[mudder]

I'm using the Analog Devices ADuM1201 for signal isolation for the first time. This isolator apparently handles very high speeds. I may try pushing the serial interface to 1 mbps and see how it goes.

ADuM isolators are incredible. We used essentially the entire lineup on various products at NI... you can push them way past their rated limits... Analog Devices is quite conservative with their minimum guaranteed specifications. Some of their ADuM modules even include an isolated, regulated DCDC, so you don't need an external one. We used that IC to achieve incredible density with a common drain NFET switch matrix design... something like QTY500 solid state DC-switched signals on a 4x3" PCB.

 

[mudder]

Finished the early stages of my triple 5v powersupply installation.

Why did you choose QTY3 5 volt supplies instead of QTY1 15 volt supply?
Series-connecting supplies in this fashion can cause instability issues if the total stack impedance (i.e. on the 15 volt output rail) differs substantially from the intermediate impedance (i.e. on any single supply). If you continue using this setup, you should consider adding additional capacitance across each individual supply, which will increase stability (assuming there's instability, which ultimately depends on each supply's control strategy).

Car idles perfect at 14.9v with blower fan on max speed, AC on, lights on, and hazards blinking.

14.9 volts is too high for a standard lead acid battery across all temperatures. Ideally you should vary the output voltage based on the battery temperature (as is done by the OEM converter).
FYI: You can force the OEM DCDC to output higher voltages by spoofing the signals on the DCDC's engine temperature line (RED/YEL, 0-5V PWM @ 10 Hz, where duty cycle reflects engine temperature). You can swap the DCDC for your QTY3 supplies (i.e. you don't need anything else from the PDU, since you already have your own rectifier).

Cleanup and redoing wiring with thicker gauge is the next step. This first step was proof of concept and function.

Please add a fuse the HVDC bus during this cleanup.

Bridge rectifier gets HOT at idle. Give the engine a buzz to around 2500 rpms, and the heat starts to go away. Idle just tries using too much amperage for this cheap chinese rectifier.

You should connect at least QTY1 of the big OEM HVDC capacitors (inside the PDU, about the size of a soda can) to the HVDC bus. This will reduce the rectifier heat at idle.

 

[M.Liston]

@mudder

In my thread (solved power supply issue) I detail what I want to do.

Better rectifier cooling, 450vDC capacitor at least 2000uF, better wiring, and a lithium battery pack.

15volts isnt going to kill the battery anytime soon, it survived 15.2-15.5v in my V8 Mercury for a couple years (previous owner installed it), before I replaced it and fixed the Mercury's voltage regulator.

It is supposed to be a maintenance free battery, but I popped the vent covers off and topped the cells off with distilled water. two were slightly low, the rest were okay. Couple good engine start zaps from a 175amp industrial charger, and the battery was once more passing load tests properly (generally a 75 amp load for 10 seconds, stayed above 11.5 volts)


If the battery was the small 51 group like factory, then I would leave the power supplies set to around 13.6-14.0 volts. Size helps shield the damage from overcharging.


For now, I plan on dialing in each power supply so they are each putting out as close to the same voltage as possible. Whether that be 4.5 to 4.8 volts each.

The overall goal (already removed the under hood battery by looping fusebox wire to charging post, and starter wire to charging post of fusebox) is to save a bit of weight, but still be able to enjoy a small stereo setup.

Going to keep it very basic. 4 speakers, 4channel amp, and simply have more bass-oriented speakers in the rear shelf. Plenty of room for deep speakers, now that the air chamber is gone.

 

[BruceinLA]

190,000 mile service. LOFR, front brake pads are down to 3mm so they will be needed next service.

 

[Atheos]

I went for a hunt in the salvage yard to scope out some roof mount antenna's, as the one I got on Ebay (for a Cadillac CTS) was missing some parts. Found a few I liked, removed and brought home two of them to choose from. I debated for a few minutes on which one to use, the black one was powered like the Insight, but it was missing the mast. That wasn't a dealbreaker as those are easy to source, but the shark-fin won me over.

Here's the result. Yes, I know my car is pretty ugly. I promise it's getting painted soon!

 

[Phantom Z3]

Bumped into this gem earlier today. Super clean but it sat in the sun for 4+ years so the RE92s are flat and the red paint is absolutely gone. Hopefully I can rescue it when the owner calls me back for the transmission I want for my other car.

