forbiddenera

First, let me be clear in that I'm talking about electronic actuation *NOT* electronic influence of pneumatic actuation or an EBC.

So, as some may know and most probably don't, I got a $50 TurboSmart ultragate38. As I suspected, it's a replica, and the diaphragm leaked from day 1. The seller (local guy) offered me a refund if it didn't work which he kept his word on. He didn't want the gate back. So I got a useless gate and my money back.

I thought about it and I have decided to modify this wastegate for electronic actuation. It should be relatively simple to convert the PWM output to something usable for this.

I picked up a pretty beefy 2" stroke, 0.5"/s actuator, mostly because it's usually $170 at Princess Auto and it was on sale for $55. I also got a linear actuator controller free that PA sells, it just has two buttons for each direction and I'm not even sure if it's outputs are compatible with the actuator but it was free so, meh . Not that I plan on controlling my WG with 2 buttons, my Civic isn't a 1940s airplane.

0.5"/s seems a touch slow maybe but should be sufficient for testing. PA will take the actuator back even after I test with it ("No sale is ever final", they return anything/everything even if its used) so I'm not really at a loss. 0.5"/s may almost be enough though. I'd estimate the stroke of the WG to be maybe 1"

The plan is to make a mount for the actuator and attach it to the top of the WG valve somehow. I haven't exactly worked out the mechanism but shouldn't be difficult.

The next step will be control, my ECU already has the familiar PWM boost output, so I'll use a micro to convert this to the appropriate drive the actuator requires.

If anyone wants to help or is interested, let me know.

Some of the possible benefits: able to hold WG closed at any boost. Able to open the WG fully on command.

Possible negatives: control too slow (different actuator would solve this), potential to fail in any position unless sprung (again, havent worked out the mechanism, first proto type will likely be unsprung bUT id eventually like this to fail open).

Once I build/test the prototype and it's functional, I plan on refining this into some sort of either DIY kit or a full blown electronic waste gate that could be sold aftermarket.

Anyone interested in helping in any way? Here's some pics of the WG, I'll upload pics of the actuator shortly.

forbiddenera

First Actuator.

2kdrift

good luck with this, keep us updated.
This is something I have thought about in my head (just the idea not actually the individual components which you have provided), the advantages I had thought of is without relying on a spring it would have the benefit of not having min/max boost limitations and also keeping the gate completely closed until a precise moment.

there has to be a reason why this is not commercially available cant wait to see results

forbiddenera

Some cars have this type of waste gate control stock, so it is commercially available.

As an aftermarket part, I'm surprised it doesn't exist but that's probably because of a few things:
Boost controllers, manual or electronic, simply control pressure to the diaphragm of a wastegate, so aside from the wastegate and actuator, you need a control system and that control system has to easily integrate with existing systems and ecus.

Simplicity, a diaphragm waste gate is simpler (and thus probably more reliable) in parts, though this makes me wonder about the prices of them. I understand R&D and materials, but we're talking $300 or more for relatively simple pneumatic valve.

Fail safe, I suppose this is moot as most waste gates fail closed but without a spring it could fail at any point open or closed.

I eventually intend on creating a custom actuator/solenoid for this purpose with a possible floating/magnetic spring (to lessen required force and fail open), I have spent about a year engineering a cam less valve actuation system on a b series head, this was a few years ago and I still plan on going forward with that project eventually although it will require a custom actuator, whereas this I can at least prototype with an off the shelf actuator.

Really I think the main reason we haven't seen this in the aftermarket is simply simplicity. Now if I (we? Anyone want to help?) can develop my prototype into a self-contained boost control/waste gate solution that can take a typical PWM input for a boost control solenoid and/or has a boost reference input and actuate the valve well, I think it may be a marketable and useful product.

Could also work as a boost controller for cars that have none, or to replace the oem system. Simply hook up the boost reference line to one of the pressure sensors and program the module for desired boost. Could alsI have RPM and VSS input for gear and RPM based boost.

forbiddenera

Let's organize this into steps.

