Honda Insight Forum banner
41 - 60 of 83 Posts

·
Registered
Joined
·
35 Posts
Discussion Starter · #41 ·
Good video. It doesn't hurt to have a spare junction board around, but I wouldn't drop $100 on parts until you know what the problem is. Additionally, forum members will know the best ways to get parts, and at a good price.

It seems like the junction board may(?) be okay. As you said in the video, the next step is testing with a volt meter. Reassemble the battery with the breaker on, and hook the negative lead to the negative battery terminal. Then you want to test the following points for voltage, and I've put the expected voltage beside it.
  • PTC strip, both ends, 0V
  • Positive terminal on top of junction board, 0V (this is because there's a contactor (relay) that doesn't enable the battery until the car ignition is on)
  • Either breaker terminal (with it flipped on), 86-100V
  • Positive battery terminal (bottom left of junction board), 144-168V
  • Ground lug coming off of capacitors, 0V
  • Each thermistor wire, 0V
As far as codes are concerned, the 59 blink code refers to P1444, High voltage short circuit, there is no 59 subcode for P1449. Perhaps the P1444 was successfully cleared by your scan tool. The service manual, posted in a sticky thread on the troubleshooting subforum, has instructions on troubleshooting this code.

Unrelated to the big shorting issues, your stick testing cycle is probably not aggressive enough to have found all of the issues in your pack. The biggest factor in stick testing is probably discharge current. It looks like you tested at 5A. My first tests were at 10A and that wasn't enough. I'd recommend 25-30A.
All PTC strips are 0V.
Positive terminal 0V.
Both breaker terminals measure 96.3 V.
Positive battery terminal 161.1V.
Ground lug 0V. When measuring this, it occasionally starts at .01V before quickly moving to 0V. Doubt that means anything, but there it is.
Thermistor wires: Here we go. I measured the voltage of the wires inside the harness. All of the wire pairs measured 0V except the two red wires from the front plate which both measure 2.5V. Is there something here? Let me know.
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #42 ·
P1542: Climate control unit signal circuit high voltage (a wire is disconnected). Check for poor connection at heater control panel and the ECM. See page 266 in the pgm_fi_system manual. Click the 'download' button on that page.

Question: At 7:36 in your video, the eyelet (wrapped in orange tape) is disconnected. Was that also the case earlier in your video (while you were driving? If so, then the HVDC system has no chassis ground reference, which can cause all kinds of problems. You might have addressed this at 7:45... I'm just making sure.

8:22 junction board wires all appear connected correctly, except that the capacitor wires aren't connected.

8:47 you're correct that none of the three capacitor wires should have continuity (when disconnected from the junction board).

Question: At 9:45, what is the resistance between the PTC screw heads (phillips) and the battery terminals (10 mm bolt)? Should be 'open' (i.e. infinite). Measure from various phillips to various bolts.

Suggestion: 10:00 measure the voltage across the four aluminum spacers. Two combinations should give voltage (e.g. 80 volts and 50 volts)... the other combinations will give floating voltage (around zero). What are the two highest voltages that you measure?

P1449(59):
As mentioned above, not a valid code. The blinking IMA code (59) is more accurate (but harder to read) than the P-code (P1449). It's possible the MCM (which generates both codes) is having multiple codes... and just reporting them at different times. Regardless, you can troubleshoot either code by following this guide.

The first test Honda recommends (on page 114 in the above link) is to disconnect the MCM'E' connector (which has two wire: RED/RED & WHT/WHT). Then measure the resistance between chassis ground and both terminals. You should measure above 300 kOhm. If so, then swap in another MCM and test for P-code again. If the resistance is less than 300 kOhm, then the issue is somewhere else in the IMA. The first thing I'd do is disconnect the two high current cables (the ones that you show sparking in the video), then repeat the test. If the resistance goes above 300 kOhm with those leads disconnected, then the issue is inside the PDU (the left box in the IMA bay). Otherwise, the issue is inside the battery module (and is probably a temperature sensor shorting out to the HVDC bus).

