ESCs for 12-16s

cmoalem

Active member
Soooo ... Alex is trying out the cool but expensive APD HV-Pro 16s https://shop.powerdrives.net/?product=hv-pro-400-16 !?
(after he toasted his old flier ... which one?)

I have found another cheap flier opto 400A with water cooling as well:

slightly smaller (300A and up to 14s) but with proper casing and some add ons:

similar to the flier (semi casing) and some add ons:
 
Last edited:
and then there is the super cool topic open ESC

http://vedder.se/2015/01/vesc-open-source-esc/
lots on youtube as well!

and here some robust, water cooled and compatible (expensive, but look at the features!) hardware (DROOL!):

has the charm of possibly integrating some combustion engine sensor data like MAP(implicitly boost seen by intake manifold:= MAP-14.7 at sealevel? or MAP - idle MAP adaptive?)/TPS/RPM/(IAT?) CAN bus splice (EDIT: already THERE!) and some smart software on the same platform 👹
EDIT of the EDIT: forget about the CAN bus on the VESC designs! a) the microcontroller currently is busy enough with high erpm motors! and b) the VESC OS is using the CAN port only for linking/syncing multiple VESCS (multi motor setup!)
 
Last edited:
Flier just changes cooling solutions for various models. The thing is, both of my Flier ESCs didin't fail from heat - they failed because the ratings are BS. My first one was the plane version of the boat one you've listed. 16s, 400 amps - 16s, yes, 400 amps - not even close. In fact, when talking with someone at Castle (a manufacturer of good quality RC stuff), he said that they would rate Flier's 400 amp ESCs at about 200 amps. After that, I went with a Flier 22s, 400 amp (450 amp) RC car ESC. That's the one that exploded spraying molten solder all over the underside of my car (some even went through my air filter).

I will NEVER use another Flier ESC for anything sizeable. Either APD or MGM (or similar) for me from now on. Besides which, with the Flier ESCs, you're flying blind - no data logging isn't a good thing. And I think the higher quality ESCs are more efficient and have better motor control. The one I'm using - APD HV-pro 16s 400 amp. The next step up would be the UHV 85 volt 400 amp model (but it' $1,700 USD). From MGM, if you can get by with 15s (the highest voltage their RC ESCs get to), For around $900, you can get this one:


But that's a bit less powerful than mine, which was about $200 more. Think in terms of watts. Peak voltage on the MGM is 63, with 400 amps continuous or 500 amps peak - so 25,200 watts continuous or 31,500 peak. But you'll never hit those numbers because of voltage sag. With a strong battery pack with minimal voltage drop (the lower voltage really makes this a challenge) - you might see a realistic 55 volts under load at the most, Giving you 22 kW continuous or 27.5 kW peak. My ESC caps at 68 volts and 400 amps continuous and 600 amps peak. In my mind, definitely worth the extra $200.

In my setup, with the RC LiPo batteries, my 16s (67.2 volts) drops down to 52 volts under load. The system consumes some 340 amps, but only about 280 amps actually gets to the motor phases. But battery draw is 340 amps, none the less. With my LTO pack, I'm hoping to see at least 58 volts under load, which should be enough for about 30,000-31,000 impeller rpm. If we can keep it at 60 volts, even better. But I've not fully saturated the motor yet, hopefully those power levels will allow us to do that.

Either way, think of it as a system. Settle on your battery chemistry, and figure out how many cells you need and what they cost. Also, keep in mind for higher voltage/lower current systems, you get less voltage drop in the system - through the cables, connectors, etc. Meaning you can use smaller cables and connectors. It's all a balance, and since this stuff is so new, we don't really know what works best yet.
 
so ... after some time we find the right info!

I like the explanations given here a LOT:


MGM esc programming problem

So ... the TP motors with their higher number of poles are more "sophisticated" (and torquey) and at the same time more finicky when run at or over the edge of their specs!

The verdict: TP motors should be watched carefully when run over 48V due to high phase currents. Most ESCs do not implement the right protection strategies (and die suddenly). Other ESCs which do protect themselves limit duty cycles "smartly".
While the TP motors seem to find lots of supporters, it is crucial to run them with a very good ESC (like the APD).
The other good ESC would then come from MGM, but there the proven LMT motors shine with better phase currents due to the simpler 2 pole construction. (but are less torquey!)

So ... TP it is ... but operated with the right voltage and with the APD HV-pro 16s! (what you save on motor price you put into the more costly ESC).

If you get LMT 3080 e.g. ... you could run it with a MGM controller. (Or even with a drastically erpm limited VESC 75/300, which due to the 2 poles of the LMT motors and reduced switching frequencies would probably be just fine!)

So ... the second best choice might be the LMT 3080 plus Trampa(!) VESC 75/300 (much cooler software and cheaper than the MGMs!) ... but nobody knows the torque characteristics/needed here. In terms of power output: The LMT3080 continuous power output spec seems to fit better for the 14kW requirement to produce 6psi boost on a 3.6!

VESCs all seem to have problems with higher pole inrunner motors and their erpm limitation. (Not fast enough processor!? or no realtime OS?)
 
Last edited:
Funny you guys are talking about this. I was just having a conversation with the guys from Flier and Rafael from ebay about this.

That stated that Flier has made some improvement to the design of the 400a unit over the last few months. On being better and upgrade mosfet. Next they have 2 new and upgraded units the 500a and 600a units now to meet the super high current demands. Also talking with the Airplane and Boat RC guys a few said to look into these units in the APD and MGM price range.


The HV300 with the built in water cooling looks interesting but they only do up to 14s

5-14 S LiPo - 20-59 V - 300 A continous current - up to 640 A peak current - BEC 10 A continous current - BEC 30 A peak current
 
Wow the Maytech 85V 300A (400 with water cooling) VESC looks outstanding. I love the robust power connections and they have an on off module for positive Battery pole.
 
Wow the Maytech 85V 300A (400 with water cooling) VESC looks outstanding. I love the robust power connections and they have an on off module for positive Battery pole.

VESC 6s have a theoretical limit @150000 erpm!
On a 2 pole motor thats good (LMT e.g.)

On a TP motor with 6 poles you would devide 150000 / 3 (number of pole pairs) => 50000 shaft rpms max!

That is "theoretical borderline" for a TP5860 with 1000KV at 48 V ... The APDs have a "limit" at 1000000 (1 million!) erpms ... thats how fast their internal controllers is.

On the vesc project forum there are reports of people having problems to set up higher erpm (>100000) motors correctly!

Nevertheless ... the Trampa VESC 6 75/300 does look "borderline ok" ... certainly more trustworthy than a Flier btw. (for a similar price!)
( I AM curious!)
interesting internals ... the motor "tests" are not so "scientific"


The Maytech Vesc MTSVESC7.5H clone is just that ... a clone ... and not fully software compatible (despite the mention of update via VESC_TOOL). I don't know if better or worse ... just much more expensive than the Trampa 75/300 (on Ali Express on APD level!)!
 
Last edited:
so ... after some time we find the right info!

I like the explanations given here a LOT:


MGM esc programming problem

So ... the TP motors with their higher number of poles are more "sophisticated" (and torquey) and at the same time more finicky when run at or over the edge of their specs!

The verdict: TP motors should be watched carefully when run over 48V due to high phase currents. Most ESCs do not implement the right protection strategies (and die suddenly). Other ESCs which do protect themselves limit duty cycles "smartly".
While the TP motors seem to find lots of supporters, it is crucial to run them with a very good ESC (like the APD).
The other good ESC would then come from MGM, but there the proven LMT motors shine with better phase currents due to the simpler 2 pole construction. (but are less torquey!)

So ... TP it is ... but operated with the right voltage and with the APD HV-pro 16s! (what you save on motor price you put into the more costly ESC).

If you get LMT 3080 e.g. ... you could run it with a MGM controller. (Or even with a drastically erpm limited VESC 75/300, which due to the 2 poles of the LMT motors and reduced switching frequencies would probably be just fine!)

So ... the second best choice might be the LMT 3080 plus Trampa(!) VESC 75/300 (much cooler software and cheaper than the MGMs!) ... but nobody knows the torque characteristics/needed here. In terms of power output: The LMT3080 continuous power output spec seems to fit better for the 14kW requirement to produce 6psi boost on a 3.6!

VESCs all seem to have problems with higher pole inrunner motors and their erpm limitation. (Not fast enough processor!? or no realtime OS?)
I have a MGM ESC comming soon. I use to have a flier water cooling 400A and it don't work animore.. I will post update when I will get to have it and programe it! they proposed to help me with the programming process and optimisation! so keep an eye open 🤟
 
Soooo ... Alex is trying out the cool but expensive APD HV-Pro 16s https://shop.powerdrives.net/?product=hv-pro-400-16 !?
(after he toasted his old flier ... which one?)

I have found another cheap flier opto 400A with water cooling as well:

slightly smaller (300A and up to 14s) but with proper casing and some add ons:

similar to the flier (semi casing) and some add ons:
if you want to check the set up I have or "had"
 
if you want to check the set up I have or "had"

yes of course I am following your project too!
I am impressed with your "fabrication skills" in your early video when you adapted the motor to the turbo ... and I liked your grommet poke stick as well ;)(y)

Regarding the MGM controllers ... they max out at 200000 rpm as well! The MGM web site does not say if erpm or shaft rpm ... but since the LMT motors they recommend are 2 pole (1 pole pair), in this case it would be the same.
The TP motors are 6 pole ... so here applies as well a much smaller upper limit of ~67k shaft rpm ... (but you should be ok, you will not reach that anyway with TP 58xx anyway). Just to show you what league the APD controllers are ...
 
