using TIP42 PNP paralleled

[boxer4]

Err... Two questions:

1. Are you controlling supplementary fuel injectors or are these solenoids going to bleed boost? Just want to be certain, for issue 2:

2. O2 sensors on startup will read 0V as if it's running really lean. O2 sensors in closed loop/idle can possibly hit fairly low voltage when operating normally. Are you going to inject more fuel or bleed boost in these situations?

Optoisolators are really weak. You may have to use another external transistor (or darlingtons as posted earlier, which is the same thing) to amplify the signal a bit so that enough base current is drawn from the TIP42 to fully turn on the solenoids. But I'm thinking there's another, better way to hook up these power transistors for solenoids without the (expensive, and probably completely unnecessary) optoisolators here. However with the first two questions I don't know yet...

Oh, I guess since people are going off topic here, I'm a DSM'er and a WRX'er... though my faith in 4G63T has diminished due to my DSM's state...

 

[TheEquineFencer]

Err... Two questions:

1. Are you controlling supplementary fuel injectors or are these solenoids going to bleed boost? Just want to be certain, for issue 2:

2. O2 sensors on startup will read 0V as if it's running really lean. O2 sensors in closed loop/idle can possibly hit fairly low voltage when operating normally. Are you going to inject more fuel or bleed boost in these situations?

Optoisolators are really weak. You may have to use another external transistor (or darlingtons as posted earlier, which is the same thing) to amplify the signal a bit so that enough base current is drawn from the TIP42 to fully turn on the solenoids. But I'm thinking there's another, better way to hook up these power transistors for solenoids without the (expensive, and probably completely unnecessary) optoisolators here. However with the first two questions I don't know yet...

Oh, I guess since people are going off topic here, I'm a DSM'er and a WRX'er... though my faith in 4G63T has diminished due to my DSM's state...

Ans for 1 &2
Supplemental fuel added under Full throttle w/boost w/engine O2 sensors heated at operating temp.
Someone metioned using MOSFETs, how can I wired these instead of the
TIP42s? I'm looking at the data sheets for a FDS6685 Made by Fairchild, would something like this work better? They said they work off voltage unlike transistors that work off current. How might I connect one of these for switching the load? I'll check back later today, I've got to read a chapter on PLCs and get to school. catch you guys later, Floyd.

 

[boxer4]

Confidence is something you get before you fully understand the situation... I see your issue now. The biggest problem is that your LEDs are adjustable brightness, and that complicates things. The Optoisolators is probably the easiest solution after all, since now you're decoupled from current and voltage (not saying there aren't other solutions.)

You will need to have multiple stages of amplification, it looks like the 4N25 can't really sink enough current to fight the pulldown. I'd probably use a secondary transistor - so 3 transistors total. Optoisolator, a GP transistor, say 2N3904, and the (crappy hFE=40) TIP42 all in common-emitter configuration (tip42 being PNP, E=VCC). The Optoisolator has a pullup of about 10K, Collector connects to base of the 2N3904. The 2N3904's collector will pull the TIP42's base down through a 390 ohm resistor, providing the current needed to saturate that transistor.

Sorry for a text description of how to connect but I think this should turn on a typical solenoid at saturation whenever the opto is "dark".

 

[TheEquineFencer]

Confidence is something you get before you fully understand the situation... I see your issue now. The biggest problem is that your LEDs are adjustable brightness, and that complicates things. The Optoisolators is probably the easiest solution after all, since now you're decoupled from current and voltage (not saying there aren't other solutions.)

You will need to have multiple stages of amplification, it looks like the 4N25 can't really sink enough current to fight the pulldown. I'd probably use a secondary transistor - so 3 transistors total. Optoisolator, a GP transistor, say 2N3904, and the (crappy hFE=40) TIP42 all in common-emitter configuration (tip42 being PNP, E=VCC). The Optoisolator has a pullup of about 10K, Collector connects to base of the 2N3904. The 2N3904's collector will pull the TIP42's base down through a 390 ohm resistor, providing the current needed to saturate that transistor.

Sorry for a text description of how to connect but I think this should turn on a typical solenoid at saturation whenever the opto is "dark".

