[drocketman]

I just ordered a few 1N4371A 2.7v zener diodes. Each diode has a voltage drop of 2.7v which is equivalent to the voltage drop of 3-4 standard diodes when wired in series. Soon I hope to install a couple of these in my failed OEM unit that I removed awhile back. The replacement (I believe it is a Dorman) is already starting to show some beginning signs of oscillation. So I should know pretty quickly whether the zeners did the trick because the old unit was completely failed to oscillation every time I shut the van down.

This is good all around.
If there truly is a software fix that works, that supports my assertion that the software transport time is involved in the instability. [What took'em so long] So the proof will be in the experience what works best. My initial mod is one data point for a hardware fix. And now that there will be several versions of that hardware fix with different forward diode voltage drops, we will be able to see where the sweet spot of bias voltage is and how wide that sweet spot will be. Personnaly, I don't trust their software fixes much anymore.

I just finished the modified actuator for @adavidw (pictured here).
Here is a writeup. More pictures to follow if others are interested.
Initial writeup is below the pictures.
So here are the 4 forward bias voltages being tested or coming out soon:

1. My initial mod, bias ~1.7 volt, modified full bridge + one more diode stage, new aftermarket actuator.
2. @cookid mod, bias about the same (~1.8 volt), three stages 1N4001, new aftermarket doorman actuator.
3. My 2nd mod for @adavidw, bias ~2.4 volt, using a pair of high power LEDs, new OEM actuator identical to the factory installed actuator.
4. @PacDave coming mod, bias ~2.7 volt, using a pair of 1N4371A zeners, rebuilt failed OEM actuator (? new Pot?)

Here is a writeup on #3
My second mod version uses only two components, not the string of diodes so it should be more reliable. The two components are high output and efficient light emitting diodes rather than simple diodes. Each have a forward voltage drop of 2.4 volts instead of the 1.7 V as before with the 3 pairs of diodes. They fit inside the case perfectly see pictures. At a test voltage of 12.87 +/-.01V volts the higher drop slows the actuator down by 26% in both the forward and reverse directions.
The two pictures attached are actually test verification photos. You will notice that one LED is slightly illuminated in each picture. A voltmeter in diode test mode is hooked up to pins 5 and 6 in each photo. One with + to pin 5 and - to pin 6. The other photo is the opposite with + to pin 6 and - to pin 5. This verifies that the diodes are operating properly in forward and reverse motor direction. The motor is not moving during this test because the voltmeter is supplying a very low current to test the diode bias which is not sufficient to drive the motor. However, the diodes are so efficient that they light up slightly, while the voltmeter displays the diode bias voltage (2.4v). Both diodes lighting up about the same and both showing about the same bias voltage verifies the circuit, motor and diodes.
The diodes light up brighter when the motor is running at steady speed (very little load) but the diodes stay cool to the touch. There is a full brightness transient during motor startup that appears as a flash of light. For this reason, I will cover the diode openings so that the light and flash will not be visible during normal operation of the blend door. These LEDs are rated for the full stall current that the motor can produce (as when the output shaft is jammed and kept from turning). But continuous operation in this condition is prevented by the PacHy's blend door drive circuit. In operation (both normal and extreme duty) the LEDs will never even get warm to the touch so they should last forever. I tested the actuator with as frequent start-stops as I could manually switch for 30 seconds or so and I could detect no sensible heat at the LEDs' heat sink surfaces, so no heat sink is necessary. I also tested the LEDs using an actuator stall torque dummy load of 33.6 ohms, (the actuator motor resistance at zero RPM) and it took more than 30 seconds to heat up the LED to the point of needing a heat sink. So this is a safe and robust design.
The initial pot resistances were measured prior to work on the actuator. This allowed me to reassembled it for delivery in the exact same position as received. (mid position of the potentiometer which corresponds to approximately the mid position of the blend door).
Remember from earlier posts not to plug in the connector until the actuator is installed to the blend door. And also that the actuator has three "mounting holes". Two of them are for the screws and the third hole is for the alignment pin. The blend door may flop around without the actuator. So rotating the actuator a little as it's installed may be required to engage the blend door. Once partially engaged with the blend door, the actuator can then be rotated slightly (moving the blend door with it) to align the actuator with its alignment pin and the two screw holes. Once aligned, the actuator will then fully seat with the door and the mounting alignment pin. Then the two screws can be installed. There should be no significant force required to seat the actuator. If properly aligned and inserted, it will pop into place easily. So don't force it. I found out the hard way that aftermarket actuators are really inferior to this high quality Swiss made actuator, and may not fit properly. The OEM Swiss actuator fits perfectly.

 

[drocketman]

BTW, Here is what I used to keep from loosing my screws or socket on install while accommodating the tight space. A little painters tape helps.

 

[adavidw]

This is great! I just picked up my van this morning, but I'm in absolutely no hurry, so test all you want.

 

[BriPac]

This is good all around.
If there truly is a software fix that works, that supports my assertion that the software transport time is involved in the instability. [What took'em so long] So the proof will be in the experience what works best. My initial mod is one data point for a hardware fix. And now that there will be several versions of that hardware fix with different forward diode voltage drops, we will be able to see where the sweet spot of bias voltage is and how wide that sweet spot will be. Personnaly, I don't trust their software fixes much anymore.

I just finished the modified actuator for @adavidw (pictured here).
Here is a writeup. More pictures to follow if others are interested.
Initial writeup is below the pictures.
So here are the 4 forward bias voltages being tested or coming out soon:

1. My initial mod, bias ~1.7 volt, modified full bridge + one more diode stage, new aftermarket actuator.
2. @cookid mod, bias about the same (~1.8 volt), three stages 1N4001, new aftermarket doorman actuator.
3. My 2nd mod for @adavidw, bias ~2.4 volt, using a pair of high power LEDs, new OEM actuator identical to the factory installed actuator.
4. @PacDave coming mod, bias ~2.7 volt, using a pair of 1N4371A zeners, rebuilt failed OEM actuator (? new Pot?)

Here is a writeup on #3
My second mod version uses only two components, not the string of diodes so it should be more reliable. The two components are high output and efficient light emitting diodes rather than simple diodes. Each have a forward voltage drop of 2.4 volts instead of the 1.7 V as before with the 3 pairs of diodes. They fit inside the case perfectly see pictures. At a test voltage of 12.87 +/-.01V volts the higher drop slows the actuator down by 26% in both the forward and reverse directions.
The two pictures attached are actually test verification photos. You will notice that one LED is slightly illuminated in each picture. A voltmeter in diode test mode is hooked up to pins 5 and 6 in each photo. One with + to pin 5 and - to pin 6. The other photo is the opposite with + to pin 6 and - to pin 5. This verifies that the diodes are operating properly in forward and reverse motor direction. The motor is not moving during this test because the voltmeter is supplying a very low current to test the diode bias which is not sufficient to drive the motor. However, the diodes are so efficient that they light up slightly, while the voltmeter displays the diode bias voltage (2.4v). Both diodes lighting up about the same and both showing about the same bias voltage verifies the circuit, motor and diodes.
The diodes light up brighter when the motor is running at steady speed (very little load) but the diodes stay cool to the touch. There is a full brightness transient during motor startup that appears as a flash of light. For this reason, I will cover the diode openings so that the light and flash will not be visible during normal operation of the blend door. These LEDs are rated for the full stall current that the motor can produce (as when the output shaft is jammed and kept from turning). But continuous operation in this condition is prevented by the PacHy's blend door drive circuit. In operation (both normal and extreme duty) the LEDs will never even get warm to the touch so they should last forever. I tested the actuator with as frequent start-stops as I could manually switch for 30 seconds or so and I could detect no sensible heat at the LEDs' heat sink surfaces, so no heat sink is necessary. I also tested the LEDs using an actuator stall torque dummy load of 33.6 ohms, (the actuator motor resistance at zero RPM) and it took more than 30 seconds to heat up the LED to the point of needing a heat sink. So this is a safe and robust design.
The initial pot resistances were measured prior to work on the actuator. This allowed me to reassembled it for delivery in the exact same position as received. (mid position of the potentiometer which corresponds to approximately the mid position of the blend door).
Remember from earlier posts not to plug in the connector until the actuator is installed to the blend door. And also that the actuator has three "mounting holes". Two of them are for the screws and the third hole is for the alignment pin. The blend door may flop around without the actuator. So rotating the actuator a little as it's installed may be required to engage the blend door. Once partially engaged with the blend door, the actuator can then be rotated slightly (moving the blend door with it) to align the actuator with its alignment pin and the two screw holes. Once aligned, the actuator will then fully seat with the door and the mounting alignment pin. Then the two screws can be installed. There should be no significant force required to seat the actuator. If properly aligned and inserted, it will pop into place easily. So don't force it. I found out the hard way that aftermarket actuators are really inferior to this high quality Swiss made actuator, and may not fit properly. The OEM Swiss actuator fits perfectly.

