(Bio)Diesel Powered

Your chance to introduce yourself and to ask your first question
Post Reply
Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

(Bio)Diesel Powered

Post by Noxz » Wed Jun 27, 2018 3:08 am

Hello All!

I am in the process of researching Open ECUs for rigging into a very modern diesel, common rail, direct injection, engine. The specific engine (LH7) that I am looking at interfacing with is loaded with sensors - and I have essentially obtained the full service manual - which is good since the wiring is basically cut at the connectors on the engine that is for sale (also missing throttle body and turbo on back).

I know that there are a few other threads on Diesel attempts.. but they really did not go anywhere, beyond this..
viewtopic.php?f=4&t=1337

Similarly, externally is a decent LT ref with http://www.analog.com/en/design-center/ ... upply.html
Simply keeping it in mind, but I may be more inclined to look into the MC33816, especially since others here have looked into it (and even designed a little proto board) already.

I need to make a few more final measurements, to verify the receiving vehicle will be able to house this newer engine without too many complications (fabrication will have to be done, either way) before I pull the trigger and commit to buying it.

Although I know how to solder, why should I waste my time with a frankenso kit? except I can tailor make it for the engine at hand? It may be worth my time to simply pickup a fully populated board/bundle already.. wasting time is my least favorite thing..

The end goal is to run biodiesel I make myself (using a more advanced method than your friendly backyard lye+MeOH version) in a modern engine swapped into a relatively small car - running an ECU that I have control of.

Anywho.. I am super excited about this!

Update, TL;DR
Using KIT33816FRDMEVM $264.57 MC33816 Development Board with FRDM-KL25Z (this board is actually 4 by 8 inches)
MC33816 is 98% compatible with PT2000
I have Rigol DS1054Z

2018 chevy cruze 1.6L LH7 diesel engine


Feb 2019 picture download/file.php?id=3910

4 wire sensor Delphi 95PP3-1 , two linear outputs, calibration see viewtopic.php?f=2&t=1393&start=120#p29860

low pressure primer pump Airtex E2000

Tools:
MC33816 Developer Studio IDE
PT2000 Dev Studio IDE
FRDMPKPT2000EVM Basic Microcode 6 cylinder + Fuel Pump + DCDC

User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: (Bio)Diesel Powered

Post by kb1gtt » Wed Jun 27, 2018 4:56 pm

I agree the money for a populated board is worth it. You'll likely need to populate the driver board. Although you might get most of that populated the just finish it up. A next step on that board is to get the components and place them on a 1:1 scale print out to make sure they fit. I've been to busy to make progress but could offer remote support.

What is your electrical back ground? Do you have a scope?
Welcome to the friendlier side of internet crazy :)

User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: (Bio)Diesel Powered

Post by kb1gtt » Wed Jun 27, 2018 5:06 pm

Would you be interested in making a flow bench, using an low cost belt driven pump attached to a drill or similar spinning device to generate pressure? If you could make the injectors spray into a bucket. That would be a huge next step.
Welcome to the friendlier side of internet crazy :)

Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

Re: (Bio)Diesel Powered

Post by Noxz » Wed Jun 27, 2018 6:16 pm

I do have a bit of electronics background, although my supplies are currently low. I've been "using" LTSpice for ~12 years or so at this point if that gives you any idea.. Haven't used KiCAD before as I learned some basic circuit design w/ Eagle - but I hope to switch over.. My mother is a stained glass artist and I've seen more solder than most - I have an okay decent soldering setup (some chinese multifunction thing, w/ hot air gun and nozzles, but it works)

I need to get a scope for this project.. would 2 channel 100mhz plenty? or what is the minimum that I should be looking for?

I could certainly create a flow bench - don't want to break any piston rods experimenting. I have all of the torque specs for the engine so removing one+ of the injectors (very little information from what I gather of them) should be nontrivial.

From an earlier thread on the MC33810 as the driver of the solenoids w/ the HV, I was somewhat looking at the LTC3862-1 (maybe even daisy chained) as the HV PSU - simply because it was returned on some diesel IC search results.

Would it be worth it to buy a used, original, ECU? to see how they do it may be meaningless in the end..

Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

Re: (Bio)Diesel Powered

Post by Noxz » Wed Jun 27, 2018 8:18 pm

for the money, likely looking at the Rigol DS1054Z, 4ch 50mhz (upgradeable/hackable to 100mhz) ..

Unless convinced otherwise..

User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: (Bio)Diesel Powered

Post by kb1gtt » Thu Jun 28, 2018 1:56 am

I have a couple DS1054Z's at work. They are a good low cost scope, and do many things for low dollars. I've had them for a couple years now. They are not hacked and probably never will be. 50MHz is fine for most applications.

Having a working engine is a good start, as it helps you validate that the engine functions. However if your harness is currently cut, it is not likely that you'll get it working with an OEM ECU. Typically the ECU has security features which prevent the ECU and harness from being moved around like that.

I dream of a cumin's 4BT. Some day I'll get around to that.

Hot air soldering is nice.
Welcome to the friendlier side of internet crazy :)

User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: (Bio)Diesel Powered

Post by kb1gtt » Thu Jun 28, 2018 2:03 am

Oh, I might as well trash talk the DS1054Z. One feature which would be really nice is being battery powered, such that it is easier to isolate. Keep in mind those GND straps on your scope are all tied to the GND in the plug. Also because the GND's are common, doing things in differential mode is a pain. However 4 channels is a good start, as it allows you to get at least 2 channels. However I've not had good luck figuring out how to make that work properly. As well I got spoiled with my DSOQuad. The open software on that is cool, as I can set it as a simple multi meter, or make it a complicated VNA network analyzer. The scope can be many things if the software allows it. I've only seen software which allows this scope to be a scope. That's not really a bad thing, but with the hacking abilities of this scope, it would be cool to see other software which allows it to do things it was not originally designed to do.

I also have the current clamps, those are nice as well.
Welcome to the friendlier side of internet crazy :)

Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

Re: (Bio)Diesel Powered

Post by Noxz » Thu Jun 28, 2018 2:39 am

neat @ scopes..

I was looking into the new Cummins R2.8 crate engine before telling myself that I need something smaller, not a rock crawler.. and those crates work out of the box for a reason :)

and yeah, hot air for SMD reflow stuffs

I need to start looking into the rusEFI codebase..

I will likely pull the trigger on buying a ton of this stuff within the week.. dont want to waste too much time..

Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

Re: (Bio)Diesel Powered

Post by Noxz » Thu Jun 28, 2018 2:45 am

Also came across this from National Instruments..
http://www.ni.com/pdf/manuals/375969a.pdf

Beyond what one person could consider cost effective..
but they show a decent equation there and sample values for determining power ratings and so forth.. I havent looked into your [kb1gtt] board in depth for various capable power, and curious if anything was calculated such as that?

User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: (Bio)Diesel Powered

Post by kb1gtt » Thu Jun 28, 2018 9:54 am

I didn't do any maths with the MC33816 board. I simply took my best stab at a layout after reading the datasheet. There is likely some room for adjustment on some of the values.
Welcome to the friendlier side of internet crazy :)

Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

Re: (Bio)Diesel Powered

Post by Noxz » Thu Jun 28, 2018 4:27 pm

Also curious, is there an IRC channel ( or slack or similar ) ?

Smaller questions (that may not warrant much archiving) isnt always meant for forums.. I usually hang out on FreeNode (recently rejoined ##electronics for the questions I will have, and idle in many more) but appears no one is in a #rusefi .. Progress should certainly be posted on forums - especially large explanations, but going back and forth could be handled elsewhere?

User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: (Bio)Diesel Powered

Post by kb1gtt » Thu Jun 28, 2018 5:33 pm

Slack is here https://rusefi.com/slack/

However it's basically never used. It's been some months sense I've last looked at it. I think that's the same for @. We are even worst with IRC. Generally the theory was that the forums were more informal, and then the wiki would be a more refined copy of what resulted in the forums. However the wiki hasn't made it into that yet. Some day, but not today.
Welcome to the friendlier side of internet crazy :)

Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

Re: (Bio)Diesel Powered

Post by Noxz » Thu Jun 28, 2018 6:00 pm

Alright, joined up, and yeah I see not much has happened in the general chat for some time.. no worries..


I was more so trying to go through a block/logic diagram of what is needed for the injector drivers (in my head)..