The LEDs are not adjustable brightness per say, the adjustment on the PCB is to get the 0-1V signal dialed in so the .5V signal is at the middle of the LEDs #1-#10 on the outputs going to the Optos and to compensate for the O2 sensor input. I do see what you mean abot the OPTOs though. I'm going to do a liitle more research into the Mosfets and maybe FETs and see if they might work better. I'll look around here and see if I have some 2N3904 laying around or can take some out of something. I've become a bit of a pack rat scavanger for electonic parts. I had a school mate over and he said I had more than Radio Shack. (US hobby electronics vendor)
Basicaly it sounds like what I need to do from what you are saying is use the 2N3904 sort of like what I'm trying to do with the TIP42 for the injector, but apply the load to the base of the added transistors because the OPTO cannot carry enough load to the base of the TIP42 I have now?

 

[audioguru]

Transistors have current gain. A small base current makes a high collector current.
Your injectors are 14 ohms so they need 12V/14 ohms= 857mA of current.
The 1.6k resistor causes a base current of 11.3V/1.6k= 7mA.
But the minimum gain of a TIP42 is only about 20.
The TIP42 amplifies it at least 20 times to only 140ma.

As i said before, use TIP125 PNP darlington transistors that have a minimum current gain of 1000. Then the circuit will be exactly the same as you have now but the injectors will operate at full current. A darlington transistor has two transistors inside connected to boost the current.

 

[Hero999]

We have the tiny SMART car in Canada now. It sells for the same price as a "normal small (much bigger) car". I laugh when I see one.

All the ones I've seen are petrol.

They are supposed have good fuel economy I've heard aroung 60MPG but they're slow and have a top speed of only 85MPH. I'd rather stick with my bike which will do 120MPH and is just as good on fuel.

 

[Nigel Goodwin]

All the ones I've seen are petrol.

They are supposed have good fuel economy I've heard aroung 60MPG but they're slow and have a top speed of only 85MPH.

Well the maximum speed in the UK is 70mph!

But they are basically a car for around town, easy to drive and cheap to run, pretty nippy but no high top speed. I've never driven one, but they are semi-automatic I think?.

As suggested they are a bit pricey though, but aren't they something to do with Mercedes?.

I'd rather stick with my bike which will do 120MPH and is just as good on fuel.

Obviously a four stroke!.

I've had a couple of Suzuki's in the past, a 350cc (twin) and a 380cc (triple), both two strokes, and both seriously fast acceleration - top speeds about 110mph. Fuel consumption though! - you might make 30mph on a good day!

An old friend of mine used to have a BMW, 900cc horizontally opposed twin, before that he bought a Yamaha 750cc four cylinder. We were talking one day and he said the Yamaha was absolutely rubbish - it didn't perform very well for a 750, and it had poor fuel economy. As a result he changed it for the BMW - this would cruise all day at 120mph, and no matter what you did still returned 75-80mph - pretty impressive machine!.

 

[audioguru]

The max allowed speed in Canada is 100km/h but I had my turbo car doing over double that and it was still accellerating. I didn't get caught speeding.
My new cars have a darn governor set to the tire spec's. It is dangerous to pass on a two-lane road when a big truck is coming at you and the governor cuts off the engine.

 

[TheEquineFencer]

Transistors have current gain. A small base current makes a high collector current.
Your injectors are 14 ohms so they need 12V/14 ohms= 857mA of current.
The 1.6k resistor causes a base current of 11.3V/1.6k= 7mA.
But the minimum gain of a TIP42 is only about 20.
The TIP42 amplifies it at least 20 times to only 140ma.

As i said before, use TIP125 PNP darlington transistors that have a minimum current gain of 1000. Then the circuit will be exactly the same as you have now but the injectors will operate at full current. A darlington transistor has two transistors inside connected to boost the current.

Thanks, that's put pretty simple that I can understand, I'll see if I have some around here if not I'll get some and try it. When looking at specs on a data sheet for something like this what do I need to compare? Just get a TIP125 and a TIP42 sheet and see what the differance is?

 

[audioguru]

When looking at specs on a data sheet for something like this what do I need to compare? Just get a TIP125 and a TIP42 sheet and see what the differance is?

It is obvious that the TIP125 darlington transistor has a current gain that is at least 50 times more than the TIP42 transistor.

 

[TheEquineFencer]

It is obvious that the TIP125 darlington transistor has a current gain that is at least 50 times more than the TIP42 transistor.

From your earlier post you said:

A TIP125 darlington PNP would work fine with the 1.6k ohm resistor to turn it on completely.
With darlington transistors, the 1.6k resistor sould be changed to 12k so that one opto can drive 8 of them in parallel.