...

FYI: The spark you're seeing in the IMA bay is you discharging the very large capacitors inside the PDU (the thing on the left). They can take quite a while to discharge. I wouldn't worry about the spark... that's expected behavior until the capacitors discharge.

Based on your video, my best guess right now is either the actual IGBT module is broken (inside the PDU, which is the thing on the left in the IMA bay), or a temp sensor is shorting out to the HVDC bus.
Got the 1542 test, I'll try it this weekend as it looks like I'll need the IMA battery back in the car. There's no 1542 when the IMA battery is disconnected.

I disconnected the eyelet before filming 7:36, so no worries there. I disconnected the main terminal bolts as well.

At 8:22 the capacitor wires are connected, but the capacitor is sitting there hanging loose because I removed its screws. The wires are still in it.

9:45, I did this by setting my multimeter to 200 ohms and I measured every single bolt to its respective ptc screw and then some random ones, all were "open". Did I do that right?

I measured the spacers seen at 10:00. If from top to bottom and left to right you'd call them 1, 2, 3, and 4, the voltage for 1-2 was 93V and 3-4 was 28.5V. No other voltages, as you said.

I know 1449 (59) is not a valid code. I don't know what the 59 corresponds to, but it's the only blink code that the IMA light shows when the OBD harness is shorted to make it do that.

I will admit that I'm completely lost when it comes to the recommended tests here. I don't have a clue what's being referred to as MCM connector E and I don't know where the chassis ground is. If you could push me in the right direction about those two things, I can perform that test and report back this weekend as well.

Thank you!
 

·
Linsight Designer
Joined
·
3,482 Posts
9:45, I did this by setting my multimeter to 200 ohms and I measured every single bolt to its respective ptc screw and then some random ones, all were "open". Did I do that right?
Repeat the test with the multimeter set to at least 300 kOhms.

I measured the spacers seen at 10:00. If from top to bottom and left to right you'd call them 1, 2, 3, and 4, the voltage for 1-2 was 93V and 3-4 was 28.5V. No other voltages, as you said.
Is the orange plate on the other side installed, with all bolts present? If so, then the pack measuring 28.5 volts could have sticks flipped the wrong way... although I'm not sure how one would do that, given that they are polarity-keyed (e.g. hexagon and square).

I don't have a clue what's being referred to as MCM connector E and I don't know where the chassis ground is.
The MCM'E' connector is plugged into the MCM, which is one of the two metal boxes bolted to the top of the battery. The MCM'E' connector is the only connector that has two wires going into those metal boxes. It's on the metal box on the right (the MCM), facing you when the pack is installed in the car. The wires on the MCM'E' connector are RED and WHT.

Chassis ground is the aluminum body of the car... When you're in the IMA bay, you can touch any of the aluminum mechanical structure that creates the IMA bay. Make sure wherever you touch ground isn't painted.
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #44 ·
Repeat the test with the multimeter set to at least 300 kOhms.


Is the orange plate on the other side installed, with all bolts present? If so, then the pack measuring 28.5 volts could have sticks flipped the wrong way... although I'm not sure how one would do that, given that they are polarity-keyed (e.g. hexagon and square).



The MCM'E' connector is plugged into the MCM, which is one of the two metal boxes bolted to the top of the battery. The MCM'E' connector is the only connector that has two wires going into those metal boxes. It's on the metal box on the right (the MCM), facing you when the pack is installed in the car. The wires on the MCM'E' connector are RED and WHT.

Chassis ground is the aluminum body of the car... When you're in the IMA bay, you can touch any of the aluminum mechanical structure that creates the IMA bay. Make sure wherever you touch ground isn't painted.
Test repeated with recommended settings, everything's still open. The plate on the other side is installed. I've triple checked the stick directions at this point and they're correct. Is there anything else that could cause this?

