work in progress Common Rail MC33816/PT2001 dev board - 4 injectors GDI

[kb1gtt]

PT2000 symbol would be useful. Once a draft schematic is made, that can be brought up in it's own thread. Lets keep this thread focused more on the MC33816.

I see the "microcode" SPI thing. This is what I do not know. It appears you might upload a small program, then after that you can do SPI com's.

The eval board uses a specific brain board, and software on the PC. This software will create the micro code and then the SPI commands. If you make the eval board, keep in mind you likely need the matching brain, and it's PC software. With out it, the brain, you will likely have trouble figuring out the SPI commands.

There is a quote in process for this board. I expect this board assembled. We'll see what pricing they come back with.

[960]

Do we have a thread suitable to rail pressure control logic discussion? I assume this thread is about the board. Am I missing something or are we supposed to start by figuring out what logic is needed to do what exacly with high pressure pump control?

What's the issue about the fuel pump?

Just download the PC-Link software and open the base map for gdi.

https://www.linkecu.com/software-support/

You will find both the settings/variables needed for E-Throttle and injectors.
If you open Help, you will also find descriptions for everything.

Link ECU are my favorite aftermarket off all times. I run it in my race-car.


GDI Fuel pump control from help file:


Fuel Pump Control provides the required functionality to drive various different types of fuel pump speed control configurations.



The following fuel pump and fuel pump speed control configurations are supported:

· Simple Fuel Pump On/Off relay.

· Dual Speed Control through relayed switched ballast resistor (most older Toyota).

· Triple Speed Control. As used on Nissan GTR.

· Three Stage PWM speed control (eg Toyota Supra, Subaru Impreza)

· Closed Loop Fuel Pressure Control. As used on some modern and GDI vehicles. Controls fuel pressure to match a desired target pressure.

· Open Loop Variable Fuel Pump Speed Control. Allows control of a fuel pump speed control module based on injector flow.



Prime Time



In all fuel pump speed control modes, the Prime Time setting is available and sets how long the fuel pump will run for when the key is first turned on. A typical value is 3 seconds. As soon as engine speed is detected the fuel pump will run continuously until the engine stops.



Speed Control Method



Sets the mode used to control an Auxiliary Output that is configured as FP Speed. See the following sections for more information on the supported modes.




--------------------------------------------------------------------------------

Simple Fuel Pump On/Off Control



This mode requires that an Auxiliary Output is wired to the fuel pump relay and configured as Fuel Pump. No other configuration is required. Speed Control Method can be set to OFF as it has no effect.




--------------------------------------------------------------------------------

Dual / Triple Speed Control



This type of control is used where two or three relays (or a control module) are used to switch the fuel pump between different speeds. The speeds are regulated by the use of current limiting resistors or special circuitry in the control module.



Two Auxiliary Outputs are required to use this mode:

1. One Auxiliary Output must be wired to the main fuel pump relay (the one that can turn the fuel pump off completely). This Auxiliary Output must be configured as Fuel Pump.

2. Another Auxiliary Output must be wired to the speed control relay (the one that switches in the resistor). This Auxiliary Output must be configured as FP Speed.



To use this mode Speed Control Method must be set to General X Over. This will allow the following settings to be adjusted:

· Eng Speed X Over - Sets the Engine Speed that fuel pump speed changes from low to high speed.

· Inj DC X Over - Sets the fuel flow that the fuel pump changes from low to high speed.



When Inj Duty Cycle is greater than Inj DC X Over OR Engine Speed is greater than Eng Speed X Over high speed pump mode is activated. Note that there is a 2% hysteresis on Inj Duty Cycle to prevent erratic switching.




--------------------------------------------------------------------------------

Three Stage PWM Speed Control



This type of control uses a Pulse Width Modulated (PWM) signal to a fuel pump speed control module to control the fuel pump speed in three steps. Only one Auxiliary Output is required to use this mode. That Auxiliary Output must be wired to the fuel pump speed control module and configured as FP Speed.



To use this mode set Speed Control Method to 80Hz 33/66/100% DC. The control will apply either 33% 66% or 100% duty cycle to the speed control module depeding on fuel flow. The following settings are available in this mode:

· FP Low -> Med - Sets the Inj Duty Cycle at which low to medium speed transition occurs.

· FP Med -> High - Sets the Inj Duty Cycle at which medium to high speed transition occurs.



Actual duty cycle applied to the fuel pump speed controller can be viewed with the FP Speed (%DC) parameter (found under Auxiliary Outputs).




--------------------------------------------------------------------------------

Closed Loop Pressure Control



This type of control continuously varies a PWM signal to the fuel pump speed control module to regulate measured fuel pressure to a target value. Only one Auxiliary Output is required to use this mode. That Auxiliary Output must be wired to the fuel pump speed control module and configured as FP Speed. An Analog Input must also be wired to a fuel pressure sensor and correctly configured to display calibrated Fuel Pressure.



Note that a special fuel pump speed control module is required for this type of control. A solid state relay will not be able to switch fast enough for variable speed fuel pump control.