Stable HV PSU, capable of, let's use some values from NI9751 doc..
75v
peak of 25A for 0.1ms
a hold of 15A ( 0.8ms @ 4500rpm )
(and some quick calc of maybe min 2ms between cylinders @4500rpm, for a metric to recharge the PSU for another peak trigger)
(This is without the idea of pre/post injection that modern Diesels use, but just to get a rough idea..)

MOSFET drivers
I see the 33816 eval board has high and low side drivers for the injectors - is this for short circuit detection/prevention?
the 33810 which was mentioned before as a driver IC only has low side

I do see a flyback zener on that board in place
(I am first trying to understand what all it is capable of before looking at your board, simply to have stuff spelled out)

physical power (heat) dissipation is commonly done through a ground plane in the middle of PCBs (or even a Metal Core PCB in the case of LEDs).. basic alternative is heatsink but that may depend more on the FETs chosen


sketching out some basics..

User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: (Bio)Diesel Powered

Post by kb1gtt » Thu Jun 28, 2018 6:17 pm

33816 is for high pressure DI while 33810 is for low pressure normal highZ injectors. Two different beasts.

The 33816 includes a high voltage generator. So no external PSU required. You have 2 different high side drives which will either connect the high side of the injector to the high voltage, or it will connect to 12V battery. The low side is what controls the on and off. Think of the older LowZ push and hold circuits. In this case you connect the injectors low side to GND. Then you connect the high voltage side to the high side of the injector. The injector will start to open and will consume many amps and many watts of energy. Then after it's open you can back down the amps, which holds it open. Then when you are ready to turn off, you open the low side which stops the injector.
Welcome to the friendlier side of internet crazy :)

Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

Re: (Bio)Diesel Powered

Post by Noxz » Thu Jun 28, 2018 6:45 pm

My only idea of going with an external HV PSU is for a more stable (multiphase) output.

Looking through the datasheets etc I do see that of the 5 highside drivers, 3 are referenced to a max of Vbatt while the other two can be referenced to Vboost..
kb1gtt wrote:
Thu Jun 28, 2018 6:17 pm
Then you connect the high voltage side to the high side of the injector. The injector will start to open and will consume many amps and many watts of energy. Then after it's open you can back down the amps, which holds it open.
back down the amps how?
are you saying HV first to open the thing, then switching to lower Vbatt to maintain the open state? it almost makes sense but at the same time not

Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

Re: (Bio)Diesel Powered

Post by Noxz » Thu Jun 28, 2018 6:50 pm

Okay, yeah, that appears to be the case.. that's actually really neat!

Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

Re: (Bio)Diesel Powered

Post by Noxz » Fri Jun 29, 2018 12:27 am

@kb1gtt
What is the purpose of the 'pump bank' ?

Electric motor that drives the pump(or electric pump itself) ?

Not sure I will need that. I will have to rig up the pump which is driven by the timing chain to a separate motor for the testing, either through a chain (requiring a gear to be made for the electric driving motor) or yank of the sprocket (a puller is required) and put on a pulley + belt (with another pulley on the driving motor).

You did mention a drill or similar, I have a router that would be easier to mount, and a few more horse power in it..

there are several sensors(2 in one, redundancy, on the common rail) and regulator(2, injection located inside the pump, and return at the end of the rail - PWM controlled) valves that make up my system, and thinking if I would need to drive such a system with logic from that board or not (needs something more)

Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

Re: (Bio)Diesel Powered

Post by Noxz » Fri Jun 29, 2018 3:45 am

Also, not a big deal that the IC can do more so why not have a PCB capable, but no need for my instance to actually populate it - at least from what I know about it right now..

User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: (Bio)Diesel Powered

Post by kb1gtt » Fri Jun 29, 2018 9:40 am

Pump bank is simply a normal fuel pump driver with some extra's. I believe you can ignore it. However the feedback might be needed to make it turn on or something like that. I'm not sure. Below are some links which might help understand what is required for this board.

https://www.digikey.com/product-detail/ ... ND/7645631
http://cache.freescale.com/files/analog ... 3816UG.pdf

Depending on how much $ is a concern, they do make an off the shelf eval board.

https://www.digikey.com/product-detail/ ... ND/7645631
https://octopart.com/search?q=KIT33816FRDMEVM&start=0

About the router, I would suggest keeping an eye on the RPM's. I don't think you need much power, any tiny motor will likely suffice. If you're considering making a flow bench, any chance you'll consider moving that content to the MC33816 thread?
Welcome to the friendlier side of internet crazy :)

Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

Re: (Bio)Diesel Powered

Post by Noxz » Fri Jun 29, 2018 4:54 pm

I will absolutely post all progress of use of the chip on that thread.. I should actually start a thread (sorta what this is - actually this was more of the 'Hello' and what I am gearing towards) for my efforts on this engine.