 

[jeffffej]

New best tank...77.5 mpgs over 723 miles, all work commute which is 10 traffic lights + 10 stop signs one way. Very proud of this car! This is with e0 gas and EFAS and tires at 45psi, no other mods/tricks. I could have gotten into the 80s I am sure, however my mostly dead battery threw two IMA codes which required one deep discharge and two grid charges during this tank, lots of miles while recalibrating which means zero assist. But whatever I'm very happy with this result.

Honestly folks I'm tired of the patience required for hypermiling, I filled the car up with e10 and I'm taking it to 6000rpms on every drive now. Gonna have some fun.

 

[Arbus]

Saw it here…

There Was No Point To These Useless Automobiles

 

[mudder]

New best tank...77.5 mpgs over 723 miles, all work commute which is 10 traffic lights + 10 stop signs one way. Very proud of this car! This is with e0 gas and EFAS and tires at 45psi, no other mods/tricks.

Awesome! I'm about to finish my best tank in at least 5 years: 65.9 mpg. That should just barely be good enough to break 700 miles on a tank, but I'll probably be running on vapor (and IMA) when I roll into the gas station.

I could have gotten into the 80s I am sure, however my mostly dead battery threw two IMA codes which required one deep discharge and two grid charges during this tank, lots of miles while recalibrating which means zero assist. But whatever I'm very happy with this result.

You should replace your battery . Are you waiting for some king of new technology to come out?

Honestly folks I'm tired of the patience required for hypermiling, I filled the car up with e10 and I'm taking it to 6000rpms on every drive now. Gonna have some fun.

Same... I stopped caring about mileage in maybe 2014. I still run e0, though.

 

[jeffffej]

You should replace your battery . Are you waiting for some king of new technology to come out?

I will probably be sending you money soon. Sourcing some batteries...

 

[bbcb]

Had to leverage the latch fix from @Jesse in the 98F heat as my daughter is borrowing it to run to Austin for the weekend.

You can see the fracture in the baby blue oem piece.

It SUCKED reattaching the long arm, but the sweat rolling down my forehead made it just that much more fun.

But oh so worth it. 30 min total down time.

If you have a whirring latch, bite the bullet and get the 3d print fix from @Jesse.

 

[*sean*]

663 miles today and more to go. Insight was checked out by this awesome looking Civic with a quick wave when they exited.

Logged BCM channel voltage, current and temperature data while deliberately discharging the pack several times - 10x/sec log rate captures data at around 5 Hz/sensor.

Logged all OBD2 ECM live stream data at about 1/5 Hz/sensor. Definitely need to log directly instead of through OBD2.

 

[Honda hybrid442]

so 15:55 is the weak link stick. Nice rig I hope to see it soon.

 

[*sean*]

so 15:55 is the weak link stick. Nice rig I hope to see it soon.

That cell actually is fine! This is a great example of how measuring the tap voltages statically will tell you nothing or worse, steer you wrong, for certain issues. There is a low capacity cell in this pack. You just can't see it when you're not actually demanding assist.

 

[*sean*]

Worked on a "lean burn" indicator for Torque Pro. The 00-01 and 02-06 ECUs require different formulas because the ECUs use different sensing technology for determining air/fuel ratio. I found different formulas for my 2000 MT and my 2004 MT. I first tried the formula I worked out back in 2018 but Torque has changed since then and didn't like it; it was too complicated anyway. Updated the post with the formulas I worked out. No idea if it will work for anyone else. That post, with instructions: #41

The result:

 

[AP1ZE1]

sean, thanks for updating your formula. On my Torque Pro, I am in Lean Burn when when O2 1x2 and 1x3 are 0.0

 

[Balto]

I will probably be sending you money soon. Sourcing some batteries...

I have 3 sets...

 

[Natalya]

Thursday I put in new sway link arms. One had the boot torn open and it was rusted inside and getting crunchy. These are just rock auto ones, I think Delphi. I would have gone Moog but theirs appear to be discontinued.

 

[Joule]

Drove 400+ miles today. Ending mpg 59.4. No A/C usage to get those numbers at highway speeds in an unmodified CVT which was a tad on the brutal side in 90+ temps. I tried periodic cool offs but I was in the hills of eastern Ct. on I84, plus I 495 and 95 on my way to Maine and I found the A/C really saps the engine power I wanted for hill climbing.

 

[julianachos]

Thursday I put in new sway link arms. One had the boot torn open and it was rusted inside and getting crunchy. These are just rock auto ones, I think Delphi. I would have gone Moog but theirs appear to be discontinued.

I have a pair sitting around that I need to install myself. Laziness is getting to me.