Most OEM vehicles that require this type of control have the correct mechanical system to suit it. This type of control does not use a fuel pressure regulator. Only a small bleed is used to keep some fuel circulating in the system. Fuel tank heating is greatly reduced due to reduced fuel circulation and reduced fuel pump current.



To use this mode set Speed Control Method to Closed Loop. The following settings will be available:

· Fuel Pressure Target Table - Sets the target fuel pressure for given conditions (typically load and engine speed). Fuel Pressure Target is the interpolated value from this table.

· Startup Override Pressure - This pressure is used instead of the Fuel Pressure Target Table during engine starting. It allows removal of air from the system, and is required for correct operation of some GDI high pressure pumps.

· Feed Forward - Used to improve the response when Fuel Pressure overshoots or undershoots Fuel Pressure Target.

· Proportional Gain - The proportional gain component in the PID controller.

· Integral Gain - The integral gain component in the PID controller.

· Derivative Gain - The derivative gain component in the PID controller.

· Pump Off DC - The duty cycle that will be applied to the fuel pump speed control module when the engine is not running.

· Max DC Clamp - The maximum duty cycle that will be applied to the fuel pump speed control module while running.

· Min DC Clamp - The minimum duty cycle that will be applied to the fuel pump speed control module while running.



Actual duty cycle applied to the fuel pump speed controller can be viewed with the FP Speed (%DC) parameter (found under Auxiliary Outputs).




--------------------------------------------------------------------------------

Open Loop Pressure Control



This type of control uses a Pulse Width Modulated (PWM) signal to a fuel pump speed control module to regulate the fuel pump speed. Only one Auxiliary Output is required to use this mode. That Auxiliary Output must be wired to the fuel pump speed control module and configured as FP Speed. Note that a special fuel pump speed control module is required for this type of control. A solid state relay will not be able to switch fast enough for variable speed fuel pump control.



This type of control is normally used on a system with a fuel pressure regulator. The idea is to maintain a constant flow in the fuel return line for all injector flows. Keeping a constant flow through the regulator prevents fuel pressure creep. Reducing fuel pump speed when fuel demand is low also reduces fuel pump current, noise and fuel heating.



To use this mode set Speed Control Method to Open Loop. The following settings will be available:

· Prime Duty cycle - Sets the duty cycle used during Prime Time.

· FP Speed Table - Sets the desired fuel pump speed depending on Inj Duty Cycle. 0% will be pump stopped,100% will be full speed, 50% will be half speed etc. Note, the number in this table is not necessarily the duty cycle applied to the fuel pump speed controller, it is the desired fuel pump speed at a particular Inj Duty Cycle. The actual duty cycle applied to the fuel pump speed controller will depend on the following settings (refer to the equation shown below). The X Axis numbers (injector duty cycle) can be adjusted as required by pressing the X key while on this table.

· Multiplier - Scales the duty cycle to the fuel pump speed control module. For example if 80% FP Speed is desired and the pump controller needs 40% duty cycle to run the pump at 80% speed then set Multiplier to 0.5 %/%. If the fuel pump speed control module outputs the same duty cycle as its input then set this number to 1 %/%.

· Offset - Corrects the duty cycle sent to the fuel pump speed control module. This will normally be zero. For example, if the fuel pump speed control module requires a 40% duty cycle input to achieve 50% speed then Offset will be -10%.

· Min DC Clamp - The minimum duty cycle that will be applied to the fuel pump speed control module. Most pumps will not turn below approx 20% duty cycle. Set this to a value well above the pumps stalling duty cycle.

· Max DC Clamp - The maximum duty cycle that will be applied to the fuel pump speed control module. Note, some modules will go into an error state if a too high duty cycle (eg above 95%) is applied. This setting can also be used to limit max pump speed where an oversized fuel pump is used.



Actual duty cycle applied to the fuel pump speed controller can be viewed with the FP Speed (%DC) parameter (found under Auxiliary Outputs).



The duty cycle applied to the fuel pump speed controller is calculated as:



FP Speed (%DC) = FP Speed Table x Multiplier + Offset , clamped between Max DC Clamp and Min DC Clamp

[AndreyB]

At the moment this board is 103x63mm (4.07x2.47 inches)

That 0.07 inch / 3mm makes the difference between $20 and $30 fabrication cost. Is it too much to ask to reduce the size by 0.12 inch / 3mm to fit in 100mm?

[kb1gtt]

It's now a little bit shorter. Let me know if that is now short enough.

[AndreyB]

It's now a little bit shorter. Let me know if that is now short enough.

New size is 101x63mm (3.97x2.47 inches)

At least for some sites it's good now since below 4 inches. Some sites could still be upset about that extra 0,033 inch if their threshold is 100mm sharp.

[kb1gtt]

My measures put it at 99.x so I stopped shrinking it. I can make it a bit shorter if the mfg's say it's too big.

[AndreyB]

jlcpcb says 101mm but gives the good under 4" price, so jlcpcb is good.

[dist01]

It will be only CR driver?

[AndreyB]

It will be only CR driver?

