27 Apr
2005
27 Apr
'05
9:50 a.m.
You can build your own rotary phase converter from an old 3 phase motor. And 3 phase
motors can be some of the cheapest to get second hand.
It seems that a three phase motor will not start if a phase is missing, but if it is
running it will continue to run if a phase is removed. The combination of the energized
windings and the spinning rotor will have the effect of creating the third phase, although
this is not an efficient state of affairs.
In my home machine shop I first used a bank of capacitors to help synthesize the third
phase. The bank of WWII tin caps was adjusted until the lathe motor was running with the
least buzz.
Then I added a three phase chipper. The first time I slammed the lathe from forward to
reverse with the chipper I was surprised that the lathe reversed and the strain of
reversing it could be heard in the chipper motor.
Then I added a three phase grinder to the setup. But when the grinder was used alone the
cap bank was the wrong size and the motor had a terrible buzz in it. So I released the
drive belt on the lathe and turned on the lathe motor. Bingo! The buzz in the grinder was
gone!
So in this configuration I am in essence using the caps to start the lathe motor as a
rotary phase converter for the grinder.
Hope this helps,
Mike Gemeny
27 Apr
27 Apr
12:18 p.m.
Mike Gemeny wrote:
You can build your own rotary phase converter from an
old 3 phase motor. And 3 phase motors can be some of the cheapest to get second hand.
It seems that a three phase motor will not start if a phase is missing, but if it is
running it will continue to run if a phase is removed. The combination of the energized
windings and the spinning rotor will have the effect of creating the third phase, although
this is not an efficient state of affairs.
In my home machine shop I first used a bank of capacitors to help synthesize the third
phase. The bank of WWII tin caps was adjusted until the lathe motor was running with the
least buzz.
Then I added a three phase chipper. The first time I slammed the lathe from forward to
reverse with the chipper I was surprised that the lathe reversed and the strain of
reversing it could be heard in the chipper motor.
Then I added a three phase grinder to the setup. But when the grinder was used alone the
cap bank was the wrong size and the motor had a terrible buzz in it. So I released the
drive belt on the lathe and turned on the lathe motor. Bingo! The buzz in the grinder was
gone!
So in this configuration I am in essence using the caps to start the lathe motor as a
rotary phase converter for the grinder.
Note that you can now buy solid state phase converters.
I have a brother that switched to one in his wood-shop, and
it has many advantages: More efficient (cheaper to run);
more motors work with it; more power is available on
created phases; quieter; etc.
It works much better than the rotery ones he used to
use; which were designed as phase converters, not just
some large three pase motors; and his equipment
is much happier, too.
7:36 p.m.
All,
On 4/27/05, Kevin Handy <kth at srv.net> wrote:
[chomp]
Note that you can now buy solid state phase
converters.
I have a brother that switched to one in his wood-shop, and
it has many advantages: More efficient (cheaper to run);
more motors work with it; more power is available on
created phases; quieter; etc.
It works much better than the rotery ones he used to
use; which were designed as phase converters, not just
some large three pase motors; and his equipment
is much happier, too.
All the solid state phase converters I have come across use a three
phase H bridge to generate three phases from a common intermediate DC
rail. The incoming voltage is converted to DC, switched, PWM
synthesising a sinusoidal at the required frequency.
The PWM works pretty well for induction motors. I would expect big
iron to have some issues.
Depending on how the three phases are used for the big iron, it may be
possible for all three phases to be wired the same. That is
electrically connect all three phases together, to a single phase
supply. This can only be done if each phase is used independently of
the others (separate transformers/SMPS). If there is anything in the
machine that wants all three phases - find a nice clean sinusoidal
source.
As a couple of characters of warning, only think about running the
three phase machine from a single phase if you understand 3 phase
systems. Things have the potential to go bang (and probably will) if
you get it wrong.
Simon
--
------------------------------------------------------------------------
"Well, an engineer is not concerned with the truth; that is left to
philosophers and theologians: the prime concern of an engineer is
the utility of the final product."
Lectures on the Electrical Properties of Materials, L.Solymar, D.Walsh
28 Apr
28 Apr
9:46 a.m.
>>>> "Simon" == Simon Fryer
<fryers at gmail.com> writes:
Simon> All, On 4/27/05, Kevin Handy <kth at srv.net> wrote:
Simon> [chomp]
> Note that you can now buy solid state phase
converters.
>
> I have a brother that switched to one in his wood-shop, and it has
> many advantages: More efficient (cheaper to run); more motors work
> with it; more power is available on created phases; quieter; etc.
>
> It works much better than the rotery ones he used to use; which
> were designed as phase converters, not just some large three pase
> motors; and his equipment is much happier, too.