So I need to replace the TIP42 with a TIP125, replace the 1.6K ohm resistor with a 12K ohm resistor, then I can install a total of 8 TIP125s paralleled, with the all the emitters connected together and then all the bases connected together, each collector connected to individual injectors and it should work, am I correct? Now to I have just (1) 12k across all the bases or one at each TIP125 and connect the output from the OPTO terminal #4 between the 12K resistor and the base? I'll order the TIP125s later tonight. Thanks, Floyd.

 

[ericgibbs]

hi Floyd,
If I undertstand correctly, you want to energise all 8 solenoids at the same time using a common signal?

Have you considered using a single power MOSFET, with all 8 solenoids connected in parallel. [dont forget the back emf clamp diode]

From a 12V supply, each solenoid has a resistance of 14R, so thats Is=12/14=0.86amps each. Times 8 solenoids, say about 7amps total.
This level of current for a MOSFET would not be a problem, you may have to use a single intermediate driver between the output of the opto and MOSFET.

This solution would cheaper, 'smaller' and easier to build than using 8 medium power transistors.

Eric

 

[TheEquineFencer]

You are corect, I want to energise all 8 Solenoids at the same time from the same signal from the OPTO. That's what I'm looking into now, do you know how I can do this? If I have it it in a running car the Alternator output will probably be closer to 14.2VDC, that would put it around 1.1A for each injector. when I tested the Injector solenoids with a 12V battery with my Fluke, max 10A capability, it went to OL, then recovered to around .400A and dropped to around .014A when the coil was saturated. I've not ordered any parts yet, I'll wait until I here back from you before I proceed. If you take my drawing and shoot me one with the plan you have in mind I'll give it a shot and see if I can get it to work. I was wondering if it would help if I put a resistor in series or parallel with each solenoid coil to change the load on the coils. Yes I know I need a clamping diode for the solenoids. I was wondering if a MOV might work better. I was thinking it might cause the solenoid to close faster by dropping the inductance spike faster.

 

[ericgibbs]

Hi Floyd,
>>If I have it it in a running car the Alternator output will probably be closer to 14.2VDC, that would put it around 1.1A for each injector.
This would be about what I would expect.

>> when I tested the Injector solenoids with a 12V battery with my Fluke, max 10A capability, it went to OL, then recovered to around .400A and dropped to around .014A when the coil was saturated.
These test results do not make any sense for a 14ohm resistance solenoid coil?

You say that the solenoid coil resistance is 14R, is that correct?
If you are not sure, use a ohm meter to measure the solenoid resistance, its value will determine the current it draws.
If you have a datsheet or part number for the solenoid, please post it.

>> I've not ordered any parts yet, I'll wait until I here back from you before I proceed. If you take my drawing and shoot me one with the plan you have in mind I'll give it a shot and see if I can get it to work.

Dont order any components until we can finalise the total current drawn by the 8 solenoids.

>> I was wondering if it would help if I put a resistor in series or parallel with each solenoid coil to change the load on the coils.

If the solenoids are designed to work from 12/14Vdc adding resistors will not help.

>> Yes I know I need a clamping diode for the solenoids.
I was wondering if a MOV might work better. I was thinking it might cause the solenoid to close faster by dropping the inductance spike faster.
I would recommend a clamp diode.

Eric

EDIT:
Important: is the solenoid coil isolated from the solenoid casing?, NOT connected to the solenoid metalwork.
Look at the datasheets for the VNP20N07 and STF20NM50D MOSFET's

 

[TheEquineFencer]

No data sheet for the injector coils, I'm 100% sure of the reading on the resistance of the coils in the injectors, all are 14 Ohms. They maybe pull and hold type injectors, that use current limiting when they are held open. Does not really matter right now, at the most they might be held for 13-20 seconds at the most, only when I'm at full throttle under boost. I can figure out how to do pull/hold current limiting later on. I've got a file saved somewhere I found on how to do it with the injectors.
Yeah, it kind of blew my mind when I saw what the meter read when I tested them, I did it several times, but they are 14 Ohms. When my meter went OL I thought I had blown that small fuse that is hard as heck to find in the back of it, then it started reading around .400A and dropped to .014A. I did this test a few times to verify the results.

 

[TheEquineFencer]

I'm heading up to the house now to get a bite of breakfast and get to school, I'l try to check back when I get out around lunch EST. I've got a job interveiw this afternoon, so it might be later on before I can check back, thanks, Floyd.

