MS/951 – Another Standalone Build - Page 2 - Rennlist - Porsche Discussion Forums

Notices
924/931/944/951/968 Forum
Sponsored by:
Sponsored by:

MS/951 – Another Standalone Build

 
Old 03-14-2019, 08:52 PM
  #16  
odonnell
Addict
Rennlist Member
 
odonnell's Avatar
 
Join Date: Nov 2012
Location: Houston TX
Posts: 4,612
Default

My idea for using the FIDLE output was this, curious to hear your opinion:

- Middle pin gets 12v
- One outer pin gets chassis ground, however, there is a resistor in series.
- Other outer pin goes to a TIP120 transistor

That transistor is pulsed by FIDLE to pass a ground to the 3rd pin. That way, one side of the ICV is always active (hardwired to ground). But, when FIDLE indirectly pulses the other side, it overcomes this because of the resistor inline with that hardwired ground. So instead of trying to control both sides, you can just control 1 side and the closed loop algorithm should be able to figure it out.
odonnell is offline  
Old 03-15-2019, 01:13 AM
  #17  
Dare
Addict
Rennlist Member
Thread Starter
 
Join Date: Apr 2002
Location: San Jose, CA
Posts: 406
Default

Originally Posted by odonnell View Post
My idea for using the FIDLE output was this, curious to hear your opinion:

- Middle pin gets 12v
- One outer pin gets chassis ground, however, there is a resistor in series.
- Other outer pin goes to a TIP120 transistor
A few thoughts...

1) You don't need the transistor if you're using a Microsquirt. The Microsquirt Hardware Manual states that the max current for the FIDLE output is 3A. My measurement of the coil resistance is about 12 ohms, which puts the max current at around 1A. (I pretty much confirmed this value by the Amps readout on my bench power supply). So the MS should be capable of driving a single coil by itself.

2) Other people appear to have had success hard-wiring one of the coils to ground through a resistor, even though this isn't how Bosch did it. But you need to use an appropriate wattage resistor, since it could be dissipating a fair amount of power depending on the value. The question then is where to put this resistor. If you put it inside the DME, then all that heat is getting trapped inside the case, which is not ideal. If you don't put it inside, then you've got to find some place to splice it into your harness, with all the hassle that entails. The value of the MOSFET driver circuit I'm using is that it has very low on resistance, meaning that most of the energy is dissipated by the coil itself, which is conveniently located in the engine bay with all the other hot things.

3) I suspect that the resistor approach reduces the effective resolution of control, because a certain amount of the ICV PWM range is taken up overcoming the force from the always-on coil. Effectively, the lower the resistor value, the more of the PWM range gets eaten up. This may not be a problem in practice, though.

4) I think you'd want to be sure to place the resistor on the close coil, not the open coil. Otherwise the MS would have to constantly drive the close coil under normal running conditions just to keep the valve closed.

5) My belief (unconfirmed) is that either circuit should work in closed-loop mode, with the caveat that, if the resolution of control is reduced too much (by having too small a resistor), the algorithm will have a hard time keeping the idle steady.
Dare is offline  
Old Yesterday, 11:42 AM
  #18  
Dare
Addict
Rennlist Member
Thread Starter
 
Join Date: Apr 2002
Location: San Jose, CA
Posts: 406
Default Correction and status

As I hate spreading misinformation, I want to correct a small misstatement I made earlier regarding the capabilities of the MicroSquirt FIDLE output. Previously I said that the FIDLE circuit built in to MS contained no mechanism to suppress flyback from the ICV coil. After digging into the MS schematics (available here), it turns out I was wrong.

The FIDLE circuit is implemented using an IPS022G, which is a dual power MOSFET with built in protection features (datasheet). One of the features it has is an active clamp, which causes the transistor to turn back on when the flyback voltage from an inductive load exceeds a threshold. This effectively limits the voltage seen across the transistor, protecting it from excessive spikes.

The datasheet lists this threshold voltage as 48-56V. In retrospect, this matches perfectly the first oscilloscope trace I posted, which showed spikes to 48V when driving a single coil without my driver circuit.

Net-net: MicroSquirt should do just fine when driving a single coil ICV without any external flyback protection. In my circuit, however, diode D2 is still needed to further reduce the voltage to protect the input to the transistor driving the second coil (Q2).

Quick status update: I’ve been putting hours on the weekends to meet a deadline at work, so progress on this has slowed. Nonetheless I’ve managed to make good progress on a PCB layout, which is helping me to better map out the physical arrangement of things inside the case. Once I validate the design with my prototype, I should be able to quickly spin a PCB which, with some luck, should drop right in and work.
Dare is offline  

Thread Tools
Search this Thread

Contact Us About Us Advertising Cookie Policy Privacy Statement Terms of Service

© 2019 MH Sub I, LLC dba Internet Brands

We are a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for us to earn fees by linking to Amazon.com and affiliated sites.
 
  • Ask a Question
    Get answers from community experts
Question Title:
Description:
Your question will be posted in: