Here is the modified schematic:

 

Hi Sgt,

I'm having a strange output behavior with the transistors.

I have a feeling it has to do with base current, maybe to Q2, but I'm not sure.

Here are the voltage values I'm getting on the collector of Q2

When the comparator non inverting pin is a higher voltage value than the inverting - 7.32vDC at Q2 Vc.

When the comparator inverting pin is a higher voltage value than the non inverting - 3.28vDC at Q2 Vc.

I believe it should be ~12vDC and 0v or .01v if the circuit is working properly, any idea what I've done wrong?

 

You show Q1 Vb, but that's pretty useless as you don't document what Ve is.

You should document Vbe and Vce instead.

I thought you measured 30mV across the 1 Ohm resistor. Is that still correct?

 

Sgt,

I made 2 different diagrams, for the 2 different states of the Opamp, as I was getting confused a little and this made it easier for me to keep track of the voltage values. I deleted the original attachment in the previous post, that had both voltage values in some cases, so that did not create any confusion.

I tried to double check each measurement, but if anything looks strange just let me know and I'll be happy to check it again. I did verify that there are .030vDC going across a 1Ω resistor, when connected to the sensor feedback wire that goes to the ecu.

 

OK, your 1st image.
Q1:
Ve=12.17, Vb=11.47, so Vbe=-0.7v.
Ve=12.17, Vc=12.15, so Vce=0.02v, or ~20mV. The transistor is well-saturated.

Now, for Q2:
Vbe is 670mV, and Vce=3.28. This indicates that the transistor is not saturated.

Since Q1's Vc is 12.15v, and Q2's Vbe is 670mV, V(R8)=11.48; R(R8)=3.3k, so Ib(Q2)= 11.48/3300=3.48mA.
(Did you follow that logic? If not, say so.)

This should be plenty to saturate Q2 (low Vce) for a load of 35mA (more than the 30mA you measured), unless perhaps you have Q2 installed incorrectly, or you did not measure the ECU load correctly.

Try measuring the ECU current to ground again using the voltmeter across the 1 Ohm resistor.

Then measure the ECU current to V+ (12v) using the 1 Ohm resistor.

Something is not quite right.

[eta]
I posted the original schematic in .png format. If you wish to edit it and add notes, that's fine - but please do so in .png format, as you went to .jpg format which is "lossy", and is not good for schematic editing.

 

Hi Sgt,

I did notice that the transistor was not saturated, but it's actually at 7.32vDC when it's supposed to be saturated (Q2 Vbe = ~.6v).

What confuses me is that when it is not supposed to be saturated and Q2 Vce should be about 12 or 13v, it's at ~3.28vDC.


End design (since I haven't been very clear on this as I've been learning over the process) goal is to have Q2 Vce = ~0 when the inverting is higher than the non inverting, and have Q2 Vce at 12 or 13 vDC when the non inverting is higher than the inverting of the comparator op amp.


What confuses me is Q2 Vce is 3.25 vDC when the inverting is higher than the non inverting? Is it because of R1?

Issue number 2 is that Q2 is not saturated when the non inverting eclipses the inverting?



(I will use png's from now on so the compression does not distort the image quality, sorry about that)

OK, your 1st image.
Q1:
Ve=12.17, Vb=11.47, so Vbe=-0.7v.
Ve=12.17, Vc=12.15, so Vce=0.02v, or ~20mV. The transistor is well-saturated.

Now, for Q2:
Vbe is 670mV, and Vce=3.28. This indicates that the transistor is not saturated.

^ I don't think it should be saturated here, I think it should be in cutoff?

Since Q1's Vc is 12.15v, and Q2's Vbe is 670mV, V(R8)=11.48; R(R8)=3.3k, so Ib(Q2)= 11.48/3300=3.48mA.
(Did you follow that logic? If not, say so.)

Yes I follow the current calculation

This should be plenty to saturate Q2 (low Vce) for a load of 35mA (more than the 30mA you measured), unless perhaps you have Q2 installed incorrectly, or you did not measure the ECU load correctly.

Try measuring the ECU current to ground again using the voltmeter across the 1 Ohm resistor.

