SwedeSpeed - Volvo Performance Forum banner

401 - 420 of 445 Posts

·
Registered
Joined
·
353 Posts
"The tests confirmed that the BA injectors are effectively an exact replacement for the Bosch 036 injectors."

Good to hear, I put a set in my 1800E and they're working just fine. Wondered about the quality, now I don't have to.

Thanks!
 

·
Registered
Joined
·
21 Posts
I am confused now.
I never had another firmware than the MS2Extra3.4.2a2 (I have even checked, not the 3.4.3 as I wrote before) on my MS2 board. I will update soon to 3.4.3. The MSExtra firmware supports also PWM injector control for my understanding. The option will be in TunerStudio visible as soon as you deactivate the additional driver card. The scope view I showed in the previous post was with the MSExtra firmware and it is a PWM signal.

That’s a good tip to check if the CPU will have resets. I searched for the signal but all I could find is a clock signal which will start with CPU start (Calculation1/Clock – clock gauge). I will ask google later for other signals.
I started with ignition only and the car was running over 1000 Km very well. My idle with the MS2 ignition and carburetor was perfect. I am using the original old coil from the Volvo Amazon and the onboard driver of the MS2 for that single coil. It is not the Bosch BIP373 driver, I still have the simple VB921 driver. The system is just prepared for WastedSpark with internal driver but so far I still have the single coil with the onboard driver!
I read also about electrical noise problems with the low-z injectors and the onboard driver how I use it. Using the P&H board should reduce the noise on the MS2 board because all the high current signals are separate from the DB37 plug and MS2 board. In the next days, I will be ready again to test the system in a running car. I do not get the injectors from the service back. I only tested the P&H board with a single injector.

I will use 30psi fuel pressure with the original Volvo D-Jet fuel pressure regulator.

About my proto area. I do not think there is something that can cause electrical noise.
This stuff I have installed:
- Bluetooth module
- Jumper for RS232 or Bluetooth connection
- Analog input circuit for the fuel pressure
- Tacho output circuit for the Rpm gauge (tested but not in use)
- Transistor circuit for cooling fan
- Pull up resistor for cam signal input
- Connector for P&H board (no high current stuff)
 

·
Registered
Joined
·
2,466 Posts
Discussion Starter #403
Thomas:

I am sorry for the confusion. Yes, MS2Extra has the ability to do PWM current limiting on the injectors when you are running with two injectors in batch fire mode. When you switch to MS2Extra in full sequential mode you have to use pins 3 and 7 on the MC9S12C64 processor chip to drive injector channels 3 and 4. Pins 3 and 7 provided the PWM signals to the mm74hc00m NAND gate on the daughter board that carried out the PWM function. When you do that modification to the daughter board connecting those jumpers to pin 3 and 7 and then set up Tuner Studio for sequential injection you lose the PWM function. I never ran MS2Extra in batch fire mode. I ran the original B&G firmware in batch fire and when I switched to MS2Extra I went directly to full sequential operation.

I clearly was not really paying attention when you said you don't have a cam signal. I was thinking that you were doing full sequential which requires a cam signal. That is my error. Since you don't have a cam signal I assume that you are doing batch fire on two injector banks and I believe you said you are doing wasted spark with with your crankshaft wheel. Or are you going to switch to full sequential with that modified distributor?

When you run batch fire on the injectors, injector timing is really meaningless (I went down that rabbit hole when I was doing batch fire). Even if you could get the injector timing on one cylinder correct the other cylinder(s) will be wrong. In batch fire, you could adjust the injector timing and you will never notice a change in operation. In order to see an improvement from timing the injection pulse you have to go to full sequential.

