We brought a couple of new electronic designers onto the ventilator project I'm working on. One of them was suggesting we control the motor with a va riable voltage rather than an H-bridge to switch the current in a PWM manne r. He mentioned some high power amps to drive this, but they are class AB and so would have high losses at anything other than the maximum output.
Is it a common practice to use a switching circuit to supply a controlled v oltage to a motor? The motor is an inherently inductive load, so a PWM dri ve at a high enough frequency would approximate a constant current to the m otor. A smoothing cap at the source (we already have 2000 uF) helps that a lot. The H-bridge would be retained to allow the polarity to the motor t o be reversed.
Is this a method used? It would lower the I2R heating in the motor with a lower continuous current rather than a higher pulsed current with a duty cy cle.
We have scope plots showing a 500 Hz pulse rate is not fast enough to make that work. I need to ask if we can get some better measurements to see how long it takes for the motor current to drop significantly when the H-bridg e removes the power connection and the reverse diode conducts.
The H-bridge has a 10 kHz max rate, but that might be good enough. I'll se e if I can get someone to make a measurement.
--
Rick C.
- Get 1,000 miles of free Supercharging
- Tesla referral code - https://ts.la/richard11209
ct I'm working on. One of them was suggesting we control the motor with a v ariable voltage rather than an H-bridge to switch the current in a PWM mann er. He mentioned some high power amps to drive this, but they are class AB and so would have high losses at anything other than the maximum output.
voltage to a motor? The motor is an inherently inductive load, so a PWM dr ive at a high enough frequency would approximate a constant current to the motor. A smoothing cap at the source (we already have 2000 uF) helps that a lot. The H-bridge would be retained to allow the polarity to the motor to be reversed.
lower continuous current rather than a higher pulsed current with a duty c ycle.
e that work. I need to ask if we can get some better measurements to see ho w long it takes for the motor current to drop significantly when the H-brid ge removes the power connection and the reverse diode conducts.
ee if I can get someone to make a measurement.
Sloman A.W., Buggs P., Molloy J., and Stewart D. ?A microcontroller
-based driver to stabilise the temperature of an optical stage to 1mK in th e range 4C to 38C, using a Peltier heat pump and a thermistor sensor? ? Measurement Science and Technology, 7 1653-64 (1996)
We used pulse-width-modulation with an H-bridge to deliver an essentially D C current - up to 3A - to drive the Peltier junction. The H-bridge switched no faster than 200kHz, and the pulse-width pattern repeated at 17.4kHz., b ut it was fudged so that a 50% duty cycle showed up as a 139Khz square wave , which made filtering out the low frequency components a lot easier. E-mai l me if you want a copy of the paper.
You can certainly use an H-bridge to produce a waveform which works as a si ne wave of lower peak voltage amplitude that the supply rail. I've done it in other applications. You do have to worry about the high frequency comp onents coming out of the H-brige, but you can filter them close to the H-br idge and stop them radiating form the motor leads and warming up motor.
ct I'm working on. One of them was suggesting we control the motor with a variable voltage rather than an H-bridge to switch the current in a PWM man ner. He mentioned some high power amps to drive this, but they are class A B and so would have high losses at anything other than the maximum output.
voltage to a motor? The motor is an inherently inductive load, so a PWM d rive at a high enough frequency would approximate a constant current to the motor. A smoothing cap at the source (we already have 2000 uF) helps that a lot. The H-bridge would be retained to allow the polarity to the motor to be reversed.
I suspect that you do not need a very constant current. Consider a AC/DC motor run on AC. Not constant current ,but the rpm is essentially constan t. And does some variation in the motor speed make any difference for a v entilator?
Just suggesting you check the requirements.