Simon> All the solid state phase converters I have come across use a
Simon> three phase H bridge to generate three phases from a common
Simon> intermediate DC rail. The incoming voltage is converted to DC,
Simon> switched, PWM synthesising a sinusoidal at the required
Simon> frequency.
Simon> The PWM works pretty well for induction motors. I would expect
Simon> big iron to have some issues.
Why? I would expect power supplies to be every bit as tolerand of
distorted waveforms as motors are, even assuming that these converters
produce major distortion.
Simon> Depending on how the three phases are used for the big iron,
Simon> it may be possible for all three phases to be wired the
Simon> same. That is electrically connect all three phases together,
Simon> to a single phase supply. This can only be done if each phase
Simon> is used independently of the others (separate
Simon> transformers/SMPS). If there is anything in the machine that
Simon> wants all three phases - find a nice clean sinusoidal source.
Motors will -- disk drives for example. The larger disk drives (DEC
RP04 class, for example) use three phase motors. They also are picky
about the phase order being right, otherwise they spin on the wrong
direction which is not a Good Thing.
paul
5:33 p.m.
Paul Koning wrote:
The problem will be that some devices work well if they have a large
amount of inductance in their power supplies, such as conventional
power supplies will have.
If they are switcher power supplies, they tend not to react very nicely
to having a solid state inverter feed them.
This is the why, and is only answerable on a case by case basis.
I didn't see a mention of the most obvious case, and that is not 3 phase,
but
that a lot of systems took in 208 three phase and did a rotating
conversion
to 400 hz.
>>>> "Simon" == Simon Fryer <fryers at gmail.com> writes:
Simon> All, On 4/27/05, Kevin Handy <kth at srv.net> wrote:
<chomp chomp>
Simon> The PWM works pretty well for induction motors. I would expect
Simon> big iron to have some issues.
Why? I would expect power supplies to be every bit as tolerand of
distorted waveforms as motors are, even assuming that these converters
produce major distortion.
Simon> Depending on how the three phases are used for the big iron,
Simon> it may be possible for all three phases to be wired the
Simon> same. That is electrically connect all three phases together,
Simon> to a single phase supply. This can only be done if each phase
Simon> is used independently of the others (separate
Simon> transformers/SMPS). If there is anything in the machine that
Simon> wants all three phases - find a nice clean sinusoidal source.
Motors will -- disk drives for example. The larger disk drives (DEC
RP04 class, for example) use three phase motors. They also are picky
about the phase order being right, otherwise they spin on the wrong
direction which is not a Good Thing.
6:42 p.m.
Jim wrote:
If they are switcher power supplies, they tend not to
react very nicely
to having a solid state inverter feed them.
May depend on the switcher. Many of them just rectify the input, so
they don't really care what the input frequency or waveform is, so long
as it's not too fast for the rectifier. So many 50/60 Hz switchers work
fine on DC, 400 Hz, "modified sine wave" 50/60 Hz, or even square wave
50/60 Hz.
But the ones that support exactly two voltage ranges (and nothing in
the middle) apparently use a voltage doubler, which won't work on DC,
and I don't know how it will react to higher frequencies or strange
waveforms.
I'm not a power supply expert, but I've definitely run some switchers
on DC or square wave input without problems. I suppose I'd just
recommend not trying it on a rare, irreplaceable supply.
Eric
8:02 p.m.
>>>> "Eric" == Eric Smith
<eric at brouhaha.com> writes:
Eric> Jim wrote:
> If they are switcher power supplies, they tend not
to react very
> nicely to having a solid state inverter feed them.
Eric> May depend on the switcher. Many of them just rectify the
Eric> input, so they don't really care what the input frequency or
Eric> waveform is, so long as it's not too fast for the rectifier.
Eric> So many 50/60 Hz switchers work fine on DC, 400 Hz, "modified
Eric> sine wave" 50/60 Hz, or even square wave 50/60 Hz.
Eric> But the ones that support exactly two voltage ranges (and
Eric> nothing in the middle) apparently use a voltage doubler, which
Eric> won't work on DC, and I don't know how it will react to higher
Eric> frequencies or strange waveforms.
Doublers won't work with DC, indeed. But they should work fine on
the other waveforms you mentioned.
paul
7:43 p.m.
>>>> "jim" == jim stephens
<jwstephens at msm.umr.edu> writes:
jim> Paul Koning wrote:
> >>>>> "Simon" == Simon
Fryer <fryers at gmail.com> writes:
>
Simon> All, On 4/27/05, Kevin Handy <kth at srv.net> wrote:
> <chomp chomp>
>
Simon> The PWM works pretty well for induction motors. I would expect
Simon> big iron to have some issues.
> Why? I would expect power supplies to be every
bit as tolerand of
> distorted waveforms as motors are, even assuming that these
> converters produce major distortion.
