Honda Insight Forum banner

181 - 200 of 219 Posts

·
Registered
Joined
·
971 Posts
Actually I was thinking about getting even better MPG out of the Insight with a much more capable motor (ie the HCH3 motor, overdriven if ambient conditions permit) and a much greater reliance on regenerative braking at low speeds which is synergistic with very high current, low capacity packs (perhaps 2P). But the braking system would have to be made electronic to do this.

This, however, would also require changes to the MCM and greater autonomy from the ECU and, mechanically, may require two clutches to keep the IMA motor connected to the drivetrain while the engine is auto-stopped. So this is a non-starter for most but we can dream (or figure out how to put an electric motor on the back wheels...)
 
  • Like
Reactions: retepsnikrep

·
Linsight Designer
Joined
·
2,172 Posts
Discussion Starter #182
Update: I've been busy with work today, so nothing to report on this project. I'll be busy with work tomorrow, too.
 

·
Premium Member
Joined
·
2,942 Posts
Actually I was thinking about getting even better MPG out of the Insight with a much more capable motor (ie the HCH3 motor, overdriven if ambient conditions permit) and a much greater reliance on regenerative braking at low speeds which is synergistic with very high current, low capacity packs (perhaps 2P). But the braking system would have to be made electronic to do this.

This, however, would also require changes to the MCM and greater autonomy from the ECU and, mechanically, may require two clutches to keep the IMA motor connected to the drivetrain while the engine is auto-stopped. So this is a non-starter for most but we can dream (or figure out how to put an electric motor on the back wheels...)
Electric in wheel motors similar to the ones Aptera uses , could also be used as a retrofit on the back wheels of a Insight.

.. but .. I personally prefer the the dual clutch with the 3rd gen IMA motor route.

That would give about ~17kw nearly continuous , similar to Gen1 ICE Lean Burn .. and if as equally over built as the gen1 IMA (peter pushed to ~3x) , then 17x3 = ~51kw short bursts , similar to the gen1 ICE rich Vtech max of ~50kw power at over 5k RPM.


89992


but .. to keep this thread on topic .. further discussion of that might best over in it's own thread .. looking back at that thread now .. wow has it been almost 5 years already , dang my progress is slow and tiny .. I am more like a snail than a turtle :( .. I should get more back into that project this year.

 
  • Like
Reactions: Natalya

·
Linsight Designer
Joined
·
2,172 Posts
Discussion Starter #184
No updates today... I've been busy tearing down a competitor's new CNC machine for the past several days. Without boring y'all with the details, my recommendation is to NOT buy the Bantam Tools Desktop CNC machine.

I should resume work on LiBCM on Wednesday.
 

·
Registered
Joined
·
92 Posts
IMO salvaging NiMH cells is a waste of time... I can't even give them away at this point. Fortunately, Chattanooga will dispose of them for free. Just last week I dropped off two G1 packs worth of untested NiMH cells at their collection center.
Ahem... I'll take them... ;)
I have some, are you buying and paying for shipping?
 

·
Linsight Designer
Joined
·
2,172 Posts
Discussion Starter #186
Boy not knowing the part numbers for these connectors is annoying. I had planned on using standard 1x2 2.54 mm pin headers, but during layout I've found that their plastic bases interfere with each other. I'm now considering machining mating connector housings from raw plastic stock... The pin holes are too small to 3DP. These are going to be annoying to manufacture.

Another option is to solder in every other pin, then remove their plastic bases, then solder in the rest. That would still need at minimum a 3DP shroud to prevent incorrect connector installation.

I'll think this over tomorrow.

...

Here's the 509 parts I've assigned footprints to so far:
90138


The remaining seven parts have unique footprints, so seven more footprints and I'm good to start layout. I've always hated the way KiCAD handles footprints, but overall KiCAD is amazing, so I just grind my teeth through the process. I certainly miss having PCB Librarians (who do all the footprint gruntwork at large companies). I tried using a 3rd party librarian on my previous board and they managed to screw up two footprints... hence I'm doing it myself this time (and probably forever).

It looks like everything will fit, but hard to tell due to the numerous 3 mm safety HVDC keepouts.
 

·
Linsight Designer
Joined
·
2,172 Posts
Discussion Starter #188
I feel your pain on the whole footprint thing.
I've never understood why KiCAD chooses to be 'different' when it comes to footprint assignment.

Alas, my set of these batteries are 1100 miles away right now and I won't going to pick them up until the first part of May.
That'll be sooner than I'm shipping this PCB. 100% certain about that.

Are there any marks/logos on the connectors?
">PBT<" (plastic type)
"2E 4" which doesn't mean anything to me.
As much as I just want to replace the connectors, I doubt many people would be comfortable with that, since any shorts would immediately destroy the sense leads.
 

·
Premium Member
Southern California
Joined
·
837 Posts
Should go with 4, always good to get that continuous ground plane, it makes routing a lot easier, and 4-layer boards are so cheap these days.
 

·
Linsight Designer
Joined
·
2,172 Posts
Discussion Starter #193
4 layers is certainly my go-to. However, in this case it's possible I can do it with 2 layers, as there isn't much signal overlap. Given how large this PCB is (355x160 mm)(88 square inches), a 2 layer PCB is a considerable cost reduction. Of course, as soon as I realize I need 4 layers I'm not going to resist that additional cost too much.

