A better Spark Out circuit.
07-01-2009, 08:55 PM
#41
Boost Pope
Thread Starter
iTrader: (8)
Join Date: Sep 2005
Location: Chicago. (The less-murder part.)
Posts: 33,023
Total Cats: 6,591
When the CPU is inactive, its output pins are hi-z. A PNP won't conduct unless it's got negative base current, and that won't happen with the CPU tri-stated.
Didn't happen.
Reply
0
0
07-02-2009, 11:03 PM
#45
In regard to using a PNP instead of the NPN transistor in the original circuit. Can't we connect the output of the CPU to the base of a PNP with its collector connected to ground and the emitter connected to the -ve side of the LED. Then if we bias the base to the ground with a resistor we should be able to drive the transistor to conduct when the output from the CPU is low/floating. The output to the ignitior should then follow the output from the CPU.
I couldn't find a tool to draw circuits on my mac, hence the wordy description of the circuit.
Disclaimer: I am not an electronics expert so pardon me if this is wrong or stupid.
I couldn't find a tool to draw circuits on my mac, hence the wordy description of the circuit.
Disclaimer: I am not an electronics expert so pardon me if this is wrong or stupid.
Reply
0
0
07-03-2009, 12:22 AM
#46
Boost Pope
Thread Starter
iTrader: (8)
Join Date: Sep 2005
Location: Chicago. (The less-murder part.)
Posts: 33,023
Total Cats: 6,591
My name's Joe, and I'm a PC. We've got an app for that. 
Seriously, I think this is what you & Abe are describing:
The reason I've had concerns here is that you need a small enough resistance to reliably turn on the transistor when the CPU is in la-la land, yet as the value of R goes down, the current which the MS is pissing away out of PTC0 / PTC2 to ground (when on) goes up. I've just been too damn lazy to look up the minimum IB needed to drive the ZTX553, which I'd assumed to be quite high, given its max IC of 1A (and the fact that PNPs usually suck in this regard). Having done so just now, I see that it is in the μA range.
Might work. Try it and find out. Take your pick- desolder a pair of transistors, clean out the holes, and replace them with two other transistors and a couple of tacked-on resistors, or add two additional transistors and resistors in the proto area.
Seriously, I think this is what you & Abe are describing:
The reason I've had concerns here is that you need a small enough resistance to reliably turn on the transistor when the CPU is in la-la land, yet as the value of R goes down, the current which the MS is pissing away out of PTC0 / PTC2 to ground (when on) goes up. I've just been too damn lazy to look up the minimum IB needed to drive the ZTX553, which I'd assumed to be quite high, given its max IC of 1A (and the fact that PNPs usually suck in this regard). Having done so just now, I see that it is in the μA range.
Might work. Try it and find out. Take your pick- desolder a pair of transistors, clean out the holes, and replace them with two other transistors and a couple of tacked-on resistors, or add two additional transistors and resistors in the proto area.
Reply
0
0
07-03-2009, 01:29 AM
#47
What program do you use PC?
That's the circuit I am talking about.
Can you calculate the resistor values please?
I am still building my MS2 for my 91.
I am in the process of trying to figure out all the different mods I want and how to assemble them.
I will try the PNP circuit while I assemble the spark output circuit.
Thanks.
Can you calculate the resistor values please?
I am still building my MS2 for my 91.
I am in the process of trying to figure out all the different mods I want and how to assemble them.
I will try the PNP circuit while I assemble the spark output circuit.
Thanks.
Reply
0
0
07-03-2009, 03:32 AM
#49
Elite Member
Join Date: Jul 2005
Posts: 6,420
Total Cats: 84
If the CPU output is 0-5V, then 470 ohms for the upper right hand (pullup) resistor is plenty, and 4.7k-10k for the base to ground resistor will be fine. The leftmost horizontal resistor is unnecessary.
Just about any TO-92 PNP will work, such as the 2N2905 or the 2N5551.
Just about any TO-92 PNP will work, such as the 2N2905 or the 2N5551.
Reply
0
0
07-03-2009, 08:46 AM
#50
Boost Pope
Thread Starter
iTrader: (8)
Join Date: Sep 2005
Location: Chicago. (The less-murder part.)
Posts: 33,023
Total Cats: 6,591
I use the free schematic editor that comes with the ExpressPCB package, for which Abe criticizes me on the grounds that Eagle is better.
The leftmost horizontal resistor is part of the standard 3.0 layout. It's R26 / R29 on page 7 (R3.0).
Bear in mind that the lower the better for the +5 pullup, since it dictates the current available to drive the igniters which as we've shown from past experience, are hugely capacitive. 220/330 is preferred here, as little as 100 is really ideal when you're dealing with a 2222 or similar.
Bear in mind that the lower the better for the +5 pullup, since it dictates the current available to drive the igniters which as we've shown from past experience, are hugely capacitive. 220/330 is preferred here, as little as 100 is really ideal when you're dealing with a 2222 or similar.
Reply
0
0
07-03-2009, 07:21 PM
#52
Boost Pope
Thread Starter
iTrader: (8)
Join Date: Sep 2005
Location: Chicago. (The less-murder part.)
Posts: 33,023
Total Cats: 6,591
In my personal opinion, the NPN design is more robust, and not significantly more complex.
Last edited by Joe Perez; 07-04-2009 at 01:50 AM. Reason: Schpelling (hey, I wrote this on an airplane while drinking scotch. Cut me some slack.)
Reply
0
0
07-03-2009, 07:22 PM
#53
Without thinking it out, basically, yeah - stuffing the different part in the same hole. Of course, a small, surface mount resistor would comfortably attach to the bottom of the board to pick up both pins, so no wires to run.
Well, the idea is for fresh builds, if it's put this part in instead of another, ok. You're right, tearing apart an existing board is more work and slightly risky (lifted pads, etc)
Actually, I've been using KiCAD - reminds me a lot of orcad. Eagle has really annoying limitations. And it's cross platform, so you can share with weird open-source/linux people. Maybe it works on mac? I dunno. Anyway - one package all the way through fab, why not?
Seriously, it's all about COPS, they are such a joy to work with. Everyone should buy them next time they need cables, costs about the same and you're set for life.
The reason I've had concerns here is that you need a small enough resistance to reliably turn on the transistor when the CPU is in la-la land, yet as the value of R goes down...
(blah blah a bunch of technical crap someone who actually knows what they're talking about would probably say)
Might work. Try it and find out. Take your pick- desolder a pair of transistors, clean out the holes, and replace them with two other transistors and a couple of tacked-on resistors, or add two additional transistors and resistors in the proto area.
(blah blah a bunch of technical crap someone who actually knows what they're talking about would probably say)
Might work. Try it and find out. Take your pick- desolder a pair of transistors, clean out the holes, and replace them with two other transistors and a couple of tacked-on resistors, or add two additional transistors and resistors in the proto area.
I use the free schematic editor that comes with the ExpressPCB package, for which Abe criticizes me on the grounds that Eagle is better.
Seriously, it's all about COPS, they are such a joy to work with. Everyone should buy them next time they need cables, costs about the same and you're set for life.
Reply
0
0
07-15-2009, 09:23 AM
#56
Elite Member
Join Date: Mar 2008
Location: Enschede, NL
Posts: 2,053
Total Cats: 12
Since my car is not driving anyway, I've decided to go ahead and order the parts I need for this mod.
2x 2N2222 TO18
Data sheet:
http://www2.produktinfo.conrad.com/d...219_2N2222.pdf
2x 2N2222 TO18
Data sheet:
http://www2.produktinfo.conrad.com/d...219_2N2222.pdf
Reply
0
0
07-15-2009, 09:40 AM
#57
Elite Member
Join Date: Mar 2008
Location: Enschede, NL
Posts: 2,053
Total Cats: 12
I've expaneded the write up from Matt. This is correct right?
We are aware of that circuit, but we haven't tested it for ourselves.
However, if you want to try putting it on an MSPNP, here's what you
would do.
1. Desolder the jumpers from PAD1 and PAD3.
2. Take a PN2222(A) transistor and connect it as follows:
* Emitter pin to ground (PAD4)
* Base pin to PAD1
* Collector pin to the jumper that previously connected to PAD1
3. Take a second PN2222(A) and
* Take a PN2222(A) transistor and connect it as follows:
* Emitter pin to ground (PAD4)
* Base pin to PAD3
* Collector pin to the jumper that previously connected to PAD3
4. Disconnect the coils and power up the MSPNP.
5. In MegaTune, change the Spark Output Inverted setting to YES and burn
to ECU.
6. You can now connect the coils.
The pull up resistor Joe mentions is already on
the MSPNP adapter card and stays in place.
However, if you want to try putting it on an MSPNP, here's what you
would do.
1. Desolder the jumpers from PAD1 and PAD3.
2. Take a PN2222(A) transistor and connect it as follows:
* Emitter pin to ground (PAD4)
* Base pin to PAD1
* Collector pin to the jumper that previously connected to PAD1
3. Take a second PN2222(A) and
* Take a PN2222(A) transistor and connect it as follows:
* Emitter pin to ground (PAD4)
* Base pin to PAD3
* Collector pin to the jumper that previously connected to PAD3
4. Disconnect the coils and power up the MSPNP.
5. In MegaTune, change the Spark Output Inverted setting to YES and burn
to ECU.
6. You can now connect the coils.
The pull up resistor Joe mentions is already on
the MSPNP adapter card and stays in place.
Reply
0
0
