DC to DC converter: what can it reasonably handle in terms of using your battery as a power source?

[FalconFour]

Having trouble finding more recent information on this. This blog post indicates that connecting directly to the DC-DC bypasses any checks on current draw. Some posts in this thread seem to indicate that large draws from the DC-DC directly are an issue. Anyone have updated info as of current software?

I don't know about current software, but I know that for a few months, last year, I was choking down about 1000w continuously from the PCS (DC-DC) with the car in half-Sentry mode (Sentry enabled but unlocked = leaves car awake, but Sentry inactive)... every night. Charging my garage battery bank

I've pulled just around 2000w out of it while I was originally playing with it. It's capable of quite a bit. But you need GOOD connections to pull high currents. No loosey goosey small gauge wire stuff. Nearing 200 amps, you're getting up into "my god, THAT is a thick wire" territory. And really tight torque. And double-checking your connections periodically for loosening from vibration.

I've still got an Anderson 150a plug hanging out of my back seat, but haven't used it in a long time. The key to using it is to have a buffer battery to prevent load spikes from shutting down the PCS, and to have Scan My Tesla or Tes-LAX available to monitor the bus and see if the PCS remains happy/operational while in use. And disconnecting that buffer battery any time you want the car to go to sleep. It'll be unhappy if it sees voltage at the PCS terminals when it's intended to shut down (to sleep).

So, two key points of PCS play:
- no load spikes (any kind of starting surge will trip the PCS) - but smooth continuous load is oh-so-powerful
- make sure voltage isn't present when the car tries to sleep (keep it from sleeping by putting it in Sentry; disconnect battery/load when you intend to let the car sleep)

 

[thefrog1394]

I don't know about current software, but I know that for a few months, last year, I was choking down about 1000w continuously from the PCS (DC-DC) with the car in half-Sentry mode (Sentry enabled but unlocked = leaves car awake, but Sentry inactive)... every night. Charging my garage battery bank

I've pulled just around 2000w out of it while I was originally playing with it. It's capable of quite a bit. But you need GOOD connections to pull high currents. No loosey goosey small gauge wire stuff. Nearing 200 amps, you're getting up into "my god, THAT is a thick wire" territory. And really tight torque. And double-checking your connections periodically for loosening from vibration.

I've still got an Anderson 150a plug hanging out of my back seat, but haven't used it in a long time. The key to using it is to have a buffer battery to prevent load spikes from shutting down the PCS, and to have Scan My Tesla or Tes-LAX available to monitor the bus and see if the PCS remains happy/operational while in use. And disconnecting that buffer battery any time you want the car to go to sleep. It'll be unhappy if it sees voltage at the PCS terminals when it's intended to shut down (to sleep).

So, two key points of PCS play:
- no load spikes (any kind of starting surge will trip the PCS) - but smooth continuous load is oh-so-powerful
- make sure voltage isn't present when the car tries to sleep (keep it from sleeping by putting it in Sentry; disconnect battery/load when you intend to let the car sleep)

Got it. All sounds reasonable. What of the talk of the PCS monitoring output current and comparing it against the sum of all of the current drawn through the e-fuses and throwing an error if it sees power going elsewhere? Is that just bogus? Maybe people pulling too much startup current and shutting off the PCS and then draining the battery?

Somewhat related, my assumption is that the PCS sits somewhere in the 13-15V range on legacy Model 3s like a standard alternator. Do we know yet what voltage range the new 15V lithium battery edition cars see? Guessing up to 20V during charging? Looks like many inverters hit over-voltage shutoff around 15V. Something to keep in mind for owners of brand new cars.

 

[FalconFour]

What of the talk of the PCS monitoring output current and comparing it against the sum of all of the current drawn through the e-fuses and throwing an error if it sees power going elsewhere? Is that just bogus?

I'd kinda wondered if they'd be doing that... but my experience, so far, has been not so much. I haven't tried it recently (past 3 months or so, probably). But before that, it never cared. As long as you don't...