I would repeat my original statement that I much prefer to use an external resistor in series with the low resistance Bosch 036 injector rather than use the PWM feature on MS2. However, you are planning to use the J&B Performance Peak & Hold board so I expect that in the future (if the LM1949 works correctly) you will disable the PWM feature

I also went looking for the reset counter. I thought there was a flag on the dashboard; but, that may have only been in Tuner Studio when you are running the B&G firmware because I cannot find that feature. I also thought you could set a field in Megalogviewer which would display a line showing the reset count; but, I also can't find that. However, I also discovered that I can no longer remember how to generate a tooth log so I have concluded that my brain is turning to mush.

You can make the original Volvo D jet pressure regulator work; but, I advise against it. The flow through the injector depends on the pressure difference between the fuel rail pressure and the air pressure at the tip. With the D jet fuel pressure regulator the fuel rail pressure is always 30 psi. That means that when your manifold pressure is 50 kPa (7.4 psi) the pressure difference across the injector is 37.4 psi. At really low MAP values the the pressure differential is even higher. The only time your injector will actually be flowing the calculated flow rate at 30 psi is when the manifold pressure is at 100 kPa (or whatever atmospheric pressure is at your location). Fuel pressure regulators that have a vacuum reference provide a variable fuel rail pressure that maintains a constant pressure across the injector. My 280 Z regulator has a base pressure of 36 psi when the vacuum reference line is disconnected. When my engine is idling at about 50 kPa the fuel rail pressure drops to about 28.5 psi to maintain a 36 psi differential across the injector.

Most people use vacuum referenced fuel pressure regulators. If you use one of those Ve table generator applications to initially fill in all your Ve values they generate the Ve values assuming that the injector flow rate is the same for every manifold pressure. Because your injector flow rate is changing for different manifold pressures, your required fuel value is actually changing. You can make the fixed pressure regulator work by adjusting your Ve values to compensate. However, that will result in a Ve map that does not look like anybody else's Ve map. If you want to be able to compare your Ve map with somebody else's Ve map at some point in the future, you have to be using a vacuum referenced fuel pressure regulator.

I looked at a log I ran a couple of days ago. At an idle speed of about 890 RPM my injector pulse width was 2.04 ms. Your injector will have about the same flow rate as mine at that speed & MAP value so you should be operating in the injectors linear range. I do have some cases where my injectors are probably dropping well into the non linear range. These occur when I let off the throttle and slow down with the car still in gear and coast up to a traffic light. The PW was dropping down below 1.0 ms and the air fuel ratio was briefly hitting around 20. It is probably creating some slight transient problems; but, the engine does not stall and the transient problems are not noticeable. The AFRs immediately recover when I shift out of gear and let the engine idle.
 

·
Registered
Joined
·
21 Posts
Hi, here you can see how the LM1949 works. I measured the injector current by using a 0,1 Ohm resistor in series to measure the voltage drop (I = V / R). The injector pulse duration is 8ms.

The P&H circuit starts with full current and reduce the current to 1A as soon as the maximum of 4A is reached. I was a bit surprised how long it takes ([email protected],2V) to magnetize the injector coil. The behavior is similar to an capacitor. The peak time will be a bit shorter with 14V supply voltage.

For my PWM settings in the previous installation I reduced the peak time to 1ms. After the 1ms the PWM mode started with the current limitation. The correct setting in my opinion would have been 2,5ms for the peak time in PWM mode.

 

·
Registered
Joined
·
21 Posts
My English is not the best. I guess some of my sentence are a bit confused J

I started with crank sensor only, inkector PWM mode and batch fire. In the last couple of days, I added the Cam signal, installed the P&H board, and did the modifications on the MS2 hardware for injector output 3 + 4. I will run the system in full sequential from now (without PWM). The ignition will be still over the single coil. The WastedSpark and PWM idle valve will be upgrade when the basics are ok.

The MS2 clock signal should be sufficient to check if the CPU got a reset by electrical noise. The timer starts from the beginning as soon as the CPU restarts.

Thanks for the explanation with the fuel pressure regulator. I did not understand the problem with the relation of fuel pressure and manifold pressure. I will still keep the D-Jet regulator even when the VE table looks not plausible. The look of my injection system should be as close as possible like the original D-Jet system. The Megasquirt should be hidden.

