Thanks, its doing crazy things on the graph can you send me all the rest of the files that worked already???

Thank you.

** never mind I fixed this....

 

Finally!!

Schematics:

Thank you Sgt!!!!

 

You still have a problem.
The saturation current for the inductors is 1.60 Amperes, and the peak current is 2.08 Amperes.

Now in the inductor setup I created, L1 represents L1 thru L3 in parallel. Inductance in parallel divides like resistance in parallel.

Nominal inductance per coil is 9.9uH, so three in parallel is 3.3uH.
Amperage capacity per coil is 1.6a to get to saturation, so 1.6*3 = 4.8 Amperes.
You really want to avoid saturation, as that will result in nothing but heat and problems.
So, the current needs to be limited to around 4.5A total.

Referring back to the datasheet:
http://cds.linear.com/docs/Datasheet/1871fe.pdf
the typical Vsense threshold is 150mV, but looking at the plot on page 6, the threshold increases slightly over temperature. So, let's just say 152mV to have some margin for error.
Since we know that the sense voltage causes the regulator to turn of the MOSFET at around 152mV, and we know that our current limit is ~4.5A through the L1 inductor, we can calculate:
R=E/I
Rsense(R1) = 152mV/4.5A = 0.152/4.5 = ~33.777...mOhms, or 0.033777... Ohms.
That's not a standard value of resistance. 33m Ohms is. Let's try that.
I=E/R = 152mV/33mOhms = .152/.033 = 4.60606... Amperes. That's still below your 4.8A peak; about 96%. However, for safety it would be better to go to 0.036 Ohms, which is a standard value.

I=E/R = 152mV/36mOhms = .152/.036 = 4.222... Amperes. This also gives the regulator more time to discharge the gate.

Tip: You can plot the power dissipation in a component by holding down the ALT key; (note the cursor changes to a thermometer) and clicking on the component.

Once the simulation has completed, you can also show the average and RMS power by holding down the CTRL key, and clicking on the name of the signal (or formula) on the plot. You can also get averages and other calculations for current and voltages by using Ctrl+Click on the name of the signal.

[eta]
The designers of the circuit wasted an opportunity to make the circuit more efficient; they really should have taken the SENSE input from the drain of the MOSFET, and used a MOSFET with an Rds(on) of around 36m Ohms. That would have eliminated the need for R1, and as an additional benefit, if the MOSFET started to get warm, the Rds(on) would go up, so the ON-time would drop because the sense voltage would build up more quickly.

 

You've copied something wrong on the Q2 gate driver circuit.

It was hard to understand with it drawn upside down and backwards, so I flipped it around in LTSpice and simulated it. See the attached.

Current through Q3 and Q4 is ridiculous when CtrlIn goes high; >900mA through Q4's BE junction and >1.1A through Q3's base, >900mA through the collector, and >2A through the emitter. They'd be fried immediately.

 

I've moved some things around (resistors) and changed some values; now it works and won't burn up.

You need to do more research on your board to figure out what the circuit actually is.

 

In the board I can see clearly that the resistor is between the emitter to base.
But I might be wrong about the type of the transistor (NPN/PNP).
this copy is "as is" it was easy to copy it, but I can be wrong about the transistors.

About R1 thank you I will change it to 0.036ohm.

I am sure that the Q3,4 circuit is for driving the contrrol pin from the uC.
How can I found out what Q3,4 are????

 

In the board I can see clearly that the resistor is between the emitter to base.

You need to use reference designators, or I don't have a clue which transistor or resistor you're talking about. Reference designators, also called "labels" are like R1, Q3, C3, L1 etc.

But I might be wrong about the type of the transistor (NPN/PNP).

You might be wrong about what terminal on the SOT-23 part is the base, emitter and/or collector.

this copy is "as is" it was easy to copy it, but I can be wrong about the transistors.

I don't think you had a straight-forward match for either transistor.

About R1 thank you I will change it to 0.036ohm.

That's the minimum standard value that I calculated would be safe. It seems to work in the simulation, too.

I am sure that the Q3,4 circuit is for driving the control pin from the uC.

It's for driving the gate of Q2.
But, I'm convinced that it is not yet correct.

How can I be sure on this?

Try some measurements.
You should be able to figure out whether it's PNP or NPN, and whether both the base-emitter and base-collector junctions are good by using a diode test function on a multimeter.

See this page in our E-book:
http://www.allaboutcircuits.com/vol_3/chpt_4/3.html

It may just be that either Q3 or Q4 are bad.