Dan
a lower continuous current rather than a higher pulsed current with a duty cycle.
e that work. I need to ask if we can get some better measurements to see h ow long it takes for the motor current to drop significantly when the H-bri dge removes the power connection and the reverse diode conducts.
ject I'm working on. One of them was suggesting we control the motor with a variable voltage rather than an H-bridge to switch the current in a PWM ma nner. He mentioned some high power amps to drive this, but they are class A B and so would have high losses at anything other than the maximum output.
ed voltage to a motor? The motor is an inherently inductive load, so a PWM drive at a high enough frequency would approximate a constant current to th e motor. A smoothing cap at the source (we already have 2000 uF) helps that a lot. The H-bridge would be retained to allow the polarity to the motor t o be reversed.
motor run on AC. Not constant current ,but the rpm is essentially constant. And does some variation in the motor speed make any difference for a venti lator?
All electric motors are essentially synchronous motors and depend on gettin g sine waves of current through the windings. DC motors just use brushes a nd commutators to do the switching.
The frequency and phase of the sine waves have to line up with the rotation of the motor - in a DC motor the commutator does that for you. Getting syn chronous motors spun up and spinning at the right speed can be tricky.
When you go away from sinewaves, you get more heating when generating the s ame torque. If you put lots of high-frequency current into the motor windin gs it doesn't do a thing to generate extra torque, but will heat up the pol e pieces, which won't help their magnetic properties.
The question didn't suggest that Rick C. hard thought all that hard about w hat sort of current he need to feed into to the motor. He didn't actually i dentify the sort of motor he was using, which wasn't a good start.
"At the source"... the power source to the circuit. The H-bridge data sheet shows a large capacitor at this point, I believe to supply the peak current and also to mitigate the dI/dt voltage spike when the motor is stopped or reversed.
--
Rick C.
+ Get 1,000 miles of free Supercharging
+ Tesla referral code - https://ts.la/richard11209
ject I'm working on. One of them was suggesting we control the motor with a variable voltage rather than an H-bridge to switch the current in a PWM m anner. He mentioned some high power amps to drive this, but they are class AB and so would have high losses at anything other than the maximum output .
ed voltage to a motor? The motor is an inherently inductive load, so a PWM drive at a high enough frequency would approximate a constant current to t he motor. A smoothing cap at the source (we already have 2000 uF) helps th at a lot. The H-bridge would be retained to allow the polarity to the mot or to be reversed.
C motor run on AC. Not constant current ,but the rpm is essentially const ant. And does some variation in the motor speed make any difference for a ventilator?
The operation of the motor is controlled to produce the desired effect. I am talking about upping the PWM rate to approximate a constant current driv e rather than the present pulsed drive we have now. The switching losses i n the H-bridge will be higher, but the losses in the motor where we are hav ing some trouble are lower. Drive with 3 amps continuously or drive with 6
es in the motor, about 2x I believe.
--
Rick C.
-- Get 1,000 miles of free Supercharging
-- Tesla referral code - https://ts.la/richard11209
This might help a bit, this guy works for Portescap and they make nice motors and generally seem to know what they are doing. (I'm assuming it is a brushed motor you want to control.)
formatting link
From what I know of your application you might need some kind of feedback control of motor speed.
roject I'm working on. One of them was suggesting we control the motor wit h a variable voltage rather than an H-bridge to switch the current in a PWM manner. He mentioned some high power amps to drive this, but they are cla ss AB and so would have high losses at anything other than the maximum outp ut.
lled voltage to a motor? The motor is an inherently inductive load, so a P WM drive at a high enough frequency would approximate a constant current to the motor. A smoothing cap at the source (we already have 2000 uF) helps that a lot. The H-bridge would be retained to allow the polarity to the m otor to be reversed.
/DC motor run on AC. Not constant current ,but the rpm is essentially con stant. And does some variation in the motor speed make any difference for a ventilator?