I will test the modified system as soon as my injectors are back. I will keep you updated.
 

·
Registered
Joined
·
2,466 Posts
Discussion Starter #406
Hi, here you can see how the LM1949 works. I measured the injector current by using a 0,1 Ohm resistor in series to measure the voltage drop (I = V / R). The injector pulse duration is 8ms.

The P&H circuit starts with full current and reduce the current to 1A as soon as the maximum of 4A is reached. I was a bit surprised how long it takes ([email protected],2V) to magnetize the injector coil. The behavior is similar to an capacitor. The peak time will be a bit shorter with 14V supply voltage.

For my PWM settings in the previous installation I reduced the peak time to 1ms. After the 1ms the PWM mode started with the current limitation. The correct setting in my opinion would have been 2,5ms for the peak time in PWM mode.

The rise time of the current in the injector is determined by the inductance of the solenoid. The current would be given by the exponential formula

I = (V/R)x(e**-(R/L)t) - the coil current would follows the same as the voltage across a capacitor as it is being charged through a resistor

where L is the inductance of the injector solenoid, R is the total resistance in the circuit and V is the supply voltage. If the inductance of the coil was constant I would expect to see the current look like a typical exponential curve up to the point current limiting occurs. However, the plunger in the injector moves slightly which alters the inductance which I think is why you see that little bend in the curve.

Your curve suggests that it takes about 1.5 ms to reach a coil current of 1.5 amps (my injector current at 12 volts). That does seem long. I did some bench tests on my injectors looking at opening times when I was attempting to assess the effect of injector temperature on opening times. I will see if I can find those Oscilloscope recordings so I can compare to your results. If I can't find them, I do have a spare 036 injector so I may set up a bench test to repeat the measurements.

If the current in the first millisecond of operation is different than from what you get with the P&H board, that may affect the injector offset value. If the current is the same, then my 0.25 ms offset value may still be valid. If the current is different, then the offset value may change.
 

·
Registered
Joined
·
2,466 Posts
Discussion Starter #407
My English is not the best. I guess some of my sentence are a bit confused J

I started with crank sensor only, inkector PWM mode and batch fire. In the last couple of days, I added the Cam signal, installed the P&H board, and did the modifications on the MS2 hardware for injector output 3 + 4. I will run the system in full sequential from now (without PWM). The ignition will be still over the single coil. The WastedSpark and PWM idle valve will be upgrade when the basics are ok.

The MS2 clock signal should be sufficient to check if the CPU got a reset by electrical noise. The timer starts from the beginning as soon as the CPU restarts.

Thanks for the explanation with the fuel pressure regulator. I did not understand the problem with the relation of fuel pressure and manifold pressure. I will still keep the D-Jet regulator even when the VE table looks not plausible. The look of my injection system should be as close as possible like the original D-Jet system. The Megasquirt should be hidden.

I will test the modified system as soon as my injectors are back. I will keep you updated.
It is better than my German!

Yes, the clock signal should be a good indicator of a reset. It bothers me that I can't find the reset counter feature that I thought existed in Megalogviewer.

If keeping the appearance of the original D jet system is important to you, then the Megasquirt can be made to work with the original D jet fuel pressure regulator. Just be aware that on trailing throttle the MAP values go very low which means that the pressure differential across the injector (and the flow rate) will increase. You will likely be running very low injector pulse widths under those conditions and you may have complete fuel shut off because you have entered the non linear region of the injector. This may cause some driving problems when coming to a stop. Try it. If it works fine you are OK. If you do have problems then you know where to look.
 

·
Registered
Joined
·
2,466 Posts
Discussion Starter #408
I couldn't find the results of the tests that I ran on the Bosch 036 injector a few years ago so I repeated them. I have a spare MS2 controller which is a duplicate ('almost' a duplicate because I had some modifications in it for a knock sensor that I was testing) that I used in test mode to control the spare Bosch 036 injector that I have. The test set up consisted of a lab power supply running at 14 volts to supply the injector with a 6 ohm / 50 W power resistor to limit the injector current. The test pulse was set to 5 ms. The injector was ground switched by the MS2 controller using VND 5N07 MOSFETS to switch the injectors.

