Hi Guys,

I am losing a lot of hair on this one.

I have two PCBs that both have MAX232 chips. One is a cpe the other
has an N suffix. That part really doesn't matter as I have swapped
the chips back and forth (they are in sockets) a number of times and
the problem doesn't follow.

Both PCBs have 10 uf caps. One boad has caps from Radio shack - it
works. The other board has caps from Jameco or Digikey, not sure.
Either way, they ARE marked 10uf 50V.

Both circuits are very close to being the same. The difference
being the caps on the working board are about 1/2" further away from
the max 232.

I have swapped out the caps on the non-working board - still didn't
fix it.

I have done a pin-by-pin comparison about 5 or 6 times now and I
can't find any differnce in the two circuits.

I tried to connect my scope up to do some signal comparisons, I have
a hunch my scope isn't working properly at all - don't want to try
and figure it out at this time - don't know how for that matter :-(

The only thing I can think of is that the caps on the non-working
board are of some other type. I don't have any others on hand so I
can't even do any more swapping.

As to what is not working, I am getting nothing out of the max232.
The Tx pin is at -9.67 volts and seems to vary .01 volts during
transmission(I am using a digital meter and can't tell for sure).

The Rx pin is at 0 volts.

From the AVRs RX pin as measured at the MAX232 I get 5 Volts steady.

From the AVRs TX pin, as measured at the MAX232 I get 5 volts, but
can see the voltage dropping during transmision - drops about .3
volts as best I can tell with the meter.

To make sure the problem isn't on the PCB somehow affecting the AVR,
I used jumper wires from the MAX232 on the non-working board, fed
the Tx / Rx from the AVR to the working board and communications is
good.

I am stumped - but then again, this is to be expected as I don't
know how to check anything else.

Any ideas? Any chance I have the wrong type of CAPS?

Chris

开云体育

In a message dated 5/4/2006 1:48:19 P.M. Central Daylight Time, lcdpublishing@... writes:

Any ideas?? Any chance I have the wrong type of CAPS?

Chris:? I have NO business responding to this inquiry as I have NO idea what your circuit is or looks like, BUT I would assume if it's the caps, the problem would "move" with the caps.? Have you tried TESTING them?? Set your DVM on the LO-ohm range and test a known-good cap with the probes ONE way, then SWAPPED, and note the changes in the reading.? Then see if the changes are similar for the suspect cap.? Better:? Use a cap. tester!? A SHORTED cap, or OPEN cap would probably be?causing your problem!?????????JRR

On Thu, 04 May 2006 15:04:28 +0200, Peter Homann <groups@...> wrote:

Vc = Vc + Kpd*E

Vc = Kpd*E

To me this seems like a PI vs. a P control.
I don't quite see what it has to do with velocity vs. position?
I would say a PID can be used for RPM control as well, what is wrong with integrating the error out to zero here?

ST

Hi Chris,

Since you are wanting to control the speed of the motor, you want to implement a velocity control loop, not position control loop. Therefore you don't have any I term, just P and D.

Thinking music . . . . . .

The difference is that in a velocity control system you are adding or subtracting the error signal to the currently set control signal.

Vc = Vc + Kpd*E

In a position control system the control signal is set to the value of error signal.

Vc = Kpd*E


A lot of the PID references are concerned with position loops and don't mention the difference in a velocity control loop.


Don't confuse the two. Also, you can make a position control loop system look like a velocity control loop system by constantly changing the desired position. This is how the Mach3 spindle step/dir signal is used with say a Gecko driving the spindle motor.

Cheers,

Peter.

lcdpublishing wrote:

--
------------------------------------------------------------------
Web: www.homanndesigns.com
email: homann@...
Phone: +61 421 601 665
www.homanndesigns.com/ModIO.html - Modbus Interface Unit
www.homanndesigns.com/DigiSpeedDeal.html - DC Spindle control
www.homanndesigns.com/TurboTaig.html - Taig Mill Upgrade board

Richard,

Yes, the net contains a ton of information on PID. I found a couple
of good pages on the subject - one I liked in particular gave good
examples of PID and PI etc. Like dummy, I didn't bookmark it so I
have to find it again.

I am looking over that Atmel app note and that looks VERY
interesting. I don't know if I would use that discrete for this
project (as I need to learn a bit more about it) but when I start
experimenting with servos, that looks like the ticket!

