little random post

[robertisaar]

not abused?

you have to put foot to floor to reach 70MPH with those 3.1/3 speed U-vans. they are scarily underpowered.

ended up not finding any (official or otherwise) LB8 dynos... didn't bother looking for the iron LH0 yet.....

maybe just finding a non terminal cancer S10 w/2.8/5 speed locally is enough of a challenge ATM... then instead of messing with the gen 1 stuff, just immediately jump up to a 3400 or 3500..... ????

[Superdave]

2 thumbs up for a 3500

This is what a local guy is doing to his... http://qcaautoforums.proboards.com/i...y&thread=12927

you'd probably have to do very similar cross member mods for a 3500.

[robertisaar]

assuming i really get into this, probably more than that...... i'm thinking Syclone killer using less displacement.

[caffeine]

you'd probably have to do very similar cross member mods for a 3500.

Any chance you could just use the 2.8 S10 oil pan?

[robertisaar]

the gen3s have the mains cross-bolt through the pan, so i wouldn't even attempt it.

[robertisaar]

got bored, decided i couldn't find a good enough version of any of this online, so i made this spreadsheet... ODS and XLS versions.



use or abuse for whatever nefarious plans you may have.

[robertisaar]

http://www.teglerizer.com/fi/megasquirt/ms_ignition_module.html

interdasting.

the second to last picture has my interest though.... VB921 ICs are being used as coil drivers(in this case, individual coil control is being employed, kind of like the newer 3500 and 3900 60V6 modules). so those are being used to do the current-heavy work, but here is where i'm drawing a blank: is the VB921 considered an ignitor? i have the datasheet for it open and i'm not seeing any implication that it steps up the voltage....

i believe that would mean that the ignitor would be internal to the coilpacks themselves? too bad i don't have a dead one around here to pull apart.

i've seen references online of how to test the ICM/etc, and almost all include a step of connecting a multimeter or testlamp to the coilpack "blades" and checking for 12V(since 12V is supposed to always be present and ground switched).

from the looks of it, the 60V6 ICM shouldn't ever have more than battery/charging system voltage available at those blades, but i think i need to investigate more to be 100% certain. i've also seen some references that the 60V6 ICM does all dwell internally, even though the PCM calculates and delivers a dwell signal to the ICM. something else i'd like to make certain of.

[robertisaar]

Linear EGR valves(96+) are operated at a frequency of 128Hz.

yep, unexciting(though well researched) plans are being had.

[pocket-rocket]

Makes sense (from a binary standpoint, lol).

[robertisaar]

I'm not entirely surprised.... all of the P4 and P6 OBD1 ECMs have the ability to operate everything PWM-capable at frequencies of 8, 16, 32, 64 or 128Hz, though nearly everything is run at 32Hz in the calibration. so, in theory, you could run a linear EGR setup on pretty much any OBD1 PCM(and there were a couple that did) so long as the program were changed to support it. where things get interesting is the current involved with driving it... the one I have sitting here measures 6.4 ohms. generally, these solenoids appear to run around a 60% duty cycle to be considered fully open, which will certainly exceed the capabilities of a single channel in the quad drivers(~.75 amps is what I would call a safe limit on them), about the only outputs that are paired up are for the EGR3 circuit.

run at a 100% DC, the EGR solenoid would run over 2 amps with a running engine, at 60%, the average current drops down to around 1.35 amps, which is within the safe limit of a paired quad-driver output. the P4 and P6 quad drivers do have current-limiting built into them, so if it's an excessive load, it should stop trying to drive it.... but from the looks of it, I'm just going to use a comparitor and a FET that can handle ~14 amps without sweating in parallel with the quad driver.... should keep the quad-driver from setting off a DTC while sending the majority of the load through the added MOSFET.

[robertisaar]

going off of this, my arduino came up with similar numbers. I wrote a sketch that allows me to command whatever duty cycle I want over a USB link(and it sends back the EGR position sensor signal), I just went through and did 0-100% in 5% increments. it took a while to find a method that allowed fully custom frequencies, but Timer1 does the job. also allows for 10-bit output resolution to match the 10-bit input resolution, which is nice.



I'm not sure if this solenoid is getting weak or the pot is flaking out, hard to say, but this certainly looks workable. the curves should change a bit once the additional hardware makes its way into the PCM, but this was entirely via the "quad-driver current booster" mentioned before.... the FET stayed right at ambient temps and the flyback diode barely got warmer than that. doing this without the flyback diode left a burn on the tip of my thumb from when I touched the FET.... the curves not exactly matching, but having the same general shape(ignoring the hangup at 45% duty cycle while closing) isn't a problem since this is done via closed loop, just have to account for additional change to the solenoid's duty cycle when changing the direction that the pintle is heading. that's assuming that this isn't happening just from being a worn out valve or that I was operating it at only ~12.4 volts.

but at least I feel confident about patching in linear EGR support now that I have the hardware side figured out.