I measured the voltage across the 6 ohm resistor as a proxy for injector current.

Injector current.jpg

The curve shows the same 'double hump' characteristic as your tests display. The start of the second hump is much more distinct in my scope trace and is around 1.7 ms which appears to be consistent with what your trace shows. My trace shows that the injector reaches 94% of its peak current at about 3 ms, peak current being 10.2 volts/6ohms = 1.7 amps. The rise is determined by the circuit R/L time constant. Yours looks like it reaches 1.7 amps in less than 2 ms. However, I believe that the hump in the characteristic at 1.7 ms is caused by the solenoid in the injector hitting its full open limit. The fact that my injector using an external current limiting resistor and MOSFET switch and your injector using the P&H board with a darlington pair switch both seem to show the injector hitting full open at 1.7 ms suggests that injector operation on your system and my system will be similar and that offset and voltage sensitivity values that I use may work just fine with your P&H board.
 

·
Registered
Joined
·
477 Posts
142G;

...good Injector test data, and insight! ...my only comment would be, if you still have the test setup and want performance data more characteristic of the Bosch ECU, you could increase the series R until you drop an equivalent Vce sat of the bipolar transistor switches of the Bosch ECU...I expect the typical very low R(on) of the MOSFET compared to the bipolars, might change the RL time constant and affect the I vs T cure a bit...

Great and competent work!
 

·
Registered
Joined
·
2,466 Posts
Discussion Starter #410
Ron:

I was initially thinking along the same lines when looking at the P&H board that Thomas is planning to use. The P&H board uses darlington pair output transistors to control the injector and darlington pairs typically have Vce sat values around 2 volts. As you note, the power MOSFET typically have very low Ron values which results in a low drain-source voltage. I did capture the MOSFET drain-source voltage during the test

MOSFET voltage.jpg

You can see that voltage across the MOSFET drops to about 200 mV when the MOSFET first switches on and then gradually increases as the injector current increases maxing out at just under 500 mv at the maximum injector current. I was thinking that the difference might alter the injector offset value. However, given that it appears that the transition in the injector current curve that I measured appears to be in the same time frame as the transition that Thomas measured, the high Vce sat may be a non issue - that is why I asked if Thomas had a higher resolution capture of the current to confirm the full open time. I interpret the current transition as the solenoid hitting its open limit.

I think Thomas did his tests at 12 volts and I did mine at 14 volts (my typical idle voltage) so there should be some difference because of the sensitivity of injector offset to voltage. If Thomas is using TIP 122 output transistors I have some of those around and could arrange to test the injector at 12 volts and compare the current rise curve to the MOSFET curve. Given the very low offset values for the 036 injectors (0.25 ms) and the offset voltage sensitivity (0.18ms/volt) I would expect that the almost to 2 volt difference between the MOSFET and darlington pair at time T0+ would alter the offset value. However, only testing would confirm.
 

·
Registered
Joined
·
477 Posts
142G;

Indeed, there are a lot of variables...buss V, output switch R(on), supply rail pressure, and about twenty others...its a game of simultaneous equations only a computer could keep track of...the important thing to know IMO is which way these parameters push the function in terms of the engine performance...then knowing and understanding that, one can make "judgement calls" when it comes to tickling the setup values for the FI CPU where you have a lot of parameters at your control...

...following.
 

·
Registered
Joined
·
21 Posts
Hi, nice to see that your test results are comparable to my. I tested the Injector current at 12,2V with a 0.1 Ohm 10% resistor. The voltage was measured on the battery so there will be a drop of +0,3V of the cables! The test was with a dry 036 injector (no fuel). I am using the TIP122 Darlington Transistor which is suggested from Jbperf.
I belief the same about the current characteristics on my scope picture. The bend in the curve comes from the movement in the injector. I read about injector current behavior and it was similar explained.