View attachment 60732

View attachment 60735

View attachment 60736

View attachment 60737

Forgive me I didn't pour over this entire post but you did all this just for that actuator?. You obviously went in there and yanked it so your almost done now anyway..
I replaced mine with a $16 one on Amazon 4 years ago and it's been just fine ever since when that's all you really need to do instead of going through all this.. Just my 2 cents..

 

[SitKneelBend]

Forgive me I didn't pour over this entire post but you did all this just for that actuator?. You obviously went in there and yanked it so your almost done now anyway..
I replaced mine with a $16 one on Amazon 4 years ago and it's been just fine ever since when that's all you really need to do instead of going through all this.. Just my 2 cents..

Make no mistake, some people just enjoy tinkering for the sake of tinkering. Improving existing designs is fun.

 

[BriPac]

Make no mistake, some people just enjoy tinkering for the sake of tinkering. Improving existing designs is fun.

Agreed, Tinkerers will inherit the earth...

 

[drocketman]

Forgive me I didn't pour over this entire post but you did all this just for that actuator?. You obviously went in there and yanked it so your almost done now anyway..
I replaced mine with a $16 one on Amazon 4 years ago and it's been just fine ever since when that's all you really need to do instead of going through all this.. Just my 2 cents..

Yea, TMI.
I get that a lot, especially from my kids.
Paraphrased history:
Some vehicles have the problem severely, some do not.
The severe ones are really annoying (6 years for me and two replacements).
Aftermarket actuators are inferior to the OEM, and ironically, poor performance helps damp the oscillation.
The oscillation eats up the actuator sensor Pot making it worse.
Some aftermarket actuators will not even fit mechanically into the blend door.
High voltage of the DC-DC convertor makes the oscillation more probable, so when it drops out after shutdown, the oscillation may stop.
The diode fix shaves off a little of the driver voltage without significantly affecting motor stall torque leaving lots of margin to operate the blend door under all battery conditions.
My 2018 that had the problem severely. I "fixed" it Apr 2025, and traded it in Aug 2025.
The problem has not appeared in my 2023 or my 2025... yet.
I will not let the knuckle draggers touch my software if the problem surfaces with my 2023 or 2025.

Yep, tinkerers will survive.

 

[BriPac]

Yea, TMI.
I get that a lot, especially from my kids.
Paraphrased history:
Some vehicles have the problem severely, some do not.
The severe ones are really annoying (6 years for me and two replacements).
Aftermarket actuators are inferior to the OEM, and ironically, poor performance helps damp the oscillation.
The oscillation eats up the actuator sensor Pot making it worse.
Some aftermarket actuators will not even fit mechanically into the blend door.
High voltage of the DC-DC convertor makes the oscillation more probable, so when it drops out after shutdown, the oscillation may stop.
The diode fix shaves off a little of the driver voltage without significantly affecting motor stall torque leaving lots of margin to operate the blend door under all battery conditions.
My 2018 that had the problem severely. I "fixed" it Apr 2025, and traded it in Aug 2025.
The problem has not appeared in my 2023 or my 2025... yet.
I will not let the knuckle draggers touch my software if the problem surfaces with my 2023 or 2025.