Maybe people pulling too much startup current and shutting off the PCS and then draining the battery?

... trip it ... Yeah, it's really easy to piss the thing off, and the way it gets pissed off is non-obvious. A fault doesn't immediately show up on the screen. If you're not watching CAN, you might miss it. And if you miss it, well, slurrrrp goes your 12v battery.

Good question re: the lithium 12v cars too. I've got a lithium 12v in my car -- but an Ohmmu one, not OEM. That battery is simply 4x LFP cells in series, and I'd imagine Tesla would do the same... it's not a sure thing, but LFP cells fit neatly in the standard voltage-curve of Pb batteries (3.6v/cell max * 4 cells = 14.4v = same exact as lead charging voltage; 3.3v nominal * 4 cells = 13.2v = perfectly happy lead battery). I'd expect the voltage to be the same. No way it'd be 20v. The 12v bus voltage propagates through everything 12v-related, so allllll the systems would have to be significantly revised to tolerate any voltage leading up to 20v. Max is 15v at the moment, and I wouldn't imagine it'd be any different for the new OEM lithium battery...

 

[miimura]

I'd kinda wondered if they'd be doing that... but my experience, so far, has been not so much. I haven't tried it recently (past 3 months or so, probably). But before that, it never cared. As long as you don't...

... trip it ... Yeah, it's really easy to piss the thing off, and the way it gets pissed off is non-obvious. A fault doesn't immediately show up on the screen. If you're not watching CAN, you might miss it. And if you miss it, well, slurrrrp goes your 12v battery.

Good question re: the lithium 12v cars too. I've got a lithium 12v in my car -- but an Ohmmu one, not OEM. That battery is simply 4x LFP cells in series, and I'd imagine Tesla would do the same... it's not a sure thing, but LFP cells fit neatly in the standard voltage-curve of Pb batteries (3.6v/cell max * 4 cells = 14.4v = same exact as lead charging voltage; 3.3v nominal * 4 cells = 13.2v = perfectly happy lead battery). I'd expect the voltage to be the same. No way it'd be 20v. The 12v bus voltage propagates through everything 12v-related, so allllll the systems would have to be significantly revised to tolerate any voltage leading up to 20v. Max is 15v at the moment, and I wouldn't imagine it'd be any different for the new OEM lithium battery...

Tesla's new lithium battery is not LiFePO4. It's standard lithium ion. That means that the normal operating range is slightly higher voltage, but always less than about 16.6VDC. The bus voltage on the car is now nominally higher and some accessories had to be revised to accommodate that. It appeared in some of the regulatory filings in Europe that were in the "Tesla news" recently.

 

[FalconFour]

Tesla's new lithium battery is not LiFePO4. It's standard lithium ion

surprised_pikachu.png

Welp! tosses_papers_into_air.png

Good to know! Now I'd really love to see more details on that. Why the heck would they do that? LFP seems like the obvious choice. Oh well...

 

[thefrog1394]

The 12v bus voltage propagates through everything 12v-related, so allllll the systems would have to be significantly revised to tolerate any voltage leading up to 20v. Max is 15v at the moment, and I wouldn't imagine it'd be any different for the new OEM lithium battery...

Maybe not quite 20V, but it's definitely higher. Enough that they had to disable 12V trailer auxiliary power for example. Based on this article full charge voltage could be close to 17V. A lot of 12V electronics are tolerant of a pretty wide voltage range (although apparently not the harbor freight inverters I was looking at haha)

 

[miimura]

surprised_pikachu.png

Welp! tosses_papers_into_air.png

Good to know! Now I'd really love to see more details on that. Why the heck would they do that? LFP seems like the obvious choice. Oh well...

Here you go.

