Controller for rear spoiler
03-02-2019, 02:21 PM
#61
Rennlist Member
I think the error in the extend frequency is due to the fact that the speed signal input is not a 50% duty cycle square wave. It is a pulse stream made up of a 2.1 ms "off" time with a variable "on" time. See the pics of the speed signal at 40 mph's and at 60 mph's. This would effect the Cx/Rx charge coming out of the MC14538 chip on the module. See hand-written schematic of the module - notes are mine. Components are in similar position as on the module. Let me know if you see something not right and I'll correct it!
I can’t set the duty cycle on my signal generator, but was having a play with the amplitude of the signal and reducing it definitely has an impact. That said I’m struggling to understand why the duty cycle changes with speed. It should just be the frequency.
03-02-2019, 03:30 PM
#62
As long as your freq generator doesn't generate a duty cycle signal which results in a pulse width less than the minimum, it'll be adequate. Ideally, a freq generator
with a constant pulse width would be ideal, it's not necessary.
03-02-2019, 03:45 PM
#63
Rennlist Member
You guys are correct, it's the freq. of the positive edge only and the duty cycle has no significance - I looked at the data sheet for the MC14538 and it's all coming back to me!
The 2 pulse streams where taken by a pc based oscilloscope connected to the module while I was driving the car on a straight section of road
So the problem with your warning light latched on could be a shorted transistor that drives the warning light .
The 2 pulse streams where taken by a pc based oscilloscope connected to the module while I was driving the car on a straight section of road
So the problem with your warning light latched on could be a shorted transistor that drives the warning light .
03-02-2019, 03:46 PM
#64
Rennlist Member
Remember, the speed signal just requires a basic trigger pulse at some minimum pulse width with a varying frequency, i.e. there's no measurement of a duty cycle. As long as your freq generator doesn't generate a duty cycle signal which results in a pulse width less than the minimum, it'll be adequate. Ideally, a freq generator
with a constant pulse width would be ideal, it's not necessary.
with a constant pulse width would be ideal, it's not necessary.
BTW to anyone reading this, quite why this matters to me is not clear, even to me. I guess I just like order in my life.
Last edited by John McM; 03-02-2019 at 04:10 PM.
03-02-2019, 04:25 PM
#65
Remember, the units you have utilizes an analog design, i.e. it has +/- 5/10% components. The later module is a digital design.
03-05-2019, 07:39 PM
#66
Rennlist Member
Originally Posted by Lorenfb
Remember, the units you have utilizes an analog design, i.e. it has +/- 5/10% components. The later module is a digital design.
03-05-2019, 07:42 PM
#67
Rennlist Member
Originally Posted by mojorizing
I think the error in the extend frequency is due to the fact that the speed signal input is not a 50% duty cycle square wave. It is a pulse stream made up of a 2.1 ms "off" time with a variable "on" time. See the pics of the speed signal at 40 mph's and at 60 mph's. This would effect the Cx/Rx charge coming out of the MC14538 chip on the module. See hand-written schematic of the module - notes are mine. Components are in similar position as on the module. Let me know if you see something not right and I'll correct it!
03-14-2019, 04:50 AM
#68
Rennlist Member
First spoiler controller fixed. Two faults. 1. The switch transistor for the up relay was faulty where it would switch on, but wouldn't switch off. A dangerous fault as it means the motor continues to drive. 2. The warning light transistor was faulty.
Thanks to the schematic diagram it was easier to fault find than it would have otherwise been. Great collaboration.
Thanks to the schematic diagram it was easier to fault find than it would have otherwise been. Great collaboration.
03-14-2019, 05:56 AM
#69
Rennlist Member
Nice!
The varistor (big yellow disc) takes care of overcurrent shutdown, so this is not a critical fault, but puts an unnecessary high stall load on the motor before the varistor disconnects.
Cheers,
Tore
The varistor (big yellow disc) takes care of overcurrent shutdown, so this is not a critical fault, but puts an unnecessary high stall load on the motor before the varistor disconnects.
