DC Motor Mystery

We have a 230V compound wound DC motor with seperate field supply. There are also two current limiting resistors totaling 1.6 ohms in
series with the armature. The field is supplied through a rheostat, the rheostat used as a speed control.
Here is the problem:
We jumped out the two series resistors to apply full voltage to the armature/series windings and dropped the rheostat to minimum. This would give us maximum torque and minimum speed. We started the motor and everthing went as expected - including the motor now running too slowly for the application. Next, we decreased the field voltage by increasing the rheostat - to increase overall speed. We started the motor and it ran BACKWARDS! We scratched our heads for a while - swapped field leads and it still ran BACKWARDS. We swapped the Armature leads - still backwards. We swapped the field leads back to original landings and it was still backwards. We put the series resistors back in the circuit and it started running forwards again.
So basically, we removed the series resistors - the motor ran forward ONE time. Then we decreased the field voltage to increase speed and it ran BACKWARD (no wiring or polarity changes, only field voltage). We put the field voltage back to where it was the first time (when it ran forward) and it still ran backwards. We swapped leads - first field, then armature, then field again - all to no avail. The motor still ran backwards. We put the series resistors back in and it is running forward again.
We are not newbies to this - we work on motor control systems for a living. The senior tech has been working on DC motors for over 40 years and has never seen this.
Any ideas?
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Mighty strange.
On 2/17/06 12:55 PM, in article snipped-for-privacy@z14g2000cwz.googlegroups.com, "slowchaos"

This should increase speed and torque.

That, as you say would lower speed and increase torque.

If the various terminals can be made available during operation, I would take a voltmeter to the field connections and armature connections just to make sure that there is no strange switching or wiring error that reverse polarity in a peculiar way. It is one thing to look a a schematic diagram on a piece of paper and deduce polarities. It is another to make sure that what you think is happening is actually taking place.
Bill
-- Ferme le Bush
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

YEP!
You have a compounded DC motor that is *differentially* compounded. That means the series field winding opposes the shunt field winding. This is often done to improve the speed regulation of the motor so it's speed doesn't vary much from the set speed as load is applied.
The resistors in the armature circuit limited the starting current (duh...). Normally you start this type of motor with full shunt field current (rheostat cut out). So the magnetic field polarity on each of the poles is set by the shunt field, even though the high starting current in the series winding is *opposite* polarity.
By reducing the shunt field some (inserting some rheostat resistance), and removing the armature circuit resistance, you now have the opposite polarity on the main pole pieces (very strong series winding, and weaker shunt winding). This results in reversed starting torque.
Swapping the armature leads doesn't help because it swaps both the current in the rotor and the series winding so direction is still the same. Normally, if the shunt winding were dominant, you would have reversed the rotation this way, but since the series winding is dominant, it has no effect.
You don't mention how long you let it run like this, but I suspect if you let it go very long you would have a problem. As starting current decays, the field strength drops rapidly and if unloaded, the motor will overspeed. Much like a series motor.
Either, change starting circuit to short the rheostat during starting so shunt field has more strength, or put the series resistors back in. Shorting field resistors when starting for maximum torque and minimum current surge is a common controller technique for DC machinery, it may already be doing that. In that case, you've little choice.
Of course, you could also short around these resistors once the motor is up to speed. Are you sure these aren't the intended starting resistors? Most large DC machines need a starting resistor in the armature circuit when starting, but have large contactors that short around them once started (various techniques used: current-decay, simple time-delay, speed switches).
Hope this helps.
daestrom (ex submarine Electrician's Mate Chief, we had a *lot* of DC machinery of all sorts ;-)
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 2/18/06 6:38 AM, in article RjGJf.12703$z% snipped-for-privacy@twister.nyroc.rr.com,

