I bought a Hobbywing ESC - Is this our Holy Grail?
- Thread starter AlexLTDLX
- Start date
Thanks much! I will most likely use the P-2 as I don't need max boost and it claims 78% isentropic efficiency. I doubt I will be running more than 10 psi boost on my Jag as I am likely too old to race it competitively. Just want to feel what it can do with a modern engine and transmission. I did race it in a slalom and beat two 911's and took 2nd place to another. In the high-speed section, drifting it was like being in a dream. I will never forget it. I will definitely get back to you when I am ready to decide if I need an intercooler. In the meantime, if you get a chance, please send me the links to the aftercoolers you are referencing.
The efficiency on a supercharger is not a fixed value. It depends on where the operating point falls on the compressor map. If you take a big supercharger and run it at low speed the isentropic efficiency suffers. A Rotrex C15-60 at 10 psi of boost has an efficiency of 60% at the surge line and 74% in its sweet spot (0.09 kg/sec) air flow. I haven't found a compressor map for the P-2. If anyone has, I'd love to see it.
Look at killerchiller.com for freon cooling of liquid. This wasn't one of the sites I found before but they are cheaper. You need the liquid to air heat exchanger for the intake air but they are not real expensive <$200. Check out intercoolers at frozenboost.com
Look at killerchiller.com for freon cooling of liquid. This wasn't one of the sites I found before but they are cheaper. You need the liquid to air heat exchanger for the intake air but they are not real expensive <$200. Check out intercoolers at frozenboost.com
Homemade WRX
New member
You are 100% correct on that. I'm also well aware of it. Belt drive also isn't an option for my project as I'm doing an electrically assisted turbo.
The biggest issue for me will be getting the alignment correct and the spark eroded shaft end that will couple to the compressor wheel. Potential wheel, slip is a concern as well, particularly in regards to load/unload. This part of why I'm so concerned about the 'punch' aspect of the ESC in regards to both applying power and braking.
So the drive unit will align concentrically via a bore in the compressor housing. It already has three threaded M6 bolt holes on the comp housing (my second concern about aggressive loading) but that makes fixturing 'easy'.
The housing itself will be fed oil for the high speed bearings that the shaft and a tunable damper (vibrations are a real concern) will ride on, as well as to lubricate the gear box. Ironically this shaft and damper layout is more like an F1 MGU-K layout all because I'm using COTS motors and asking 120k+ RPM for any BLDC is a big ask.
Homemade WRX
New member
If doing a centrifugal, an air to air would also be another potential option and then just less mess with a heat exchanger, water pump, hoses, etc...
Where I live the ambient day time temperature is greater than 90 'F six months of the year. An air-to-air intercooler can never cool the intake air to less than ambient. The difference between the intercooler outlet temperature and ambient is called the approach temperature. The surface area of the intercooler goes up exponentially to get closer approaches. I have no personal experience but from what I have read an approach temperature of 50 'F is typical for an air-to-air intercooler. So that would mean 140 'F charge intake temperature where I live.
By chilling a liquid coolant (glycol water) with freon the coolant temperature can be brought down to about 30 'F. A liquid-to-air heat exchanger with the same air side surface area will achieve a better approach temperature than an air-to-air. This is because the heat transfer coefficient for a liquid to a metal surface is much greater than air to a metal surface. So let's assume the approach is 30 'F for liquid-to-air. That would mean a charge intake temperature of 60 'F or an 80 'F improvement over air-to-air. The car's A/C system does not have enough capacity to keep if the chiller is used all the time so a small tank (say 2 gallons) of chilled liquid would provide enough capacity for a short run like 15 seconds.
I've taken this thread off topic. Let's start a new topic thread for any further discussion.
Homemade WRX
New member
I mean if you're that worried about your charge air temps being at or less than ambient, yes, you'll need some thermochemical one way or another 
Plumbing your AC into your intercooler system or even just inlet path has been done many times before. I'm reminded of a particular Dakar project that did it.
Just curious, are you chasing the cold charge air temps for air density or because of knock?
BTW, your 'approach' temperature will be heavily dependent on how much boost you run, but you seem to be aware of that. Using some random 50* rule of thumb is also silly/dangerous.
Plumbing your AC into your intercooler system or even just inlet path has been done many times before. I'm reminded of a particular Dakar project that did it.Just curious, are you chasing the cold charge air temps for air density or because of knock?
BTW, your 'approach' temperature will be heavily dependent on how much boost you run, but you seem to be aware of that. Using some random 50* rule of thumb is also silly/dangerous.
Thanks much for the info! I've been tied up for a bit. Looked at Killerchiller, very interesting. I have a universal AC cert. so could DIY. Was planning to add AC to Jag, so I could do some specific heat calcs to see how to size it to accommodate intercooler for intermittent e-turbo use as I am planning. May be able to put an accumulator for liquid refrigerant just upstream of the intercooler to take advantage of the large latent heat of vaporization. May be able to set it up to completely divert R-134a from AC during hard acceleration.
I had a question about the direct drive bearing concerns you have - were you talking about a splined or hexagonal motor drive like Alex used on Sledgehammer that accommodates some misalignment?
Both density and knock. I said 50 'F approach temperature because I got that number from a website. The approach temperature depends on the size of the heat exchanger (surface area) and the heat transfer coefficients (dependent on several things). But a fact is that as the approach temperature approaches zero the heat exchanger size approaches infinity. With more boost you need a larger heat exchanger to keep the same approach temperature all else being equal.I mean if you're that worried about your charge air temps being at or less than ambient, yes, you'll need some thermochemical one way or another
Just curious, are you chasing the cold charge air temps for air density or because of knock?
BTW, your 'approach' temperature will be heavily dependent on how much boost you run, but you seem to be aware of that. Using some random 50* rule of thumb is also silly/dangerous.
I started a new thread let's take the discussion there.
There are both alignment and balance issues to consider. I ran this by a rotating equipment enginner that I work with and he was pretty skeptical that a DIY coupling would give reliable performance. Still the proof is in the pudding so we'll see how the sledgehammer and other direct coupled projects hold up. Maybe I'm overly paranoid. My current plans are to use a Rotrex C15-60 supercharger with a 10,000 rpm motor. McMaster-Carr sells a fit-for-purpose coupling good for 12,000 rpm.
I was watching a video by Peter Bjorck and he mentioned an off the shelf interchiller kit that uses the factory AC and just diverts it to a 2nd heat exchanger at the press of a button. seemed like an awesome solution for cooling on demand.
https://fiinterchillers.com/ is the product
is the video. he has some awesome projects.
https://fiinterchillers.com/ is the product
Please see the thread entitled charge air cooling with freon. The second heat exchanger you mention does not directly cool the charge air with freon. It cools a non-freezing coolant (likely water/glycol mixture). This chilled liquid is then pumped to an liquid-air intercooler heat exchnager that cools the intake air. Most cars shut-off the A/C compressor at wide open throttle so there needs to be a reservoir of chilled coolant for wide open throttle bursts.
soooo ... the combo
castle 2028 800KV sensored motor behind a HW max4HV esc as underdog solution
vs
LMT 30100 / MGM gorilla ESC as premium sledgehammer solution.
It occurs to me that the 2028 is quite close to the HW 56113 800KV in dimensions (minus sensor). The HW 56113 is "rated" 8s max and popular behind the HW max5 ESC ... The CC 2028 800KV is rated up to 12s and comes as sensored version. GOOD!
(there IS rumour out there that HW 56113 has e.g. sealed bearings vs the open ones on the 2028 etc ... thats why HW appears to cost more to the end user)
Now to the smaller hammer (ball peen) :
The HW Max4HV is being sold as nice combo FROM HW with a monster 70125 sensored 4pole 560KV motor ... rated for 12s as well. As you can imagine this from factory tuned combo would have gobs of torque ... and would deliver 15kW (not just burst) stable.
Me seriously looking at this "ball peen" combo now. With the torque reserves gearing slightly up with a 24-28 teeth pulley (@12s/560KV) to get the right input rpms with a belt driven P2 seems doable as well without shooting amps through the roof...
as for a geared SC with a 3.6 step up e.g. ... the 70125 seems big enough to me to drive with the right motor to input underdrive 8rib belt pulley combo to have about 12000 inputshaft rpms at 12s ... to be tested as well.
(oh ... and the HW ball peen combo would still be half the cost of the LMT/MGM sledgehammer combo)
castle 2028 800KV sensored motor behind a HW max4HV esc as underdog solution
vs
LMT 30100 / MGM gorilla ESC as premium sledgehammer solution.
It occurs to me that the 2028 is quite close to the HW 56113 800KV in dimensions (minus sensor). The HW 56113 is "rated" 8s max and popular behind the HW max5 ESC ... The CC 2028 800KV is rated up to 12s and comes as sensored version. GOOD!
(there IS rumour out there that HW 56113 has e.g. sealed bearings vs the open ones on the 2028 etc ... thats why HW appears to cost more to the end user)
Now to the smaller hammer (ball peen) :
The HW Max4HV is being sold as nice combo FROM HW with a monster 70125 sensored 4pole 560KV motor ... rated for 12s as well. As you can imagine this from factory tuned combo would have gobs of torque ... and would deliver 15kW (not just burst) stable.
Me seriously looking at this "ball peen" combo now. With the torque reserves gearing slightly up with a 24-28 teeth pulley (@12s/560KV) to get the right input rpms with a belt driven P2 seems doable as well without shooting amps through the roof...
as for a geared SC with a 3.6 step up e.g. ... the 70125 seems big enough to me to drive with the right motor to input underdrive 8rib belt pulley combo to have about 12000 inputshaft rpms at 12s ... to be tested as well.
(oh ... and the HW ball peen combo would still be half the cost of the LMT/MGM sledgehammer combo)
Last edited:
as for the VESC:
as long as the relively complicated (trigonometry! yeah I know you can work with approx tables and do int arithmetic vs float but still "expensive") FOC calculations are completely done in software and the VESC compute power is what it is right now (and us running 4 pole motors or in case of TP Power 6 pole motors with high erpms) ... don't rely on that!
