Alfa Romeo Forums banner
1 - 20 of 32 Posts

·
Registered
Joined
·
549 Posts
Discussion Starter · #1 · (Edited)
Hello all,
I started working with Richard (gprocket) on an electronic Thermostatic Actuator to replace the rebuilt and ever more scarce hydraulic ones in SPICA (all models) equipped Alfas. Besides replacing stock components, the eTA could also be designed (in a final, or one of many iterations) with a selectable curve for tuned SPICA engines. It is a simple, logical, device.

I was working on this project when I was not employed (but was looking) and put together an Arduino Nano based prototype. This prototype took an input from a Bosch Temperature sender mounted to the intake manifold using a custom 3D printed adapter. I worked on improving the design, including trying to reduce the influence of noise caused by (I believe) the old points ignition and distributor. I created a second prototype, which is all but finished.

I realized recently that our world is changing with regards to projects like this. 3D printing and other rapid prototyping technique, along with specialized software, has rendered projects like this doable for even the amateur ‘Millenial' Mechanic. In the spirit of the forum, and of this community in general, I would like to share my work with all the people here so that we may work together towards a common goal. This information is released here under the Creative Commons License.

===========SYSTEM DEFINITION===============
The system as a whole contains the following components:
•A CPU, which includes a ‘brain’ and a chip capable of moving a stepper driver
•A Stepper Motor capable of moving the SPICA fuel/throttle spring/piston. The Motor that I have used in testing is the Haydon Kerk 25000 series 25443-05-912 captive can-stack stepper motor. It is a four pin (bipolar) motor.
•A Thermistor (temperature sensor/sender) that is mounted to the intake manifold. The Thermistor that I have used is the Bosch 0280130026
•An intake manifold adapter for the Thermistor
•A Stepper Motor adapter for the SPICA pump
•ARDUINO code that brings everything together and processes the algorithm for determining what the motor's extensions should be at any given temperature.


=======eTA CIRCUIT DESCRIPTION==============
The CPU circuit I’ve designed consists of the following parts:
An ATMEGA328P microprocessor
An Allegro A4988 Stepper Driver microprocessor
A 12V to 5V step down circuit, using a LM2596S
A 12V to 20V step up circuit to power the stepper motor, using an XL6009

The Allegro A4988 is being fed 20V, but a simple change in a specific resistor can either increase or decrease this value.

Additionally, the circuit is designed so that the ATMEGA328P can be coded with a shut-off timer. Once the boot sequence is complete on this eTA, the microprocessor can maintain itself ON using a micro-Relay, and the 4N35 OPTOISOLATOR circuit. This circuit keeps watch on the ignition signal from the car. The reasoning behind this is so that the eTA can keep going its position as the car cools down. A loss of power would require the motor to cycle back to home and re-extend to the correct point.

The circuit board was designed using EagleCad (a new skill!) and printed by OSH Park. It was designed to be mostly surface mount, so I had to learn how to hot-air solder components! Some through hole components exist, such as the Relay and the Optoisolator socket.

High resolution link:
https://www.dropbox.com/s/7izlt1l8j7vymq6/eTA-board-circuit.pdf?dl=0





=============STEPPER MOTOR================
The Haydon Kerk motor has plenty of power to push on the SPICA throttle/plunger. A drawback to this kind of motor is that it does not really know where it is. This means we apply 20V and expect it to extend to a specific point. If it slips, we wouldn't know. 20V seems to be a good point, where we aren't going to see much slippage. An upgrade to this component would include a way to know where the motor is at all times.

The motor has been augmented with a 6x6x6mm button-type microswitch at its base. The microswitch is fed to the CPU and we can tell when the stepper motor reaches its zero point. At start up, we go through a sequence where the motor retracts until the microswitch is tripped – and then extends out to the proper distance.

==============THERMISTOR=================
The Bosch thermistor is a perfect fit for this application. It fits within the adapter and extends into the intake manifold. Add in the fact that it's an automotive temperature sender and we’re golden. The only work we have to do on the thermistor is a quick calibration. Once hooked up to a CPU, we record the ADC values (resistances) it measures at 0*C, 100*C and somewhere in between. These resistances are then fed into the following calculator, and a new set of coefficients are generated. The code takes advantage of these for the most accurate calibration.
The thermistor uses the EV1 connector, which are actually widely available, so creating a wire-harness is not a difficult endeavor.

SRS Thermistor Calculator

For more information on thermistors, see the following:
Calibrating Thermistor Sensors



===========THERMISTOR ADAPTER==============
The thermistor adapter was created in Solidworks (a new skill!) and 3D printed using my Flashforge Dreamer (a new skill!). I 3D printed a Jig that allowed me to perfectly tap the adapter with a M12X1.50 tap. The STL file for the adapter and Jig can be found below:





https://www.dropbox.com/s/8q4v8kz1c8bkpdk/bosch-flange-1-12-2016.STL?dl=0
https://www.dropbox.com/s/qjaf0bg84ypt9vo/flange-jig.STL?dl=0


===========ARDUINO CODE==============
I’ve uploaded the source code…
https://www.dropbox.com/s/ux3welvz6zqfltg/electronic_TA_March-27.zip?dl=0

I just took Alfa’s published TA extension guides (what people use to troubleshoot), created a function from that curve, and used that function as the basis of this code. The code is in a way, the most complicated portion of the device. It must know whether the key is connected and begin the timer. It must read the temperature, and correctly determine what the extension of the stepper motor should be. For the prototype, it must also output to the LCD.