Yep, tinkerers will survive.

Well this proves how weird and electrically connected the the universe is. Days after you sent me your last message & right after I said my $16 aftermarket actuator is still fine after 4 years, it started constantly clicking & oscillating after the van was turned off🤪....Also while on another thread after I said my ESS battery was still fine after 4 years days later I get the Circle A light now it's dead as a doornail🤣...I'll either throw another aftermarket actuator in or spend more then twice as much on an oem hoping it lasts longer.. For right now I just reached back dropped the glove box and disconnected the plug so it doesn't run

Moral of the story the less people talk about their vehicles the longer it will last..😛

 

[holycannoli]

I think the gist of it is friction reduces after break-in as factory lubrication spreads.

Has anyone tested the addition of plastic-safe, high viscosity grease to slow down RPM instead of adding diodes?

 

[SitKneelBend]

Has anyone tested the addition of plastic-safe, high viscosity grease to slow down RPM instead of adding diodes?

You thinking just to pack it with dielectric grease?

 

[holycannoli]

You thinking just to pack it with dielectric grease?

I am going to explore the software update firstly but if I can't find a cooperative technician, I think I'll experiment with progressively larger amounts of grease.

 

[SitKneelBend]

I am going to explore the software update firstly but if I can't find a cooperative technician, I think I'll experiment with progressively larger amounts of grease.

I think I might try that too.

 

[drocketman]

Has anyone tested the addition of plastic-safe, high viscosity grease to slow down RPM instead of adding diodes?

Good point. You may have noticed in the picture of the gears (brand new OEM) in my post 241. grease is visible on the gears. I had taken some of the "displaced" grease and redistributed it to the gears prior to final reassembly of that modified new actuator. That grease is pretty high viscosity (looks like dielectric grease) but still redistributes, being pushed aside as the actuator ages. Here are some pictures of new and failed old actuators that may shed some light on the grease theory.

New OEM actuator (just out of the package), manufactured Dec 06,2024. All the OEMs that I have look exactly like this one, Swiss made, with different manufacture dates.

The above new actuator opened up, having not been installed or run. Notice the grease in the case cover indicates that the grease was in contact with the first and second stage gears, leaving an imprint in the grease. Also notice the grease in the first stage worm gear interface. When the actuator is new like this one, the grease slows the actuator down, but it is still fast with more than 14 volts delivered to the driver from the DC-DC convertor. This new actuator is the one modified and shown in post #241.

Here is the original factory installed OEM actuator manufactured on Nov 22, 2017 (identical to the 2024 OEM above) that began to oscillate in the first year of ownership, accumulated damage to the POT contact, and I replaced in September 2020. See posts starting with #54 in this thread. Notice the grease is now discolored with age and has moved away from the primary worm gear interface. This lets the actuator speed up. Also, when grabbing the actuator output shaft and twisting back and forth, the gear lash is noticeably increased. I believe this is mostly due to the absense of grease in the gear interfaces and not so much a product of gear wear. I removed the second nylon gear for better grease visibility. Notice the shiny brass worm gear with very little grease at the wear points.

Here is the failed OEM actuator that I installed new in Sept 2020. It was also identical to the original Swiss made actuator and was manufactured on Sept 18, 2019. It began to oscillate within a month of installation (see post 63 in this thread). I finally removed it in April 2025 after listening to the oscillation for 3-1/2 years (see post 198). It also accumulated damage to the POT contact, shows grease discoloration and movement from the gear interfaces, and had increased apparent gear lash. I replaced it in April 2025 with a modified aftermarket actuator as described in post 198. Again notice the shiny brass worm gear with very little grease at the wear points.

