di/dt

[ILCML]

I am going to rectify 24v to supply a 24VDC motor with an FLA of 57A. The startup current for the motor is around 120A I'm a little confused on how to calculate the inrush capability the rectifier would need to be able to handle.

The spec sheet I have on the diode module (Crydom M50 series) I'm looking at says

di/dt =
"Critical Rate of Rise of On-State @ TJ=125 C (A/uS)" = 100

Additionally, it says that the

"Maximum Non-Repetitive Surge Current (A) [1/2 cycle, 60Hz]"

is 800 for the 60A module and 1500 for the 100A module.

So for a motor pulling roughly 120A at startup, a 60A rectifier should have no problem at all with it....

 

[Skogsgurra]

Probably no problem. The di/dt of a diode is seldom critical. It *could* be critical in an application where a capacitor is connected to the rectifier - but even then, the reactance of the transformer will keep di/dt down.

The time constant of a 24 V DC motor is probably in the 10 ms range, seldom less. The locked rotor current is usually 10 - 20 times FLA (LV DC motor, not ASM). That would mean a di/dt around 57*20/10 = 114 A/ms or .114 A/us - a lot less than maximum di/dt. I have used very pessimistic data here. And you still have a ridiculous safety margin with regard to di/dt.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[itsmoked]

I agree with skogs. I'd probably go with the 100A for a 57A FLA.(depending) Remember those buggers will be dropping some voltage at that current and so will be dumping waste heat that must be contended with. Probably around 25-50W which is non-trivial in cooling aspects.

Are you using a full-wave or a bridge setup?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[ILCML]

the diode module is a FWB in a pretty sturdy little package (not sure what it's called)-little square potted unit with 4 faston type tabs sticking out. About 1.5" square.

What is the "maximum non-repetitive surge current" value of 800A for?

 

[Skogsgurra]

It is what you need when you connect a rectifier with a smoothing capacitor. You then charge the capacitor in the first half-cycle and then, in the following half-cycles, the rectifier only has to supply load current. This surge can be very high. If you look at the spec, it usually says 8.3 or 10 ms. That is how wide half a cycle is in a 60 Hz and 50 Hz grid.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[ILCML]

I just realized there is a cap across the rectifier output....47000uF electrolytic...

 

[itsmoked]

Ouch.. You must walk the line between impedance available and the inrush to charge that cap and the efficiency of the supply. You will explode the diodes if you execeed that 800A limit. Which is possible. Often a resistor is put in series between the bridge and the cap to limit the inrush to say 750A. This isn't as bad as it seems since the resistor is there only doing much voltage dropping when the current is up in the hundreds of amps.

However! Often you can get by without any series resistor limiting because supplies usually have a transformer in front of them. A transformer provides a large amount of impedance to large currents. In most cases just having one will keep your inrush well below those hundred amp levels.

The nasty problems come when you are just rectifying the mains voltage with no transformer and charging a front-end capacitor. All the switching supplies have to contend with this. That is why all switching supplies list their inrush currents.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[ILCML]

oh, no, nothing near that high of current! Max current will be around 60A (other than inrush)....The 24vac is off a 3kVA step down transformer.

 

[itsmoked]

Max current will be around 60A (other than inrush)

I don't get it. We are talking [red]inrush[/red] here. It doesn't matter what steady state is if your circuit never gets there!

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

A 3 kVA transformer usually has something like 5 % regulation from idling to full load. If this is a single-phase transformer at about 24 V (neglecting the sqrt(2) now), the rated current will be like 100+ A. A 5 % regulation (voltage drop at full load) means that current can reach 20x100+ or around 2 kA if you switch on at 90 degrees (top of sine). Wiring resistance will reduce that peak current, but you are still quite close to what this rectifier can take - maybe even above that limit.

Has the rectifier worked so far?

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[sreid]

Two additional thoughts.

1) 24 VAC full wave rectifies to about 33 VDC.

2) As Skogsgurra says, the starting inrush current is usually about 10 times the full load current. A peak current of 120 amps may be "limit the start up current to 120 amps." And although the starting current decreases as the motor RPM increases, it may take hundreds of milliseconds for the motor to get to full speed. The current may be limited by the transformer saturating but the diode bridge will probaably be stressed.

 

[LionelHutz]

Well, as already pointed out, 24VAC rectified with a capacitor filter is not 24VDC and charging the 47000uF capacitor is not trivial.

I would use about 19VAC which when rectified and accounting for the diode drops of a full-wave rectifier gives about 25VDC.

I think you need to pre-charge the capacitor. Use a resistor in series with the AC input to the rectifier. Then, use a relay (contactor) to bypass the resistor after a certain amount of time.

I didn't look at it but you should be calculating the ripple voltage and current on your capacitor bank. The closer you push an electrolytic close to it's ripple current rating the shorter it's life.

Why do you need the capacitor to drive this DC motor? Most of the time you can rectify directly to the motor without any filtering for the DC.

Finally, to me that Crydom module look really light for the currents you're expecting it to handle. I'd be using something like 2x Eupec DD104 (dual diode) modules. Better surge and I2t ratings.

 

[Skogsgurra]

I added the "(neglecting the sqrt(2) now)" in my post to avoid further discussions on selection of transformer secondary voltage - that part is another discussion. But, as we are at it, the idling voltage and the full load voltage may be a surprise to some. I recently actually had to go a bit higher than common wisdom says in order to move a heavy load on a 24 V DC motor.

And, the motor starting current is not the problem either. The problem is when you switch transformer primary on top of the sine and need to charge your 47 mF capacitor. Do you really need a 3 kVA transformer for that? Is your motor running fully loaded all the time? If not, you could certainly go down to something like 1.5 or 2 kVA. That would increase transformer impedance and reduce your half-wave peak.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[TurbineGen]

Where are you getting your 120A inrush current value from? Is the motor going to be loaded when it is switched in? If the motor is going to be loaded when switched in, it will take some time to speed up so you might be over the 60 AMP rating for quite some time, which will stress the device. Additionally the 47mF Capacitor is going to draw high currents as well when switched in, which will add to the stress on the rectifier.

Additionally, as it was posted above, brushed DC motors seldom require a filter capacitor. Why are you using such a large capacitor? If the capacitor is necessary, is it possible to place a switching device between the motor and the rectifier? By switching the motor in after the rectifier has been energized, you won't have the high inrush currents from both the motor and the capacitor at the same time.

 

[electricuwe]

Regardless that there are already a lot of answers you all missed on important point:

additionally to the requirements from charging the capacitor and the time constant of the motor you also should consider the mechnical load. If the motor has to speed up load with an high inertia, it may take several minutes until the the current approaches stedy state.

Basicly there are two solutions to this problem:

- calculate the DC-output current for the recifier and use a thermal model of the rectifier including heatsink to estimate rectififier juction temperature after start-up

- if you really have high inertia load you will not be able to solve the problem just by oversizing the rectifier. In this case you need an approack like it is done in large DC-drives: controlled recifier with current limit