Prev: Dumbed down consumer electronics: Adding DTV channels
Next: SCR and snubber LTSpice simulation
From: Tim Wescott on 2 Aug 2010 23:51
On 08/02/2010 06:48 PM, Ignoramus18921 wrote:
> On 2010-08-03, Tim Wescott<tim(a)seemywebsite.com> wrote:
>> On 08/02/2010 04:50 PM, Ignoramus18921 wrote:
>>> Lots of new pictures here.
>>>
>>> http://igor.chudov.com/projects/Bridgeport-Series-II-Interact-2-CNC-Mill/24-Troyke-Rotary-Table-U12PNC/
>>>
>>> This is a servo motor driven rotary table Troyke U12PNC.
>>>
>>> I thought that it was a simple servo motor with a tachometer and
>>> encoder on the back. Nothing could be further from the truth. When I
>>> opened it up, the tach was on the back alright, but no encoder.
>>>
>>> I took off a box hanging on the front of this contraption and I think
>>> this is where the encoder is. There are two cylinders.
>>>
>>> One is called "Electro-craft moving coil tach generator". Part
>>> 0100-00-022.
>>>
>>> The other is "Summit engineering, Boseman MT. Model 573-211-10, 2500
>>> Hz, rotor 1 phase, stator 2 ph, spec code H.S.C.T., 11BRW-300-70/10."
>>>
>>> Does anyone have AN idea just what are they and whether I can use it
>>> as a quadrature encoder.
>>>
>>> On the rear of the motor, there is not enoug free hanging shaft to
>>> mount a modern encoder.
>>>
>>> My uneducated guess is that it is a "resolver".
>>
>> My educated guess is that it is a resolver. The frequency is right, the
>> labels on the rotor and stator are right.
>>
>> Do you have a signal generator and an O-scope? Feed it with 2500Hz to
>> the rotor, and see if you get 2500Hz out the two stator windings, with a
>> coupling that depends on the shaft position.
>>
>> There may be industrial resolver to encoder converters out there -- it
>> would be something that a machine designer or retrofitter might need.
>>
>
> There are some converters out there, Jon has one for sale too.
>
> Tim, how much angular accuracy could I get from this resolver, in
> pulses per revolution?

Um -- pretty damn good?

It's hard to say, because it varies with the resolver. But I'd expect
anywhere between one and ten minutes of an arc (no, you don't get
dimensions in degrees, or counts, when you're dealing with resolvers).
Resolvers that do better than that generally go to "multi-speed" units,
with a "high-speed" resolver that repeats itself 8, 16, or 32 times
around the circle, and a "low-speed" (1x) resolver to tell you what
quadrant of the high-speed resolver you're looking at.

> If this is too complicated, I may just look for the right sized modern
> servo motor on ebay.

Aw, what's the fun in that?

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" was written for you.
See details at http://www.wescottdesign.com/actfes/actfes.html
From: Ignoramus18921 on 3 Aug 2010 01:01
On 2010-08-03, Tim Wescott <tim(a)seemywebsite.com> wrote:
> On 08/02/2010 06:48 PM, Ignoramus18921 wrote:
>> On 2010-08-03, Tim Wescott<tim(a)seemywebsite.com> wrote:
>>> On 08/02/2010 04:50 PM, Ignoramus18921 wrote:
>>>> Lots of new pictures here.
>>>>
>>>> http://igor.chudov.com/projects/Bridgeport-Series-II-Interact-2-CNC-Mill/24-Troyke-Rotary-Table-U12PNC/
>>>>
>>>> This is a servo motor driven rotary table Troyke U12PNC.
>>>>
>>>> I thought that it was a simple servo motor with a tachometer and
>>>> encoder on the back. Nothing could be further from the truth. When I
>>>> opened it up, the tach was on the back alright, but no encoder.
>>>>
>>>> I took off a box hanging on the front of this contraption and I think
>>>> this is where the encoder is. There are two cylinders.
>>>>
>>>> One is called "Electro-craft moving coil tach generator". Part
>>>> 0100-00-022.
>>>>
>>>> The other is "Summit engineering, Boseman MT. Model 573-211-10, 2500
>>>> Hz, rotor 1 phase, stator 2 ph, spec code H.S.C.T., 11BRW-300-70/10."
>>>>
>>>> Does anyone have AN idea just what are they and whether I can use it
>>>> as a quadrature encoder.
>>>>
>>>> On the rear of the motor, there is not enoug free hanging shaft to
>>>> mount a modern encoder.
>>>>
>>>> My uneducated guess is that it is a "resolver".
>>>
>>> My educated guess is that it is a resolver. The frequency is right, the
>>> labels on the rotor and stator are right.
>>>
>>> Do you have a signal generator and an O-scope? Feed it with 2500Hz to
>>> the rotor, and see if you get 2500Hz out the two stator windings, with a
>>> coupling that depends on the shaft position.
>>>
>>> There may be industrial resolver to encoder converters out there -- it
>>> would be something that a machine designer or retrofitter might need.
>>>
>>
>> There are some converters out there, Jon has one for sale too.
>>
>> Tim, how much angular accuracy could I get from this resolver, in
>> pulses per revolution?
>
> Um -- pretty damn good?
>
> It's hard to say, because it varies with the resolver. But I'd expect
> anywhere between one and ten minutes of an arc (no, you don't get
> dimensions in degrees, or counts, when you're dealing with resolvers).
> Resolvers that do better than that generally go to "multi-speed" units,
> with a "high-speed" resolver that repeats itself 8, 16, or 32 times
> around the circle, and a "low-speed" (1x) resolver to tell you what
> quadrant of the high-speed resolver you're looking at.

