ANA 787 makes emergency landing due 'battery fire warning'
Variable Frequency Drives (VFD)
Typically, these take a HVDC input (+/- 270 Vdc) and chop it into three phase, pulse width modulated square waves. Motors are happy enough with square waves, so there is no real need to produce clean sine wave power like inverters feeding some avionics loads do.
While the typical aircraft electric motor has been designed to run on 400 Hz, this has been due to the past use of CSDs driving AC generators at that fixed frequency. Now, with DC distribution, that limitation has been removed. It would be possible to optimize motors and their mechanical loads for higher frequencies (and speeds). Or vary their speed to match mechanical load demands.
The other advantage of a VFD is to 'soft start' a large motor load. By modulating the pulse width and frequency, one can get higher torque out of a motor at zero or lower RPM and then ramp up the frequency as the motor accelerates. This also reduces the voltage drop that would be experienced on a common AC bus by all the other loads sharing it with a line operated motor.
While the typical aircraft electric motor has been designed to run on 400 Hz, this has been due to the past use of CSDs driving AC generators at that fixed frequency. Now, with DC distribution, that limitation has been removed. It would be possible to optimize motors and their mechanical loads for higher frequencies (and speeds). Or vary their speed to match mechanical load demands.
The other advantage of a VFD is to 'soft start' a large motor load. By modulating the pulse width and frequency, one can get higher torque out of a motor at zero or lower RPM and then ramp up the frequency as the motor accelerates. This also reduces the voltage drop that would be experienced on a common AC bus by all the other loads sharing it with a line operated motor.
Chris, that diagram is VERY simplified. Eeng has it about right. The CSMCs alter the output to suit demand. EG. Engine starters. Initial output would be low frequency and as the motor speeds up the frequency gets gradually higher to suit. It can also alter the speed of the hydraulic pumps. So in stable cruise the pumps slow down until a control input is made, the CSMC gets a signal to throttle up the pumps and delivers the required output. Very clever but damn complicated. Posted from Pprune.org App for Android
Last edited by TURIN; 24th Jan 2013 at 04:47.
270v Bus
Am familiar with vfd's and they have been used for industrial motor
control for decades. Also pwm, which is a standard technique for
synthesising waveforms from a dc supply.
The question was over your assertion that the 270v dc bus was square
wave, not rectified ac. Usually, individual vfd control electronics
are local to each motor, to provide independent speed and other
control functions. There's no real advantage to have a square wave
on the bus and it would take a lot of added electronics and weight
to generate that from a tru output. Assuming that it's accurate, the
system layout shown in the block diagram is pretty unambiguous.
So, where did the square wave 270v bus come from ?...
control for decades. Also pwm, which is a standard technique for
synthesising waveforms from a dc supply.
The question was over your assertion that the 270v dc bus was square
wave, not rectified ac. Usually, individual vfd control electronics
are local to each motor, to provide independent speed and other
control functions. There's no real advantage to have a square wave
on the bus and it would take a lot of added electronics and weight
to generate that from a tru output. Assuming that it's accurate, the
system layout shown in the block diagram is pretty unambiguous.
So, where did the square wave 270v bus come from ?...
Square wave is how it was explained to us. In truth it will always be rectangular.
And yes, the CSMCs are very, very heavy, large and get very hot if the liquid cooling system fails.
And yes, the CSMCs are very, very heavy, large and get very hot if the liquid cooling system fails.
If the atru is handling 1Mw, then yes, it would need very extensive
cooling. Liquid cooling allows the cooled unit to be compact and
installed in areas that are not otherwise convenient for the heat
disposal.
Not being pedantic, but trying to understand the system. I still don't
have a definate answer on this 270v bus question.
Anyone else have any info on this ?...
cooling. Liquid cooling allows the cooled unit to be compact and
installed in areas that are not otherwise convenient for the heat
disposal.
Not being pedantic, but trying to understand the system. I still don't
have a definate answer on this 270v bus question.
Anyone else have any info on this ?...
Join Date: May 2008
Location: USA
Posts: 44
Likes: 0
Received 0 Likes
on
0 Posts
Back in the dark ages (when I had a brain that worked) I used to design AC-DC-AC power conversion systems for very large wind generation systems which were by their nature variable speed/frequency devices like the gen-sets on jet turbines.
I think the +/- 270 VDC buss is rectified, but un-filtered DC, probably 3 wire, bipolar buss derived from a Wye connected gen-set/rectifier. This would have a pseudo AC appearance on a scope. The cost/weight penalty to condition this high power buss would be huge and is really not necessary for driving large motor loads. PWM VFDs are more than capable of dealing with a "dirty" input and large motors are more than happy to run on a dirty output. A bipolar input would reduce the parts count for the VFDs as well.
I think the 115 VDC buss and the 28 VDC buss are both filtered/conditioned (with 28 VDC being the most so), as they serve smaller, more sensitive loads. The cost/weight penalty would be much smaller for theses busses.
One thing that is interesting to me is that all six AC gen sets appear to be paralleled. Phasing gen sets on separate shafts can be a real pain. It would interesting to see how that is handled.
I think the +/- 270 VDC buss is rectified, but un-filtered DC, probably 3 wire, bipolar buss derived from a Wye connected gen-set/rectifier. This would have a pseudo AC appearance on a scope. The cost/weight penalty to condition this high power buss would be huge and is really not necessary for driving large motor loads. PWM VFDs are more than capable of dealing with a "dirty" input and large motors are more than happy to run on a dirty output. A bipolar input would reduce the parts count for the VFDs as well.
I think the 115 VDC buss and the 28 VDC buss are both filtered/conditioned (with 28 VDC being the most so), as they serve smaller, more sensitive loads. The cost/weight penalty would be much smaller for theses busses.
One thing that is interesting to me is that all six AC gen sets appear to be paralleled. Phasing gen sets on separate shafts can be a real pain. It would interesting to see how that is handled.
Good discussion going--particularly since we have some electrical engineers weighing in.
Back up to the view from FL390 but the NYT put this out this morning as to the interplay between Boeing and the FAA:
http://www.nytimes.com/2013/01/24/bu...M84Yu4eo7N2Cpg
Worth a look.
Back up to the view from FL390 but the NYT put this out this morning as to the interplay between Boeing and the FAA:
http://www.nytimes.com/2013/01/24/bu...M84Yu4eo7N2Cpg
Worth a look.
Chris,
That +/-270vdc is not a single bus supplying the CSMCs. The ATRUs (4 of them) supply various motor controllers.
The output from the motor controller varies depending on user.
I now have access to a better schematic which shows, for example the output to the starter/generator (VFSG) is actually 235vAC. My mistake, I'm not a conehead remember.
However, that AC is still variable frequency.
Hope this helps, I'm still trying to understand it myself.
That +/-270vdc is not a single bus supplying the CSMCs. The ATRUs (4 of them) supply various motor controllers.
The output from the motor controller varies depending on user.
I now have access to a better schematic which shows, for example the output to the starter/generator (VFSG) is actually 235vAC. My mistake, I'm not a conehead remember.
However, that AC is still variable frequency.
Hope this helps, I'm still trying to understand it myself.
areobat:
Thanks for that. From what I remember, 3 phase rectified output looks
like dc with superimposed ripple on top. As you say, it would be fine
for input to a VFD and the bipolar configuration would reduce parts
count elsewhere. Could also cancel any magnetic fields that could affect
things like magnetic compasses, assuming the +/- were tracked together
in the loom.
I guess initially, the shaft speed control must be pretty close. Then
when online, they would tend to lock to each other, just as with the
generators in the national grid. For each shaft, power output is varied
by by fine adjustment of fuel and generator exitation, via some smart
electronics. If you've done this stuff before, correct me if i'm wrong
there.
Wind Generators: With such low blade speeds, do they use step up gearboxes
to get the correct speed for the alternator, viscous couplings etc and how
is it connected to the grid ?. Direct connection as an induction generator,
or via an inverter ?. Off topic, sorry, but just curious...
I think the +/- 270 VDC buss is rectified, but un-filtered DC, probably
3 wire, bipolar buss derived from a Wye connected gen-set/rectifier.
This would have a pseudo AC appearance on a scope. The cost/weight
penalty to condition this high power buss would be huge and is really
not necessary for driving large motor loads. PWM VFDs are more than
capable of dealing with a "dirty" input and large motors are more than
happy to run on a dirty output. A bipolar input would reduce the parts
count for the VFDs as well.
3 wire, bipolar buss derived from a Wye connected gen-set/rectifier.
This would have a pseudo AC appearance on a scope. The cost/weight
penalty to condition this high power buss would be huge and is really
not necessary for driving large motor loads. PWM VFDs are more than
capable of dealing with a "dirty" input and large motors are more than
happy to run on a dirty output. A bipolar input would reduce the parts
count for the VFDs as well.
like dc with superimposed ripple on top. As you say, it would be fine
for input to a VFD and the bipolar configuration would reduce parts
count elsewhere. Could also cancel any magnetic fields that could affect
things like magnetic compasses, assuming the +/- were tracked together
in the loom.
