Fan Stuff Again...Amplifier Issue And Temporary Remedy
06-25-2014, 04:07 PM
#31
Electron Wrangler
Lifetime Rennlist
Member
Lifetime Rennlist
Member
High voltage from the alternator is possible if the regulator was damaged...But you also noted very high voltage even in the battery only case - this is not feasible with any loading (like the ignition circuits on).
19.2V would cause significant lifetime reductions on basically everything in the car - you would rapidly fry circuits... so make sure its not real - cheap insurance to buy another DMM at Harbor Freight.
Alan
06-25-2014, 04:09 PM
#32
Electron Wrangler
Lifetime Rennlist
Member
Lifetime Rennlist
Member
"I thought that the final stage fan unit was an amplifier?"
Good point...most decent consumer grade stereos are using MOSFET controlled amps. There are some good sized capacitors in the final stage controller. I noticed my original final stage doesn't spin the right fan at all. Putting in a used final stage I saw the right fan try to start. Perhaps the capacitors failing can be another reason a fan won't turn. I'm guessing the capacitors deliver a higher voltage and current needed to start the fans??? Makes sense in terms of simple physics - it's like mashing the accelerator to get up to speed (over come inertia when fan is at rest), then nominal system voltage is enough to keep 'em spinning???
Good point...most decent consumer grade stereos are using MOSFET controlled amps. There are some good sized capacitors in the final stage controller. I noticed my original final stage doesn't spin the right fan at all. Putting in a used final stage I saw the right fan try to start. Perhaps the capacitors failing can be another reason a fan won't turn. I'm guessing the capacitors deliver a higher voltage and current needed to start the fans??? Makes sense in terms of simple physics - it's like mashing the accelerator to get up to speed (over come inertia when fan is at rest), then nominal system voltage is enough to keep 'em spinning???
Alan
06-25-2014, 07:00 PM
#34
Electron Wrangler
Lifetime Rennlist
Member
Lifetime Rennlist
Member
My real concern is in having the fans bypassed from the controller and amplifier circuit (lines to the fans isolated, disconnected, and protected), will there be any effect(s) on the AC circuitry for some components of the AC system are tied into the fan/cooling circuitry ?
Gurus...?
Gurus...?
Alan
06-25-2014, 07:12 PM
#35
Nordschleife Master
Thread Starter
Alan,
Is Pin 7 of Plug II of the controller the output signal line from the controller that feeds Pin 2 on the amplifier ?
And, thanks for that info. above.
Is Pin 7 of Plug II of the controller the output signal line from the controller that feeds Pin 2 on the amplifier ?
And, thanks for that info. above.
06-25-2014, 08:36 PM
#36
Electron Wrangler
Lifetime Rennlist
Member
Lifetime Rennlist
Member
Yes it feeds final stage pin 2 - but be aware it is a PWM signal - you cannot use it to trigger a relay - at least not directly.
Alan
Alan
06-25-2014, 08:47 PM
#37
Nordschleife Master
Thread Starter
If I wanted to trigger a relay, what is involved ?
I'm not trusting in an amplifier with the FET at 26A and the fuse at 30.
One always hears their failures.
06-25-2014, 09:08 PM
#38
Electron Wrangler
Lifetime Rennlist
Member
Lifetime Rennlist
Member
PWM = Pulse Width Modulated. It cycles on & off very rapidly with a varying duty cycle - so if you feed a relay coil with this the coil will likely buzz (and also become damaged quite quickly). In steady state full speed the signal is probably continuous, but certainly not at half speed.
There is no simple solution to this - I even wonder if the controllers output is capable of driving a relay coil (that is likely ~100mA which is likely a lot compared to the final stage input requirements, and you also probably need 2 relays...). For temporary summer duty it may be easiest to just run them all the time based on the ignition (minus while starting)...
