surface and it does not handle the one/off operation as smoothly.
Wheel Automotive tire Tire Wood Tread

Wheel Tire Automotive tire Motor vehicle Tread

Light Automotive tire Rim Automotive wheel system Auto part
 

· Registered
Joined
·
1,413 Posts
@sinistriel , that's a good explanation. Have you had one apart? Looks like it.

Your explanation re the reduced amount of room on the CD cars goes 'round with the notion of that pendulum-type additional clutch- (or sprag?) type device taking-up space in the CD cars' flex-disk / bell housing area.
 

· Registered
Joined
·
1,413 Posts
The variability in on/off throttle roughness between CD cars and non CD cars - @sinistriel - you figure is due to lock-up clutch differences... but in all generality -re the operation of the injectors - from dormant (on the over-run) to active (when foot is again applied to the throttle pedal) - could that also figure-in?
 

· Registered
2021 Mazda3 Turbo Sedan
Joined
·
46 Posts
@sinistriel , that's a good explanation. Have you had one apart? Looks like it.

Your explanation re the reduced amount of room on the CD cars goes 'round with the notion of that pendulum-type additional clutch- (or sprag?) type device taking-up space in the CD cars' flex-disk / bell housing area.
I don't have a CD one open right now but these pics are from another site. First pic is where the lockup clutch resides, its underneath that cover vs a stacked drum. Second pic is showing the back of the damper and the pendulum damper is inside. The pendulum mass moves back and forth slightly to take up unwanted vibrations when in 2 cylinder mode.
 

Attachments

· Registered
2021 Mazda3 Turbo Sedan
Joined
·
46 Posts
The variability in on/off throttle roughness between CD cars and non CD cars - @sinistriel - you figure is due to lock-up clutch differences... but in all generality -re the operation of the injectors - from dormant (on the over-run) to active (when foot is again applied to the throttle pedal) - could that also figure-in?
Torque converter operation is mostly to blame, when the car coasts it unlocks the converter, the revs will rise, when throttle is applied the converter locks back up, it causes a fair amount of tach movement. The car is slipping the converter as well when reigniting the dormant cylinders so there is some from that as well.
 

· Registered
Joined
·
1,413 Posts
I don't have a CD one open right now but these pics are from another site. First pic is where the lockup clutch resides, its underneath that cover vs a stacked drum. Second pic is showing the back of the damper and the pendulum damper is inside. The pendulum mass moves back and forth slightly to take up unwanted vibrations when in 2 cylinder mode.
@sinistriel , do you work for or own a transmission shop, where you work on Mazda Skyactiv-AT's? If so, that'd be great for the Forum to know of, as cars with this trans, where serious malfunction occurs, could be rendered Total Losses without an economic means of transmission rebuild. Obviously this all depends on knowledge/experience with the transmission, and price and parts availability... but just 'sayin...
 

· Registered
2021 Mazda3 Turbo Sedan
Joined
·
46 Posts
@sinistriel , do you work for or own a transmission shop, where you work on Mazda Skyactiv-AT's? If so, that'd be great for the Forum to know of, as cars with this trans, where serious malfunction occurs, could be rendered Total Losses without an economic means of transmission rebuild. Obviously this all depends on knowledge/experience with the transmission, and price and parts availability... but just 'sayin...
I'm the service manager/parts procurement/technical reseracher/tuner of a high volume drivetrain repair facility that also does R&D. I am a member of ATSG and ATRA and go to regular schoolings/briefings about everything drivetrain related. I've personally been involved in the automotive industry my entire career starting in a machine shop rebuilding cylinder heads. I'm active in a few of the Mazda specific fb groups. I try and correct info and help people when they genuinely want to learn. We rebuild these transmissions but do not see them regularly as they are one of the best engineered and built units on the market. They do fail but very infrequently. I have an upgraded torque converter I can sell now for some of the supercharged builds that are hitting the market. Just trying to decide how I want to go about it. I'm working with a few vendors on upgraded friction sets but those are further off.

I do have a few youtube videos I put together on a teardown of the medium case version(2.0 and 2.5 non turbo cars) as well as the torque converter I had for the pictures I posted. If you're interested

Part of what makes these transmissions so good is also what makes them expensive to overhaul. You will never be able to cheaply fix them in the event of an actual failure.
 

· Registered
Joined
·
144 Posts
I'm the service manager/parts procurement/technical reseracher/tuner of a high volume drivetrain repair facility that also does R&D. I am a member of ATSG and ATRA and go to regular schoolings/briefings about everything drivetrain related. I've personally been involved in the automotive industry my entire career starting in a machine shop rebuilding cylinder heads. I'm active in a few of the Mazda specific fb groups. I try and correct info and help people when they genuinely want to learn. We rebuild these transmissions but do not see them regularly as they are one of the best engineered and built units on the market. They do fail but very infrequently. I have an upgraded torque converter I can sell now for some of the supercharged builds that are hitting the market. Just trying to decide how I want to go about it. I'm working with a few vendors on upgraded friction sets but those are further off.

I do have a few youtube videos I put together on a teardown of the medium case version(2.0 and 2.5 non turbo cars) as well as the torque converter I had for the pictures I posted. If you're interested

Part of what makes these transmissions so good is also what makes them expensive to overhaul. You will never be able to cheaply fix them in the event of an actual failure.
Fascinating. Thank you for sharing. If there's any way of telling, what's the typical cause of those rare failures?
 

· Registered
2021 Mazda3 Turbo Sedan
Joined
·
46 Posts
Fascinating. Thank you for sharing. If there's any way of telling, what's the typical cause of those rare failures?
We don't really see a common failure due to a design issue like we do on some other transmission families. The transmission in the first video was out of a 2017 CX5 that was used as a rental car for 125k miles. It lived a very hard life and the torque converter failed. The transmission that the torque converter came out of in the second video lost first gear due to abuse. It was out of a 2014 Mazda 3. The customer would not come to a complete stop before shifting from reverse/drive and it broke the one way clutch. We are currently doing a 2015 CX5 that suffered the same fate.
 

Attachments

· Registered
Joined
·
112 Posts
Between 35-40 is a shift point, I know if I'm on the gas between those speeds and say going up a hill it'll up/down shift If I'm right around the 37mph point, it kind of doesn't know where to go depending on how much gas and how much incline it will switch in and out.
 

· Registered
Joined
·
144 Posts
We don't really see a common failure due to a design issue like we do on some other transmission families. The transmission in the first video was out of a 2017 CX5 that was used as a rental car for 125k miles. It lived a very hard life and the torque converter failed. The transmission that the torque converter came out of in the second video lost first gear due to abuse. It was out of a 2014 Mazda 3. The customer would not come to a complete stop before shifting from reverse/drive and it broke the one way clutch. We are currently doing a 2015 CX5 that suffered the same fate.
Hm, now I have proof that it really does matter whether you come to a complete stop. I once saw a former teacher of mine doing this in her CX-5 at about 5mph and it was painful to watch. The thing must have traveled a good yard or two before it finally lurched forward. She was not a hugely pleasant person so maybe the Mazda will repay her someday.
 
  • Like
Reactions: brassrox
1 - 17 of 17 Posts
Top