 

[rrolsbe]

Got it. All sounds reasonable. What of the talk of the PCS monitoring output current and comparing it against the sum of all of the current drawn through the e-fuses and throwing an error if it sees power going elsewhere? Is that just bogus? Maybe people pulling too much startup current and shutting off the PCS and then draining the battery?

Somewhat related, my assumption is that the PCS sits somewhere in the 13-15V range on legacy Model 3s like a standard alternator. Do we know yet what voltage range the new 15V lithium battery edition cars see? Guessing up to 20V during charging? Looks like many inverters hit over-voltage shutoff around 15V. Something to keep in mind for owners of brand new cars.

With the cars telemetry, it sure seems probable that Tesla could deduce energy was drawn out of the traction battery for ancillary purposes (especially if it was many KWH). Currently the DC-to-DC voltage output is around 14.7V while topping off the 12V battery and around 13.4V (float voltage) at all other times while not sleeping. Around a year ago the DC-to-DC converter would output at least 14.7V ( if memory serves) anytime the car was not sleeping, which many people attributed to early demise of the 12V battery (including me); likewise, at least for my Model 3, the car would almost always stay awake for six to eight hours anytime the car woke from sleep state for any reason. Tesla now seems to be handling the charging of the 12V battery in a better way and the phantom drain is much less than before. My PCS board and 12V battery were replaced under warranty so glad I did not connect anything to the terminals under the rear seat. My last three charging sessions had some strange behavior during the first hour of charging (it was not warming the battery cells per SMT). I am currently monitoring the charging sessions sure hope the new PCS board is not having any problems! This charging behavior started after the last firmware update. It is possible the UMC2 delivered with the car is developing a problem?

 

[thefrog1394]

With the cars telemetry, it sure seems probable that Tesla could deduce energy was drawn out of the traction battery for ancillary purposes (especially if it was many KWH). Currently the DC-to-DC voltage output is around 14.7V while topping off the 12V battery and around 13.4V (float voltage) at all other times while not sleeping. Around a year ago the DC-to-DC converter would output at least 14.7V ( if memory serves) anytime the car was not sleeping, which many people attributed to early demise of the 12V battery (including me); likewise, at least for my Model 3, the car would almost always stay awake for six to eight hours anytime the car woke from sleep state for any reason. Tesla now seems to be handling the charging of the 12V battery in a better way and the phantom drain is much less than before. My PCS board and 12V battery were replaced under warranty so glad I did not connect anything to the terminals under the rear seat. My last three charging sessions had some strange behavior during the first hour of charging (it was not warming the battery cells per SMT). I am currently monitoring the charging sessions sure hope the new PCS board is not having any problems! This charging behavior started after the last firmware update. It is possible the UMC2 delivered with the car is developing a problem?

Yea, I'm definitely leery of doing something like this for fear of voiding my warranty (if you're listening Tesla, I haven't actually touched my car! I'm just thinking about it, which last time I checked, shouldn't impact my warranty ). Stories like this don't help. It's really insane that Tesla spends their time monitoring forum posts to try and deny warranty claims.

My question would be, does the PCS really know how much current it's outputting? Presumably the 12V state of charge logic, dying battery warnings, etc are all handed on the path from the battery to the PCS which is why tapping the 12V battery directly causes so many issues. For those that have done something like this with ScanMyTesla connected (@FalconFour ) do any of the measured metrics indicate the additional power draw on the PCS?

 

[FalconFour]

Oh god yes It absolutely does, that's where the importance of watching PCS comes from. It shows how many watts, amps, volts the PCS is outputting. It's when it goes to 0 / drops off the bus, that you have to worry.

There is a lot of stuff Tesla knows/gets, but doesn't ever look at/care about. Remember, warranty is not "all or nothing". Your modifications to the PCS don't affect the warranty of the seats or MCU, for example. They'd even have to stretch pretty hard to shake their finger at it for a charge port fault or something. Much of Tesla's problem is incompetence/lack of communication in the service department (I've had relatively good service, but so many stories of bad service)... you'd honestly be lucky to get good warranty service even on a new car, tbh

They collect a ton of data, but don't have the systems/people to look at all of it. So yeah, the car knows, but there's too much noise in the world. They don't care.

 

[thefrog1394]

Oh god yes It absolutely does, that's where the importance of watching PCS comes from. It shows how many watts, amps, volts the PCS is outputting. It's when it goes to 0 / drops off the bus, that you have to worry.

There is a lot of stuff Tesla knows/gets, but doesn't ever look at/care about. Remember, warranty is not "all or nothing". Your modifications to the PCS don't affect the warranty of the seats or MCU, for example. They'd even have to stretch pretty hard to shake their finger at it for a charge port fault or something. Much of Tesla's problem is incompetence/lack of communication in the service department (I've had relatively good service, but so many stories of bad service)... you'd honestly be lucky to get good warranty service even on a new car, tbh

They collect a ton of data, but don't have the systems/people to look at all of it. So yeah, the car knows, but there's too much noise in the world. They don't care.

PCS is $1,500 so that's an expensive mistake if that's no longer covered. That being said, it appears to not be covered under the battery warranty, so for those of us past 50k miles it's less of an issue. I'd mainly be concerned with Tesla trying to avoid a warranty claim on the high voltage battery. Presumably the reason they note stationary power specifically in their warranty documentation is b/c additional charge cycles could wear out a battery faster. For what I'd be using this for (emergency power or the occasional tailgate) the argument that an extra few charge cycles killed the battery would obviously be absurd. But Tesla service doesn't have a particularly stellar reputation these days so who knows

Biggest risk is probably a future software update adding more checks and throwing errors. Aftermarket subwoofer users would be screwed then too.

 

[FalconFour]

PCS is $1,500 so that's an expensive mistake if that's no longer covered

Ha! PCS is inside the sealed battery, so you're boned either way -- good luck getting that thing open . In my experience, the thing is a tank. You can't hurt it. You can only piss it off, mildly. If it doesn't like what you're doing, it shuts off. And it shuts off at the drop of a feather out of line with operational limits. It's also liquid-cooled, so it can take a constant load. A 12v-pull, wait 30 seconds, make sure none of your accessories are still connected to PCS (it's sensitive and does interesting self-checks), and plug it back in... that fixes everything. I've had some scares. haha

To me, PCS replacement has never been a valid consideration... it's either gonna be way more, or way less than that. Not at all likely to be the "way more" problem, though, so thus I persist. Risk balance.

 

[thefrog1394]

Ha! PCS is inside the sealed battery, so you're boned either way -- good luck getting that thing open . In my experience, the thing is a tank. You can't hurt it. You can only piss it off, mildly. If it doesn't like what you're doing, it shuts off. And it shuts off at the drop of a feather out of line with operational limits. It's also liquid-cooled, so it can take a constant load. A 12v-pull, wait 30 seconds, make sure none of your accessories are still connected to PCS (it's sensitive and does interesting self-checks), and plug it back in... that fixes everything. I've had some scares. haha

To me, PCS replacement has never been a valid consideration... it's either gonna be way more, or way less than that. Not at all likely to be the "way more" problem, though, so thus I persist. Risk balance.

Charger failure within the PCS seems to be a thing at least. It would be pretty dishonest to claim that an inverter caused premature charger failure IMO, but given it's all within the same device, it wouldn't completely surprise me to see Tesla try and pull that. I'm not suggesting that hooking up an inverter would actually cause HV battery damage or PCS damage, just that Tesla might claim it did. In any case, social media posts or physical evidence of modification are probably the biggest things to get you in trouble on that front vs some sort of analysis of historical power draw from the system.

 

[rrolsbe]

Ha! PCS is inside the sealed battery, so you're boned either way -- good luck getting that thing open . In my experience, the thing is a tank. You can't hurt it. You can only piss it off, mildly. If it doesn't like what you're doing, it shuts off. And it shuts off at the drop of a feather out of line with operational limits. It's also liquid-cooled, so it can take a constant load. A 12v-pull, wait 30 seconds, make sure none of your accessories are still connected to PCS (it's sensitive and does interesting self-checks), and plug it back in... that fixes everything. I've had some scares. haha

To me, PCS replacement has never been a valid consideration... it's either gonna be way more, or way less than that. Not at all likely to be the "way more" problem, though, so thus I persist. Risk balance.

My PCS was replaced because the onboard charger portion on the board was flaky, the DC-to-DC converter had no problems (at least to my knowledge).

 

[FalconFour]

Wow, well, this thread has been a double "ackshuelly" for me. Hey, I'll take it. So, the PCS can be replaced Interesting. I never thought that was even possible or necessary. Good to know...

OK, so maybe don't beat it to a bloody hell, but I think it's safe to bash it on the head a few times. It's a hefty boi.

 

[smatthew]

Wow, well, this thread has been a double "ackshuelly" for me. Hey, I'll take it. So, the PCS can be replaced Interesting. I never thought that was even possible or necessary. Good to know...

OK, so maybe don't beat it to a bloody hell, but I think it's safe to bash it on the head a few times. It's a hefty boi.

The PCS is in the penthouse - right below the rear seat. It's actually pretty easy to get to and definitely not sealed in. Mine was replaced around 50K miles under warranty.

 

[jjrandorin]

Oh god yes It absolutely does, that's where the importance of watching PCS comes from. It shows how many watts, amps, volts the PCS is outputting. It's when it goes to 0 / drops off the bus, that you have to worry.

There is a lot of stuff Tesla knows/gets, but doesn't ever look at/care about. Remember, warranty is not "all or nothing". Your modifications to the PCS don't affect the warranty of the seats or MCU, for example. They'd even have to stretch pretty hard to shake their finger at it for a charge port fault or something. Much of Tesla's problem is incompetence/lack of communication in the service department (I've had relatively good service, but so many stories of bad service)... you'd honestly be lucky to get good warranty service even on a new car, tbh

They collect a ton of data, but don't have the systems/people to look at all of it. So yeah, the car knows, but there's too much noise in the world. They don't care.

They would likely only care if someone had enough loss in the battery capacity to actually trigger the battery degradation warranty. I am not inside tesla, dont have any specific knowledge of their inner workings etc, however, it is extremely common for "the process" around any sort of part failure etc to have someone "look at the logs / data".

TL ; DR , Tesla wouldnt look at the data until a battery claim, but a battery claim would have them look at the data 100% of the time as part of the process of processing / approving such a claim. They would look if there was an issue.

 

[megabyte]

The DC-DC converter is capable of producing 193Amps of 12Vish power. Unfortunately, there's no way to access all that power without the computers freaking out. The car measures current delivered to different sources, and if the charging current and current used are significantly different, it freaks out. So connecting directly under the rear seat is a no-go.

If you could ensure that the fridge and freezer don't operate at the same time, you could use a SmartSine 300w inverter connected to the 12V power socket in the center armrest. The limit there is 12 Amps continuous, 16 amps peak.

Sorry this isn't correct. The issue is not the car monitors what is used versus what is put out. There is large capacitors at the input of most inverters to smooth the power coming into it. When connected to the 12v output of the Penthouse, these are viewed as very close to a short circuit, so the efuse kicks in to prevent damage to wiring, etc. The below guide has been followed by myself, and on 4 days in a remote cabin my wife could dry her hair through a 1800W inverter for a decent 10mins a time, of this DC to DC converter, with use of the slow charge (of the capacitors) circuit.

Model 3 Inverter Power Connection

Not ideal, but it works with some babysitting. For more AC power than a standard 12V Power Socket inverter could provide, I tapped into the high current DC-DC output from the Tesla PCS (Power Conve…

www.cafeelectric.com