=========IMAGES of eTA Prototype 2==============

eTA brains inside of 3D printed case:



"under the hood"



Shielded cables with 'overmoulded' style ends




Manifold adapters in different materials (Nylon was tested with prototype 1)


Motor assembly. Note that the motor we used was the longer version - while the shorter version would have worked and fit better. Its ugly for now. I didn't focus on the casing much beyond the molex connector I needed on top.

 

·
Premium Member
Joined
·
6,779 Posts
This is awesome. I think my biggest worry about the project from the engineering view is the reliability of the stepper extension / retaining its position mechanically. To that end a disc with an optoencoder on the back end will satisfy both the need to know where 'home' is and also provide positive feedback that the stepper shaft is turning (which means that the stepper is extending UNLESS the threads are stripped, very low failure rate I would expect).

The locator disc can be quite simple actually and the optoencoder is a LED on side and a photodiode on the other wherein the encoder disk passes through and breaks/opens the light beam

Yet another embodiment might be to add a multi- turn potentiometer and read off the wiper setting via voltage drop with an A/d converter. This assumes there is more than one 360 revolution required to go from full cold to full hot on the stepper. Theres a ton of ways to skin this cat.

A question-- starting from the 'cold' setting -what extension are you setting as 'cold?' -- and how many stepper motor revolutions are required to extend to ?31? mm ie full cold to full hot?

thanks, bob
 

·
Push hard and live
Joined
·
11,617 Posts
Assuming there is a future for this, and I hope there is, you might want to make sure the same unit will fit the Montreal. The TA is essentially the same, except the top of the extension end is at 90 degrees in order to clear the air cleaner assembly. So long as your stepper assembly is as short as the original Montreal TA body, it should be fine. What is your dimension from mounting flange face to the top?
 
  • Like
Reactions: stefmiesterb

·
Registered
Joined
·
1,140 Posts
. . . Theres a ton of ways to skin this cat.

General Bonkers says, "Soldier, secure that $hit. You'll need to find an alternate expression if you want to remain under my command."

To VXer: Just awesome. You've done yourself proud with this multi-discipline development project. With your decision to go open source, it would seem you do not have the time nor desire to take it to a product on your own. Although this is an area that I play in myself, ironically I'm quite the Luddite, hence attraction to old Alfas. My position has been, "Nice project . . but", a tendency to go with the original TA. The prices of these just keep climbing though. Heck, $500 is starting to become real money. This approach is rising on my long list of priorities.

Thank you for your excellent efforts, and of course, sharing it will all of us.
Salute!
Gerry
 

·
Premium Member
Joined
·
6,779 Posts
^^^^^ Gerry that's spot on!!! ^^^^^ Thanks for that!!! Gen. Bonkers not real happy with me but I think going to Board for your promotion!
 

·
Registered
Joined
·
549 Posts
Discussion Starter · #8 ·
Thank you, Gerry - I appreciate the kind words.
I have had some time-crunch issues, but the move to open-source is a personal one. I've created several different 3D printed components that I've shared with the BB (maybe in the Spider forum). Things like a winshield wiper-blade adapter for older wipers, and the cam-cover breather tube for Weber conversions.

With this project, I don't really see a logical end point as a company of one. These things need to be tested, refined. It would take a team to work out all the bugs. I figured that it was best to put this stuff out there!

I know of another project, having to do with an electronic replacement for another component that may have sputtered out. I'm hoping that maybe this can keep going in one intereation or another - even while I'm working on it.

As for the other questions - I will add more detail that should help when I get a chance tonight/this weekend.

Joe
 

·
Registered
Joined
·
319 Posts
I've been using the Shankle Sure-Start's for maybe 3 decades, and for my purposes, it's basically a binary device. Either full-on or full-off works just fine, and I don't see why any intermediate steps are necessary. With the original TA, once the engine is fully warm, the piston is fully extended, and that's all you need. I do the same with the Sure-Start; engine fully warm and I push the knob all the way in. The warm engine temp doesn't change, the warm-up period is just a few minutes, so why is anything more complex needed? Why not a simple 2-position solenoid device?
 

·
Premium Member
Joined
·
24,122 Posts
Talk to Brian Shorey. He has an electro-mechanical one on his Spica Berlina, made in San Diego in the 70s/80s. I have a picture on my other computer, can't recall the maker, but I recollect something "CAT" related. I can find the pic on Monday. Pre-digital, I'd gather.

My own experience is a good rebuilt actuator lasts a heck of a long time and doesn't bleed down much. I drove a Berlina for five years without touching the Spica once set up, and a GTV for 15, same actuator whole time.