I am talking about upping the PWM rate to approximate a constant current dr ive rather than the present pulsed drive we have now. The switching losses in the H-bridge will be higher, but the losses in the motor where we are h aving some trouble are lower. Drive with 3 amps continuously or drive with
roject I'm working on. One of them was suggesting we control the motor wit h a variable voltage rather than an H-bridge to switch the current in a PWM manner. He mentioned some high power amps to drive this, but they are cla ss AB and so would have high losses at anything other than the maximum outp ut.
lled voltage to a motor? The motor is an inherently inductive load, so a P WM drive at a high enough frequency would approximate a constant current to the motor. A smoothing cap at the source (we already have 2000 uF) helps that a lot. The H-bridge would be retained to allow the polarity to the m otor to be reversed.
C/DC motor run on AC. Not constant current ,but the rpm is essentially co nstant. And does some variation in the motor speed make any difference fo r a ventilator?
I am talking about upping the PWM rate to approximate a constant current drive rather than the present pulsed drive we have now. The switching loss es in the H-bridge will be higher, but the losses in the motor where we are having some trouble are lower. Drive with 3 amps continuously or drive wi
losses in the motor, about 2x I believe.
hed-dc-motors-using-pwm
Thank you. Yes, we have a PID control loop. There are a number of issues in the thing being controlled starting with the motor itself that we don't have characterized. The guy doing most of the motor work is a mechanical e ngineer who seems to be happy with a hunt and peck approach to finding a wo rkable solution. He has created a rather large number of hands to press on the bag, each time changing it without properly understanding how it might interact with the bag it is pressing on... resulting in yet another attemp t needed. He's also the project leader so not much I can tell him.
We are making progress though and may have something ready for a manufactur er to look at by the end of the year. Not sure about the software. That's not really being worked on at the moment while the hardware is being bange d on. Not sure why we can't come to terms on an interface spec and work on both in parallel. Too many members seem to have a hobbyist approach using the REPL method I guess, not that REPL is bad, just not the only approach.
I've been trying to explain the nature of PWM control to these guys, but it isn't sinking in. Your link should help a lot. But my biggest problem is getting people to stop experimenting and start doing thoughtful, rational design.
Even when Larkin is doing his experiments, it's not just to find out if som ething will work or not. He has a plan to take measurements that will give him information. The garage mentality is the pits.
--
Rick C.
-+ Get 1,000 miles of free Supercharging
-+ Tesla referral code - https://ts.la/richard11209
ct I'm working on. One of them was suggesting we control the motor with a v ariable voltage rather than an H-bridge to switch the current in a PWM mann er. He mentioned some high power amps to drive this, but they are class AB and so would have high losses at anything other than the maximum output.
voltage to a motor? The motor is an inherently inductive load...
Motors have iron in the windings, and there's leakage inductance, BUT that' s at audio and below. At typical SMPS frequencies, that iron is resistive, and rotor inertia is capa citive...
Yes, switching is a way to supply a motor, but it's better to consider usin g a HF-capable choke in series, rather than modeling the motor as an inductor, if you want to keep your de sign on-track. Why does one want variable motor control, can't you just do ON/OFF?
ject I'm working on. One of them was suggesting we control the motor with a variable voltage rather than an H-bridge to switch the current in a PWM ma nner. He mentioned some high power amps to drive this, but they are class A B and so would have high losses at anything other than the maximum output.
ed voltage to a motor? The motor is an inherently inductive load...
t's at audio and below. At
pacitive...
ing a HF-capable choke in series,
design on-track. Why
The motor turns an arm with a "hand" that pushes on a bag that expels air t hat inflates a patient's lungs. The mode of operation is constant pressure . There is a feedback system to maintain the pressure during the inspirato ry portion of the cycle. None of that is the target of this matter. I'm j ust looking to reduce the losses and noise of operating the motor.
It would be useful to find a H-bridge motor controller that is rated to wor k at PWM rates more in the 20 kHz range with 15A+ loads. The VNH3SP30-E we are planning to use has an upper limit of 10 kHz.
--
Rick C.
+- Get 1,000 miles of free Supercharging
+- Tesla referral code - https://ts.la/richard11209
You are trying to inflate and deflate the patients lungs slowly and steadily with gentle periodic movements not pump up a car tyre quickly.
It should be possible to use a relatively modest motor and gearing to convert that basic rotary motion into whatever movement is required.