Then measure the ECU current to V+ (12v) using the 1 Ohm resistor.

Something is not quite right.

[eta]
I posted the original schematic in .png format. If you wish to edit it and add notes, that's fine - but please do so in .png format, as you went to .jpg format which is "lossy", and is not good for schematic editing.

I remeasured the voltage across the 1ohm resistor a couple of times. I checked and Q2 is not installed backwards. The voltage measurement from the ecu sensor out wire to the ecu + voltage wire caused smoke and honestly I can't believe I didn't damage the ecu ... I'm guessing the current just went straight through the resistor (I have burned finger tips again even with that monster resistor) and through the wiring to the engine compartment sensor? Before I realized what was going on, it measured -6vDC?

On Q2, the emitter is to dc ground, the base is to Q1's emitter (through a 3.64k resistor) and the collector the ecu wire.

 

Small update:

I had an epiphany and decided to unplug the sensor (path to dc ground) portion and then compare the voltage difference between the ecu input voltage and the sensor output voltage (duh!) and it literally read 0.00v and -0.00v and it oscillated back and forth every few seconds at a DMM setting of 20vDC. If I had set it on the 200mV setting I may have seen something very minuscule show up, and I can do that if you'd like me to just let me know.

 

If you disconnect the ground to the sensor, there WILL be no difference between the two points.

 

Sgt,

I'm not sure I understand what you are asking, as with the sensor connected it offers a direct path to ground, which is why the jumper wire insulation started smoking and the resistor got so hot it burned my fingers.

Is there another way to test that without damaging my electronics?

If you disconnect the ground to the sensor, there WILL be no difference between the two points.

 

It sounds like you may have fried your ECU when you attempted to use the resistor to measure the current to +12v.

Did you try it on a wrong input, or what happened?

Smoke from the ECU is the last thing you want to see.

You thought that the input was supposed to go from 0v to +12v. Is that what you were seeing before, or is that just what you thought would be the correct input?

 

Fortunately I did not fry it, and the smoke was from the alligator clip wire/insulation I used. When thinking about the current path, current would actually go from the supply through the wire and right to the sensor (dc ground) and never actually go through the ecu because of its high resistance and pull up resistor, so I lucked out there.

I didn't use a wrong input (yet ), but I probably have explained how I thought it worked before, which may be different then how I've seen it work now.

Hopefully these 2 diagrams help describe what I've learned about this circuit? Let me know if this is how you've understood what I've tried to explain thus far.

*edit - the black rectangle at the bottom is the ecu*

It sounds like you may have fried your ECU when you attempted to use the resistor to measure the current to +12v.

Did you try it on a wrong input, or what happened?

Smoke from the ECU is the last thing you want to see.

You thought that the input was supposed to go from 0v to +12v. Is that what you were seeing before, or is that just what you thought would be the correct input?

 

I have some spare time tonight so I'm going to try and measure milliamperage going to Q2 base and maybe lower the value of R8.

My logic is that since it is a BJT it operates off of current and it should be getting 3.68mA in the scenario drawn below at Q2's base. With a beta of 100+ on the 2n2222, it should get 36+mA of current and completely saturate the Vce of Q2.



What has baffled me is that I have some voltage drop across Q2 when I should not, and not enough when I should have ~ all voltage dropped.

• What I see happening in my 2 examples are as follows: a small amount of current/voltage is getting to Q2 when the inverting pin is higher than the non inverting pin
• Not enough current is getting to Q2 base when the non inverting pin is higher than the inverting pin

I see two parts to fixing this, putting a diode in between Q1's emitter and Q2's base, as well as replacing R8 with a resistor value of 2k. My hope is that it will now fully saturate Q2 with 6 milliamps at the base of Q2 and that it will resist the .565vDC coming out of Q1 Ve to Q2 Vb that is allowing Q2 to operate in the active region when it should be in cutoff.

Any thoughts?

.

 

Well for some reason when I disconnect the base of Q2 and just apply 13v (on the test bench) to the collector and supply dc ground to the emitter, the transistor goes into the active area?