Thanks for explaining the E connector and chassis ground like I'm a 5 year old, it's unfortunately necessary. Though I can promise that I won't forget about it now that I know. I'll put the battery back in Saturday and try those tests as well as with the "new" junction box to see if there's any difference. Interesting thought that the problem could be with the DC-DC. At least by the time this is all over, everything will either be new or refurbished to the point where I shouldn't have to worry about it for a while.
 

·
Registered
Joined
·
72 Posts
I saw this and came here thinking it's would be a simple fix. But after I read what you did, I think there are several problems at hand here and definitely not as simple as I thought it was.
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #46 ·
I saw this and came here thinking it's would be a simple fix. But after I read what you did, I think there are several problems at hand here and definitely not as simple as I thought it was.
Someone decided when I was born that nothing in my life is allowed to be simple or easy in the ways that it is for other people. I have demons that chase me around and possess all of the electrical devices I ever come to own. What I really need is a priest, not a mechanic.
 

·
Registered
Joined
·
49 Posts
Reading this thread I am totally nostalgic for my 01 Insight (sold to another member several years ago after upgrading to a 2011 Volt) and more appropriately this forum and it's myriad patient, knowledgeable and helpful experts. The GM-Volt forums are fine but this one is a whole other level of knowledgeable/helpful!

I think this is due to several things including that Insight owners were earlier adopters of HV/EV technology and the hybrid/analog nature of Honda's rushed-to-delivery design makes them more amenable to "hacking". Also, the ER/EV nature of the Volt makes it higher power/voltage and therefore "yet more dangerous" to fiddle with. As a "series hybrid", there is no option for laming/removing the EV components as most of us have done with our Insights, so harder to take "bricking" risks with. Also, the Honda IMA system vehicles are significantly less expensive (especially after being well-used) and therefore more likely to be donors for experiments. I got my 2011 Volt with 166k miles and an end-of-life traction battery for $5k from a young woman who had paid $6k from an inscrupulous dealer who represented it as having "lots of life left"... when the GM Records ultimately showed it had had 2 traction battery failures before she bought it (and two while she owned it for about a year). I don't find other GM-Volt members with vehicles that "used up". I (my local GM service, $1k) swapped in a 2011 pack with 98k miles I bought online ($1.6k).... most folks probably wouldn't put out that kind of cash with that much uncertainty. Im not sure I would again, but I got through it and have a pretty capable PHEV/EREV as a result.

But to the point of this thread, I really miss having people with significantly more hands-on experience and willingness to share. The Volt forum has one guy who did his own swap in his garage, very few others come close to that level of interest/willingness. The My Green Volt app developer has also been a stunning resource, especially for monitoring the traction pack health. I wish I'd had it before I bought the car and before I swapped out the pack.

The point someone made here about the logic of walking 100 miles in bad shoes is very apropos for my own experience in life, as well as these vehicles. By often owning vehicles that need a certain amount of maintenance, I have had my hands and eyes under the hood and/or under the carriage enough that if/when a warning light or strange noise or behaviour or smell springs up, I'm halfway there to figuring it out on the fly. To the analogy with shoes, I own a pair of moccasins I made myself (decades ago) and have repaired some of my older work boots/shoes/sandals by hand (awl/thread and glue) so when my fancy new shoes/boots from Eddie Bauer or REI act up, I can make sense of whether/how to stretch/re-affix/repair them if needed, even if they rarely need that kind of attention.

I built up my old mountain bike (circa 1985) into an EV (using used Insight Sticks for battery) but eventually gifted it forward and bought a modern, integrated foldable one for lots of reasons. Moral of the story is that I know how most if not all of the components of the beast are supposed to work and won't be afraid to troubleshoot/repair anything that goes wrong. If I hadn't walked 100 miles in ill-fitting shoes (old clunker bike with dynamite colored sticks strapped to the luggage rack) I might be like the rest of the reviewers online of these bikes complaining because they misunderstand the theory of operation or limitations of materials of the bikes they bought for very low (by many standards) prices.