Step 1: Acquire and turbocharge a vehicle for testing. DONE.
Step 2: Acquire a crappy broken replica wastegate for testing. Why mangle something that actually costs money/already works? DONE.
Step 3: Acquire a roughly suitable linear actuator for testing. DONE.
Step 4: Acquire a suitable programmable microcontroller, ADC, map/pressure sensor, driver module and powersupply. IN PROGRESS.
Step 4a. Test microcontroller control of solenoid. NOT STARTED.
Step 5: Program microcontroller to accept a PWM input normally for a boost solenoid. NOT STARTED.
Step 6: Connect map sensor to analog input, make plug for connecting VSS, TPS and tach to an input. NOT STARTED.
Step 7: Program microcontroller to control linear actuator, either using PWM input from ECU or EBC or 0-5v reference from the pressure sensor as input against user programmable tables which can be based on gear or rpm or throttle or any/all of the above. NOT STARTED.
Step 8: Design a way to program the options/tables on the microcontroller, I am thinking of incorporating a Bluetooth module and having the programming done via an app on a phone/tablet, simple and cheap but also sexy and could data log map, tps, VSS, RPM and heck, any obd2 thing if we decided to tap the obd2 bus for input instead of just having analog inputs. NOT STARTED BUT BASIC DESIGN PLANNED.

I want to keep this simple but functional:

If you have existing EBC, you should be able to bolt the gate on, hook up your solenoid lines and boost reference to the gate and pair the gate, open the app, set the gate to EBC control mode, which will allow you to specify a spring rate to simulate.

Otherwise, or if you want more control, gate is bolted on, boost reference is hooked up and optionally OBD2 or analog TPS/RPM/VSS for throttle based boost, RPM based boost or gear based boost respectively. Then you open the app, set your desired targets and go.

Thoughts? I'm liking the simplicity of the Bluetooth thing. It's a lot easier to develop an app than design an on-dash controller with its own screen and buttons and it would be pretty nice/easy to modify the tables on a touch screen..

Infinitely controlable boost control, 0psi to the limits of your turbocharger/engine, closed or open loop, based on RPM, throttle position, gear, any or all of the above or simulated spring.

Zero boost on cruising, hold wastegate closed on WOT until 14psi then start opening (or be fully open by 14psi, your choice) unless you're in first gear then open it at 5psi, or unless you're lugging at low RPM open by 5psi, hold gate open for fuel economy, pair it with a vacuum compressor bypass inlet (engine doesnt habe to suck thru turbo when not boosting), hold gate closed until 5000rpm or 14psi which ever comes first, only close wastegate above 50% throttle, password protect gate closing?? (let your gf or valet drive youe car n/a and save the boost for yourself).

These are all relatively easy possibilities with this system.

If people are truly interested and excited about this, I'm willing to put in the time/effort to develop this into a working product. The biggest challenge (for me, personally, based on my experience and skillset) would be the design and manufacture of the actual wastegate valve parts, while the prototype can be based off a broken knockoff as proof of concept, it's hardly practical for an aftermarket product, something would need to be designed incorporating the actuator and valve into one unit as somethingthat can be manufactured easily and survive the conditions. While I have experience in all other areas this project requires, the actual engineering and design a valve is something I have little experience in, if someone has knowledge in this field, help would be appreciated

forbiddenera

So since my local junkyards didn't have the engine I wanted this weekend, I spent the money on an Arduino development board a few prototyping things, an lcd shield and a motor driver.

I don't have the specs of my linear actuator on me, it may be a bit big for my driver, it's a bit big for the project anyway but it's mostly for testing.

The final project should be buildable for under $100 (excluding existing wastegate)

While I'm using an arduino mega for dev, a mini should be enough for the project.. we need 1 or 2 pins for pwm input, 1 or 2 pins for pwm output to the motor driver, and a few gpio for talking to the Bluetooth board to communicate to the control app.

I don't yet have the Bluetooth board, they didn't have any suitable ones at my electronic shop. Actually, I do have a bt to ttl module I could use.

More soon when I tinker.

forbiddenera

I guess I lied, the Bluetooth board I got for my Honda ecu data logging is the same one they use for these arduinos usually, nice.

