Honda Insight Forum banner

1 - 12 of 12 Posts

·
Registered
Joined
·
6,319 Posts
Discussion Starter #1
Over the last couple weeks or so I've been tinkering with steps toward installing a DIY lithium titanate 12V battery in my car. One option that comes up is that I can likely use 5 cells or 6 cells:

-The 5 cell battery would have the DCDC output voltage set as low as it can go, which at this point looks like about 13.6V.
-The 6 cell battery would have the DCDC output voltage set as high as it can go, which I've tested to be about 15.1V.

I'm trying to think of pros and cons to one or the other.

Assuming the 5 or 6 cell versions are equally viable as far as the battery is concerned, are there other reasons why I would choose to have the Insight 12V system running at ~13.6V or 15.1V?

The only thing that really comes to mind is that it's nicer to have a higher voltage and brighter headlights. Other than that I'm thinking it might be better for the DCDC to run at the low voltage rather than the highest voltage it can, maybe less stress. But I don't really know. It might be that it's more efficient at one or the other voltage??. Or, 13.6V is closer to the normal operating voltage - so maybe it's best to stick as close to normal as possible? Etc etc...

Any insights, opinions, etc?
 

·
Premium Member
Joined
·
764 Posts
15V, in my opinion, is starting to get into the territory where automotive systems aren't made to normally run. They can tolerate it, of course, but they'll use a bit more power, your bulbs will probably burn out sooner, etc. Personally, I'd stick closer to the normal operating voltage of 13.6V.
 

·
Administrator
Joined
·
10,958 Posts
I agree with Mario 13.6V. (5 Cells.)
It makes your pack lighter as well and less liable to burn out awkward to get too bulbs like the dash cluster or aircon display etc etc
 

·
Super Moderator
Joined
·
7,510 Posts
-The 5 cell battery would have the DCDC output voltage set as low as it can go, which at this point looks like about 13.6V.
-The 6 cell battery would have the DCDC output voltage set as high as it can go, which I've tested to be about 15.1V.
I'm aware of how to "command" the dc-dc high output by cutting the green/white wire, but how does one command the low level?

I drove this morning and didn't see the same binary levels that you mentioned. I saw more like 12.2-14.2V, but the battery appeared to be pretty low when I turned the key. My lighter gauge may be off a bit. Should probably set up the OBDII C&C to read the 12V level. mudder spoke of the two binary levels at InsightFest2019, but I don't remember his exact numbers. Anyone?

I'm trying to think of pros and cons to one or the other.
It might be instructional to consider the spec operating range of a battery of each cell count. The "spec" range of a 5S is 7.5-13.5V, while the "spec" range of a 6S is 9-16.2V. I don't think the 6S would rise above the dc-dc charge level, so not a problem, IMO. The 5S would be operating most time at fully charged and maybe very slightly overcharged.

Seems that the 5S would be best simply because it would be fully charged most of the time, at least at a pegged operating voltage of 13.6V.

I'm also interested in doing this conversion. I have a pack with a damaged strap as the result of a careless manual balancing effort. I think I can make two 12V LTO batteries from it:)

edited
 

·
Registered
Joined
·
6,319 Posts
Discussion Starter #5
Thanks, all, for the suggestions. That's exactly what I needed...

I'm aware of how to "command" the dc-dc high output by cutting the green/white wire, but how does one command the low level? I drove this morning and didn't see the same binary levels that you mentioned. I saw more like 12.2-14.2V, but the battery appeared to be pretty low when I turned the key. My lighter gauge may be off a bit. Should probably set up the OBDII C&C to read the 12V level. mudder spoke of the two binary levels at InsightFest2019, but I don't remember his exact numbers. Anyone?
I'm working on a way to use the "DVCT," RED/YEL DCDC wire to set output voltage, which is different than what the WHT/GRN, "DVC" wire does. DVC sets either a low or high power mode, either about 12.2V at battery or about 13.8-14V, respectively. (If you wanted to manually drop to 12.2V, you can short DVC to ground - I did this a few times with a paper clip between DVC and the BLU/WHT 'SG10' at DCDC grey connector.)