I think that what daestrom is saying is that the combination of the combination of the series winding and the armature is making the motor behave as a series motor. That could be checked by reversing the connection to the series field winding with respect to the armature. Again, you would have the behavior of a series motor but with reversed rotation.
What bothers me about such a situation is that you would be getting a stronger magnetic field from the compounding winding than from the shunt winding. Moreover, starting with the series armature resistors shorted out would cause an enormous inrush current until a back emf gets established. Those resistors are there just to prevent such high currents during startup. During normal startup, I would expect the currents to be low enough so that the shunt field winding establishes the magnetic field and that armature currents would be too low to have the field reverse sign from that established by the field winding.
Again, a voltmeter can be your friend.
Bill -- Ferme le Bush
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

You and Daestrom have it right. There is nothing else that explains this behaviour. Fortunately it appears that there was sufficient load on the motor or sufficient residual magnetism to prevent even stranger behaviour such as rather unstable oscillations.
--

Don Kelly @shawcross.ca
remove the X to answer
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
I believe that you all must be correct on this. I was thinking that it had to be something with the series winding. What had us really stumped was that the first time we started this motor with the resistors shorted it ran forward. Thereafter, it ran in reverse...with no changes made on our part. So let me run this conclusion by you...
We had run the motor in the forward direction previous to shorting out the resistors. The back emf of the armature/field caused the motor to come to a stop in a position that was - shall we say, preferential to forward motion. So when we started this motor the first time with the starting resistors shorted out it still ran forward due to the fact that the dominance of the series field was very small and the position of the windings therefore made it more likely for the motor to turn forward. However, when it stopped it had a tendency to stop in a position where that would favor reverse starting torque the next time due to the slight dominance of the series coil. So, thereafter, it ran in reverse until we put the starting resistors back in and gave the shunt field dominance once more....
Oh, and to answer a couple of the other questions...
-------------snip-------------------------------- You don't mention how long you let it run like this, but I suspect if you let it go very long you would have a problem. As starting current decays, the field strength drops rapidly and if unloaded, the motor will overspeed. Much like a series motor. ----------------------------------------------
We allowed the motor to run for 15 to 30 seconds. The equipment was not designed to run backwards so we didn't dare let it run longer. During this time though, it showed no overspeed or unstable characteristics. It reached a stable speed with no problems.
-------------------------------snip------------------------------------- Fortunately it appears that there was sufficient load on the motor or sufficient residual magnetism to prevent even stranger behaviour such as rather unstable oscillations. ------------------------------------------------------------------------------- Yes, The motor had a load but no where near its full capability.
---------------------------snip-------------------------------------- Moreover, starting with the series armature resistors shorted out would cause an enormous inrush current until a back emf gets established. Those resistors are there just to prevent such high currents during startup. During normal startup, ------------------------------------------------------------------------ Well, yes. But believe me when I say that it is quite common within an industrial setting to have no series resistance inline with the armature - for starting or otherwise.
Thanks to everyone for their input.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 2/18/06 9:20 PM, in article snipped-for-privacy@g47g2000cwa.googlegroups.com, "slowchaos"

This last sentence does not make sense to me. The direction of motion should be determined by the net field direction in the magnetic circuit.

This also makes little sense to me. The only way the motor can run in reverse is because it works like a series motor with little back emf. You do not indicate how high the inrush current went with the starting resistors shorted. I would expect that current to be so high that the armature could be damaged.
If the motor was under light load and started in reverse because of series motor action, I would expect the back emf to drop as the motor sped up. Then, the shunt field will become predominant so as to slow down the motor to where it works like a series motor again. This could lead to a hunting condition. If the load is heavy (high torque), there is the possibility of it continuing to operate as a reverse series motor with excessive armature current.