Also got a Canadian made sbc (made by Chinese people who can't speak english) that plugs into the arduino. Great idea, poor execution. Only TX, rx, 5v and gnd are connected, rest just pass thru..still n2b but no support means I said **** it and grabbed a Pi.

The Pi is running the Arduino IDE for fun.. (will be mostly using it for developing and other projects, eg, I plan on making my own custom systems that are like android decks based off sbcs. So easy dev.)

As of tonight, my chosen linear actuator is moving, controlled by the Arduino as well as position feedback is working with my actuator, programmed from my Pi, wee. Actuator is maybe a bit slow but I don't think super fast is required for this though a certain speed will be and only testing/datalogging will show what.

I've also got PWM output at 30.64 hz (closest I could *easily* get on my arduino to 31hz) so my thing could be used as a boost controller on its own with a standard solenoid.

I'm still working on quantizing PWM input and calculating the duty cycle and frequency, it needs to auto detect like 10 to 40hz (average bc range) and as close to 0 to 100% duty as possible, I think it should be possible to quantize down to 0% duty making it respond to an ecu or bc output more accurately than a solenoid (some noids don't even activate til 10 to 20% duty depending on frequency)

Once I can properly sample the PWM input (would be easy with the Pi but that's not REALLY part of this project) I'll build a circuit to buffer the 12v PWM to 5v max and test my ecus output.

Once it can read the pwm from my ecu, I can convert duty cycle to position, move the actuator watching it's potentiometer feedback to match the actuators position with the input.

Then I need to fabricate a mount for the actuator and attach it to the valve of my wastegate somehow. At this point I may return my test actuator and order something more specific.

I had another idea today, and perhaps I should make a thread, but, would anyone be interested in a device that can connect to an obd1 chipped Honda ecu (crome, ectune, hondata) and convert that to OBD2 output? You could install a working OBD2 port in your car and use apps like Torque, might even be a way for people to pass CARB w OBD2 car and OBD1 ECU..?

Because I'm considering doing this for my own uses already (seems OBD2 graphing/gauging/logging apps are >> tunerview app) and if it's useful to others, I don't mind building a kit and selling it for a reasonable price!

There, I'm like 80% done prototyping an electronic wastegate control system.. I've spent like $140 so far including the arduino, h bridge and actuator. This doesn't of course include the cost of the Pi, my kits of transistors/resistors/capacitors/jumper wires/breadboards.

Fot positional feedback, at 10bit input, the 2 inch actuation is divided into about it 1000 "steps" or .0.05mm sensitivity. I'm not sure how sensitive the potentiometer is inside, the specs say approx voltage for 0-full actuation but not sensitivity. If required, I do have a 12bit adc which will give me 4096 steps of input or 0.012mm distance feedback. I'm not even sure the potentiometer can be trusted that accurately, and besides if it's tethered to an existing ebc (which shouldnt be a requirement, though would be the easiest setup), chances are the ebc's output pwm is maybe 8bit or 0-255, if I can even measure that accurately. Ultimately, I think 100 steps over the actuation length should be plenty. While my current actuator has a 2 inch stroke, I don't even think my WG valve moves that much, maybe an inch.

I may also try and experiment with a voice coil actuator. Should be smaller and faster and could be controlled by pwm anyway, though likely at a much higher frequency than a standard solenoid valve.

I have a video of the actuator moving and graphed feedback, I'll probably YouTube it soon here.

forbiddenera

I just found my local hacker space has a full machine shop pretty much, so, I'm probably going to create a wastegate for this project from scratch and/or hardware to easily adapt it to an existing one.

My main concern is theasy metallurgy of the valve itself, I want to build something that will last. I could potentially build on top of an existing valve or valve and housing as well but if I were to go that route I'd rather make it adaptable to any gate instead of having to buy an existing one to adapt to?

Need to find a better actuator. Will be a few months before I get full access to the machine shop as new members don't immediately get access to all tools (regretting not going years ago as I had planned but I didn't realize there was so many resources available there)

Feedback appreciated

forbiddenera

Getting more excited about this project and I'm going to start finalizing the software base for it pretty soon.