Bottom line(s):
I believe the grease is already pretty high viscosity. But unfortunately it is eventually mostly cleared from the high speed worm gear set where it, when new, can slow down the actuator speed. However, increasing the grease viscosity may only be a temporary fix and may also thicken up in cold temperatures depending on the grease used. The aftermarket actuators are only about $10 cheaper than the OEM and are inferior in every way so I would not use them again. The root problem is the combination of the high voltage that the DC-DC convertor delivers to its driver circuits and the engineering failure of not matching the driver to the actuator characteristics. I highly doubt that Chrysler will lower the DC-DC output voltage. Instead, I suspect that they considered multiple ways to affect this driver/actuator mismatch, and that the options considered are only software changes. I also suspect that there is not a direct timer or POT limit value setting in the software, so the fix might involve "gaming" the software: Something like changing the vehicle shutdown timing or sequence to stop oscillation early (something that does not prevent the oscillation, just makes it harder for the user to notice). Please consider the history on this forum of Chrysler's track record of Band-Aid software fixes. From catalytic convertor fires caused by software fixes of the Automatic cruise control to battery fire fixes that fail to stop battery fires, to bricking radios, to breaking vehicles with battery software fixes, etc... I do not trust their software fixes if I have an effective work around. The root cause is that they installed a very reliable tried and true Swiss made actuator that has a very wide acceptable operating voltage and is reliably used in countless other vehicles and brands, probably used MOSS type H-bridge drivers that have a very low voltage drop, paired them to a slow response driver algorithm and upped the DC-DC output voltage to compensate for high 12v battery drains late in the design process. All these together made the oscillation a problem. The easiest fix is to shave off a little of the driver voltage at the actuator. So my recommendation: wait until and if the oscillation is noticed, modify and install a new OEM actuator with voltage dropping diodes and go on with your life.

 

[drocketman]

Not sure if this has been posted, but they released a TSB to update the HVAC software for this issue: https://static.nhtsa.gov/odi/tsbs/2025/MC-11019765-0001.pdf

(TSB 08-193-25 for 2022 model year)

Yep, I did find 08-194-25 and 08-195-25 for 2023 and 2024 respectively, but nothing for older models.

Maybe Chrysler is only issuing the 2022, 2023 and 2024 software fixes because:

• There are relatively few 2022 and later occurrences of the problem so fixing them first will reduce their liability for replacement of actuators that have not failed yet.
• They may have already incorporated the fix into 2025 model year vehicles.
• By my count there are 73 owners that have had this problem on this thread alone. The vast majority of failures are on earlier models. It is not hard to imagine that the actual number of failed actuators is much much larger and many have been replaced 2 and 3 times. Since the TSBs may allow for reimbursement to owners who have spent big bucks on a fix, they may be hoping to defer this expense or hoping the problem will just go away as vehicles age out. So a fix for earlier models may be “less financially feasible" to rush out there.
• The 2021 and earlier Pacificas may have significantly different software or may not be as simple to fix as the 2022 and later models.

 

[jeffwb]

They may have already incorporated the fix into 2025 model year vehicles.

If I had to wager on this I'd go against the actuators being 100% fixed in the '25 vehicles. Of the four reasons you listed, only the 4th is legitimate when talking about quality customer service IMHO. Who knows if they are not beta testing the fix itself thus limiting the sample size. But I do think your reasons are well thought out and make sense when considering them.

 

[drocketman]

If I had to wager on this I'd go against the actuators being 100% fixed in the '25 vehicles...

I agree. Time will tell. I only have about 700 miles on the 2025 and only about 13k miles on the 2023. If either starts oscillating, I will do the hardware mod with a new OEM actuator, NOT a software fix.

I've got plenty of 1N4001 diodes and plenty of power LEDs to mod more than a few dozen actuators and I am getting pretty good at non-destructively opening the OEM actuators for modification.

I am moving on to a new project. A 12v battery regulated charge controler with disable if the battery drops below ~10v; Fail safe so no disable possible when running; Manual reset.
12v battery ugly failure modes and 12v battery early failures are two things that have bothered me since 2018 and are overdue for corrective action IMHO. I am convinced that there is no charge control of the 12v battery other than a mild compensation for temperature. There is no battery age compensation. There appears to be a variable in the software recording run time or watt hours since last reset, but it never seems to get reset no matter what procedure one uses to replace the 12v battery. The ground work for tracking battery age appears to be in the software, but there does not seem to be any age compensation code or any charge voltage age compensation being done that I can see. I will start a new thread when I implement it.