------------------------------------------------------------------------------> -
-- Ferme le Bush
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
I understand what you are saying but the problem still exist. Why did the motor change direction?
I know that this makes little sense but this IS what happened..
1) Turn on motor - motor runs forward.
2) Stop motor.
3) Short out starting resistors and move shunt rheostat to minimum - maximum shunt field.
4) Start motor and it runs forward - witnessed by many and impossible to get wrong.
5) Stop motor and turn shunt rheostat up to about 45% - to increase speed by decreasing shunt field (normal method of controlling speed for this equipment).
6) Start motor and it runs backwards - again, this is impossible to mistake.
7) Stop motor and move armature back to position in #3 above.
8) Motor still runs backwards.
9) Motor continues to run backwards - no matter what is done (due to diffential winding I guess) until start resistors are put back in.
So, I was just trying to come to some understanding of why the motor ran forward that ONE time. The only thing I could come up with was that the shunt and series windings are very close in value/control over the motor direction. Without the starting resistors to lower the current on the series winding at start up the deciding factor for direction is the position of the windings in relation to each other when power is applied.
I am not an engineer, obviously (LOL), but as a technician I would like to have some understanding of this strange behavior.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

According to your statements, it ran forward when the shunt field was strengthened (moved rheostat to minimum). This raises the strength such that it is stronger than the series field, even during the high starting current.
When you reduced the shunt field current ('turn shunt rehostat up to about 45%'), you weakened it to the point that it was no longer stronger than the differential series field. So now it started like a series motor.
Does that answer your question?
daestrom
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Well, yes and no. I was just trying to understand why this motor ran forward the first time and then backwards thereafter with no difference in the circuitry. No one has replied in a way that gave any hint as to a reason for this.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Again, this is only half correct. It ran forward when the shunt field was strengthened - yes. The we weakend the shunt field and it ran backwards. Then we strengthened the shunt field again - back to where it was to begin with - and it STILL ran backwards. This is the mystery that I am trying to answer.

Not fully.
I really appreciate your patience. I am just trying to get a grasp on this.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
The inrush current in the first case may have driven the field to saturation and the shunt field, opposed by the series field was not sufficient to demagnetise or reverse this. If it had been, you would likely have seen the motor start to slow down, until the series current rose again, then a speed up follwed by a slow down and some continuing but unstable oscillations.
--

Don Kelly @shawcross.ca
remove the X to answer
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Agreed. That's why I suspected that when it was running reverse, they didn't run it for very long. As armature current drops during acceleration, the field weakens a great deal available torque drops off. If it were driving a considerable load, it would not get to full speed in the reverse direction and would draw more than normal currents. If it were unloaded (or the load is such that running in reverse doesn't present much of a load), it could be very unpredictable.
daestrom
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Ah, I didn't see that in your earlier statements. You had said you....

By 'move armature back', I took it you rotated the shaft to some particular position for some reason, not 'returned the rheostat to the zero resistance position'.
At this point, I don't have a 'pat answer' for you. It may be what Don suggested, the pole pieces were magnetized to the point that the shunt field couldn't overcome it. It's clear that the shunt and series fields are precariously balanced when starting this unit. *Something* should be done to 'tip the scales' in favor of the desired direction of rotation more consistently. Either starting resistors, shunt field boosting, or both would seem in order.
Curious, what type of load is this unit driving? Could the load have anything to do with this? If it was rotated backwards for a short time in the first 'reversal start', could it in anyway affected later starting?
daestrom
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

This motor is driving a metal plate belt at the bottom of a hopper. The operation generally unloads coal from ships with a 30 ton crane. The crane dumps its load into the hopper and the hopper belt moves the coal onto the conveyor. They are currently in the process of speeding up all the conveyors and the hopper - along with other mods - to get max capacity. The senior tech wanted to remove the series resistors to put max voltage on the armature windings for max speed but then we had this problem.

I'm not sure what you mean here. I think the first forward run and then the reversal, after removing the series starting resistors, must have to do with the magnetization of windings. I think that this is similar to what I was proposing earlier - but relies on amount of magnetization rather than physical position of the rotor at start. If the shunt and series windings are closely balanced without the series starting resistors in place then the direction of rotation could hang in the balance as well?
Anyway, thanks for all your input on this. I believe I have my general answer to this problem. If you want to know more, feel free to ask.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Well, I wasn't sure what type of load you were driving. Take for example a reciprocating compressor. Once it is run for a few minutes, there may be residual pressure in the cylinders that would resist starting again in the same direction (barring some form of cylinder unloaders). Or a loaded conveyor belt running up-hill. Depending on the gear drive, when an electric brake is released, the weight of the load would tend to reverse the system and apply a torque that would tend to reverse the motor. But I see by your description that this is probably not at all the case here.

Yes, I would tend to agree. Apparently, the series winding without resistors, and the full current shunt field (with rheostat at minimum resistance) are very closely balanced. So only residual magnetism is left. Frankly, this is a terrible way to start a DC motor. It will draw a lot of current, and generate very little torque until the current decays enough for the shunt winding to establish a magnetic field.
Another 'trick' to get better starting with differentially compounded DC motors can be used if the series and armature connections are both taken out to the motor starter. During starting, have a contactor short around the series winding so that the motor starts as a simple shunt motor. Then when it's up to speed, open the contactor to put the series winding back into the circuit. This gives the motor much higher starting torque.
If you want to look into it further, find what your shunt field current is, and the winding data for the motor. The number of turns of the shunt field times the amps gives you the MMF (magneto-motive force, in ampere-turns) that it generates. Then find the number of turns in the series field. Divide the shunt field's MMF by the series field number of turns, and that gives you the number of armature amps (starting current) where the two fields will cancel out. I'll bet a beer that this current is more than the amount you get by taking the line voltage divided by the starting resistors (i.e. the resistors were sized to prevent the series field's MMF from cancelling the shunt field's MMF during starting).
Something we haven't talked about yet is whether this motor has inter-poles. They can sometimes be confused with a series field winding since they also carry full armature current. But they have a completely different function and can't really generate any torque on the rotor. It is important to keep the magnetic polarity of these correct for the direction of rotation, as reversing their magnetic polarity can cause severe arcing at the brushes when loaded (when reversing the rotation of a machine with inter-poles, you can *not* do it by simply reversing the shunt field connections). You didn't mention if your machine has them, but I thought I'd ask since it is common for *large* DC machines to have them.
Anyway, glad I could help, good luck.
daestrom
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Not sure where you got that idea. Universally, every DC motor I've seen larger than a kw (and I've seen thousands) has some form of starting controller. And in that controller is either simple resistor bank, or solid-state feed for the armature. If it has a resistor bank, it is only in circuit during starting and cannot handle continuous duty. Also applying full shunt field current during starting is very common. Propulsion, winches/hoists, drags, you name it.
One rare situation I've seen is a resistor in series with the armature and parallel with an auxilary field winding. So the voltage drop in the resistor is used to power the auxilary field winding. Gives you the same effect as a full series winding, but is more easily adjusted and uses less copper in the winding. Of course, the resistor stays in the circuit all the time, and that means some other issues.
daestrom
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Well, about the reistors being commom - I'm sure you are right. I work with drives usually and the DC drive acts as the starting controller. I guess I'm just not use to working on DC motors without a drive - and I haven't seen thousands....
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Well, about the reistors being commom - I'm sure you are right. I work with drives usually and the DC drive acts as the starting controller. I
guess I'm just not use to working on DC motors without a drive - and I haven't seen thousands....

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
----------------------------

Actually a set up similar to what you mention is (or was) not uncommon. The resistor acted as a shunt to decrease the current in the series winding and essentially weaken the series field. This allowed some control of the level of compounding (slope of the speed torque curve). Losses with this resistor in service would be less than with the series winding only. Full series compounding would only occur with the resistor open circuited and the series winding would be a normal series winding.
Use of a higher resistance auxiliary winding with any shunt resistor would be, I should think, counterproductive. However, people have tried many combinations and then went on to design(?) wiring for VW vans.
--

Don Kelly @shawcross.ca
remove the X to answer
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.