I am old now and HATE changing tires or even lifting the hood, but am not intimidated by the very idea of it and would rather do either than wait around for a rescue. I'd never be in this position if I hadn't chosen to own/maintain various vehicles that needed/deserved hands-on attention over the decades.
 

·
Linsight Designer
Joined
·
3,482 Posts
Based on your testing, it sounds like there aren't any shorts from inside the battery pack to the low voltage system.

At this point I would focus on figuring out why the voltages between pack halves are so different (93 vs 28.5 volts). With the main IMA switch off (on the junction board), the pack is split into two sections: one section has QTY8 sticks, while the other has QTY12. Therefore, you should measure 40% of the total pack voltage on one half, and 60% on the other. In your case, you are measuring 23.5% of the pack voltage on one side (28.5 volts), and 76.5% on the other side (93 volts). You need to figure out why that's the case before you continue. I would start by focusing on the side that measures 28.5 volts, which should measure more like 62 volts.
 

·
Registered
Joined
·
8,247 Posts
^ I agree with that. That's pretty weird... The top half is normal, but the bottom is way off. If the sticks are oriented properly as you say, it seems like the only way you could measure such a low value is if the sticks really do have a low value, like they've been discharged. The actual value's a little suspicious, as it comes out to about 0.6V per cell, which is a kind of typical low-end voltage plateau (if discharged at a really low current, a cell's voltage will tend to stall-out around there, for a while)...

Haven't been following closely, but maybe this is related:
...Thermistor wires: Here we go. I measured the voltage of the wires inside the harness. All of the wire pairs measured 0V except the two red wires from the front plate which both measure 2.5V. Is there something here?
The two red-clad wires are the ends of the PTC strip system, the strips on each stick. Maybe there's a short somewhere in that second half of the pack between the PTCs and the cells?
 

·
Linsight Designer
Joined
·
3,482 Posts
Measure resistance (300 kOhm or higher range) between:
-the PTC positive lead (the red wire that's inside a red conduit), and;
-the high voltage bus (e.g. at several different 10 mm bolts on the cell ends). You should measure 'open'.
 

·
Registered
Joined
·
8,247 Posts
I've been thinking a little bit off and on about how the PTC strips can/would be shorted to drain cells or sticks, particularly where the whole '2nd half' of pack gets discharged, and I can't say I've got answers... Maybe someone could fill-in blanks? I've heard of PTC shorts around here, but I've never actually had one or thought about it too deeply...

The strips themselves stretch across each stick, their ends connect to busbars enmeshed in the orange end board, and the orange end board is ultimately connected to the BCM via the red-clad wires. Once everything is connected it's just one long circuit...

Now, if the covering on a single strip were breached in two spots, bridging one or more cells' positive and neg ends, that cell or those cells would be shorted and they would be discharged. That seems obvious. If you measured voltage at "the two red wires from the front plate," as notkyle did, would you measure + voltage? I'm assuming notkyle stuck his probes on the connection points on the orange end board, where the red wires connect, and I guess the red wires were connected(?) to the BCM...

It seems like you would measure voltage, like if it were one cell being shorted you'd measure about whatever the voltage is of that one shorted cell, or maybe a little lower because of resistance in the PTC strips along the circuit... Since notkyle measured 2.5V, maybe at least two cells are shorted?

Now assume this is all true: How does that whole half a pack get discharged? If only two or a few cells are shorted, i.e. if a scathed PTC strip bridges between two or more cells, can it discharge more than just the bridged cells?

If the 'bridging', the shorting, were between the end cells, certainly you'd discharge all the cells. But then the voltage you measure at the ends of the PTC circuit wouldn't be only 2.5V, I'm pretty sure...

If it were some middle cells? I don't see how it'd drain all the cells, it'd just drain the shorted ones, all else being equal...

So it seems like some other kind of short would have to exist. I think that's about as far as I can take this at this point...
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #55 ·
Measure resistance (300 kOhm or higher range) between:
-the PTC positive lead (the red wire that's inside a red conduit), and;
-the high voltage bus (e.g. at several different 10 mm bolts on the cell ends). You should measure 'open'.
Okay. I think we might be close here. After doing what was directed, I realized I screwed up one of the original resistance tests; my multimeter goes in factors of 2, so I tested the bolts and PTC resistances at 200kOhms. I set the multimeter to 20mOhms and am consistently getting a reading of around 14-15 on, if you're looking at the battery array from the junction side and going left to right, bolts 4, 5, and 6. All 17 others are "open". To test if this was a stick problem, I replaced the sticks in those slots with 3 spares I have and with 3 from other positions that were measuring correctly. Same result. I don't know what's causing this, but I have a feeling that with this info, you may.
 

·
Linsight Designer
Joined
·
3,482 Posts
That's at least one of the issues. You need to inspect the heat shrink for cuts and scrapes that are causing the temperature sensor(s) to contact the metal cylindrical cell housing.
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #57 ·
That's at least one of the issues. You need to inspect the heat shrink for cuts and scrapes that are causing the temperature sensor(s) to contact the metal cylindrical cell housing.
The heat shrink on the sticks? I'm not sure I follow. I've swapped multiple sticks into those same slots and they're still reading resistance where they shouldn't be.
 

·
Linsight Designer
Joined
·
3,482 Posts
Background:
FYI: "20 'em' Ohms" is pronounced "20 mega Ohms", which is shortened as 20 MOhm. In the 20 MOhm range, your DMM is likely outputting 100 nA or so, and then measuring the resulting voltage. Then, using V=IR, the DMM is telling your the observed resistance (e.g. 1.234 volts = 100 nA * R, so R = 1.234 MOhm).

Analysis:
You are measuring 14.7 MOhm, which is 49 times higher than the minimum allowed resistance value - between the HVDC system and chassis ground - that will immediately cause a P-code. This is why you aren't immediately getting a "HVDC resistance fault" P-code.

However, on a stick that doesn't have a short between the PTC temperature sensors and outer cylindrical metal wall, the resistance should be well over 1000 MOhm (i.e. 1 GOhm). Given that the measured resistance is 66x lower than what it should be, I would suggest that you have an intermittent short somewhere between the PTC temp sensors and one (or more) of the cell outer cylindrical walls.

If you remove the cells that indicate 14.7 MOhm from the pack, do they still indicate 14.7 MOhm? If not, then the leakage path is one of the two black rubber grommets. These grommets come into contact with every single stick, and for some strange reason, are actually conductive! If the heat shrink on any two cells (in the entire pack) is damaged in such a way that the rubber grommets contact the PTC strip (on any stick) AND the outer cylindrical cell wall (on any stick), then that will cause a high impedance short... I suspect that's what's happening here.

So the next question is: Will a 14.7 MOhm leakage path from the HVDC bus to the temperature measurement circuitry cause problems? I modeled this in LTSPICE and I don't think it will. The temperature sensors are biased hard enough that even the strongest RF noise - leaking through the 14.7 MOhm leakage path - is at most -96 dB from DC to 10 MHz. Put another way, the noise injected by this short is 63,000x smaller in amplitude than the measured temperature... which means the temperature signal will dominate:
Azure Rectangle Slope Font Parallel


So then I would guess that this isn't the issue. However, the IMA system is a strange beast, and RF noise injection is difficult to model. Notably, if the AC decoupling capacitors (which we were discussing previously) aren't connected (or are broken), then the 14.7 MOhm return path is going to inject much 'harder' a frequency increases.

I propose we perform the following test:
0: The overall goal is to put the pack into the car, but with a slight (testing) modification.
1: Disconnect the two cables that connect to the orange bus bar plate (they have red insulation jackets around them). They are shown at 0:16 in the video you just posted.
2: Unwind these two cables so that you'll have access to them after the pack is entirely reassembled.
3: Reassemble the pack entirely, except for the two cables mentioned previously.
4: Place a 22 Ohm resistor between the two wires mentioned previously.
5: Drive the car. See if issue persists.
 
41 - 60 of 83 Posts
Top