Chris

PID control of motors is a favorite topic for Google. I just did a
search on 'pid control of dc motor' (without the quotes) and came

up

with a bunch of relevant hits.

There is the Atmel app note AVR221 'Discrete PID

controller'.
pdf

The source code is available and is suited to all AVRs according to
Atmel

It is writen in C but the DOXYGEN documentation is excellent.

I would start with this and see where it goes.

Richard

On Wednesday 03 May 2006 06:07 pm, Chuck Merja wrote:

A while back this list discussed a benchtop multiple voltage supply. I am
a farm kid who likes gadgets and am frustrated by all the individual
voltage adapters I need to plug into the cig lighter, so I am looking for
something to probably hardwire into a rig that will supply these various
voltages.

Right now I use PDA (5v 2A), Panasonic Laptop CF27 (10.8 V at 1? Amp) and
GPS 12 v at an Amp. I would also like to have a small diode protected
backup battery, so that when I start the engine the GPS won't drop out.
Both the laptop and PDA have bats in them, so I'm not too worried about
them. I've used several separate devices, but none is very satisfactory,
and the 5V deal is now supplied by 12v to 110v inverter, then through a
wall wart back down to 5V - well, it works and better than the variable
voltage converters that I've let the smoke out of for one reason or
another.

I see several voltage converters, but nothing that fits the whole bill, by
itself.

Anyway, I guess I could build something, but I'm really a mechanical guy
not an electronics guy - but am kinda an electronics wannabe. Any
suggestions - volunteers?

For the 5V supply that's easy, you can do it with a single part more or less,
but for 2A the common 7805 won't cut it -- you'll need a bit more oomph. Try
the LM323 or LM350, both of which are in TO-3 cases, and which should
probably be mounted on a heatsink, since dropping 7V at 2A means it's going
to be dissipating 14W, more if the charging system on the vehicle pushes the
voltage up a bit.

For the 10.8V, I'd suggest the LM317, which is an adjustable regulator that
you set up with a couple of resistors to set your output voltage. This is
also available in the TO-3 case though for only around an amp or so you can
probably get away with the TO-220 case version, but I'd still use a
heatsink, since it's dissipating a couple of watts.

The 12V? That depends, you can perhaps use the incoming DC power as is from
the vehicle with some filtering and conditining, sometimes the charging
voltage of a vehicle will push that upwards some and you may run into the
limits of some of the components. That could get a little bit tricky, but
maybe a zener diode and a series resistor would take care of it.

With the regulator chips, you will want some capacitors at the input and
output, and with the LM317 you need to keep the case of the device isolated
from the vehicle ground, but that't not hard to do. For details of the use
of these parts the datasheets are usually pretty good with all sorts of
examples, and I'd recommend that you search them out and see what all is in
them. The ones from National Semiconductor are pretty good, if you have a
choice of brands.

If you get stuck on any of this, feel free to post in here and one or another
of us will have something to say on the subject, I'm sure. :-)

--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, a critter that can
be killed but can't be tamed. --Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin

--- In Electronics_101@..., "lcdpublishing"
<lcdpublishing@...> wrote:

After pondering the thought about current sensing, the only thing I
could come up with is that I could use that to detect if the motor
is nearing stall. Whereas with only the tachometer feedback, I
wouldn't know it stalled till the RPMs kept dropping when the PWM
was at 100%.

Not sure yet how I will sense the current though. A few months ago
I started a discussion here about that using resistors and such. So
I need to go backwards in time a bit and search the archives to see
what everyone told me and re-apply it to this new idea.

All in all, I don't really think this is at all practical for this
machine - sort of overkill, but I really hope to learn a lot from
it. Then again it could all backfire and confuse the hell out of
me ;----)

Chris

PID control of motors is a favorite topic for Google. I just did a
search on 'pid control of dc motor' (without the quotes) and came up
with a bunch of relevant hits.

There is the Atmel app note AVR221 'Discrete PID
controller'

The source code is available and is suited to all AVRs according to
Atmel

It is writen in C but the DOXYGEN documentation is excellent.

I would start with this and see where it goes.

Richard

开云体育

After pondering the thought about current sensing, the only thing I
could come up with is that I could use that to detect if the motor
is nearing stall. Whereas with only the tachometer feedback, I
wouldn't know it stalled till the RPMs kept dropping when the PWM
was at 100%.

Not sure yet how I will sense the current though. A few months ago
I started a discussion here about that using resistors and such. So
I need to go backwards in time a bit and search the archives to see
what everyone told me and re-apply it to this new idea.

All in all, I don't really think this is at all practical for this
machine - sort of overkill, but I really hope to learn a lot from
it. Then again it could all backfire and confuse the hell out of
me ;----)

Chris


--- In Electronics_101@..., "Stefan Trethan"
<stefan_trethan@...> wrote:

It depends on what you want to keep constant. If you want constant

RPM the

current does not really give you any valuable information for this

loop.

It would give you information about the torque though.

Some small grinders, like those made by proxxon that have

adjustable

speed, have a current feedback speed control. Basically they

notice a

higher current means more load so it will slow down so they apply

more

voltage to keep the speed up. This is done because it is much

cheaper to

sense current than to sense RPM, but you can get a better system

if you

really measure RPM.

One of my power drills has a magnetic sensor on the shaft, that is

really

a good one, stable RPM and lotsa torque even if you set it to the

slowest

speed and load it badly like for driving in screws. If you try

that on a

cheap drill without the sensor feedback you will get far too many

RPM with

no load and it will stall if loaded.

ST

Hell Roy, if it involves electronics, I know it ain't gonna be
simple!

Although there isn't a whole lot of magic as to "what to do if this
occurs".

With only about 10 minutes of thought into it so far, this is what I
am thinking...

I tell the controller what RPM to run at and start the motor

Based on Pre-set parameters, ramp up PWM to desired speed based on
parameters.

After accel time, check for current speed
If too slow - speed up
If too fast - slow down
(don't do this too much or the hunt will never end!)
If stalled - cause a fault - can't cut if it ain't spinin!

Factoring in the load based on current draw is a secondary aspect of
what I was thinking about. Not sure exactly what I would do with
the feedback of the current, but I am sure I could make the machine
a bit smarter with that information.

Won't know much of anything till I actually make the spindle, get
some gears, make a motor driver for the little feller, and give it a
go. Only way I know how to learn about something - try it and if it
don't work - scream help :-)

Chris

It's not quite as simple as that, though. What do you do when

the motor sees

a bit more of a load for some reason (cutting forces?), and slows

down a bit

-- push harder? How about if things jam, and it stops totally,

then what?

There's probably more to consider here than what I'm thinking of

at the

moment.

--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, a critter that

can

be killed but can't be tamed. --Robert A. Heinlein, "The Puppet

Masters"

-
Information is more dangerous than cannon to a society ruled by

lies. --James

M Dakin

On Wed, 03 May 2006 21:30:33 +0200, Roy J. Tellason <rtellason@...> wrote:

It's not quite as simple as that, though. What do you do when the motor sees

a bit more of a load for some reason (cutting forces?), and slows down a bit

-- push harder? How about if things jam, and it stops totally, then what?


There's probably more to consider here than what I'm thinking of at the

moment.

But that's basically what you do. If you still don't get the RPM up with pushing lotsa power in you either stop and shut down right there or you rely on the current limit to prevent the motor making smoke signals, or both.

You can calculate the motor losses from RPM and voltage, so you can theoretically set a limit to these losses that may not be exceeded.

ST

It depends on what you want to keep constant. If you want constant RPM the current does not really give you any valuable information for this loop. It would give you information about the torque though.

Some small grinders, like those made by proxxon that have adjustable speed, have a current feedback speed control. Basically they notice a higher current means more load so it will slow down so they apply more voltage to keep the speed up. This is done because it is much cheaper to sense current than to sense RPM, but you can get a better system if you really measure RPM.

One of my power drills has a magnetic sensor on the shaft, that is really a good one, stable RPM and lotsa torque even if you set it to the slowest speed and load it badly like for driving in screws. If you try that on a cheap drill without the sensor feedback you will get far too many RPM with no load and it will stall if loaded.

ST

On Wednesday 03 May 2006 02:06 pm, lcdpublishing wrote:

Hi Guys,

For this little machine I will be going with a variable speed
spindle driven by a small DC motor. I will be using an AVR to
control it's speed via PWM. I had planed on putting a photo
interrupter and wheel on the spindle somewhere so I could have a
digital tachometer. My original plan was to simply feed the speed
back into the control software and display the speed.