Say, ten minutes of arc, gives you 360*6 = 2160 units of resolution
per circle. Very good in my view. Certainly enough in a geared way
down rotary table.

>> If this is too complicated, I may just look for the right sized modern
>> servo motor on ebay.
>
> Aw, what's the fun in that?
>

I am having enough fun as it goes. I want some uncomplicated
solutions.

I have seen people having too much fun with unfinished machines
sitting for years. The second addiction is being too cheap. $30 servo
drives are not cheap enough for them, so they sit and wait until they
can find $5 servo drives or "make their own" from stuff they find in
dumpsters.

The first approach (too much fun) has some advantages, but
disadvantages are big enough, in my view, to not go there.

The second approach is uneconomical, if you are smart to make your own
servo drives it is better to buy one and make money doing something
more economically valuable.

i
milling out a bigger mold than yesterday and standing in front of EMC
From: Karl Townsend on 3 Aug 2010 03:54

....
> I would bore the hole for the threads, then counter-bore and ream to a
> light press fit of some handy drill rod. Then I'd make a bushing out of
> the handy drill rod and press it (lightly) into the hole. Then I'd say
> "whoa, look! a stepped shaft!" and I'd bolt an encoder disk to it.
>

This is the best route to go. I did this to a 7.5 hp lathe spindle motor.

Karl


From: Lloyd E. Sponenburgh on 3 Aug 2010 08:54
Ignoramus30076 <ignoramus30076(a)NOSPAM.30076.invalid> fired this volley in
news:ltydnfJvi6Ojl8XRnZ2dnUVZ_oadnZ2d(a)giganews.com:

> So, .2495 reamer, 1/4" dowel pin, and what drill should I use for
> initial drilling of the hole?

Unless that's a really short-shafted "pancake" motor, you're going to
play hell setting it up to run true for the initial spotting, unless
you've got enough shaft sticking out to mount the outboard end in a
steady rest.

For that purpose, you should use a short, stiff "spotting drill". A
center drill will work. But if you want the hole not only true in size,
but also perfectly centered in the work, you should bore it to reaming
size, not drill it. (pre-drill, of course...) That's a teensy boring
tool.

Somehow, I envisioned turning a "hub" type shaft extension, and pressing
it ONTO rather than INTO the motor shaft. That would be a lot more
accurate to set up and make, and wouln't involve modifying the motor
itself (along with the concomittant problems of getting a heavy weight on
a relatively long spindle to run true in total overhang.

If you did it the way I envisioned, then you could even mount the whole
affair between centers after joining, to true everything up and finish
the stub shaft to size.

LLoyd


LLoyd
From: Ignoramus30076 on 3 Aug 2010 09:03
On 2010-08-03, Lloyd E. Sponenburgh <lloydspinsidemindspring.com> wrote:
> Ignoramus30076 <ignoramus30076(a)NOSPAM.30076.invalid> fired this volley in
> news:ltydnfJvi6Ojl8XRnZ2dnUVZ_oadnZ2d(a)giganews.com:
>
>> So, .2495 reamer, 1/4" dowel pin, and what drill should I use for
>> initial drilling of the hole?
>
> Unless that's a really short-shafted "pancake" motor, you're going to
> play hell setting it up to run true for the initial spotting, unless
> you've got enough shaft sticking out to mount the outboard end in a
> steady rest.
>
> For that purpose, you should use a short, stiff "spotting drill". A
> center drill will work. But if you want the hole not only true in size,
> but also perfectly centered in the work, you should bore it to reaming
> size, not drill it. (pre-drill, of course...) That's a teensy boring
> tool.

I can have up to 0.01" shaft error.

> Somehow, I envisioned turning a "hub" type shaft extension, and pressing
> it ONTO rather than INTO the motor shaft.

I cannot do it, the shaft sticks out only by 1 mm or so.

> That would be a lot more accurate to set up and make, and wouln't
> involve modifying the motor itself (along with the concomittant
> problems of getting a heavy weight on a relatively long spindle to
> run true in total overhang.

I would do it on a mill, not on a lathe. 0.01" is not space age
accuracy, I think that I can do it.

> If you did it the way I envisioned, then you could even mount the whole
> affair between centers after joining, to true everything up and finish
> the stub shaft to size.
First  |  Prev  |  Next  |  Last
Pages: 1 2 3
Prev: Dumbed down consumer electronics: Adding DTV channels
Next: SCR and snubber LTSpice simulation