Once they have upgraded their compute part I might look at it again ... (The mosfet final stage looks ok in terms of dimensioning and heat dissipation ... it is just this hard realtime compute and control requirement which is "killing it" for me now)
So basically super low KV (with reltively low rpms as a result) is the domain of the VESCs!
Just look at where they are popular (for a reason) : eBoards !
LOL ... just logged into my vesc project forum account ... to find out "You are not allowed to post" ... read only for me ... must be an old post from one year ago addressing exactly above issue which made me "persona non grata"?!
as long as the relively complicated (trigonometry! yeah I know you can work with approx tables and do int arithmetic vs float but still "expensive") FOC calculations are completely done in software and the VESC compute power is what it is right now (and us running 4 pole motors or in case of TP Power 6 pole motors with high erpms) ... don't rely on that!
Once they have upgraded their compute part I might look at it again ... (The mosfet final stage looks ok in terms of dimensioning and heat dissipation ... it is just this hard realtime compute and control requirement which is "killing it" for me now)
So basically super low KV (with reltively low rpms as a result) is the domain of the VESCs!
Just look at where they are popular (for a reason) : eBoards !
LOL ... just logged into my vesc project forum account ... to find out "You are not allowed to post" ... read only for me ... must be an old post from one year ago addressing exactly above issue which made me "persona non grata"?!
Last edited:
Homemade WRX
New member
I would run cogged and not a ribbed drive. The tension to make the ribbed belt work is something these motors (specifically bearings) weren't designed for.soooo ... the combo
castle 2028 800KV sensored motor behind a HW max4HV esc as underdog solution
vs
LMT 30100 / MGM gorilla ESC as premium sledgehammer solution.
It occurs to me that the 2028 is quite close to the HW 56113 800KV in dimensions (minus sensor). The HW 56113 is "rated" 8s max and popular behind the HW max5 ESC ... The CC 2028 800KV is rated up to 12s and comes as sensored version. GOOD!
(there IS rumour out there that HW 56113 has e.g. sealed bearings vs the open ones on the 2028 etc ... thats why HW appears to cost more to the end user)
Now to the smaller hammer (ball peen) :
The HW Max4HV is being sold as nice combo FROM HW with a monster 70125 sensored 4pole 560KV motor ... rated for 12s as well. As you can imagine this from factory tuned combo would have gobs of torque ... and would deliver 15kW (not just burst) stable.
Me seriously looking at this "ball peen" combo now. With the torque reserves gearing slightly up with a 24-28 teeth pulley (@12s/560KV) to get the right input rpms with a belt driven P2 seems doable as well without shooting amps through the roof...
as for a geared SC with a 3.6 step up e.g. ... the 70125 seems big enough to me to drive with the right motor to input underdrive 8rib belt pulley combo to have about 12000 inputshaft rpms at 12s ... to be tested as well.
(oh ... and the HW ball peen combo would still be half the cost of the LMT/MGM sledgehammer combo)
Further looking at the numbers, it seems the 70125 could run 15kw at about 340 amps. Would be interesting. I had looked at it but it's kV is far too low for my application. It is also a big can but if you have the room for the gearing, awesome!
I'm looking at the Max4 with a 56123 from SSS Motors for my own application. I just wish it was sensored. I do have some concerns with eRPM accuracy with the 6 pole architecture.
good point re tension required for rib belts and motor bearings!
I can't find anything on the HW site that mentions a kW rating. Their sizing information seems to be "good for 1/4 to 1/5 scale cars". What they do give is:
560 KV
70 mm diameter by 125 mm long
2 kg weight
8 mm shaft
By comparison a Neumotor 3030 which is rated for 15 kW continuous and 30 kW peak:
466 KV
90 mm diameter by 128 mm long
2.9 kg weight
12 mm shaft
If HW does say that the 70125 is good for 15 kW continuous it would seem that the rating is optimistic by comparison to a Neumotor. 15 kW is 20 hp and an 8mm shaft just does not feel right.
ok ... I just multiplied 300A (the esc rating) with 50V(~12s lipo esc and motor rating) since it is sold as a combo as is from HW ... maybe I have used too much "over the thumb" estimates
I also think that the jury is still out on whether the Max 4 ESC is really capable of 300 A continuous. Admittedly in the test Alex did at a little over 300A (for 14 seconds) the fan was not running but the ESC temperature was increasing at 2 'F per second. Really need to see if the temperature reaches steady state with the fan on but I think you cannot datalog and run the fan simultaneously (that's seems very weird).
I've actually purchased the Max4 and 70125 combo on those exact assumptions @cmoalem i think its a pretty good option for the price point.
I'm in the process of making a new shaft to suit a 15mm HTD belt and pulleys. I've got a 20t pulleyfor the shaft and 24t, 28t and 32t options for the motor.
i'm probably a month or so off doing some testing but will definitely update when i do.
I'm in the process of making a new shaft to suit a 15mm HTD belt and pulleys. I've got a 20t pulleyfor the shaft and 24t, 28t and 32t options for the motor.
i'm probably a month or so off doing some testing but will definitely update when i do.