 

Ok I have "diode check" but I didn't understand how to analyze the results.

The results have been checked on the board and may be effected by the components on it.
when probes shorted : 2
Q3:
(+)B (-)E: 334
(-)B (+)E: 373
(+)B (-)C: 106
(-)B (+)C: 104
(+)C(-)E: 320
(-)C (+)E: 360

Q4:
(+)B (-)E: 675
(-)B (+)E: 1550
(+)B (-)C: NOP
(-)B (+)C: 1551
(+)C(-)E: 586
(-)C (+)E: NOP

checked twice same results.
Can you see which of them PNP/NPN/burned??

Thank you.

 

That's pretty strange.
I expected to see the base-emitter and base-collector junctions to measure ~500-750 one way, and open the other.

Q3's acting like it's been overheated or something, and is basically shorted.

Q4 isn't acting properly either. Not sure what to tell you there.

 

Why don't you power it up, and measure the gate voltage on Q2 to ground? (source terminal should be grounded).

You might find that Q2 is turned on all the time (gate voltage >5v; if <2v, then the MOSFET would be off.)

Then power it down, and try measuring the resistance from Q2's source terminal to the gate, and the gate to the drain. If Q2 were blown (shorted drain to gate, a frequent failure mode), that could take out Q4.

[eta]
Just some more notes..

Generally, bjt's (bipolar junction transistors) test out like two diodes joined at the anodes for NPN, and cathodes joined for PNP.
So, if you put the positive lead on the base of an NPN, you should measure around 630 or so mv from the base to the emitter OR collector, and if the negative (black) test lead on the base, you should read infinite (open) from the collector or emitter.

 

Ok I did measurments and here the results.
Control Pin is OFF
Q3:
Base-0v
Collector-12v
Emitter-0v

Q4:
Base-12v (connected to Q3 collector)
Collector-0v (connected to Q2 gate)
Emitter-12v

Control Pin is ON
Q3:
Base-5v
Collector-0v
Emitter-0

Q4:
Base-0v (connected to Q3 collector)
Collector-12v (connected to Q2 gate)
Emitter-12v

I think Q3- is NPN, and Q4 is PNP as the schematics.
Please analyze this and tell me which is NPN/PNP.
Thank you.

 

Define "ON" and "OFF" for the control pin.

What do you do, apply 5v for ON, 0v for OFF, or short two pins together? I really don't know.

 

I give 5v to the Controll pin from uC. (ON-5V, OFF-0V)
This pin open the Q2 gate for discharging the capacitors on the solenoid.

 

OK, looks like Q2's gate is getting controlled properly.

If you are not getting 100-110v on the HV caps yet, then there is a problem with the boost converter.

If you are getting 100-110v on the HV caps but the solenoid won't fire, then Q2 is no good.

I really don't know how that circuit is working the way you have it drawn..
I think there is a resistor between the base of Q3 and your control input, but instead you have a 510 Ohm resistor connecting the base and emitter, and placing control in directly on the base; that would lead to VERY high base currents for both Q3 and Q4.

If the 510 Ohm resistor were between the emitter and ground (instead of between the emitter and the base) then it would work properly.

 

The Input circuit is working correctly.
The capacitors charge up to 100v-110v.
RR is from the base to the ground, but the emitter is also connected to the ground.
I think they connected ressistor from the base to ground for real "0v" (if not connected then 0v, and not somthing else)

I have checked the Q2 on breadboard and it's not working correctly it is always open the gate with 5V or not, it is burned.

Can you find for me equivalents for the 62N25C??
Thank you.

 

Here's a datasheet for it; Fairchild part:
http://www.fairchildsemi.com/ds/FQ/FQA62N25C.pdf
VDss=250v, Rdson=35m@Vgs=10v, Id=62A, Qg=100nC, TO-3PN case.
Fairchild Semiconductors' page for that part:
http://www.fairchildsemi.com/pf/FQ/FQA62N25C.html

Mouser stocks them; $5.58:
http://www.mouser.com/ProductDetail/Fairchild-Semiconductor/FQA62N25C/?qs=TpDLOUGfCJw8LHwIBQacaQ==

Newark has them in stock, ~$6.50+shipping:
http://www.newark.com/jsp/displayPr...mi&s_kwcid=TC|14641|fqa62n25c||S|b|5706369141

 

Ok That's it, I am going to buy it...
Thank you for all your help!!! you are the best!!
The schematics is the most important thing, now I can put it in my project.

And again Thank you.