Brushed DC servo question

[Black Forest]

Torque is current - rpm is voltage...

Danke!

[polaraligned]

Not quite: the capacitor will charge to 1.41 times the AC RMS value only if no current is drawn. As soon as current is drawn, the (DC) voltage drops & the ripple starts. How much ripple depends upon the load current and the capacitor size. Also if the load is disconnected, the capacitor's voltage will rise towards its peak until the load is reconnected and the voltage drops again.

I know well how it works. 1.41 is a reasonable estimate of the DC output of a well filtered supply. My response was to you saying that the DC would be down around the RMS of the AC, which it is not.

[MaxHeadRoom]

What actually is dictating what speed the DC servo is running? Is it the amount of volts being applied by the drive to the motor? Or something else?

The typical servo drive uses PWM drive with PID feedback or similar control of the servo motor to maintain the programmed motion, the motor is subjected to a mean voltage level throughout the control range.
Even your typical Treadmill uses either 90v DC or 180DC motors on 120vac or 230vac rectified supply for either.
The control limit is a subject of the drive.control.

.

[darryl]

Just to add some unwanted confusion to the mix- the act of rectifying ac and smoothing it with a filter capacitor is a mixture of high currents flowing for part of the time of the ac cycle, and low or no current flowing for the remainder of the cycle. Take the typical arrangement of a full wave bridge rectifier, filter capacitor, and ac supply voltage. When you first power it up, the voltage on the capacitor rises until it reaches the peak value of the ac voltage, minus rectifier losses and resistive losses. With no load, the peak voltage value remains on the capacitor, and the current drawn from the ac line drops to zero. As soon as you apply a load, the voltage on the cap starts to drop and will continue to drop until the voltage on the ac cycle rises to that level- at which point it passes current to bring the cap voltage back up to the peak value. During this part of the cycle, the current flowing to recharge the cap can be (and must be) much higher than that drawn by the load. Ignoring resistive losses and rectifier losses, this current could be very high indeed. The higher the capacitor value, the higher it will be- and the higher will be the average dc voltage on the cap. The outcome of this, and the point I'm making, is that the ac input WILL supply more current than is drawn from the output, and DOES result in being able to draw more power from the output when the capacitor is included in the circuit.

This is not multiplying power, it is making the average output voltage of the power supply higher, and thus maximizing the power that can be delivered to the load. A pm motor in particular likes the ripple voltage to be as low as possible, since it's own 'back emf' serves to counteract the voltage supplied to it, and limits the rpm it can achieve.

As to the relevance in this discussion, the ac derived voltage that runs the typical 90 volt dc treadmill motor does not use the capacitor, and therefore the motor runs from a voltage which cycles from zero to 150 volts pulsating dc. The motor itself acts as its own capacitor by virtue of the inertia of the armature. The control element, whether an SCR or a TRIAC, is not subject to the high peak charging currents of a capacitor, but is only subject to the medium running currents of the motor. This makes all the power supply circuitry cheaper and smaller, and does require the motor to have a lower rated voltage- 90 as opposed to 150 or so, so it can actually achieve the desired rpm. The motor is running on little 'packets' of voltage, which does nothing for it's smoothness- witness the often included flywheel on these things. This is not ideal for a servo motor where you may want fast response. In total contrast is the power supply using a filter capacitor, which means dc output with a ripple factor, and control by a transistor which can now be operated at any suitable frequency and not limited to power line 60 or 120 hz. In fact you can choose a frequency at which the noise produced by the motor is minimized, and response times are improved. Now you have something which is more truly a servo system.

[PStechPaul]

For perhaps a bit more confusion, there are PFC correction circuits which are basically PWM boost converters, which draw higher current during the low voltage parts of the mains supply, and less current toward the peaks. This does require more complex circuitry as well as possibly expensive magnetic and solid state components, but the functionality can be realized with a single low cost IC. Voltage regulation and current limiting are bonuses.

[MaxHeadRoom]

J
As to the relevance in this discussion, the ac derived voltage that runs the typical 90 volt dc treadmill motor does not use the capacitor, and therefore the motor runs from a voltage which cycles from zero to 150 volts pulsating dc. The motor itself acts as its own capacitor by virtue of the inertia of the armature. The control element, whether an SCR or a TRIAC, is not subject to the high peak charging currents of a capacitor, but is only subject to the medium running currents of the motor.

The older SCR bridge TM's have been replaced by PWM versions, For e.g.now the popular MC2100 versions that uses PWM has AC direct into a bridge and a 2000uf capacitor..
It is microprocessor controlled that monitors the subsequent voltage and current to the motor. Just as any high quality servo drive does.
The strict rule of using the motor rated voltage for the supply level is confined to direct - uncontrolled DC to the motor.
All of the high quality servo drives I use operate with PWM to the motor.

[MaxHeadRoom]

One paper from Electronic Design

https://www.electronicdesign.com/technologies/electromechanical/article/21751588/powersupply-considerations-for-servo-amplifiers

[Bob Engelhardt]

... 1.41 is a reasonable estimate of the DC output of a well filtered supply. ...

Suit yourself.