DVCT allows you to adjust the output voltage to be within a range of what at this point I think will be between a low of about 13.6V and a high of 15.1V. As far as I can tell, all that's needed is a signal voltage that's adjustable between 0 and 5V, hooked-up to DVCT. Or you can intercept the source signal from the ECM and use an Arduino or similar to alter those values, send along to the DCDC what you want. Etc. There's probably various ways to achieve adjustability, but I think they all boil down to being able to adjust the voltage at the DCDC DVCT pin to between 0 and 5V. From what I gather, this sort of adjustability is common on DCDC converters...

DVCT is a PWM signal from the ECM, but as far as I understand, PWM is usually used or generally meant to simulate analog signals. When I measure DVCT as a voltage with my DMM, I get a pretty steady voltage. I grounded DVCT, which pulls the pin as low as it goes, to get the max voltage value (pin pulled low=output voltage high). I'm going to try a voltage converter set to output 5V to see what the minimum DCDC output voltage will be - minimum in terms of the lowest voltage to which DVCT can adjust, since we know the DCDC can go even lower, to 12.2V with DVC...

DVCT is supposed to adjust DCDC output voltage based on engine coolant temp (hence the "T" in DVCT). But, frankly, I'm kind of wondering either how much that is true, or how much it actually matters. For instance, at an ambient temp around 50 degrees F, at cold start, I'll see an output voltage at about 14V (DVCT voltage is around 3.5V). Once the car warms up fully, I see a drop to 13.8V (DVCT voltage at about 4.5V). What's strange is that I haven't seen DVCT voltage changing in lock step with engine coolant temp, though I admit my observations aren't the greatest.

What's more, I have to wonder about the correlation between battery temp and engine coolant temp: DVCT can almost certainly tell when the battery is cold, but engine coolant warms up fast and stays there - and that would be the end of the usefulness of DVCT in the OEM manner, for the most part. Basically, I think DVCT is meant to make minor adjustments to DCDC output voltage based on broad climate and seasonal differences - if you live in a cold area, engine coolant temp would behave more erratically and would be colder, colder longer, more often, etc. And that's likely when it would matter. Ditto on the hot climate side. But the impact in temperate climates is likely negligible... And this is all in the context of using a lead acid battery anyway, which falls away if you want to change chemistries... etc etc.

It might be instructional to consider the spec operating range of a battery of each cell count. The "spec" range of a 5S is 7.5-13.5V, while the "spec" range of a 6S is 9-16.2V. I don't think the 6S would rise above the dc-dc charge level, so not a problem, IMO. The 5S would be operating most time at fully charged and maybe very slightly overcharged. Seems that the 5S would be best simply because it would be fully charged most of the time, at least at a pegged operating voltage of 13.6V.
Yeah, I've been working with this. I left the battery out of my original query so the question would be easier for people to respond to. But, yeah, there's a lot to think about in terms of whether a 5 cell or 6 cell LTO would be best for the battery itself, as well as performance. Up until I started this thread, I had been wedded to 5 cell. But then, seeing how easy it is to boost DCDC voltage (basically ground DVCT, no cutting even), I had to question whether 6 cell might be better. Note that I'm using SCiB 2.9Ah high power cells; these will be slightly different than the high energy cells, plus there's considerations for their small size.

6 cell would basically operate in the 0 to 55% SoC range - if I left DVC intact - so 12.2V to 15.1V, 2.03V to 2.52V at cell level. 2.52V isn't bad, a little low for my small capacity cells, while 2.03V is kind of taxing the bottom. The discharge voltage curve for my cells basically tanks at about 2.3V.

5 cell would operate in the ~35% to 100%+ SoC range, 12.2V/2.44V to whatever minimum DCDC output voltage I can achieve. The OEM DCDC voltage is generally too high for 5 cells, which is why I've been trying to figure out how to adjust it lower. 14V/5 cells=2.8V per cell. I've done some tests, and without any balancing method, that high voltage has the potential to drive an individual cell well over 2.7V (even with only very slight imbalance, I've seen about a high of 3.05V). I actually don't think this kind of overcharge is that big of a deal for these cells, but I'd rather err on the side of caution at this point - if I can lower the DCDC output voltage, then I'll do it, as low as I can. Ideally I'd be able to keep the cells to below about 2.65V, but that would mean a DCDC output voltage of only 13.25V, and I'm not sure whether I can use DVCT to get it that low, it's likely the answer is 'No', but I'll find out soon enough...

[Keep in mind, any random lay-reader of this stuff, that unlike other lithium cells, these LTO cells can be floated at 2.7V with little to no unusual degradation. Etal. We'd have a totally different situation if we were talking about normal lithium cells.]