Alan
There is no simple solution to this - I even wonder if the controllers output is capable of driving a relay coil (that is likely ~100mA which is likely a lot compared to the final stage input requirements, and you also probably need 2 relays...). For temporary summer duty it may be easiest to just run them all the time based on the ignition (minus while starting)...
Alan
06-25-2014, 09:56 PM
#40
Former Sponsor
Not in this sense... I guess you could say its a current amplifier. At the kinds of currents this sinks you would need a very large & fancy circuit to increase the voltage too (very large!)
High voltage from the alternator is possible if the regulator was damaged...But you also noted very high voltage even in the battery only case - this is not feasible with any loading (like the ignition circuits on).
19.2V would cause significant lifetime reductions on basically everything in the car - you would rapidly fry circuits... so make sure its not real - cheap insurance to buy another DMM at Harbor Freight.
Alan
High voltage from the alternator is possible if the regulator was damaged...But you also noted very high voltage even in the battery only case - this is not feasible with any loading (like the ignition circuits on).
19.2V would cause significant lifetime reductions on basically everything in the car - you would rapidly fry circuits... so make sure its not real - cheap insurance to buy another DMM at Harbor Freight.
Alan
I've always called this unit a fan amplifier, because it has the ability to make one fan run at a higher speed if one of the fans is bad (or run both fans at a higher speed, if the car gets too hot.)
Have I been fooled, all these years, into thinking the voltage to the fans is lower than 12 volts when they are running at their "normal speed" and the voltage is 12 volts to the fans when they are running at high speeds?
Seems like Porsche could have done this with a resistor, like they did on the 924 and the 944 models, if this was the case.
Maybe I should call this unit the "fan resistor unit", instead of a "fan amplifier unit"?
06-25-2014, 10:06 PM
#41
Nordschleife Master
Thread Starter
I'm confused, as always.
I've always called this unit a fan amplifier, because it has the ability to make one fan run at a higher speed if one of the fans is bad (or run both fans at a higher speed, if the car gets too hot.)
Have I been fooled, all these years, into thinking the voltage to the fans is lower than 12 volts when they are running at their "normal speed" and the voltage is 12 volts to the fans when they are running at high speeds?
Seems like Porsche could have done this with a resistor, like they did on the 924 and the 944 models, if this was the case.
Maybe I should call this unit the "fan resistor unit", instead of a "fan amplifier unit"?
I've always called this unit a fan amplifier, because it has the ability to make one fan run at a higher speed if one of the fans is bad (or run both fans at a higher speed, if the car gets too hot.)
Have I been fooled, all these years, into thinking the voltage to the fans is lower than 12 volts when they are running at their "normal speed" and the voltage is 12 volts to the fans when they are running at high speeds?
Seems like Porsche could have done this with a resistor, like they did on the 924 and the 944 models, if this was the case.
Maybe I should call this unit the "fan resistor unit", instead of a "fan amplifier unit"?
I measured the control signal from amplifier to controller at the specified engine temperature for the test, 175 deg F. The normal value according to the WSM is 7v.
Amplifier is actually a resistor.
06-26-2014, 12:08 AM
#42
Electron Wrangler
Lifetime Rennlist
Member
Lifetime Rennlist
Member
Well no its not a resistor or an amplifier.
I can see why they might use the name but its not really at all accurate.
Fan Final Stage is an accurate description. The controller in the cabin is the low power control stuff and the final stage is the high current drive stages located close the the fans.
The feedback mechanisms from the final stage to the controller allow it to modulate the drive on Pin 7 (pin 2 on the final stage) in case either fan fails (e.g. go to full speed mode) OR to just go to high speed for more cooling. The controller also handles the periodic restart/retry mechanism.
The final stage never works like a resistor - that would dissipate way too much power - it works more like a very high speed switch.
