Oooh, this is a Model X thread. In most respects, things would be about the same as a Model S, but I think I figured out how it's different in this case with HVAC. The Model X has a bigger cabin inside, and in the 6 or 7 seat versions, they do have an extra air conditioning unit to help with cooling the back half. So that does make sense that two air conditioners running about 3.5kW each would show a 7kW power draw.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Charging Raven Long Range on HPWC
- Thread starter platylover
- Start date
I'm smart enough to seldom say "not possible" because I have not seen the design documentation for the car and I was not present when the observation was made but I will say "not very likely" because 7 kW doesn't seem to make much sense. I have measured, using accurate instrumentation, the current drawn by a Model X 100D (5 seater, white on white) under sun load on a warm day and posted the results. The peak draw (see graph in No. 15) was 4.5 kW and that persisted for only a few seconds. Now I recognize that if the OAT is higher than that at which I experimented (80's) head pressure will be higher and compressor load will go up but I have never seen it jump as much as 55% (7/4.5 -1) as temperature varies between the 80's and 90's. As an example of this, the draw of an A/A heat pump in Virginia increased by 11% as the temperature rose from 78 °F at about 8 AM to 97 °F at the hottest part of the day at around 3 PM. Another heat pump, also A/A and at the same location, showed an increase of 12%. Applying that figure to the 4.5 kW measured at 82 °F for the X, we could estimate that the peak A/C load might be 5 kW in the high 90's and, assuming the increase to be linear with OAT, about 5.6 kW. Now that's getting close to 6 kW but 7 is quite a reach and the quote says "near triple digits" meaning that the estimate of 5 kW would apply.
Second, I don't know anything about Tesla Remote. A quick check on the internet shows that it is another app that uses the API to display data to a user on his smart phone. Since, AFAIK, the API only provides charger current readings, since it only does that when polled, since I don't know what the polling interval is and since we know that peak draw only persists for a few seconds, I don't have much confidence in the 7 kW reading. But if you can furnish more detail about how this reading is obtained I'll certainly re think this.
Finally, 7kW will, in a modern air conditioner, which we certainly assume the air conditioner on the X to be, produce about 7 tons of cooling! Yes, 7 tons which is 84,000 BTU/hr. My experiment of the other day showed that 4.5 tons is sufficient to drop the temperature of a Model X cabin by 25 °F in a few minutes and about 1 ton is sufficient to maintain a 15 ° drop against afternoon sun loading. For perspective: my whole house in Virginia is cooled by a heat pump (W/W) rated 8 tons. The new addition to my summer house is served by a 5 ton (W/W) unit. Seven tons in a car? It doesn't compute.
Regarding the theory of another compressor in the 3 row models: I don't think that's very likely either. The cabin in those cars isn't any bigger (though the thermal mass is larger by the thermal mass of that 3rd row of seats which isn't much compared to the total thermal mass). The second row of seats, however, does block air flow to the rear of the car so that additional ducting is required to get A/C to the third row when it is installed. That additional ducting might have its own blower and maybe even its own evaporator but certainly not a separate compressor.
Your chart says it's from a 90F cabin. That's not really a good test for summer in most places. I'm not surprised you didn't see the full draw the system is capable of - a 90F cabin in 82F environment doesn't need the full power of the system, the compressor and the condenser fans don't need to work nearly as hard as they would with hotter temps inside and outside the car.
With near triple digit outside temps, the cabin would be ~150F if cabin overheat wasn't on. It is in my case, so in theory the cabin is around 105-110F, but elements inside are heat soaked above what the mere temperature would suggest. And since this is Philly, the RH is up to the 80s or 90s, with a dewpoint higher than the AC target temps, which means a lot of water that has to come out of the air, too.
I can't speak to the methods used by Tesla Remote or their accuracy (used to be Remote S - it isn't "another" app using the API - it's the original 3rd party app, and has long has capabilities other apps didn't,) but the X certainly has more cooling capability than my Volt did, and I know from OBDII instrumentation that that car could exceed 5 kW on a really hot day (all up, compressor, HVAC fan, condenser fan, computers,) during initial cooldown, so the 6-7 kW seemed quite plausible to me.
Actually it is if you know how to interpret the data and, while I would love to test under conditions closer to what you guys are currently experiencing the reason I am in Quebec at the moment is so that I don't have to experience it! Hope you all get relief soon Apparently it is on the way.
Yes, that's right with respect to the steady state but when you ask a PID system to start up with a SP outside the proportional band it is going to go full bore until you get within the proportional band. As I said, I don't have the details of the system design so I don't even know that the car is using PID control. It's pretty clear from the way the autopilot behaves and given Tesla's love of AI, that there is lots of fuzzy control on board but even with fuzzy control there is going to be a rule that says "if cabin is way too hot WRT to SP THEN cooling is maximum". The plot of current draw indicates this behaviour whether the algorithm be PID or fuzzy.
The load on the compressor is determined primarily by the OAT as that is the main determinant of the head (high side) pressure against which the compressor must work. Initial cabin temperature doesn't matter that much in this regard. Yes, you'll have high super heat which raises suction pressure but that actually reduces the compressor's load. What it does influence, of course, is the length of time it takes to cool the cabin down to the steady state. If I park my white on white outside in full sun around noon the interior air temperature will zoom up over 100 in a couple of minutes. But the 4.5 tons A/C will cool it down even quicker. OTOH if I leave it outside for hours the hot air has the opportunity to "charge" the more massive (than the air) parts of the car (seats, dashboard, floor...) with heat and it will take the A/C appreciably longer to remove that heat than if it is just the air that has been warmed.
Condenser fans actually do less work as OAT goes up as the air is less dense. In small systems like these they generally are not speed controlled except when it gets cold out and head pressure gets too low in which case they are slowed down. That's because as OAT and head pressure rise so does the temperature of the exhaust gas so that difference between ambient and exhaust gas temperature does not drop at the same rate as OAT rises. But this is Tesla. They may regulate condenser fan speed to maintain a desired head pressure in the warm region too. In any case fan current draw is a small fraction of compressor draw.
I did make estimates based on OATs in the high 90's and high 110's based on observed change in compressor load with OAT for a couple of A/A heat pumps. This comes under interpreting the data I collected. The largest draw for my X at near 110 °F would be 5.6 kW based on that data and what I measured. They are estimates. I recognize that.
I'm not quite sure what you are trying to say. Temperature is temperature.
The temperature of an object and its thermal mass determine its heat content. If you "heat soak" something its temperature rises until it gets hot enough that no more heat flows into it. I discussed this a little above. If the car is at equilibrium in the garage at 80°F then everything in it is at 80 °F. If I pull it out of the garage into the sun the temperature of the air in the car will be 110 °F in a few minutes but the temperature of a floor mat will be perhaps only 82. It has only absorbed 2*S units of heat where S is its specific heat. Two hours later the temperature of that mat may be 90. It has absorbed 8*S more units of heat. That is heat which would have to be withdrawn when pre-conditioning is used. But, as mentioned above, this determines how long the high current draw persists. Not the magnitude of the peak draw.
Again, not likely as 90% relative humidity is nigh on to impossible in a weather system at high temperature. The highest dew point I actually ever measured was 21 C when the dry bulb was a relatively modest 27.8 °C for an RH of 67% and a water vapor pressure of 25 mb. According to the Wunderground website Virginia beach has seen dew points as high as 30 °C which for a dry bulb of 35 °C corresponds to RH of 75% and a water vapor pressure of 42 mb. Again, I suppose it's possible to see RH's that high but it's more likely that your RHs are well below 80%. Yesterday the temperature at the Philly airport peaked at 97 °F in the late afternoon. The RH was 44 - 47%. RH's only rise to 70 - 80% at night when it is much cooler. Beyond that, there simply isn't that much water to remove because there simply isn't that much air trapped in your car. Yes, part of the peak load of 4.5 kW is going to remove latent heat but in the few minutes it takes 4 tons to get the air temperature down to SP it has removed much of the water in it. Open the doors and let in a new mass of damp air and its a different ball game.
Had you posted that you put an ammeter on the circuit feeding your charger, turned on climate in the car and observed the equivakent of 7 kW increase in measured current then your assertion would be more credible. As it is I won't say "impossible". Just "unlikely".
Your expectations for how an HVAC system works are very different from how Tesla's system appears to work. Tesla's compressor is inverter controlled, and ramps up and down progressively - and the condenser fans do, too. If you've ever supercharged in a warm environment, you've heard them slowly ramp up progressively. Neither one has the sort of on/off behavior you're suggesting.
If your data showed it going straight to 4.5 kW and sitting there for a several minutes before dropping off, I'd be more inclined to believe your conclusions - but it doesn't. It shows the system building progressively to 4.5 kW, then dropping again sharply as it starts getting to temps it wants.
I'm quite certain that if you'd been in more trying environments, you'd have seen that peak continue to build progressively to higher numbers before leveling off in the 6-7 kW neighborhood for a while and finally heading down as it gets cooler.
I cannot see how you concluded that I am suggesting on/off control. Do my data show on/off control? Do any of my words? If so please point out which ones and I'll correct them.
It's a minor point but the motors are not controlled by the inverter. The inverter is nothing more than a set of 6 switches which are turned on and off in response to the controller's commands.
The data shows behaviour typical of a PID system with continuous output. That's why I interpret it as being a PID or PID-like fuzzy system. With respect to the progressive buildup: that's the "soft start" feature incorporated in most all modern compressor (and other type load) controllers. Even though I expect this is a scroll compressor there would be high inrush current if the response to a controller command to "floor it" resulted in flooring it. Soft start is a lot easier on the motor, the compressor and the electronics.
Well as I've stated before if the OAT is 10 ° hotter then I would expect the peak draw to be 12% or thereabouts higher. Were it 20 ° hotter out I'd expect perhaps 25% higher. I would expect to see the high current draw last a little longer for a brief exposure and appreciably longer for a "soak" in the sun.
Note that the 12% estimate is based on fixed systems using R22. I expect (but, again, do not know) that the cars use R134a. I don't think that would make a lot of difference but as the two have different PH characteristics it's going to make some. In any case, the 12% number is a bit of a WAG.
I may regret this but I'll pray for an unusually hot day so I can do another experiment.
Please note that I'm not trying to sell you on anything or really convince you of anything. My motivation is that I haven't had an HVAC failure recently and these discussions/experiments keep rust from forming in the mental hinges so I'll be better prepared when I do.
It's not a theory. There is, and it can be turned on and off. When driving, it will default off if it does not sense people sitting in the third row. Here are a couple of threads about control of the rear A/C unit.
Turning rear A/C off for better range?
Reat seats AC - Model X
Judging from those two threads it certainly is a theory. They are devoted in large measure to trying to figure out whether there is a separate compressor or just a separate evaporator. Common sense certainly suggests the latter. Clearly adding some piping to/from a remote evaporator (with its own blower and TXV) is simpler than having to add all of those plus a new compressor, filter dryer and condenser (with its fan) and is thus more cost effective. Beyond that there is no need for a separate compressor. Addition of two seats is a small fractional increase in thermal mass. Two more bodies adds 1365 BTUh i.e 0.11 ton to the steady state load. The compressor (single) delivers about 1.2 tons at steady state in a modest solar loading situation (afternoon sun, 82 °F OAT, 65 ° cabin) and about 4.5 tons at turn on. I don't run an HVAC company but if I did and you proposed adding a separate circuit for an additional 2% load (most of the load is solar - not people) you wouldn't be working for me much longer. It just doesn't make sense. All that needs to be done is to deliver some of the cooling to the rear of the car where the extra seats are and which area does not, evidently, receive air flow from the front system. If it's not convenient to run ducts (which are quite large) the next best thing is to run liquid and suction lines to separate restriction device (TXV) and evaporator and that is evidently what Tesla has done. I note that several posters in the referenced threads came to this same conclusion. The rear system can easily be turned off by means of a liquid line valve and the amount of heat it extracts controlled by varying blower speed across the evaporator.
So far, we have common sense on our side but do we have any real evidence? Well there was one piece in the reference threads: this picture:
What does this label tell us? With certainty it tells us only that there are two configurations for the cabin A/C named Single and Dual and that the latter sends cold air to the last row of seats. It suggests, but does not clearly state, that there is only one compressor because if there were two there would be two separate circuits and the amounts of refrigerant and oil in each would be listed separately. If I shipped a 1 L bottle and a half liter bottle of something to someone he'd want to be told that I shipped 1 bottle containing 1L and one bottle containing 0.5L rather than that 1 sent 1 bottle containing 1 L and that the total shipment was 1.5 L.
The most telling thing here is that if there were a separate circuits the tolerance for the dual configuration would be ±28 grams as the uncertainly associated with charging each is ± 20 g.
Some of the extra 260 g of R134a (36% increase) is needed to fill the relatively long liquid line back to the evaporator and some more resides in the low pressure (evaporator and suction line) return as well. And some oil is transported through the circuit with the refrigerant but not an extra 90 g (64% increase) I wouldn't think, This suggests that the dual system has a different compressor (i.e. one with a larger sump) than the single. Is it, therefore, bigger? Does it move more refrigerant per hour in order to produce that extra 0.11 ton of cooling and thus draw more power? I have no idea. But it would seem from the available evidence and common sense that there is only one compressor.
A relevant question now arises: are the guys seeing 6 and 7 kW draws driving 3 row cars?
So far, we have common sense on our side but do we have any real evidence? Well there was one piece in the reference threads: this picture:
What does this label tell us? With certainty it tells us only that there are two configurations for the cabin A/C named Single and Dual and that the latter sends cold air to the last row of seats. It suggests, but does not clearly state, that there is only one compressor because if there were two there would be two separate circuits and the amounts of refrigerant and oil in each would be listed separately. If I shipped a 1 L bottle and a half liter bottle of something to someone he'd want to be told that I shipped 1 bottle containing 1L and one bottle containing 0.5L rather than that 1 sent 1 bottle containing 1 L and that the total shipment was 1.5 L.
The most telling thing here is that if there were a separate circuits the tolerance for the dual configuration would be ±28 grams as the uncertainly associated with charging each is ± 20 g.
Some of the extra 260 g of R134a (36% increase) is needed to fill the relatively long liquid line back to the evaporator and some more resides in the low pressure (evaporator and suction line) return as well. And some oil is transported through the circuit with the refrigerant but not an extra 90 g (64% increase) I wouldn't think, This suggests that the dual system has a different compressor (i.e. one with a larger sump) than the single. Is it, therefore, bigger? Does it move more refrigerant per hour in order to produce that extra 0.11 ton of cooling and thus draw more power? I have no idea. But it would seem from the available evidence and common sense that there is only one compressor.
A relevant question now arises: are the guys seeing 6 and 7 kW draws driving 3 row cars?
Somewhat to my surprise, I think we have a winner. The public version of the EPC shows four versions of the AC compressor, two marked MX Compressor, and 2 marked No RHVAC:
https://epc.teslamotors.com/#/systemGroups/66340
'
The line setup seems to clearly suggest only a single compressor system:
Oddly, I couldn't find the evaporator cores on any of the various group screens...
And yes, I had a 6 seat and now have a 7 seat that I was reporting 6 kW and in extreme temps 7 kW provided by Remote S/Tesla Remote (still don't know how the App is developing that information, either.)
Oh, so you're saying single unit either way, but there is a smaller version (lower power) and a larger version (higher power)? Same concept I was thinking of then (that the ones with rear A/C have something "extra" that can draw more power), but not actually done in two separate bodies.
Last edited:
That's what the post above proposed based on the refrigerant datasheet, and it seems to match what I'm seeing in the EPC.
Perhaps a quick review of how an automobile A/C works might be in order. Such a system is a closed loop containing a volatile substance, in this case, R134a (Suva, 1,1,1,2 fluoroethane). We have to start somewhere so consider the compressor inlet (suction port). The refrigerant is a gas at that point. It might be at a pressure of 30 psig and its temperature might be 45 °F (these numbers are pulled off the R134a P-H diagram and are just examples not to be interpreted as what is found in any particular system). The compressor is just a pump and might compress the gas to 200 psig. Compressing the gas makes it hot. It might leave the compressor at 150 °F or so. This hot gas is piped to the condenser over which ambient air is passed. If that air is cooler than 135 °F the hot gas will condense to a liquid. There is some resistance to flow in the condenser but not much so that the liquid leaving the condenser is still at a pressure of around 200 psig. If the air is at, say 90 °F, the liquid gets cooled further (called "sub cooling")but obviously can't cool below the ambient air temperature. The warm liquid is piped to the point where cooling is wanted at which point some kind of flow constrictor is installed followed by an evaporator which is another liquid to air heat exchanger similar to the evaporator (in a heat pump their roles are reversed by reversing the direction of refrigerant flow). In passing through the constriction the pressure of the liquid stream is reduced and the liquid converted to a spray of droplets which enter the evaporator. There they boil and in so doing absorb heat. The resulting cold gas is returned to the compressor suction inlet via larger (than the liquid line) tubing which is usually insulated as the gas is cold. If the pressure is 30 psig as we supposed originally the refrigerant boils at 35 °F. Just as the liquid can be cooled by the evaporator to a temperature below the condensation temperature so can the gas be warmed to a temperature higher than the boiling point. This is called "super heat" and as we want to be very sure that liquid is never returned to the compressor, systems are usually designed for a few degrees of super heat. That's why we supposed the return gas to be at 45 °F rather than at the boiling temperature corresponding to 35 °F.
Now if a new load is introduced, e.g. if you put two chubby kids in the back of your X, more heat reaches the evaporator and the liquid boils faster (e.g in the first quarter of the evaporators tubing) than it would with out the extra load and the super heat rises as does the temperature in the cabin. The evaporator is said to be "starved". The solution to this is to open the restricting valve wider letting more liquid enter the evaporator and making it able to absorb more heat. Note that this is often done automatically by a Thermostatic Expansion Valve (TXV). Really the point of all this is that the extra heat load will cause the TXV to allow more refrigerant to flow around the loop per unit time. This, of course, means the compressor must do more work and, therefore, draw more power from its energy supply. Thus the second evaporator is the "something 'extra'" that can draw more power, if rather indirectly.
I argued yesterday that a couple of chubby kids would not increase the load (0.11 ton) appreciably relative to the apparent solar load in an X. But today It occurred to me that the back end of my 2 row X isn't really a conditioned space as little blower air reaches back there. Put the extra row of seats in and now you want it to be conditioned space and so you install the extra evaporator to make it conditioned space. The dual system has 36% more R134a than the single and if we argue that 36% more refrigerant maps directly to 36 % more power requirement we'd have explained peak power demand of 6.12 kW in 6 and 7 seat models.
Now if a new load is introduced, e.g. if you put two chubby kids in the back of your X, more heat reaches the evaporator and the liquid boils faster (e.g in the first quarter of the evaporators tubing) than it would with out the extra load and the super heat rises as does the temperature in the cabin. The evaporator is said to be "starved". The solution to this is to open the restricting valve wider letting more liquid enter the evaporator and making it able to absorb more heat. Note that this is often done automatically by a Thermostatic Expansion Valve (TXV). Really the point of all this is that the extra heat load will cause the TXV to allow more refrigerant to flow around the loop per unit time. This, of course, means the compressor must do more work and, therefore, draw more power from its energy supply. Thus the second evaporator is the "something 'extra'" that can draw more power, if rather indirectly.
I argued yesterday that a couple of chubby kids would not increase the load (0.11 ton) appreciably relative to the apparent solar load in an X. But today It occurred to me that the back end of my 2 row X isn't really a conditioned space as little blower air reaches back there. Put the extra row of seats in and now you want it to be conditioned space and so you install the extra evaporator to make it conditioned space. The dual system has 36% more R134a than the single and if we argue that 36% more refrigerant maps directly to 36 % more power requirement we'd have explained peak power demand of 6.12 kW in 6 and 7 seat models.
If the HPWC is set up to deliver 48A or more (not guaranteed,) then the difference is 16A - 32A with the 14-50 and 48A with the HPWC.
If the station delivers 240V, with my Ludicrous Raven X's 324 rated Wh/mile and assuming 85% charging efficiency, that's 381 Wh/mile from the wall, and 7680 W vs 11520 W, so 20.1 rated mph vs 30.2 rated mph.
You haven't given enough information to provide an answer to that question, because you haven't specified how many amps. A wall connector can be installed on many different levels of circuits, from 15 amp to 100 amp. And the actual current available will always be 80% of what the circuit size is. So what size circuit are you talking about?
The mobile charge cable has a maximum current level of 32A, so that is what you can get from a 14-50 outlet. So if you had a wall connector on a 40A circuit, it would also pass 32A to the car, and there would be no difference at all between the wall connector or an outlet. But if you had the wall connector on a 50 or 60 or higher amp circuit, then it could get faster charging speeds. The Model X now has a 48A onboard charger, though, so that is a limit there, where a wall connector at 60A circuit or higher is going to be capped from the car's internal charger. Here's a table to get an idea of what charging speeds you can get from different sized circuits.
Wall Connector
In an X 100D it's pretty simple to figure approximate charge rates from AC chargers. At 240 Volts you'll get about 0.25%/hr/Amp charge. The rated miles for an X are very close to 3 per % of the battery so that corresponds to ~ 0.75 mi/hr/Amp. Thus OP's charge rate at 48 A is going to be about 12%/hr corresponding to 36 mi/hr which is quite close to his observed 34 mi/hr.
I reiterate that these are approximations. Their value is in that mental calculations using these round factors is very easy. If you are thinking about putting in a 60 amp breaker rather than a 50 it is easy to determine that it will get you 0.8*10*.25 = 2%/hr faster charging equivalent to about 6 mi/hr. Even an old fart like me can do that in his head.
I reiterate that these are approximations. Their value is in that mental calculations using these round factors is very easy. If you are thinking about putting in a 60 amp breaker rather than a 50 it is easy to determine that it will get you 0.8*10*.25 = 2%/hr faster charging equivalent to about 6 mi/hr. Even an old fart like me can do that in his head.
That's rounding up after assuming 100% efficiency in the first step.
You'd probably be closer jumping down to .2%/A/Hr, which would be 83% efficiency with no rounding.
That'd be .6 miles per Ah, and give ~29 mph at 48A, much closer to the official figures.