I think that is the lowest ERPM i have ever seen. Id keep away from that POS

Trampa VESC 75/300:

Looks like their FOC control loop is taking up too much processing time ... for 60kHz switching frequencies (to make ~200000 erpm) they use some tricks to skip one cycle via flag setting in their software. So either APD is using similar "tricks" ... or Vedder and community have software issues (which I do not really beleive)

But in terms of powerstage here in excerpt from their forum :
"
18x 75v Direct FETs IRF7759L2TRPBF 1.8mOhm (0.6mOhm per phase)
18x 4.7uF 100V 1210 Ceramic capacitors (5.2uF @ 18S, 6uF @ 16S per phase)
6x 470uF 80V 16x30mm KZN United Chemi-Con Electrolytic capacitors
9x 0.5mOhm 9W Bourns Current Sense Resistors (0.16mOhm per phase)
3x INA240A1 20V/V (+/-495A measurement range)
3x TI Gate driver

I heard Benjamin in the video saying he is using 75v fets? That seems silly when you could easily support higher voltages by just changing to 100v fets. (obviously this falls outside of the low voltage directive in Europe, but it doesn't stop people using 18S and you just label it as a 16S controller)

If I simulate at 16S and 18S and don't let any spikes go above 80V (Since there are 80v electrolytics and INA240 rated is 80v) I get
(18S) 75v 200A
(16S) 67.5V 550A
"
not too shabby ... and the VESC MCUs are STM32F464 ... not shabby too (32bit, up to 180MHz) ... Firmware is running ChibiOS (RT capable).

The VESC project has put a lot of emphasis on vector technology control ... or Field Oriented Control (FOC). Works best/smooth on UAVs and watercrafts trying to get a very smooth acceleration and deceleration with mostly sensored big outrunners!
They have even a motor detect / parameter "auto learn" functionality!
(FOC is where the computing time gets burned!!!)
We are having some small (BIG amps though) simple but 6 pole BLDC motors with relatibely high rpm targets... and APD seems to be appliying a sensorless trapezoidal control which is more tolerant (and less compute heavy) to motor parameters changing while in operation (heat, inductivity, ... until something gives!).
So ... pick your poison: pay for something that just works (APD!) ... or get a geeky VESC and learn how to tweak the software settings right without burning something! 🙊

Not sure I can look away from the VESC though 🤣
 
Last edited:
In addition to ordering the APD HV-Pro 16s.

I did this:

Started talks with techs at Flier. Gave them an idea of what my/our requirements are and told them 400a units just don't seem to hold up to the 400 amp rating under high stress/load conditions. He admitted that the older units didn't meet specs they had intended for them and that it do to some bad choices of parts from a few manufactures that didn't meet specs. He also said they received more than a few emails and even a few calls about the 400a units more than any other unit.

They have since started to make changes and upgrades to some of components to improve and make that unit more stable and reliable.

So;

I gave them my requirements and in what application/manner the unit will be used. He said ok and offer to build me a custom 22s 500 continuous / 800 peak amp unit with watering cooling. They said the new high amp units a very much improvements over the 400a units as they have made great improvements over the past 7 months to the designs.

I asked about rpm rating as i couldn't find and info listed any where. He said on the new 500 and 600 amp units with a 6 pole motor such as the TP5860 750K it is rated 80k rpm max, with a 4 pole unit 120K rpm max and 2 pole units 240K rpm max.

Hmmmm we will see:

He said this unit shouldn't even break a sweat under my conditions. Drawback is these units are big. Nearly twice the size of the 400 units. So large like the MGM just not as smart and at the price point i sure parts quality will be not as good.

I only did this Because I just want to see if we could get a reliable yet a bit Raw, not as smart as the APD and MGM or the more open source units for affordable price. As I did this after one of the guys from facebook message me and told me to contact them directly and tell them what I wanted and about the application. In all honesty the price was still 1/3 of the cost of what I paid for the APD 16s even with express shipping.

We will see how it holds up. I will be opening this unit up and take a look at the parts caps, mosfet etc... To see what parts that can be improved. I do and have done more than few mods to electronics.
 
They have them on their site now, but no specs.

How I do believe R-Mamba on ebay has specs for the now water cool units.
 
I know you guys are aware of this, but I would never go with another Flier ESC. Although Rafael, the R-Mamba brushless guy, is great to deal with. An excellent source for the TP Power motors, too. But I can't help but feel that not only are the MGM/APD controllers more advanced as far as datalogging capabilities (the fliers have none - so you're literally flying blind as to what's going on with your e turbo), I believe they're also more efficient - but without a side-by-side comparision, that's just my belief. I wouldn't push that "500 amp" esc past about 300 amps and 70 volts. Even then I wouldn't trust it. When my second Flier blew up, it actually did some damage to the car - molten solder sprayed all over my shiny powder coated chrome-moly k member and a bunch of holes in the air filter. In other words, the money I saved cost me more money in repairs after the failure.
 
here a 3 D schematic of the APD HV_pro 16s:
BAD ASS ! (12 mosfets per phase!, I did not even start to count the power caps!)

whatever : this thing is worth it!
 

Attachments

  • HV_Pro16S.pdf
    159.4 KB · Views: 5
The only problem is it's off the planet expensive, especially when converted to my local currency ... i mean they are like 1/2 the price of a torque amp ? (less than 1/2 but you get my point)
 
Top