One thing that is interesting to me is that all six AC gen sets appear
to be paralleled. Phasing gen sets on separate shafts can be a real
pain. It would interesting to see how that is handled.
to be paralleled. Phasing gen sets on separate shafts can be a real
pain. It would interesting to see how that is handled.
when online, they would tend to lock to each other, just as with the
generators in the national grid. For each shaft, power output is varied
by by fine adjustment of fuel and generator exitation, via some smart
electronics. If you've done this stuff before, correct me if i'm wrong
there.
Wind Generators: With such low blade speeds, do they use step up gearboxes
to get the correct speed for the alternator, viscous couplings etc and how
is it connected to the grid ?. Direct connection as an induction generator,
or via an inverter ?. Off topic, sorry, but just curious...
Questions
syseng68k
Its rectified AC from the engine generators provided by the ATRUs (see #244). By definition, DC doesn't have a waveform, its just a flat line (ignoring some inevitable ripple, of course).
The question was over your assertion that the 270v dc bus was square wave, not rectified ac.
Join Date: Jun 2009
Location: somewhere
Posts: 451
Likes: 0
Received 0 Likes
on
0 Posts
Large Motor Controller Schematic
One thing that is interesting to me is that all six AC gen sets appear to be paralleled.
For a general understanding:
Last edited by A33Zab; 24th Jan 2013 at 17:27.
Yup each engine gennie supplies a separate bus.
BTBs allow the APU gennies to connect to any bus (almost).
BTBs allow the APU gennies to connect to any bus (almost).
Join Date: Mar 2011
Location: southern England
Age: 66
Posts: 139
Likes: 0
Received 0 Likes
on
0 Posts
Watches use silver oxide batteries. Lithium polymer is a more recent technology LiPO4 is more recent still and arrived probably too late to be used on the 787 programme.
Of all the opinions aired here, I would suggest that a software "upgrade" would be the most likely cause because the affected aircraft are distant in production terms but failed almost simultaneously. software upgrades are probably sent directly to the aircraft without any human intervention so they can all fall out of the sky at once...
Of all the opinions aired here, I would suggest that a software "upgrade" would be the most likely cause because the affected aircraft are distant in production terms but failed almost simultaneously. software upgrades are probably sent directly to the aircraft without any human intervention so they can all fall out of the sky at once...
Usual disclaimers apply!
Join Date: Nov 1999
Location: EGGW
Posts: 843
Likes: 0
Received 0 Likes
on
0 Posts
software upgrades are probably sent directly to the aircraft without any human intervention so they can all fall out of the sky at once...
Join Date: May 2008
Location: USA
Posts: 44
Likes: 0
Received 0 Likes
on
0 Posts
O.T. reply
syseng68k:
Except for some very small wind generators and a few newer designs, they use step up transmissions to match rotor speed to motor speed. At the time I was working with them, induction motor/generators were used almost exclusively because of their mechanical and electrical simplicity. The downside to this design is that the blades are most efficient when the tip/speed ratio is constant, e.g., as the wind speed increases, so must the rotational speed. Since the rotational speed for the simple induction design can vary by only +/- 5% or so (the "slip" frequency) of the line synchronous speed of the induction generator, the blades have to be designed to an "average" condition and therefor lose efficiency at other wind speeds.
The generation systems I was working on used the AC-DC-AC converter/inverter design to allow for for a highly variable blade speed, thus raising the overall efficiency of the system. This topology was superseded, for the most part, by direct electronic control of the induction generator (cycloconverter and other controls) which has permitted variable speed operation. So today, there is a mix of fixed speed induction and variable speed induction generation.
I sure wish IGBTs had been around back then - we did it all with SCRs and had constant problems with commutation failures.
Wind Generators: With such low blade speeds, do they use step up gearboxes to get the correct speed for the alternator, viscous couplings etc and how is it connected to the grid ?. Direct connection as an induction generator,
or via an inverter ?. Off topic, sorry, but just curious...
or via an inverter ?. Off topic, sorry, but just curious...
The generation systems I was working on used the AC-DC-AC converter/inverter design to allow for for a highly variable blade speed, thus raising the overall efficiency of the system. This topology was superseded, for the most part, by direct electronic control of the induction generator (cycloconverter and other controls) which has permitted variable speed operation. So today, there is a mix of fixed speed induction and variable speed induction generation.
I sure wish IGBTs had been around back then - we did it all with SCRs and had constant problems with commutation failures.
Join Date: Mar 2012
Location: West of the rockies...
Posts: 18
Likes: 0
Received 0 Likes
on
0 Posts
NTSB: 787 batteries - thermal runaway, internal short circuit...
Join Date: May 2008
Location: USA
Posts: 44
Likes: 0
Received 0 Likes
on
0 Posts
Should I go chew some one out?
BTW, large, fixed speed wind generators tend to have a low plant capacity factor - I'll bet that it is not running more often than it is running when you pass it.
As it a wild frequency AC it can't posibly be in phase