F1technical.net

It's been my understanding that the MGUs are 3 phase, i.e. 3 wires with phases 120 degrees apart, no neutral.
As described here:
http://thewptformula.com/2014/03/26/ana ... tem-mgu-k/

No - it will be a three phase motor. 1 wire per phase - no neutral or earth.

gruntguru wrote:No - it will be a three phase motor. 1 wire per phase - no neutral or earth.

So what are they doing for the return path?

The MGU motors are AC 3 phase. I'm pretty sure in harvesting mode the power goes through the 3 cables through a series of bipolar transistors and capacitors and converted to DC then to the batteries. The process reversed to power the motors. I suspect there is tremendous amount of heat generated during this AC to DC to AC conversion process and where many failures occur.

djos wrote:So what are they doing for the return path?

In an engine phases ale balanced so there is no return current in neutral.
https://en.wikipedia.org/wiki/Three-pha ... tric_power

Higher voltages result always in better over all efficiency. Higher current means always higher losses, being it induction losses, resistive losses, whatever. Higher voltages and constant cable cross-section results in less losses, and with a given loss reduced copper weight.
Limits are the semiconductor voltage threshold and insulation materials, but 700 V is not anything high.

MGU-K (120kW) at 690V means about 100A current...
At 400V AC, it's already 170A.

Big cable car motors (in Europe) run often on 690V AC (3 phase) instead of 400V AC, this results in 400V per phase towards "ground" and around 565V peak, i.e. the battery pack (DC) would be on a voltage of around 570 V DC.

piast9 wrote:

djos wrote:So what are they doing for the return path?

In an engine phases ale balanced so there is no return current in neutral.
https://en.wikipedia.org/wiki/Three-pha ... tric_power

Cheers for that, I only studied basic electronics and never looked at AC power in depth.

Wazari wrote:The MGU motors are AC 3 phase. I'm pretty sure in harvesting mode the power goes through the 3 cables through a series of bipolar transistors and capacitors and converted to DC then to the batteries. The process reversed to power the motors. I suspect there is tremendous amount of heat generated during this AC to DC to AC conversion process and where many failures occur.

IGBT's are often used, they are a mix of Mosfet drive and bipolar transistor power switch.
The losses are on one side the diode junction losses or the on-resistance losses (in case of MOSFET's), and the losses during switching, as switching cannot be performed with infinite speed.
The diode junction losses or on-resistance losses increase linearily with current. And yes, they are relevant, not just some 100 Watts in cases of 100+ kW motors.

Wazari wrote:...bipolar transistors and capacitors... ...I suspect there is tremendous amount of heat generated during this AC to DC to AC conversion process and where many failures occur.

Actually it will be more like FET, SiC transistors (you can go up to 1.2kV) also the heat generation depends on what topology you chose/frequency and size of the power components. There are topologies which are quite effective but they need bigger or more components and more measurement/control but you can go up to 96% of efficiency maybe even more (98%) if you make the setup work only in narrow operating window (voltage range).

Abarth - IGBT are usable for high voltage but they are bit slow... I mean we use them but not always... In F1 I think they go for SiC as they are the best you can get right now...

Abarth wrote:Higher voltages result always in better over all efficiency. Higher current means always higher losses, being it induction losses, resistive losses, whatever. Higher voltages and constant cable cross-section results in less losses, and with a given loss reduced copper weight.
Limits are the semiconductor voltage threshold and insulation materials, but 700 V is not anything high.

MGU-K (120kW) at 690V means about 100A current...
At 400V AC, it's already 170A.

Big cable car motors (in Europe) run often on 690V AC (3 phase) instead of 400V AC, this results in 400V per phase towards "ground" and around 565V peak, i.e. the battery pack (DC) would be on a voltage of around 570 V DC.

Higher voltage require thicker insulation of the copper wires which means less density of the winding.
Less density of the winding, means bigger motors with less effiency. It must be a certain point which is the best compromise.
690V, as you say, seems to be standard figures. Also, when you rectify 3-phase 400VAC, you got 690VDC which you then chop to your frequency that you want.
I work with 3phase synchrounus motors in the automation industry so I have some experiance with this kind of controlling a motor.
It works quite good in both way also, a 3phase synchrounous Ac-motor is a perfect generator also which is excatly what the MGU-H and MGU-K work as.

I think we have confirmed the voltage/current paths are AC-DC-AC via an inverter and with very little cutting edge technology, After all AC synchronous motors have 96% efficient for 10 years and inverter drives are around 98%, so there's very little to gained. I would like to know more about the batteries? I'm guessing at graphene nano lithium polymer 3.7 - 4.2 volt cells about 140 cells with a C rating of 240. Can anyone shed some light on this, in particular the C - rating

Doesn't they use super-capacitors instead of batteries.

Mr.G wrote:Doesn't they use super-capacitors instead of batteries.

I thought super capacitors were ten times the size of equivalent batts and although would solve charging cycle discharging problems, would create a packaging problem?

Mr.G wrote:
Actually it will be more like FET, SiC transistors (you can go up to 1.2kV) also the heat generation depends on what topology you chose/frequency and size of the power components. There are topologies which are quite effective but they need bigger or more components and more measurement/control but you can go up to 96% of efficiency maybe even more (98%) if you make the setup work only in narrow operating window (voltage range).

Abarth - IGBT are usable for high voltage but they are bit slow... I mean we use them but not always... In F1 I think they go for SiC as they are the best you can get right now...

I'm pretty sure they are using IGBT in F1 to switch the current.

If so, they have space for improvement... SiC are better...