Motenergy ME1304 motor with SEVCON size 6 on Dyno

[e-vektor]

66 mH vs 69.9 mH seems ok
Ke=0.021 vs 0.029 has more discrepancy.
I measured the Ke value with DVT (VCI 3, Tool 25.6, Tcl patch level: 8.5.18)

how should I verify Ke?
it is supposed to be the V RMS voltage / rad/s, isn't it? For instance (95V / 1,42) / 5000 RPM / (60 / 2 PI) ? (if 5000 RPM is reached at 100% mod. index)

Motenergy gives Ke=0,021 for ME1304, Ke=0,032 for ME1302, Inductance 66 mH for ME1304, 75 mH for ME1302.
It's data from official motenergy dcf files.
Logic tell to us that with the same magnets ke must be proportional to inductance. My DVT only check inductance of ME1302 and it's matched to 75 mH. If Ke also right, with inductance 66mH of ME1304 Ke must be 0,028.
In the drawings of ME1302 and ME1304 I have found different data Inductance 66 mH for ME1304, and Inductance 140 mH for ME1302.
ME1304 data.JPG
ME1302 data.JPG
Something wrong in this chimistry!
I know only way how to check Ke with oscilloscope.
What version of DVT do You use?

 

[sanfox]

As I understand sevcon's Ke use electrical radian per second speed vs usually uses mechanical angular velocity.

 

[e-vektor]

As I understand sevcon's Ke use electrical radian per second speed vs usually uses mechanical angular velocity.
Ke.JPG

ah, ok. It makes sense if I do the calculation for W rad dividing by 2 (4 poles?) and Veff / 2, it shows about 0.031, but it is approximated
when reviewing old logs the Ke changes across the log, but values are close to 0.029 most time
Last time I measured it it showed 0.0299, but I can measure it again

I think the problem should be more at the % modulation headroom, as it is 85% it stops increasing Veff 15% before of the "maximum", maybe I can try 90 or 95% in case it provides more power in the non-FW zone (up to about 5000 RPM), which is the one I am mainly going to use in the kart

 

[sanfox]

ME1304 has 8 magnets (4 poles). It's possible to try to rise up headroom factor to 95%.
The field wekeneeng must begine to work when headroom factor reach up set value. You can get 10% plus power if you encrease headroom factor.

 

[e-vektor]

ME1304 has 8 magnets (4 poles). It's possible to try to rise up headroom factor to 95%.
The field wekeneeng must begine to work when headroom factor reach up set value. You can get 10% plus power if you encrease headroom factor.

ok, I'll try next Monday. Today I expent almost all day charging the batteries with the "Lipro" chargers (12 x 6S x 16Ah), and tomorrow is a holliday here. But I think it should give some extra power.

 

[sanfox]

Can you make test of rated power of your motor? An example, can you rotate your motor about 15 minutes at half of maximum power?

 

[spinningmagnets]

Speaking of Motenergy, is there a pattern to their names? 1304 = xxxx?

 

[e-vektor]

Can you make test of rated power of your motor? An example, can you rotate your motor about 15 minutes at half of maximum power?

Wow, it is complicated. No enough battery and no coolling yet. At the moment it gets hot in seconds
I will do a log when I test it in the kart. Mainly for temp

 

[sanfox]

Speaking of Motenergy, is there a pattern to their names? 1304 = xxxx?

It's look so, but some motors named with another way like my DLC-28 (ME1302). The production line of Motenergy motors not so big, so I think that thay don't encrease four digits!

 

[sanfox]

Can you make test of rated power of your motor? An example, can you rotate your motor about 15 minutes at half of maximum power?

Wow, it is complicated. No enough battery and no coolling yet. At the moment it gets hot in seconds
I will do a log when I test it in the kart. Mainly for temp

What battery do You plan to use for the kart racing?
I have Toshiba LTO cells for my new battery 9kWh, and after assambling it I will try to test maximum rated power of the motor on the straight road.

 

[e-vektor]

Can you make test of rated power of your motor? An example, can you rotate your motor about 15 minutes at half of maximum power?

Wow, it is complicated. No enough battery and no coolling yet. At the moment it gets hot in seconds
I will do a log when I test it in the kart. Mainly for temp

What battery do You plan to use for the kart racing?
I have Toshiba LTO cells for my new battery 9kWh, and after assambling it I will try to test maximum rated power of the motor on the straight road.

8 x lipo 16ah x 22,2v 10c. Total 100v, 3kwh, max 320 a

 

[spinningmagnets]

Spoke to John at Motenergy, the 1304 number means it was designed in 20"13", and it is the 4th unique configuration of a product that they designed that year.

 

[sanfox]

Spoke to John at Motenergy, the 1304 number means it was designed in 20"13", and it is the 4th unique configuration of a product that they designed that year.

So what? How it's can help for us? Number ME913 mean that in 2009 year John has work more productivetly and made 13 configurations of one motor?

 

[e-vektor]

ME1304 has 8 magnets (4 poles). It's possible to try to rise up headroom factor to 95%.
The field wekeneeng must begine to work when headroom factor reach up set value. You can get 10% plus power if you encrease headroom factor.

ok, I'll try next Monday. Today I expent almost all day charging the batteries with the "Lipro" chargers (12 x 6S x 16Ah), and tomorrow is a holliday here. But I think it should give some extra power.

Finally I did the "last" test, 100V just charged and cool morning (and motor):

I increased the mod. index to 95%, and it gave 39.9 HP at brake, and about 42.5 HP at motor (estimated). I is about 2.5 HP more than the last test, but the torque and HP graphs changed a lot. Actually TQ has the same pattern (first fast decay, then slow), but at least it happens a bit latter (higher rpm). Probably this could be patched at F.W. settings, but by now I am not going to continue as my target for race setup is 28-30 HP. This is only for demo mode.
Note: the motor can reach more than 5000 RPM, but with a "fixed brake" test it starts to be too slow, although I prefer the "fixed brake" for this motor because the resulting graphs are much smoother than with the PID control

Regarding the efficiency, it is important that in the karting race track the gear allows the motor to be always over 1000 or 1500 RPM to have at least 60% efficiency (full torque: 550A, with less torque eff. is better, see the thin graph for 100A)

So the best results would happen when the kart is running between 1500 and 4000 RPM in the corners area, and then it can reach 5000 RPM or more in the finish line (with less torque at the end)

There still are things that could be improved: in all tests battery voltage decreases up to 89V at full power (green graph), so with better batteries and wiring maybe it can have 5V more and have more power. Also the coefficients could be adjusted a little, but I think 41 HP is a good value for a "demo" setup, while for racing setup I probably limit power to 28-30 HP, but with a flat response to keep battery consumption "low"
(Iq is pink: max 550 A)

 

[sanfox]

E-vector, please repost your latest pictures in smaller screen, it's not opening.

 

[e-vektor]

E-vector, please repost your latest pictures in smaller screen, it's not opening.

cannot open those links?

https://www.sportdevices.com/uploads/pics/ME1304_40HP.png

 

[sanfox]

E-vector, please repost your latest pictures in smaller screen, it's not opening.

cannot open those links?

https://www.sportdevices.com/uploads/pics/ME1304_40HP.png

I don't know reason, but I can't open this link!
My friend whant to test ME913 motor performance for his motorcycle and whant to make himself stand. Can you make photo of your stand as example for him?

 

[e-vektor]

and with http? (instead https)

here is a video with the GM (induction) motor:

[youtube]PES5hT-3nF0[/youtube]

it is basically a truck eddy current brake (small size, less than 1000 N*m) with two semi-axles installed and connected to the motor with sprockets and a chain. For measuring torque the system uses a load cell which holds the brake (the brake is floating over bearings)
For electronics I use a SP5 kit (from my company)

The chassis must be very rigid. We had to install a tensioner between the motor and the brake, the chassis bends a little when applying torque

E-vector, please repost your latest pictures in smaller screen, it's not opening.

cannot open those links?

https://www.sportdevices.com/uploads/pics/ME1304_40HP.png

I don't know reason, but I can't open this link!
My friend whant to test ME913 motor performance for his motorcycle and whant to make himself stand. Can you make photo of your stand as example for him?

 

[sanfox]

Very good! I can see all pictures! Thank you!

 

[SevconGuy]

by now the ME1304 has been taken apart to tune another motor in the dyno.
EDIT: my 2 goals with the ME1304 are:
1) get the maximum performance at about 100V, I would expect 30-35 HP
2) get a high effciency setup for about 28-30HP for the karting application


offtopic: I developed a CANopen to OBDII adapter with bluetooth (ELM327 protocol) to be used with Torque App or similar (OSD and Telemetry)

looks very promising, but I like the original Sevcon SpeedApp EV more, which I recently bought from hsct in the dutch.
it also shows how many kilometers or miles you can drive off your battery.
https://photos.app.goo.gl/iFWFLU41ZmKM5np76

 

[e-vektor]

looks very promising, but I like the original Sevcon SpeedApp EV more, which I recently bought from hsct in the dutch.
it also shows how many kilometers or miles you can drive off your battery.
https://photos.app.goo.gl/iFWFLU41ZmKM5np76

It sounds good, do you have more information? involved hardware, etc
anyway my device also acquires data from the BMS (CAN) to send to the mobile phone, but I admit that it a bit heavy (too much BMS channels) for the way that 'Torque' works. It would need a special bars-array to show the whole BMS cells at same time (as in the video from Kolme Faasiline, youtube)

Also I noticed that 'Torque' App has some small bugs in the video mode, which make much more difficult to use it as a serious racetrack-video-recording system (they are mainly related with the storage of gauges scales but the result is more ugly than if the scales are recorded)

 

[frodus]

Concerning the Speedapp, I emailed the hsct company about the hardware they were selling multiple times earlier this year without any response. Nothing was listed on their website at the time. I saw the Speedapp on google play, but it said you needed their dongle. No other information was given, so I gave up and started working with e-vektor.

It looks like they now have more information:
https://www.hsct.nl/winkel/speedapp-ev-for-sevcon/

I think Torque is more flexible for DIY guys who may want to tailor the dash to their specific use. Yes, speedapp looks great, but there are a lot of abilities and skins for Torque that allow some great looking dashboards. It can also send write commands, which could be implemented to send a command to switch drive profile, or adjust regen.

Here's some skins:
https://www.google.com/search?q=torque+android+skin&source=lnms&tbm=isch&sa=X&ved=0ahUKEwiygp7_0JPeAhXFTd8KHQ_aA6sQ_AUIDigB&biw=1920&bih=948

 

[e-vektor]

Offtopic:

Testing an ebike with Cyclone 3000W motor and VESC (50V, 60A limit) on dyno:
comparison of FOC (2.2 KW / 3 HP) vs BLDC (1.7 KW / 2.3 HP). It seems like the battery current limit is applied to BLDC mode (flat section), but not to FOC mode (any ideas?)

any thread to post this?

 

[sanfox]

I have made some modification of DLC-28 (ME-1302).
In stock condition it's can't hold enough load.
Terminals burns inside connection box.
I have made additional oil cooling (silicone oil) like at the scheme.
It's require hermetisation of the stock motor.
After modification this motor can hold more heavy load than stock one.

 

[j bjork]

I am back again to the testing. After my friend luigidesign finished his tests in the dyno he has been adjusting my SEVCON for the ME1304 motor. Efficiency has been improved a lot at low torque areas. It reached about 86% at the brake (after the chain), and I estimated the losses of chain about 6%, so it should be about 92% which is very close to what Motenergy declares (for low power)
The readings were: 9.9 HP and 21 N*m @100A RMS

After that I did a test at 400 A RMS, but I have to admit that I cannot go much further with this configuration (4S x 2P) as the max discharge current is 320A DC

Here you can see 25 HP at brake (26.5 HP estimated without chain losses), 71 N*m and about 78% Eff (84% without chain) which is pretty good. Note that during tests the voltage drops as low as 84V, so there still is some marging if voltage is kept close to 100V and motor current is increased to 660A (for a few seconds)

Now the next step is doing a test with new batteries (100C !) to see the actual max value that this setup can give (95V & 660A target)


And Finally, for the 400A RMS test, here is the Bat Voltage, Bat Current, Motor Temp chart: (read directly from SEVCON with IXXAT)

Very interesting thread :thumb:
I am trying to make sense of all the sevcon settings myself at the moment.
What did you adjust to gain efficiency at low torque?