Cheers,
Tore
03-14-2019, 03:38 PM
#70
Rennlist Member
The schematic is written up using https://www.expresspcb.com/expresspcb/ if you want to download the file and edit the circuit directly it's on my dropbox HERE
The schematic for the digital version of the module was written out using https://diptrace.com/download/download-diptrace/ and the file is HERE Traces have my shorthand naming scheme....Rennlist member RonCT generously loaned out his digital module to me to figure out how to accomplish the RUF code modification to the extend/ retract speeds on the microprocessor. I'd like to find out how they accomplished this because the processor is obsolete, 4 bit, one-time programmable. I figure you'd have to remove the processor and put in a add-on circuit that has a current chip. For various reason I didn't pursue that, but the main reason was the fact that original porsche components, modules, etc have too much value now, and I still have about 8 of the digital modules of my design stuck away somewhere - BOM cost was around $50 for those.
The schematic for the digital version of the module was written out using https://diptrace.com/download/download-diptrace/ and the file is HERE Traces have my shorthand naming scheme....Rennlist member RonCT generously loaned out his digital module to me to figure out how to accomplish the RUF code modification to the extend/ retract speeds on the microprocessor. I'd like to find out how they accomplished this because the processor is obsolete, 4 bit, one-time programmable. I figure you'd have to remove the processor and put in a add-on circuit that has a current chip. For various reason I didn't pursue that, but the main reason was the fact that original porsche components, modules, etc have too much value now, and I still have about 8 of the digital modules of my design stuck away somewhere - BOM cost was around $50 for those.
Last edited by mojorizing; 03-14-2019 at 04:04 PM.
03-15-2019, 12:50 AM
#71
Rennlist Member
Originally Posted by mojorizing
The schematic is written up using https://www.expresspcb.com/expresspcb/ if you want to download the file and edit the circuit directly it's on my dropbox HERE
The schematic for the digital version of the module was written out using https://diptrace.com/download/download-diptrace/ and the file is HERE Traces have my shorthand naming scheme....Rennlist member RonCT generously loaned out his digital module to me to figure out how to accomplish the RUF code modification to the extend/ retract speeds on the microprocessor. I'd like to find out how they accomplished this because the processor is obsolete, 4 bit, one-time programmable. I figure you'd have to remove the processor and put in a add-on circuit that has a current chip. For various reason I didn't pursue that, but the main reason was the fact that original porsche components, modules, etc have too much value now, and I still have about 8 of the digital modules of my design stuck away somewhere - BOM cost was around $50 for those.
The schematic for the digital version of the module was written out using https://diptrace.com/download/download-diptrace/ and the file is HERE Traces have my shorthand naming scheme....Rennlist member RonCT generously loaned out his digital module to me to figure out how to accomplish the RUF code modification to the extend/ retract speeds on the microprocessor. I'd like to find out how they accomplished this because the processor is obsolete, 4 bit, one-time programmable. I figure you'd have to remove the processor and put in a add-on circuit that has a current chip. For various reason I didn't pursue that, but the main reason was the fact that original porsche components, modules, etc have too much value now, and I still have about 8 of the digital modules of my design stuck away somewhere - BOM cost was around $50 for those.
I fixed the second controller. It was another faulty transistor. This time the warning light BC337. If there is a next time it would be the first thing I'd replace. It's cheap and easy to do.
03-15-2019, 12:49 PM
#72
Rennlist member RonCT generously loaned out his digital module to me to figure out how to accomplish the RUF code modification to the extend/ retract speeds on the microprocessor. I'd like to find out how they accomplished this because the processor is obsolete, 4 bit, one-time programmable.
The last 4 bit microprocessor available/used in designs was in the mid to late '70s, e.g. Intel 4004. Remember, the 911 3.2 (intro 1984)
DME ECM was probably designed in about '82-'83 and used an Intel 8 bit controller.
2. RUF most likely had the manufacturer of the digital version, who had a source for the uP chip, revise the source code and then
supply the new chip to RUF or have that manufacturer update the units RUF supplied. It's also possible that RUF obtained the
source code and bought some old stock of the original chip and programmed them for the new up/down settings, but unlikely.
03-15-2019, 03:24 PM
#74
Rennlist Member
The unit also has timer functions that will shut off the motor when the spoiler is expected to be fully up or down.
Cheers,
Tore
Cheers,
Tore
03-15-2019, 03:58 PM
#75
Rennlist Member
On the matter of timing, it might be good to add to your arsenal of tools is a Capacitor checker to verify that the cap is within spec. with regards to leakage and ESR not only capacitance.