 

[TheEquineFencer]

Replacing the Tip42s with maybe a FDS4935A

Hi Floyd,
>>If I have it it in a running car the Alternator output will probably be closer to 14.2VDC, that would put it around 1.1A for each injector.
This would be about what I would expect.

>> when I tested the Injector solenoids with a 12V battery with my Fluke, max 10A capability, it went to OL, then recovered to around .400A and dropped to around .014A when the coil was saturated.
These test results do not make any sense for a 14ohm resistance solenoid coil?

You say that the solenoid coil resistance is 14R, is that correct?
If you are not sure, use a ohm meter to measure the solenoid resistance, its value will determine the current it draws.
If you have a datsheet or part number for the solenoid, please post it.

>> I've not ordered any parts yet, I'll wait until I here back from you before I proceed. If you take my drawing and shoot me one with the plan you have in mind I'll give it a shot and see if I can get it to work.

Dont order any components until we can finalise the total current drawn by the 8 solenoids.

>> I was wondering if it would help if I put a resistor in series or parallel with each solenoid coil to change the load on the coils.

If the solenoids are designed to work from 12/14Vdc adding resistors will not help.

>> Yes I know I need a clamping diode for the solenoids.
I was wondering if a MOV might work better. I was thinking it might cause the solenoid to close faster by dropping the inductance spike faster.
I would recommend a clamp diode.

Eric

EDIT:
Important: is the solenoid coil isolated from the solenoid casing?, NOT connected to the solenoid metalwork.
Look at the datasheets for the VNP20N07 and STF20NM50D MOSFET's

I've done a lilttle bit of research into Mosfets, do you think I can use (4)
FDS4935A Dual 30V P-channel PowerTrench MOSFETs in place of the TIP42?
If I'm correct I would take Pin #1 on the FDS4935A and connect it where the Emitter is now, #2 pin to where the Base is now, and #8 to where the collector is. If I'm correct in the operation of this Mosfet it should replace the TIP42. Then I would add 3 more of these and tie all four Source terminals, to the same 12V coming in, all four Gate terminals to the same base from the OPTOs and then connect a single injector to each of the Drains coing out of the FDS4935A dual gate Mosfets. Then the Mosfets should stay on until the optos put 12V to the Gate. Am I correct in my theory?

I just read the EDIt, yes it appears the injector is isolated, I've downloaded a data sheet for the VNP20N07, I'm waiting for the other to download. The first it looks like I might have to buy 50 minumin order from Digi-Key, I'm not sure of the other one, Digi-Ket did not list it by that number. It'll be late Friday night before I can check mail again< i've ben just getting 3-4 hours sleep between working on this project and shool work, I've got to get some sleep tonight and be back in class again at 8am EST, USA, Say tuned, this looks like we're getting somewhere! Thanks, Floyd.

 

[boxer4]

Yeah they should work. You will likely need to increase that 1.6K resistor much higher to ensure that the MOSFETs turn off properly. IIRC at 1.6K the optoisolator is struggling to pull up all the way, and the MOSFET will be partially on, which you don't want (you'll probably kill them especially the ones w/o heat sinks.)

Since it's a fairly slow circuit (where milliseconds is probably OK) a pulldown of something in the tens of Khm: should allow the opto to pull the gate high enough to shut off the FET.

But yes, Source <-> Emitter, Drain <-> collector, and gate <-> base. Just watch for ESD on the device!

 

[ericgibbs]

hi Floyd,
Looking thru your drawing, one or two queries.

What current are you expecting thru opto TX diode when LED11 is lit?.

What current are you expecting thru the opto RX transistor when LED11 is lit?
Check the opto current transfer function.

You do realise that the TIP/Solenoid will be OFF when only LED11 [last LED] is lit.

If you plan to switch 8 solenoids, ie about 10Amp, have you checked that the two switches in series
with solenoids can handle 10Amps?. [ 'full-opon' and 'on-rise' contacts]

I believe that pin #3 of the LM3914 is going to the wrong end of R1 [10K] ,also whats the purpose of R1?
The LED current sinks of the LM3914 are constant current, fixed by R2 [680R]

Where is the 0V/Gnd connection for pin #8/#2 of the LM3914?

Who drew the circuit?

Eric
EDITED: viewed the pcb layout you posted on another thread, from which I was able to clear some of the questions in my post.
re-font as italic

 

[ericgibbs]

hi Floyd,
Have redrawn your diagram, with P MOSFET [1 stage shown]

Eric