hi agu,
I believe the OP has now decided to go the P MOSFET route.???

I have indicated to him that the opto LED current should be in the order of 20mA, if is to be sure of say 4mA opto collector current. Using a 10K0 collector resistor should enable it to drive a P MOSFET.

Eric

 

I was wrong because I have two broken calculators.

The third calculator shows the LED currents are 18mA when the resistor from pin 7 to ground is 680 ohms and the currents are 38mA (way too high) if the resistor is 330 ohms.

When the supply voltage is 13.8V, the LED currents are 18mA and they are all on, then the LM3914 dissipates about 1.8W and it will smoke and melt.

I can't remember if the 10 ohms resistor is in series with the LEDs' power line. Its value could be increased to 33 ohms then the resistor will share half of the heat with the LM3914 when all the LEDs are on.

__________________
Uncle $crooge

 

hi,
As you can see I did edit my 330R back to 680R, like you I did my sums again!.
LED current approx = (1.25/680) * 10 =18mA

 

I just ran a test as theunit is presently configured. The Supply voltage from the 12V 2A charger I'm using was 14.69VDC
I connected the Fluke 87 set for Milli-amps record, between the LED and the Opto in series and energized the board.

Results: Ampreage spiked to 11.99ma and drop to a low of .5 after I left it on a while. Input voltage droped to 10.60VDC. I cycled the 0-1V siganal to drive the LED#10 (not LED#11 as in the misprint on the drawing) OPTO on/off several times with the same results. I also checked the LED current when I cut if off and got a sharp rise in current to around 11.33ma. The injector did energise and cut on/off. I hope this helps with what we need to know as far as what the current draw is on the LED side of the OPTO. I'll get a 12VDC battery to connect and retest later today, I do not think using the 12V 2A is working too well. I hope this does not smoke something on the board side of the O2 panel.

 

I think your circuit design has many parts on the edge of working or not working.
I wouldn't use it to protect the expensive pistons of my car from blowing up.

__________________
Uncle $crooge

 

hi Floyd,
Noted the test result on the opto driver.
As the PMOSFET and Solenoid are switching when you expect them to, it sounds as though you have enough opto drive output.

While you are getting a 12V battery [car type??] I would recommend that you get a vehicle type fuse holder and say a 15A or 20A fuse, connect the fuse in series with wire going to the positive terminal of the 12V battery.

I'm sorry if I appear to be telling you how to suck eggs, BUT lead acid batteries will not tolerate shorts, be careful.

Lets know how it goes,
[ I'm going to put my fingers in my ears, close my eyes and hum very loudly]

Eric

 

I realize at the present time it's "on the edge", my plan is when I get it working to try it w/o the turbocharger and such, natrually asparated until I'm comfortable it works.
As far as telling me how to suck eggs, there's nothing wrong with telling me or anyone else what you think if it's ment to be productive or related to safety. It shows me you care! Thanks! I'm working on a new drawing with the info provided to see if it looks like were on the same page, I'll post it just as soon as I finish it. Thanks.

 

Your circuit will be a lot more reliable if you replace the 4N25 opto-coupler that has a minimum current transfer ratio of only 0.2 with a 4N35 that has a minimum current transfer ratio of 1.0.

Then replace the TIP42 transistors with TIP125 darlington transistors or P-channel Mosfets.

__________________
Uncle $crooge

 

I agree with you guys on using 4N35s over the 4N25s, I've got 47 more of them right now, later I will replace them because of the reliability issue.
Alright, here's what I've come up with using Mosfets, check it out and see if it looks about right. Is there supposed to be a blocking diode across the Mosfet source and drain as it looks in the drawing posted by ericgibbs?

I'll post the other one with the 8 Mosfets in a little while.

Attached Images

Modified_injector_with_1_MOSFETs.JPG (80.2 KB, 8 views)

 

hi,
The diode across the Source/Drain is part of the internal structure of that MOSFET.
Looking forward to the final drawing.
Signing off for today.
Eric

 

Since MOSFETs are voltage controlled and not current controlled, the 4N25s can be just biased for voltage swing, and a really high pulldown resistor can be used. You'll still get the voltage swing needed to turn on/shut off the MOSFET (only issue being turn on/turn off time to charge/discharge the gate cap, but that's usually in the few pF range and virtually negligible (calculate the RC constant with the pulldown and the gate cap to doublecheck, I doubt it would be anywhere in the "care" region.)

 

No.
A power Mosfet has a total gate capacitance (including the Miller effect) of many thousands of pf. The 10k gate resistor will cause it to turn on slowly.

__________________
Uncle $crooge

 

OK, here's what I've come up with. My only other question is where the 10K R3 is located, do I use just that one or do I need to have one across each M1-M8 Mosfets? ALSO NOTE that in this drawing I did not add the "S" select switches but I do plan to add them as in the drawing with the one Mosfet design. The program I'm using is a student version and will not let me use more than 50 componets doing the drawings.

Attached Images

O2 w-8 Mosfets.JPG (89.5 KB, 8 views)

 

How slow is slowly? I want things to happen really fast. The faster the better I would think.

 

Here's a picture of the PCB w/o the transistors or Mosfets being in the picture, as it is presently assembled. I realise the 10R looks a "little used", that's what happens when you land a wire wrong! But it still works and is within specs, I'll replace it later on.

Attached Images

O2 MonitorDriver newer pic.jpg (546.9 KB, 8 views)