>
jim> The problem will be that some devices work well if they have a
jim> large amount of inductance in their power supplies, such as
jim> conventional power supplies will have.
jim> If they are switcher power supplies, they tend not to react very
jim> nicely to having a solid state inverter feed them.
I don't believe that.
Without PFC, a switcher begins with a full wave rectifier. If you
feed a square wave into that, it will be happier than it was with a
sine wave, not less happy.
With PFC, the analysis requires more knowledge than I have.
paul
5:55 p.m.
All,
On 4/28/05, Paul Koning <pkoning at equallogic.com> wrote:
>>>>> "Simon" == Simon Fryer
<fryers at gmail.com> writes:
[...]
Simon> All the solid state phase converters I have
come across use a
Simon> three phase H bridge to generate three phases from a common
Simon> intermediate DC rail. The incoming voltage is converted to DC,
Simon> switched, PWM synthesising a sinusoidal at the required
Simon> frequency.
Simon> The PWM works pretty well for induction motors. I would expect
Simon> big iron to have some issues.
Why? I would expect power supplies to be every bit as tolerand of
distorted waveforms as motors are, even assuming that these converters
produce major distortion.
Two areas of initial concern.
1) Transformers in linear supplies seem to be very well suited for
50/60Hz. Trying to feed them with a square wave with frequencies of up
to 30KHz + would probably not do them a lot of good. They may be able
to handle it just fine, but then again they might not.
2) Recovery times for the diodes in the rectifier. I initially though
this would be a problem. The datasheet for a 1N4001 specifies
capacitance at frequencies of 1MHz. I guess the 30KHz + of a phase
converter is well within the stride of rectifier diodes.
Simon> Depending on how the three phases are used
for the big iron,
Simon> it may be possible for all three phases to be wired the
Simon> same. That is electrically connect all three phases together,
Simon> to a single phase supply. This can only be done if each phase
Simon> is used independently of the others (separate
Simon> transformers/SMPS). If there is anything in the machine that
Simon> wants all three phases - find a nice clean sinusoidal source.
Motors will -- disk drives for example. The larger disk drives (DEC
RP04 class, for example) use three phase motors. They also are picky
about the phase order being right, otherwise they spin on the wrong
direction which is not a Good Thing.
MMMmmm, out of phase wiring! I have heard of people who keep three
phase extension leads wired out of phase to solve these sort of
problems.
Some machines just take the three phases and split them out to
multiple single phase devices. Kind of like having a local three
single phase distribution boards at the back of the rack.
Simon
--
------------------------------------------------------------------------
"Well, an engineer is not concerned with the truth; that is left to
philosophers and theologians: the prime concern of an engineer is
the utility of the final product."
Lectures on the Electrical Properties of Materials, L.Solymar, D.Walsh
7:40 p.m.
>>>> "Simon" == Simon Fryer
<fryers at gmail.com> writes:
Simon> All, On 4/28/05, Paul Koning <pkoning at equallogic.com> wrote:
>> >>>>
"Simon" == Simon Fryer <fryers at gmail.com> writes:
Simon> [...]
Simon> All the solid state phase converters I have come across use a
Simon> three phase H bridge to generate three phases from a common
Simon> intermediate DC rail. The incoming voltage is converted to DC,
Simon> switched, PWM synthesising a sinusoidal at the required
Simon> frequency.
>
Simon> The PWM works pretty well for
induction motors. I would expect
Simon> big iron to have some issues.
> Why? I would expect power supplies to be every
bit as tolerand of
> distorted waveforms as motors are, even assuming that these
> converters produce major distortion.
Simon> Two areas of initial concern.
Simon> 1) Transformers in linear supplies seem to be very well suited
Simon> for 50/60Hz. Trying to feed them with a square wave with
Simon> frequencies of up to 30KHz + would probably not do them a lot
Simon> of good. They may be able to handle it just fine, but then
Simon> again they might not.
Yes, but do solid state converters put out square waves, or rounded
waves? I'd expect the latter, or certainly nothing with realy fast
rise times. Don't those things ofen use IGBTs? Those don't switch
all that fast as I understand it.
In any case, I don't see why a transformer would be any less happy
about harmonics than a motor; they are both windings on laminated
cores.
Also, square waves have theoretical spectra out to infinity, but the
amount of energy above a few kHz for a 60 Hz fundamental is very low.
I don't have the Fourier transform memorized, so I can't be more
specific...
paul
7024
days inactive
7026
days old
9 comments
6 participants
participants (6)
-
eric＠brouhaha.com -
fryers＠gmail.com -
jwstephens＠msm.umr.edu -
kth＠srv.net -
MGemeny＠pgcps.org -
pkoning＠equallogic.com