Voltage drops from 13v to 9v consistently without anything to the base of Q2 and with the emitter straight to ground? Maybe somehow it's acting as a voltage divider with the 3.9k resistor I have shunting the amperage going to Q2's collector? I tried a second 2n2222 and it did the same thing. Q2 Vc = Vce = 9.04v , Q2 Vbe = .57v (and that's with nothing connected to it) ???

How could Vbe have any voltage value without anything connected to Q2's base? Whatever is causing this is definitely the final piece to this puzzle and keeping me from actively testing this in the car.

 

Connect a 10k resistor from the base of Q1 to +V.
Connect another 10k resistor from the base of Q2 to GND.

That will help to ensure they are off when they are supposed to be off, and on only when they are supposed to be on.

 

Ok, new schematic of which I am excited to try tonight Thanks Sgt!

Connect a 10k resistor from the base of Q1 to +V.
Connect another 10k resistor from the base of Q2 to GND.

That will help to ensure they are off when they are supposed to be off, and on only when they are supposed to be on.

 

Well, I have gotten it working (yay!) but I'm a little confused.



Here is the back of the package, but I had to make pin 1 the collector and pin 3 the emitter?

Is this possible that the package was a misprint or I just lack some understanding about transistors (which is very possible)? I've quadruple checked it in both scenarios?

I thought that an NPN transistor will not conduct current until .5-.7v are applied to the base?

 

Don't believe ANYTHING written on a Radio Shack bubble package until you have found the manufacturer's datasheet for the specific product that you have in your hand.

You may have become confused about viewing it from the bottom vs viewing it from the top; I don't know.

You need one of these meters from Harbor Freight, or something similar:
http://www.harborfreightusa.com/usa...o?itemid=92020&CategoryName=&SubCategoryName=
It has a built-in transistor checker.
That particular unit isn't my favorite, but they go on sale from time to time for about $3, and that's hard to beat.

Here's my favorite for "little meters": http://www.apogeekits.com/digital_multimeter_dmm810.htm
I've had one of these for a decade, and have really liked it. The transistor test feature is much easier to use than the Harbor Freight type meters.

By the way, they might say they are rated for up to 500v - but I would never use a cheap meter like this for anything over 120v. It would be foolish to do so.

If you want to measure voltages over 120v, go buy a Fluke. They are a remarkable bargain compared to the cost of a funeral.

 

Hi Sgt,

I'll be buying one of your little meters tonight, < $10 is a bargain!



I've ran into something a little confusing ...

While monitoring the ecu sensor input signal wire I have noticed that it sometimes will be 13v instead of 0v upon startup.

Normally it is 0v from 0rpm to 6799rpms and 12~14v from 6800rpm to 8400rpm.

Three times I've noticed that it will be 12v as soon as I turn the key to on and it will stay at 12v for a few minutes and then go back to the normal pattern of 0v from 0rpm to 6799 and 12~14v from 6800rpm to 8400rpm.

I thought maybe it was engine temperature, but the car sat for 8 hours from last night until 6am this morning and it was 65F outside when I turned the key into the on position. It was at 0v.

I then let it sit for 8 hours and turned the key to the on position again at 2pm when it was 90F outside and it read 12v and stayed that way for approximately 2 minutes, until it went back to the normal behavior while I was trying to test for different things with my DMM. Can you think of why the ecu would do this? Maybe this is something normal a manufacturer would do? Can you theorize a reason for this?

I do not have a video of it, but I did take a video of it later with my phone, I remember you saying you thought there was a relay and based on this I think there might be because of what I've read about when the coil unloads, it causes a spike? Maybe that is related?

Here is the video


http://www.youtube.com/watch?v=xZuJam-4MpE

 

Scratch that, it appears I had a poor connection for R8 and somehow it manifested itself as intermittent ... hard lesson to learn, but definitely learned

I've been driving around for about 2 days and this has been working without issue now! Another 5 days and I'll consider it good to go.

 

Well, I'm glad that you figured out where you went wrong.

Frankly, I was at a loss to explain why it wasn't working.

This thread has gone on for (now) 120 posts, and it will be good to put it to bed.