I already have low speed PWM-input taken care of (10-40hz from an existing boost controller) and high speed PWM-output for controlling the speed of the linear actuator. I've also got analog potentiometer input solved for monitoring motor (valve) position.

I think I'm going to build the whole thing from scratch now, it doesn't make sense to build off an existing housing or valve, even if it solves my metallurgy problems - though, at the same time, if I use dimensions of an existing valve, I can use either my own or an existing aftermarket manufacturers, there's some that change color based on exhaust temp even which is neat.

Mostly it just needs to hold up, survive and last a long time.

That said - other than my own vehicle, I will be looking for testers for my electronic wastegate system.

Requirements:
- 38mm wastegate mounting (or adapter, your choice but my setup is 38mm now and thus the first gates will be built for 38mm)
- turbocharged car (doesn't have to be a honda but only single wastegate for now) in running/driving condition, turbo system has to work and be reliable as is as we need a good base (doesn't have to be street legal, but, I need feedback from actual usage/driving, dyno time is great too if you want to)
- reasonable experience with turbocharged vehicles and building turbocharged systems and experience with typical wastegates
- ECU with boost cut. I'm not going to be responsible for anything. If it locks up and over boosts and blows your engine, it's your fault because you should've had a reasonable boost cut set! This is a beta test that only very few will be able to participate in and could help shape a product I hope will be quite popular.
- Android phone or tablet with Bluetooth - the app will be developed multi-platform but Android will be ready for testing first, once the Android is going well, as it'll already be written on a multi-platform base, the iOS version will be perfected.
- ability to datalog - this is an absolute must. Doesn't matter whether it's Hondata, Neptune, K-Tune, AEM, Motec, or even if it's a Honda, but I need logs I can read.

I will want at least 1 person with an existing boost controller to test interfacing with an existing system (eg. as a drop in replacement wg) and 1 person willing to experiment with having my boost controller fully in control.

If you don't understand the potential benefits of such a system then this beta test is not for you. If you don't understand engines, it's not for you. I will be doing a wider-scale beta test with more average users in the future as a direct drop in ewg or boost control system. Also will need someone with an iPhone/iPad to test once the system is running good w/Android.

Please apply via PM if you're interested in testing my electronically actuated wastegate - if you're accepted, a test version of the system will be provided for a period of time for testing with the option to purchase (a deposit will be required on the hardware)

The beta will probably start in late spring/early summer to allow time for me to finish prototyping and produce at least 5 of the final design.

While this project will still have some DIY-flare and parts will likely be made open-source, I've decided to take this project more seriously than just DIY/hobby stuff and pursue actually building it with my business.

I've recently put a website up for my company, I'm going to make a sub-domain for this project pretty quick and add a blog (relating to this and my other projects). I'm not sure if it's appropriate for me to post the link, I'll have to check the rules / receive advice from a mod (mods?) but in the meantime if you're curious feel free to PM me.

I'm guessing it's probably fine if I post a link to the project page as it's still a project and not a product but I want to be sure. Moreover, I want to benefit the community..sure I'd like to make some money to eat as well, but I've always wanted to combine my expertise with electronics and computers with my passion for automotives and this is the exact type of project I want to bring to the community, along with the OBD>OBD2 thing - I love old vehicles and making them new again is great.

93h2beg

This is awesome.
If I had a bit more turbo knowledge I would be very interested in helping you out as a tester.

forbiddenera

If you like the idea, spread the word. Of course you can also learn more between now and when we actually start testing.

I can't wait to go for the first test drive with this.. :-)

If I didn't bust my control arm, I'd already have a prototype running in my car but since I'm fixing that I've decided to take some time to really engineer this project instead of just hacking it together with my busted fake wastegate.

forbiddenera

Besides it'll be a while before I can dig out my turbo car (I had just cleaned it off and moved it from across the street early last week, this is *entirely* from *ONLY* the most recent snowfall):
I <3 Canada