Andrew
 

·
Premium Member
Joined
·
5,932 Posts
Andrew. Your are thinking of the "Alfacat" I believe.

Yes, you can use a sure shot, maybe a binary solenoid or spend $400 periodically on a rebuilt TA. But the point of the eTA is to mimic the SPICA TA in function for a fraction of the cost and greater reliability. Joe is very close to achieving these goals.

You really start to appreciate the simple elegance of the TA when you try to replicate it.
 

Attachments

·
Premium Member
Joined
·
6,779 Posts
Dear Howard,

I believe most here are well aware of the shankle device and some probably the tomcat device, and I especially appreciate your through knowledge on said device. But why can't you leave these guys alone and let them try to advance our options? Do you own rights to the shankle device? If so just come out and say so. I believe innovation should be rewarded and celebrated as we have far too little of it in Alfa land in my view.

In other words from me to you, why can't you find it in yourself to offer encouragement ?
 

·
Registered
Joined
·
319 Posts
Dear Howard,

I believe most here are well aware of the shankle device and some probably the tomcat device, and I especially appreciate your through knowledge on said device. But why can't you leave these guys alone and let them try to advance our options? Do you own rights to the shankle device? If so just come out and say so. I believe innovation should be rewarded and celebrated as we have far too little of it in Alfa land in my view.

In other words from me to you, why can't you find it in yourself to offer encouragement ?
You caught me - I'm heir to the Shankle fortune.:grin2:

Seriously, I'm not trying to discourage/encourage anyone from doing as they please, but the original TA, as mentioned, basically operates as an on/off device, so why complicate things? If I still had a Spica car, and someone offered a solenoid-type device to replace a defective TA, that might be an interesting product. An over-complicated, possibly over-priced, not-as-reliable replacement, not so much. How many potential customers do you think exist for such a device, realistically?

That's all. I'll let others continue this discussion and bow out now.
 

·
Premium Member
Joined
·
24,122 Posts
My own experience with Sure Starts, admittedly limited, is that they tend to slip from whatever setting they're at. Seems like the designed spring is stronger than the designed friction to hold a setting.

I buy decent used actuators when I find them, typically $100-150. Plus I like non-digital, but that's just me.

Andrew
 

·
Registered
Joined
·
549 Posts
Discussion Starter · #17 ·
You caught me - I'm heir to the Shankle fortune.:grin2:

Seriously, I'm not trying to discourage/encourage anyone from doing as they please, but the original TA, as mentioned, basically operates as an on/off device, so why complicate things? If I still had a Spica car, and someone offered a solenoid-type device to replace a defective TA, that might be an interesting product. An over-complicated, possibly over-priced, not-as-reliable replacement, not so much. How many potential customers do you think exist for such a device, realistically?
Please note that the original TA is NOT an on/off device. It is not a solenoid-type device and the extension curve over time is designed to help the engine come to temperature correctly, and keep the fuel mixture correct for that engine temperature. This cannot be accomplished with a solenoid. Additionally, you may not see an issue if you live in an area that does not vary greatly in temperature.
 

·
Premium Member
Joined
·
24,122 Posts
Definitely. The system had starting trouble in very cold climates, and there was a mechanical "super-cold start" lever for the Canadian and northern market that moved the mixture richer, by a cable, than was otherwise possible. I got one for giggles, but it's a shelf item, not in-use item, in coastal Calif. Also, I've seen folks plumb the cooling system into the logic section cover to heat it, presumably when they had a block heater installed to warm the coolant. Again, a northern climes/Canadian thing most likely.

Andrew
 

·
Registered
Joined
·
319 Posts
In actual use, I think it is a binary device. I've never seen the water temp gauge on one of my Spica cars have any variance once the engine is fully warmed-up. That stability is the function of the engine's thermostat, and the needle just isn't moving. How much variance have you seen? Maybe with some inadequacy in the cooling system, severe track use, etc. you might see some, but not in a properly-maintained street-driven car. Same apples to the 5 other cars in my garage - I just don't see any variance in their temp gauges (any time of year), and without that variance, the TA isn't changing its piston extension.

If you want to use your device to adjust the mixture curve for other purposes, that's a different circumstance.
 

·
Registered
Joined
·
549 Posts
Discussion Starter · #20 ·
I've never seen the water temp gauge on one of my Spica cars have any variance once the engine is fully warmed-up.
well yeah.. thats the point... If you want to sit in your garage as the car warms up go right ahead! I mean.. its only 75 degrees year round in florida. But when I have to sit at 32F a try to start my Spider w/Webers in the MN cold I empathize with people who also suffer the luxury of seasons.

I can tell you that the space between the 0 and 1 in your binary device can be VERY COLD.

I think all electronic devices can become customizable. Are you still using points? Do you have the 1-2-3 selectable curve distributor (electronic)? An eTA could help w/performance. It could help with emissions testing..

This is just step 1.
 
1 - 20 of 32 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top