This board would handle 4 injectors - renaming the thread

https://rusefi.com/forum/viewtopic.php?f=4&t=1511 would handle six

[dist01]

Yes and you will use your mcu from gasoline ecu?


Sent from my iPhone using Tapatalk

[AndreyB]

Yes and you will use your mcu from gasoline ecu?

At the moment we are thinking to use same stm32 running rusEfi for GDI - gasoline direct injection.

[dist01]

you rework sw for rail pressure some pilot and main injestion? turbo pressure? same pressure/ fuel map? from edc15c?

[AndreyB]

you rework sw for rail pressure some pilot and main injestion? turbo pressure? same pressure/ fuel map? from edc15c?

I believe you are looking to chat and not about MC33816 dev board. I would prefer not to discuss these offtopic issues within this MC33816 dev board forum thread.

Please see https://rusefi.com/forum/viewtopic.php?f=13&t=1198 for slack registration, relevant room in chat is https://rusefi.slack.com/messages/CFP3QAJCW/

[kb1gtt]

I hear it's small enough, so I do not currently plan to make it smaller, unless someone tells me it should be a bit smaller.

[AndreyB]

At the moment BOM includes IHLP6767GZER100M01 FIXED IND 10UH 16.5A 12 MOHM SMD which is $5.37@10

What do we think about srp1265a-100m FIXED IND 10UH 10A 16.5 MOHM SMD which is $1.57@10?

[960]

Bom from the original board shows SRP1245A-100M, so that sounds close enough

[kb1gtt]

Hmmmm, how did that get in there. The SRP1250-100M should be just fine. I wonder why I didn't use the one in the datasheet, perhaps it was not available at that time, or perhaps it some how got copied from another reference design. I'm not sure, but as far as I can tell, the SRP1250-100M should be fine assuming it fits on the pads.

[AndreyB]

look what was posted at https://rusefi.slack.com/messages/CFP3QAJCW/details/

[kb1gtt]

Looks like someone is slacking

[AndreyB]

https://github.com/rusefi/rusefi/tree/master/hardware/Common_Rail_MC33816/gerbers

[kb1gtt]

Oops cap and inductor consume the same space also how did I forget a place for 12V and Gnd. Cool to see the first spin is being built.

[mck1117]

Oops cap and inductor consume the same space also how did I forget a place for 12V and Gnd. Cool to see the first spin is being built.

Thankfully the 12v/gnd omission isn't too big a deal, because the input filter caps have nice big pads to solder wires to.

[AndreyB]

Great news - all we need to know about MC33816 is spelled out for us between https://www.nxp.com/files-static/training_pdf/FTF/2012/americas/WBNR_FTF12_AUT_F0098.pdf and https://www.nxp.com/docs/en/application-note/AN4849.pdf

TL,DR: MC33816 has four independent microcores - four "threads" of execution.

One thread has to be controlling variable frequency modulation (VFM) DC-to-DC converter, whatever it means.

Two more threads are running two banks of two injectors each.

Each injector has it's own low-side driver.
Each bank/pair of injectors share a "boosting" - i.e. high-voltage opening high-side driver, and a "holding" - i.e normal battery voltage high-side driver. (page 59)

[AndreyB]

Trying to place jlcpcb order with 47 populated components - resistors and capacitors on the back side of the board.

I've created some scripts to help with format changes - https://github.com/rusefi/rusefi/blob/master/hardware/Common_Rail_MC33816/gerbers/kicadBom2jlcpcb.bat and https://github.com/rusefi/rusefi/blob/master/hardware/Common_Rail_MC33816/gerbers/kicadPos2jlcpcb.bat

C17415 part number was not accepted maybe because it's "Minimum : 100 Multiples : 100"? I've selected C17403 which is 10.2Ω(10R2) ±1%.

C17724 is same "Minimum : 100 Multiples : 100" issue - I've selected C247080 5.11Ohms ±1% 1/8W 0805

[AndreyB]

Your https://www.patreon.com/rusefi dollars at work - 0.2 PCBs have arrived

[AndreyB]

Missing R12 0.01 OHM so we are a bit stuck, to be continued...

[AndreyB]

Code: Select all

 8)                  J15 needs silk and larger hole. J20 needs silk and should use same larger terminal as J8-J9-J10
 9)                  J15 & J20 silk ^^^ should be on both top and bottom sides of the board
10)                  power LED. back side is pretty much free pre-assembled, front side is manual assembly. maybe one LED on each side?

[AndreyB]

No smoke with just +12v
Once I've enabled SPI communication and reset MCU with +12v applied to the test board I got a bit of noise and a small puff of smoke from the J20 corner of the board.

At this point my strong preference would be to ship both assembled boards to Seattle.

[960]

Have you checked all diodes have right orientation?
(If there are any)

[kb1gtt]

I understand this board is being sent to Seattle to get a proctology exam. In Seattle there is a OEM eval board which can be side by side compared. This will help find the issue. It seems like something is wrong with Q2. Perhaps a pin-out issue, perhaps a drive issue, we don't currently know. Hopefully we'll know soon-ish when the good Dr's in Seattle can probe it with there tools.