The voltage you will measure with a DMM will be something different than the actual peak voltage of the waveform. The waveform is ~a square wave with varing duty cycle. Full speed has the output stage always on so ~12V all the time - duty cycle 100%). In the slow fan speed (normal) case the duty cycle is approx 50%. Since it is changing so fast between 12V & 0V you will get a DMM reading of a sort of average of the voltage over time (depending on the duty cycle). You'd need an oscilloscope to see what is really happening.
The fans also see this waveform so they actually still see 12V but not all the time - only about half the time (there is likely some smoothing here too), and so it acts as a lower effective average voltage in the low speed mode.
Now since the actual drive stage is generally either on or off at any given time point, and its on resistance & voltage drop is very low - it dissipates very little power relative to the power it controls. In fact most power is dissipated in the very short periods when it is transitioning from off to on and when it is transitioning from on to off. Its a fairly clever design for this usage model - but perhaps a bit under rated. Today faster switching, lower on resistance and higher current/power rated FET parts are now available.
I do think mechanical issues over time (and heat cycles) are as much to blame as the original design e.g. case seal compromised -> corrosion, thermal conduction path damaged due to aged/corroded mechanical attachments and long term breakdown of the thermal conduction media. General semiconductor wearout mechanisms are also in play - high current, high operating temperatures and long operating life (the oldest of these are now well over 25 years old) all contribute to a difficult environment.
Over temperature of the FET's will cause faster wearout and damaged/worn fans that consume too much power will accellerate this.
Alan
I can see why they might use the name but its not really at all accurate.
Fan Final Stage is an accurate description. The controller in the cabin is the low power control stuff and the final stage is the high current drive stages located close the the fans.
The feedback mechanisms from the final stage to the controller allow it to modulate the drive on Pin 7 (pin 2 on the final stage) in case either fan fails (e.g. go to full speed mode) OR to just go to high speed for more cooling. The controller also handles the periodic restart/retry mechanism.
The final stage never works like a resistor - that would dissipate way too much power - it works more like a very high speed switch.
The voltage you will measure with a DMM will be something different than the actual peak voltage of the waveform. The waveform is ~a square wave with varing duty cycle. Full speed has the output stage always on so ~12V all the time - duty cycle 100%). In the slow fan speed (normal) case the duty cycle is approx 50%. Since it is changing so fast between 12V & 0V you will get a DMM reading of a sort of average of the voltage over time (depending on the duty cycle). You'd need an oscilloscope to see what is really happening.
The fans also see this waveform so they actually still see 12V but not all the time - only about half the time (there is likely some smoothing here too), and so it acts as a lower effective average voltage in the low speed mode.
Now since the actual drive stage is generally either on or off at any given time point, and its on resistance & voltage drop is very low - it dissipates very little power relative to the power it controls. In fact most power is dissipated in the very short periods when it is transitioning from off to on and when it is transitioning from on to off. Its a fairly clever design for this usage model - but perhaps a bit under rated. Today faster switching, lower on resistance and higher current/power rated FET parts are now available.
I do think mechanical issues over time (and heat cycles) are as much to blame as the original design e.g. case seal compromised -> corrosion, thermal conduction path damaged due to aged/corroded mechanical attachments and long term breakdown of the thermal conduction media. General semiconductor wearout mechanisms are also in play - high current, high operating temperatures and long operating life (the oldest of these are now well over 25 years old) all contribute to a difficult environment.
Over temperature of the FET's will cause faster wearout and damaged/worn fans that consume too much power will accellerate this.
Alan
06-26-2014, 01:15 AM
#43
Three Wheelin'
Join Date: Feb 2004
Location: Jacksonville and sometimes St. Aug Beach, FL
Posts: 1,727
Received 342 Likes
on
171 Posts
FWIW, I struggled with this issue and ended up swapping my fans for the Dorman units. Optimal solution in my scenario and I even have two spare final stages 
06-26-2014, 01:30 AM
#44
Nordschleife Master
Thread Starter
06-26-2014, 09:03 AM
#45
Team Owner