Let me know if I shall repeat the test with exact 14V on the injector. I have the scope and lab power supply at my home.
 

·
Registered
Joined
·
477 Posts
Thomas;

You must know this...but I didn't see it in your posting, so it scared me, and it bears repeating...you must have 6Ohms of current limiting in series with the Injector current!, which with an Injector R Value of around 1.3Ohms drops the Injector voltage to around 3V...you cannot put 12 or 14V directly on the Injectors!

Cheers
 

·
Registered
Joined
·
21 Posts
Thomas;

You must know this...but I didn't see it in your posting, so it scared me, and it bears repeating...you must have 6Ohms of current limiting in series with the Injector current!, which with an Injector R Value of around 1.3Ohms drops the Injector voltage to around 3V...you cannot put 12 or 14V directly on the Injectors!

Cheers
🙂

Of course I will connect the injector directly to 14V. But the ground will be controlled by the P&H board.
The opening behavior will be different if the supply voltage is higher. My first test was with 12,2V battery voltage. I guess the real supply voltage on the injector was 11,9V during the peak sequence. The P&H board is limiting the current flow to 1A after it reached 4A.
 

·
Registered
Joined
·
477 Posts
Thomas;

"Of course I will connect the injector directly to 14V. But the ground will be controlled by the P&H board. "

That is called "low side switching" and that is OK...a switch to control current can be in the high side OR the low side, but not what I am referring to...I am pointing out that you MUST limit the current in (each) Injector with a 6Ohm resistor in the current path. You may NEVER connect a full 12V across the Injector (which has a resistance of only 1.3Ohms), because without current limiting limiting, the current would be upwards on 9Amps, and the Injector would surely not survive!

Here is an excerpt of the output circuit with Injectors showing voltages:
View attachment 67995

Source: https://www.sw-em.com/bosch_d-jetronic_injection.htm#Dynamic_and_Static_Injector_Test_Notes

Cheers
 

·
Registered
Joined
·
2,466 Posts
Discussion Starter #416
Thomas;

"Of course I will connect the injector directly to 14V. But the ground will be controlled by the P&H board. "

That is called "low side switching" and that is OK...a switch to control current can be in the high side OR the low side, but not what I am referring to...I am pointing out that you MUST limit the current in (each) Injector with a 6Ohm resistor in the current path. You may NEVER connect a full 12V across the Injector (which has a resistance of only 1.3Ohms), because without current limiting limiting, the current would be upwards on 9Amps, and the Injector would surely not survive!

Here is an excerpt of the output circuit with Injectors showing voltages:
View attachment 67995

Source: https://www.sw-em.com/bosch_d-jetronic_injection.htm#Dynamic_and_Static_Injector_Test_Notes

Cheers
Switching the ground connection on the injector is the norm for most modern applications. The hot side is switched on the D jet system. I don't know whether Bosch maintained that arrangement with the later LH digital systems where the injector resistors are mounted on a cluster usually on an inner fender wall.

Thomas doesn't need external current limiting resistors because the P&H board provides a current limiting function. In his scope plot you can see the continuous current to the injector is limited to about 1 amp. On my car with external 6 ohm resistors the continuous current is limited to about 1.7 amps at 14 volts if the fuel PW is long enough for the current to reach the maximum value defined by the R/L time constant.

I have a scope plot stored someplace which shows the voltage across the injector on my set up. In a series R-L circuit like the injector arrangement, when the injector is first switched on the Injector looks like an infinite impedance and all of the voltage (minus the drop across the MOSFET or switching transistor) is applied directly across the injector. As current flow starts to increase through the injector voltage drops across the external resistor and the voltage applied to the injector also drops. The external resistor combined with the injector internal resistance and inductance does create a form of peak and hold. The peak initial voltage applied is established by the supply voltage and the hold voltage is determined by the external dropping resistor and the drop in voltage applied to the injector is determined by the R/L time constant. With the P&H board the peak initial voltage is established by the supply voltage; but, the control transistor does not operate as a simple on-off switch. The transistor operates in its linear range controlled by the LM 1949 chip which monitors the current through the injector. Based upon the timing circuits in the chip and the current measurement the chip reduces the base current to the control transistor which increases the voltage across the transistor reducing the current through the injector to the hold value.