That is more the sort of motor power control kit I would expect to see on electric scale model trains on 3.5" or 5" gauge. They can typically pull three or four bogeys loaded with kids round on a track. Or at least they could until Covid restrictions came into play. eg.
formatting link
RFI is something else to look out for motor power control in a hospital setting. It caught out Crossrail in the UK badly as their signalling system was confused by dynamic interference from the rolling stock motor controllers. They are not the only ones Hitachi's Azuma trains didn't get on with the signalling system north of York when first delivered.
roject I'm working on. One of them was suggesting we control the motor with a variable voltage rather than an H-bridge to switch the current in a PWM manner. He mentioned some high power amps to drive this, but they are class AB and so would have high losses at anything other than the maximum output .
lled voltage to a motor? The motor is an inherently inductive load, so a PW M drive at a high enough frequency would approximate a constant current to the motor. A smoothing cap at the source (we already have 2000 uF) helps th at a lot. The H-bridge would be retained to allow the polarity to the motor to be reversed.
C motor run on AC. Not constant current ,but the rpm is essentially constan t. And does some variation in the motor speed make any difference for a ven tilator?
ing sine waves of current through the windings. DC motors just use brushes and commutators to do the switching.
on of the motor - in a DC motor the commutator does that for you. Getting s ynchronous motors spun up and spinning at the right speed can be tricky.
same torque. If you put lots of high-frequency current into the motor wind ings it doesn't do a thing to generate extra torque, but will heat up the p ole pieces, which won't help their magnetic properties.
what sort of current he need to feed into to the motor. He didn't actually identify the sort of motor he was using, which wasn't a good start.
He is asking the right question. The inductance of a motor winding is very high, and a 10kHz to 20kHz PWM is typical to control the motor winding curr ent. it will not generate a lot of losses since the inductance reduces the ripple current to very low amplitude
ject I'm working on. One of them was suggesting we control the motor with a variable voltage rather than an H-bridge to switch the current in a PWM ma nner. He mentioned some high power amps to drive this, but they are class A B and so would have high losses at anything other than the maximum output.
ed voltage to a motor? The motor is an inherently inductive load...
t's at audio and below. At
pacitive...
That is not correct. At switching frequency, the winding is inductive. Roto r inerita has almost no effect on the winding. If you have burried versus s urface magnets, the winding inductance varies with respect to rotor orienta tion, but that is another issue, relevant for the motor control algoritm
ing a HF-capable choke in series,
esign on-track. Adding more components that is not needed
Rotor/load inertia gets interesting when you need deceleration. That stored energy has to go somewhere. An H-bridge will happily extract the mechanical energy from the system and stuff it back into the power supply.
project I'm working on. One of them was suggesting we control the motor wit h a variable voltage rather than an H-bridge to switch the current in a PWM manner. He mentioned some high power amps to drive this, but they are clas s AB and so would have high losses at anything other than the maximum outpu t.
olled voltage to a motor? The motor is an inherently inductive load...
that's at audio and below. At
capacitive...
otor inerita has almost no effect on the winding. If you have burried versu s surface magnets, the winding inductance varies with respect to rotor orie ntation, but that is another issue, relevant for the motor control algoritm
Correct, but that is the same case if he used a H bridge with PWM, or a con verter in front
tirsdag den 20. oktober 2020 kl. 23.43.40 UTC+2 skrev John Larkin:
project I'm working on. One of them was suggesting we control the motor wit h a variable voltage rather than an H-bridge to switch the current in a PWM manner. He mentioned some high power amps to drive this, but they are clas s AB and so would have high losses at anything other than the maximum outpu t.
olled voltage to a motor? The motor is an inherently inductive load...
that's at audio and below. At
capacitive...
otor inerita has almost no effect on the winding. If you have burried versu s surface magnets, the winding inductance varies with respect to rotor orie ntation, but that is another issue, relevant for the motor control algoritm
if you you keep switching so it works like a boost converter
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.