...

It's too soon to tell for sure, but it's possible I might be able to layout 60S on this PCB. The HVAC & HVDC sections laid out more compactly than I had estimated, and this just might give me enough room for a 5th LTC6804. No promises. Keep in mind that even if 60S fits, it will not be a drop-in solution (like 42S & 48S are). Specifically, you'll need to place a 12S pack outside the OEM IMA battery bay, and then manually wire each cell tap to said 5th connector. A 60S pack will also require replacing the OEM DCDC converter (e.g. with the one Peter is using in his high voltage setups). As always, I do not intend to support the 60S configuration until the 42S & 48S configurations are sorted out.
 

·
Linsight Designer
Joined
·
2,172 Posts
Discussion Starter #194 (Edited)
Update:
-Routed the most complicated LTC6804 section (top left)... it routes to two separate connectors (for 12S or 18S pack in bay 3).
-Routed entire HVDC section and power supplies (top right)
-Certainly 48S fits... maybe 60S will, too?
-Still trucking along with two layers. It looks like I'll be able to do nearly every low voltage signal on just one layer... leaving layer 2 for power/ground.
90154
 

·
Registered
Joined
·
971 Posts
A 60S solution potentially creates a path for Fit EV LTO pack owners willing to split a module, or those wishing to assemble a pack from loose LTO cells. I mention the latter because apparently a new generation of Toshiba SCiB cells is right around the corner. 60S appears to be a nice number for LTO except for the risk of overvoltage. I guess the question is which ECUs are responsible for the graceful fading of regen as pack capacity is reached, and what would be needed to cause this same behavior to occur at a lower voltage?
 

·
Linsight Designer
Joined
·
2,172 Posts
Discussion Starter #196
A 60S solution potentially creates a path for Fit EV LTO pack owners willing to split a module, or those wishing to assemble a pack from loose LTO cells. I mention the latter because apparently a new generation of Toshiba SCiB cells is right around the corner. 60S appears to be a nice number for LTO except for the risk of overvoltage. I guess the question is which ECUs are responsible for the graceful fading of regen as pack capacity is reached, and what would be needed to cause this same behavior to occur at a lower voltage?
Anyone using this PCB with a different battery will need to solder leads directly to the PCB.
The BMS connector pitch on these Honda lithium batteries isn't readily available.
I imagine the best way to do that is to order inline connects and connect them to the PCB with 6" leads.
This isn't something I'm going to support, but for those that want to use different batteries, I'll sell a PCB without the connectors installed.

...

The MCM ultimately controls IMA charging. However, the MCM makes decisions based on the following data:
-BATTSCI (data generated by BCM)
-MCM Connector 'E' (measures IMA battery voltage)
-VPIN (~50x reduced battery voltage)
-Battery Current Sensor

LiBCM controls the above signals as follows:
-BATTSCI: Full control (LiBCM generates all data on bus).
-MCM Connector 'E': On-the-fly voltage adjustment from 88% to 100% actual pack voltage.
-VPIN: On-the-fly voltage adjustment from 10% to 100% actual pack voltage.
-Battery Current Sensor: On-the-fly current adjustment from 10% to 150% actual current.

Given the above, LiBCM is able to completely control charging, albeit indirectly; the MCM ultimately makes the decision... LiBCM just rigs the deck. The major caveat here is that LiBCM cannot spoof an MCM'E' voltage higher than the actual pack voltage. However, a 60S LTO pack is nearly discharged at 125 volts, which is high enough that the MCM won't try to background charge. LiBCM can sense a background charge and disable it (e.g. by sending a flag to the MCM on BATTSCI).
 

·
Engine-Off-Coast
Joined
·
2,001 Posts
... However, a 60S LTO pack is nearly discharged at 125 volts, which is high enough that the MCM won't try to background charge. LiBCM can sense a background charge and disable it (e.g. by sending a flag to the MCM on BATTSCI).
I've got a 72 cell LTO pack, voltage always above 155, and my car tries to background charge all the time. If SoC is below like 65% I can expect it to background charge, especially while in lean burn.
 

·
Linsight Designer
Joined
·
2,172 Posts
Discussion Starter #198
Clarification: Sean's original question was about overcharging the IMA battery. I wandered off topic with my answer, ultimately making the unqualified statement "the MCM won't try to background charge." What i meant to say is the MCM won't force charge the pack (e.g. even when Calpod is active). IMO the OEM background charge behavior is prudent, given that the car as-shipped doesn't allow grid charging.
 

·
Linsight Designer
Joined
·
2,172 Posts
Discussion Starter #199
Update:
There's room for 60S:
90169

It's comfortably two layers:
Signals on layer 1 (red).
Power on layer 2 (blue), plus localized signal routing. I haven't drawn in planes yet; that happens last.

My goal is to order the PCB tomorrow night... I'll be out of town for the next couple weeks after that.
 

·
Registered
Joined
·
2,069 Posts
Thanks for the update.
Nice to hear boards are being ordered.

So tempting to always put off ordering just because want to add "one more thing" first.

And too often I've let feature creep be the sand that stops the gears of progress.

Exciting times!
 
181 - 200 of 219 Posts
Top