The P&H board has the theoretical advantage that it can hold the injector supply voltage at a high value until it switches to the hold current. With my external resistor arrangement the voltage across the injector starts to drop immediately after being switched on as defined by the R/L time constant of the circuit. The fact that the P&H board applies a high voltage for longer up front might have the theoretical advantage of reducing the injector offset time. However, the Bosch 036 injector already has an exceedingly low injector offset value. The reaction of a lot of people when I tell them what I measured has been that the value is impossible for a low Z injector. Because of the very low offset time I think the P&H board will not result in a reduction because there is not much of a reduction to be had. The one area of concern might be the voltage sensitivity factor. Because the P&H board can hold the applied voltage to the injector at a high value during the initial opening, the offset time may have a lower voltage sensitivity than what I measured. Thomas will have to sort that one out for himself because I don't have a P&H board to test plus doing all the flow measurements to see how the X intercept changes with supply voltage is a lot of work.

The Bosch 036 injector's very low offset is interesting. I was discussing this with my Son last night who was over and is also an Electrical Engineer (my Wife just rolls her eyes and gets up and leaves) and speculated that it may have been because the D jet was an analog system. Analog systems don't handle discontinuities / offsets easily so the Bosch engineers might have gone to extraordinary efforts to create an injector with a flow curve that crossed the X-Y axis as close as possible to (0,0) which would probably be easier to model in an analog system. To be sure you would need to parse the timing circuits in the D jet controller. However, figuring out what is going on in the D jet is much much harder than in a digital system where the fuel equation is written out and is actually quite simple.
 

·
Registered
Joined
·
477 Posts
142G;

OK...I've got it...the drive circuit of the MS Sys throttles the current back (with PWM?) and doesn't allow unlimited current for a prolonged time...my comments were based on the D-Jet control sys...

Cheers
 

·
Registered
Joined
·
353 Posts
Hello, I have to hand it to you both (Ron and 142G) , in another life I may have been an engineer, but was a lowly physics major... I find all this fascinating, interesting and informative, and admire the depth of intricacies you speak about. Not complaining in the least, rather I find this eminently interesting and along the way I pick up a thing to two. My 1800E runs great by the way.
 

·
Registered
Joined
·
21 Posts
142G;

OK...I've got it...the drive circuit of the MS Sys throttles the current back (with PWM?) and doesn't allow unlimited current for a prolonged time...my comments were based on the D-Jet control sys...

Cheers
Hi Ron,

I tried two different modes to control the injectors. The first method was by using the PWM function of the MS2 V3 board.
You have different parameters for that mode like peak time, current limit and PWM frequency. You can see the scope (Injector voltage) in post 391.

The P&H board is quiet different. It does not work with PWM. It is, let me say a closed loop system. The LM1949 is measuring the injector current flow by a 0.1 Ohm resistor. The chip is reducing the base current of the TIP122 (Power Transistor) depending on the injector current and a time function.
The TIP122 is working like a throttle and not like a on/off switch. You can see the scope (injector current) in post 406.

Thomas
 

·
Registered
Joined
·
477 Posts
Craig;

My former high school physics teacher is still a good friend who is credited with directing the youthful energies of many now quite successful individuals...many into science and electronics, like me, and some even into the medical field...and we are forever thankful for it! I still remember his teachings like "there is no such thing as centrifugal force...its centripetal!" When it comes to these car projects, I had better get the science right, or never live it down!

Cheers
 
401 - 420 of 445 Posts
Top