I figured that is what the pump bank was for, was simply trying to understand the usecase. The other solenoids in the engine that I am aware of, such as for oil circulation and what not - all valves(?), are all Vbatt (from memory of service manual). Of course you can boost/peak the voltage initially to get them to trigger faster, but no need for something that is on for an extended period of time.

My request for samples of the MC33816 was denied for some reason, maybe 2 sample requests was one too many, I may call them to ask why, I know they have stock of the chip.

I was actually going to bring up the fact that eval boards exist, but appears KIT33816AEEVM is no longer being offered, for preference of KIT33816FRDMEVM which appears to come with their 'Freedom' ARM dev platform - isn't much of an issue if the main thing I am sending high on START1..5 and setting up the code on the actual MC33816 (which could easily do pre/post injections)

Money isnt much of a concern (right now, but I will soon be quitting my job in IT to go back to school for a career change). Sure everyone wants to save where they can, but I am more worried about time and the added cost for a proven board is worth it - but it doesnt really help the development of a board specifically to our needs, really only how to interface (easy) and programming the driver..

I am likely going to buy the engine today, ebay is having a 15% off everything up to $100 off - save where you can, right?

User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: (Bio)Diesel Powered

Post by kb1gtt » Fri Jun 29, 2018 5:12 pm

The KIT33816FRDMEVM is likely the way to go. I have some experience with these FreeScale specialty driver chips. The experience I had was that the datasheet from top down was written by one person, then from bottom up was written by someone else. In the middle, they simply did not jive and it was nearly impossible to figure it out. I believe the intent was that you'd get an eval board, which would them make the datasheet make sense, then you could develop your board. It was a pain. So if you are looking to shorten the development cycle, and don't mind some extra $'s, I'd suggest the dev kit to start. My layout is handy to have such that we can eventually make our own board which does exactly what we need in a nice case and all. However for now, I think getting it going is a big step in the correct directions. Once we have that then we can make it pretty.

Also a bit of a side note. My understanding is that the MC33816 knows when you are firing the injector and starts pumping energy into the HV rail much quicker than what a normal power supply would do. Normally a power supply needs something to react to, while the MC33816 reacts before the voltage starts to drop. So an external supply might be less desirable than you might think. If one needs a higher or lower voltage, I can think of some ways to make that happen. However for now probably best to just make it go, then look for performance tweaks later.
Welcome to the friendlier side of internet crazy :)

Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

Re: (Bio)Diesel Powered

Post by Noxz » Fri Jun 29, 2018 6:00 pm

That kit does seem like a good way to go - I would have preferred the other one as it didnt have the Freedom board and I could have just used the Frenkenso directly (gaining more exp with interfacing with it), but hey, no worries, not like they are charging more..

I have seen some datasheets that seemed multi-author and hard to comprehend.. Not sure if some brands do it more than others.

I would think if they have a dc-dc converter on same chip they could at least send the other die a signal to let it know that energy from it will be going away and start to PWM again.. unless it's 'dumb' and simply measures usage as it goes along..

And yup, of course just have it go first, as there will be tweaks later on..

Also, from what I can gather from the service manual, there is no calibration code (printed ontop of every injector, typically) enterable to account for differences, but rather, the engine is smart and can learn which injectors are putting more/less and calibrate them itself - it can "Learn" the calibration codes, possibly from a difference in speed output from each cylinder firing until they go into equilibrium. I tried looking for any documentation of decoding the numbers but it all seems fairly proprietary.. maybe contacting Denso (my injector mfgr) would give some insight, but likely all closed source and no luck there..

Just bought the engine.. before the Level3 outage really shook things up too much (half of the internet is down for me).. Starting to get fairly excited about this :D :D

Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

Re: (Bio)Diesel Powered

Post by Noxz » Fri Jun 29, 2018 6:19 pm

Just purchased the KIT33816FRDMEVM through NXP directly..

was a couple bucks cheaper than through Mouser (thanks to added tax).. Stock shows as Pending for NXP, so hopefully they'll report they have it and ship it out, otherwise I'll go through Mouser..

User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: (Bio)Diesel Powered

Post by kb1gtt » Fri Jun 29, 2018 6:32 pm

That's some good news.

Do you have a logic analyzer? I think you can do this with the DS1054Z scope, but I forget if that was an purchased upgrade. Any how, we'll likely be interested in the SPI coms. Do you have the ability to capture SPI com's? For the short term, it wold be a great step to adjust things with the dev board, then connect to Frankenso, then fill a graduated cylinder. Once you have that step, then a next step is likely to capture the SPI coms such that we can replicate those steps with the Frankenso SPI. Once we have SPI and a working eval board, then the board I did will likely be of more interest.
Welcome to the friendlier side of internet crazy :)

Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

Re: (Bio)Diesel Powered

Post by Noxz » Fri Jun 29, 2018 6:45 pm

the DS1054Z can read SPI, along with a few other protocols

I am starting to sketch out the flowbench design..

I know it can sometime be pulling teeth to get people to commit to a project, but certainly glad to be a part of it.

User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: (Bio)Diesel Powered

Post by kb1gtt » Fri Jun 29, 2018 8:36 pm

Noxz wrote:
Fri Jun 29, 2018 6:45 pm
I know it can sometime be pulling teeth to get people to commit to a project
Pulling teeth is fairly easy...

About the scope, it can physically decode all sorts of things. I recall the base model has some features turned off. So it may or may not happen to decode SPI. I should pull it out and check if it appears to be an option. It's sitting just some feet away.
Welcome to the friendlier side of internet crazy :)

User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: (Bio)Diesel Powered

Post by kb1gtt » Fri Jun 29, 2018 9:06 pm

On my base model scope, Under MATH --> Decode#, the only option I have is "parallel" I expect this is the place where you would normally get to choose SPI, or other protocols. It appears that it is not an option for the base model with out perhaps hacking or similar.
Welcome to the friendlier side of internet crazy :)

Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

Re: (Bio)Diesel Powered

Post by Noxz » Fri Jun 29, 2018 9:15 pm

Yeah, upon some searching it appears it was available as a trial or similar (it is an upgradable "Option" in the manual), but unlocking it wont actually be an issue..

Should have it on Monday and can verify..

Noxz
Posts: 130
Joined: Wed Jun 27, 2018 2:40 am
Soldering skill: yes
Coding skill?: yes

Re: (Bio)Diesel Powered

Post by Noxz » Mon Jul 02, 2018 6:17 pm

scope came in, "installed options" had spi, i2c, and rs232 decoder as 'official' were there on powering up

but I put in the key (firmware seemed to be latest) to get that 100mhz and such

installed options only altered as showing 100M, but under MATH->DECODE, I have 4 options, Parallel, RS232, I2C, and SPI

So if I didn't have it (under trial?) before, I do have it now.


I suspect that because the eval board works w/ the SPIgen software, that no decoding will be needed to interface with the 33816 as it should all be known/available.. The kit is being sent from direct from NXP in Malaysia (which is why I think the sample request was denied, overseas/availability), though I should have it by Friday - sooner than the engine which was shipped today.

I also have a dedicated 12v 30A PSU for the eval board..
Attachments
DSC_4936.jpg
DSC_4936.jpg (620.57 KiB) Viewed 1915 times
DSC_4935.jpg
DSC_4935.jpg (520.22 KiB) Viewed 1915 times
DSC_4934.jpg
DSC_4934.jpg (502.21 KiB) Viewed 1915 times

User avatar
kb1gtt
contributor
contributor
Posts: 3489
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA
Contact:

Re: (Bio)Diesel Powered

Post by kb1gtt » Tue Jul 03, 2018 12:28 am

That's some good news.

Do you have particular bio diesel desires? So far it sounds like your making a bench top diesel engine. Where does the bio work into the mix?
Welcome to the friendlier side of internet crazy :)

Post Reply