However, being all caught up in all this nonesense of electronics
these days, I got to thinking.....

If I take the tachometer signal, and put that into the AVR creating
the PWM signal, I could then have a pretty accurate speed controller
that actually runs at the speed I tell it to.

I believe that this would all be a mater of twiddling bits
(programming) and no additional electronics at all other than what I
mentioned above.

Am I on track here for a closed loop speed controller?

It's not quite as simple as that, though. What do you do when the motor sees
a bit more of a load for some reason (cutting forces?), and slows down a bit
-- push harder? How about if things jam, and it stops totally, then what?

There's probably more to consider here than what I'm thinking of at the
moment.

--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, a critter that can
be killed but can't be tamed. --Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin

On Wednesday 03 May 2006 08:45 am, Barry Savage wrote:

You might find the Dremel drill press adapter at micro-mark.

Somebody pointed me to Lowe's as having one of those a while back. I haven't
gotten around to getting it yet, though, and I probably should. :-)

--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, a critter that can
be killed but can't be tamed. --Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin

Ah but this is a great chance to learn a lot for me too!

While at first glance, you wouldn't think a PCB dril would cause a
load on a spindle motor, however, when that motor is little, the
load is relative to the motor, a small drill and a small motor
should make for a good experiment!

In reality, other than for knowing if the motor actually stopped
turning (Stalls), it isn't needed at all for the little machine.
But for working out many of the details of that process, this very
well could be a great test bed.

Hmmm, thinking about this a bit more, what if I also bring into the
loop the current draw of the motor? If the current is near peak,
then I know the motor is actually stalling.

Having only stepper motors to compare against, I don't know about
this at all. When using PWM for motor speed control, the voltage is
constant, but the current is switched on and off. Would I still be
able to use a current sensor in such a circuit like this?

Hmm, this is getting more and more interesting by the minute!
Balancing the feedback of current and speed against what the motor
should be doing - I like!

Chris



--- In Electronics_101@..., "Stefan Trethan"
<stefan_trethan@...> wrote:

A process (motor), with control input (PWM) and feedback output

(opto

tach) and a controller hooked up to the two.
Well, yes!
With the right bit of software that is.

Feedback control can greatly increase the usefulness of a tool. In

the

case of the spindle it would be useful where you have relatively

high

loads on the spindle that would like to slow it down and where you

would

like relatively constant RPM. For PCB drilling the loads would

probably be

low, and the tolerance to RPM very high, so i would not see it as

a

requirement for that. But if you are going to mill plastics that

may smear

or something it could be very useful.

ST

A process (motor), with control input (PWM) and feedback output (opto tach) and a controller hooked up to the two.
Well, yes!
With the right bit of software that is.

Feedback control can greatly increase the usefulness of a tool. In the case of the spindle it would be useful where you have relatively high loads on the spindle that would like to slow it down and where you would like relatively constant RPM. For PCB drilling the loads would probably be low, and the tolerance to RPM very high, so i would not see it as a requirement for that. But if you are going to mill plastics that may smear or something it could be very useful.

ST

Hi Guys,

For this little machine I will be going with a variable speed
spindle driven by a small DC motor. I will be using an AVR to
control it's speed via PWM. I had planed on putting a photo
interrupter and wheel on the spindle somewhere so I could have a
digital tachometer. My original plan was to simply feed the speed
back into the control software and display the speed.

However, being all caught up in all this nonesense of electronics
these days, I got to thinking.....

If I take the tachometer signal, and put that into the AVR creating
the PWM signal, I could then have a pretty accurate speed controller
that actually runs at the speed I tell it to.

I believe that this would all be a mater of twiddling bits
(programming) and no additional electronics at all other than what I
mentioned above.

Am I on track here for a closed loop speed controller?

Thanks

Chris

开云体育

You might find the Dremel drill press adapter at micro-mark. They used to carry it, but a quick search of their site just now didn't turn up anything. That doesn't mean it isn't there though, the site is large, and I was in a hurry.. They have a lot of stuff that builders in electronics might find useful. Here's the link:



Barry

ctall31724 wrote: