ToreB

Hi folks.
I've digged deep into the HVAC servos lately, I have been very curious on the real cause of the burn marks and failing of these electric devices.
A typical burnt device is shown below, and I've been intrigued by the damage and obvious currents involved, thinking it was the potentiometer. But, such high currents are not part of the pot at all!
The servo unit have a large area of rotational feedback system with printed carbon tracks. One half is the potentiometer, (a varable resistor) and the other half is a rotating limit switch. The motor current runs through these printed tracks, (no resistor, conducting tracks) and shut down the motor if the defined maximum travel is reached. I would guess this is a safety measure if the CCU servo control should fail and try to run the servo beyond it's limits.
These tracks can't cope with the motor current, especially if the connected mechanical flaps or maneuver arms runs a bit sluggish.

A failing servo could then probably be revived by bypassing the limit switch, a mere short of the two diodes. This is especially interesting since the price of these servos are now rocketing.

I have no candidates for testing this yet, but I'm sure there's a lot of people with failing servos around. Anyone up to this and have a soldering iron available?

EDIT: I've digged deeper into this now, and it is NOT recommended to short the diodes to disable the end swithes. The Climate Control unit rely on these to set the servos to the extreme positions. The only lasting remedy to burnt servos is to change the PCB and electric motor.

Cheers,
Tore

Typical servo failure, a fix would be to short the two smaller solder pads near the white gear, top left corner:


My schematic of the servo:

993/907

Hi Tore.
I have (2) failed servos for drivers side footwell; neither of them have the burn marks in your pic, so suspect they are not getting good signal from the CCU?
I can dink with them and get occasional/intermittent action from one; the other is non-responsive.
Gordo

ToreB

If a whack to the servo makes it work, I would think that it's got weared potentiometer or limit switch tracks, or a failing motor. You could try to bend the whiskers out a bit to get a higher contact pressure to the tracks.
I don't understand your note on having two servos in the driver's side the footwell? There's a servo in both left and right footwells in the 993, behind the outer carpet walls. These are the mixer servos, and are driven by separate outputs in the CCU. The most common problem is failing servos, but I have seen some cases of failing CCU servo drivers as well.
Have you tried a new servo in the two mixer positions? If still inert it's a CCU failure, perfectly repairable. See my CCU details page on www.porschehvac.bergvill.com
Cheers,
Tore

993/907

Hi Tore....that was a misleading statement on my part.
I have the original (failed) servo and bought a used spare servo off this board.
The spare is intermittent, and I get no response from the original when I slide the CCU controls.
I have tried bending the whiskers (great description!) and cleaned all the contact points.
Winter is coming fast, and I have the mixer flap set in open position for now, gets me by.
Gordo

x50type

Torre, Is this one too far gone?

ToreB

As far as I can judge by my findings, no burnt servos could be too far gone. The limit switches is a safety feature one can decide to live without by shorting the two diodes as described.
On second thoughts, there's a risk of damage to the climate control unit if the CCU try to run the servo beyond it's mechanical limits. Any connected rods or flaps could stop the movement due to mechanical constraints, and there would be a risk of frying the CCU output drivers due to increased motor current when stalling.
This could be solved with a fuse on the #4 or #5 wire to the servo. A low price alternative to buying a new servo.
Cheers,
Tore

IXLR8

Or if something is stuck.

Tore, have you measured the motor currents upon start-up with no resistance on the motor and also with slight pressure on the motor?

x50type

The 5 servo units [at $300+ each] and the CCU [at $2000] in the HVAC system are now getting old and eventually will no longer work .
I am thinking about not renewing the left and right foot well servos when they let go, and instead installing short extension rods through the carpeting and afixing hand controls ***** so they can be opened and closed manually.

Possibly the CCU could be removed and simple switches used to turn things off and on. In this case there would be no dial up temperatures and it clearly would not be automatic - just a much simpler system as we had in the 'olden days'...... hot, cold, on, off.
Possible?

Torre,
Looking from the front of the 993, can you let me know what each of the 3 servos do?

Lorenfb

"A failing servo could then probably be revived by bypassing the limit switch, a mere short of the two diodes."

That is a minor problem and rare occurrence. The majority of the servo
motors fail because of commutator wear and its contact (essentially the
brush in a conventional motor).

ToreB

Loren, I've never seen any failing servo motors, but must have seen at least 150 servos with burnt limit switches. To me, that's pointing to what's the most common fault in these servos.
You may have other evidence than me on this?
Cheers,
Tore

x50type

Agree, ToreB,

The tiny motor is not the problem [they probably cost a few dollars wholesale.]

The weak link is the circuit board/whisker interface IMHO.

993/907

"I am thinking about not renewing the left and right foot well servos when they let go, and instead installing short extension rods through the carpeting and afixing hand controls ***** so they can be opened and closed manually."

That's my current set up at driver's side footwell....I left carpeted cover off and have the flapper housing oriented so I can simply reach down and open/close using my finger inserted in the star shaped indent. Get a bit of a whistle when closed though, as the flapper is forced open just a little by the pressurized duct.

gordo

ToreB

X50type, my HVAC page contains most details on the servos. Pictures, explanation and all...

I'm still thinking on this end switch delete. I would guess the Porsche engineers constructed the CCU to not exceed the limits of the servos, thus tripping the limit switch. If so, it would produce a fault code in the CCU in normal operation. I am almost certain they would not let this happen. Makes me even more certain that shorting/disabling the limit switches will not hurt anything at all. I've seen very few cases of servo fault codes in the CCU unless the servos are about to give up the ghost. (motor signal interrupted by the damaged limit switch) I will start gathering damaged HVAC servo's, that's for sure.

Cheers,
Tore

ToreB

That's an Interesting question Alex.
The unloaded motor current is about 80mA, rising to about 430mA at full stall. The TCA 2465 servo drivers inside the CCU are cabable delivering of 2.5A, so we could be well below this. The datasheet also mention short circuit and thermal protection, but you can't always rely on that.
I've seen TCA's that have been very hot inside the CCU, but do not know the cause. A fully stalled motor could heat up things if the internal protection does not work as fast as it should.
Cheers,
Tore

Lorenfb

"I've never seen any failing servo motors, but must have seen at least 150 servos with burnt limit switches"

Well that's great, but you obviously haven't seen the typical failure here
in the USA and haven't read many of the posts where Rennlist members
have opened the motors and found bad burnt commutators.

Bottom line:
If repairing the motors were that simple, someone would have easily
figured it out and the replacement cost for used motors wouldn't be so
costly. (Basic Economics)

"I've seen TCA's that have been very hot inside the CCU, but do not know the cause."

It's simple. That's what happens when the commutator fails, i.e. the motor fails to turn
and acts as a basic short to the driver.

Read here under 'Generic' for more info: http://www.systemsc.com/problems.htm