Anyway, that about covers it. A note on various voltage readings: my OBDIIC&C reports 12V system voltages about 0.3V higher than what I measure at the battery. I see a fluctuating reading of 14.2-14.4 when it's 14.03V or so at the battery; I see 14.0-14.2 on OBDIIC&C when it's 13.83V or so at battery.
 

·
Administrator
Joined
·
10,958 Posts
A note on various voltage readings: my OBDIIC&C reports 12V system voltages about 0.3V higher than what I measure at the battery. I see a fluctuating reading of 14.2-14.4 when it's 14.03V or so at the battery; I see 14.0-14.2 on OBDIIC&C when it's 13.83V or so at battery.
Various reasons for this.

1) You get voltage drop in the various wires and circuits that feed the ECM or MCM depending on the other items currently drawing power in the car, so the voltage they see and report is different to the voltage at the 12v battery terminals.

2) The ECM/MCM adc data used is only 8bit with a wide range 0-25.5V IIRC so resolution is only to the nearest 100mv at best.

3) IIRC There might be a compensation factor in the OBDIIC&C 12v reading code, check the menus.

An decent voltmeter actually connected to the 12v battery terminals is the only way to get a reliable and accurate indication.

The 12v cig lighter socket it suffers from the same problems as the ECM/MCM.

However if you only want to know if the DC_DC on high/low then it's good enough.
 

·
Super Moderator
Joined
·
7,510 Posts
Semi-permanently install your Fluke to read the 12V battery, wrap the leads around the door seal, and tape it to the dash, until you get an idea of the various meter offsets/resolution/accuracy ;)

I've thought this through a bit. When I test 6S multiple cells on the PL8 and look at balance, I don't see cell divergence voltage wise at full charge. I frequently/usually see some small divergence in the 1.5-1.7V range. So the top is probably going to stay balanced, even without BMS, but not the bottom.
 

·
Registered
Joined
·
6,319 Posts
Discussion Starter #8
I'm working on a way to use the "DVCT," RED/YEL DCDC wire to set output voltage, which is different than what the WHT/GRN, "DVC" wire does....DVCT allows you to adjust the output voltage to be within a range of what at this point I think will be between a low of about 13.6V and a high of 15.1V. As far as I can tell, all that's needed is a signal voltage that's adjustable between 0 and 5V, hooked-up to DVCT....There's probably various ways to achieve adjustability, but I think they all boil down to being able to adjust the voltage at the DCDC DVCT pin to between 0 and 5V....I'm going to try a voltage converter set to output 5V to see what the minimum DCDC output voltage will be - minimum in terms of the lowest voltage to which DVCT can adjust, since we know the DCDC can go even lower, to 12.2V with DVC.
This didn't work. If you pull the RED/YEL wire terminal out of the connector, so DVCT doesn't connect back to the ECM, the voltage from DCDC side to ground is already 5V. It appears that 13.8V is the lowest that DVCT can adjust output voltage... If one were to do an LTO 12V battery though, I'd still do this - cut DVCT - because then the highest voltage will always be 13.8V, rather than DVCT-attenuated 'cold' voltages that can get up into the 14.Xs...
 

·
Registered
Joined
·
671 Posts
It appears that 13.8V is the lowest that DVCT can adjust output voltage...
You could always connect a honking big diode to the output and drop the voltage a half volt or so. You can do the math to determine how much heat sink you might need.

This was the result of a 15 second Google search, and I have no idea if it is sized appropriately for what you want to do.

 

·
Registered
Joined
·
6,319 Posts
Discussion Starter #11
I guess mostly for 'fun'... I bought a bunch of cells a long time ago to make an IMA pack, but turned out there were problems to work through, work around. So, I have extras and thought it'd be helpful to work with the cells, putting them together/'packaging', use them as a 12V and see how they hold up, before I move forward with the IMA pack. Plus, when I was making the move to LTO, I was having a hard time with my NiMH IMA pack - so I felt I needed to get away from that. But since then, I tried some new stuff and ended up with a good NiMH rebuild, so I don't really need an alternative IMA pack now, not as much incentive to push forward with the LTO IMA build. Which leaves me with all these cells just taking up space.

My NiMH stick 12V has worked great, I have a lot of confidence in that. I don't need to replace it... I bet it would last a lot longer than normal lead acids.
 
1 - 12 of 12 Posts
Top