Regenerative power dump circuit

[rokag3]

i build this circuit below made by Mariss Freimanis of gecko
http://www.hv4all.com/cnc/Power%2520...%2520Bleed.pdf
it work fine to discharge the capacitor (i use 2 bulb 220v 100W in // r=40ohms per bulb =20 ohms)

[Mariss Freimanis]

While on the subject: A brush-type servo doesn't need returned energy protection because it cannot generate any in normal operation. The motor 'back EMF' is proportional to speed, speed is limited by the supply voltage, therefore all voltages are equal to or less than supply voltage. No 'returned energy'.

Two exceptions:

1) There is energy stored in the motor's inductance. This is a minuscule amount compared to stored mechanical energy and can be safely neglected.

2) During an e-stop you do want to stop the motor quickly instead of leisurely coasting to a stop. The original circuit doesn't address this while all three of the circuits I posted do. The 'dump' circuits dynamically brake the motors to a stop when AC is removed to the power supply. The motor becomes a generator working into a large load (dump resistor) which quickly extracts stored mechanical energy from the motor and mechanism.

Mariss

[Mariss Freimanis]

rokag3,

Remember a 220VAC 100W light bulb has a 'hot' resistance of 20 Ohms. The 'cold' resistance is much lower, typically 2 Ohms. Use an Ohmmeter on this light bulb to verify my comment.

Mariss

[Mariss Freimanis]

NC Cams,

It is a perfectly valid method. The currents and voltages look impressive but the amount of energy involved is relatively small meaning the dissipation event lasts a short time. As an example, a 1,000 lb mass moving at 500 IPM contains only 10 Joules of kinetic energy. A Joule is 1 Watt for 1 second; decelerating this mass to a stop in 1/10 of a second means the dump resistor dissipates 100 Watts for 1/10 of a second.

The dump resistor can be as small as 5W for this example provided it's a wire-wound resistor. Wire-wound resistors have the highest pulse power dissipation rating, typically 100:1. This means a 5W wire-wound resistor can dissipate 500W peak power for a short period of time without any damage.

Mariss

[H.O]

Hi guys,
Since I'm the one who originally designed and published the circuit that Irfan has attached to the first message in this post I thought I'd explain a bit why I did it the way I did....

Before designing the circuit I looked at both of Mariss' circuits that he attached to his post (they've both been available on Geckodrive's Yahoo-group for quite some time).

Dump1 was discarded because I couldn't source a PNP transistor rated at 150-170V and ~30A. Dump2 was also discarded because I'm using a three-phase rectifier and I honestly couldn't figure out if that circuit would work with a three phase rectifier. (I confess, I'm embaressed). I did post a question on the Geckodrive Yahoogroup about it but it was/is unanswered so I set out to design my own.

Also, on the schematic you see two connections marked Bleeder+ and Bleeder-. In my own system these are connected to a 7.5ohm resistor via the normally closed contacts of the main contactor. When entering E-stop the contactor deactivates and connects the resistor across the powersupply capacitors, discharges them and "breaks" the motors. Agreed, it requires a second resistor....


Mariss, I'm a bit confused that you say a revese energy dump isn't needed under normal conditions. When testing my system I set the dump-circuit to activate at 15V above nominal powersupply voltage (nominal is 130VDC) and it clearly activates during decceleration from high speed. There's even a video on Youtube (not mine but he's using "my" dump circuit) where you can see the difference with and without the circuit. I'll see if I can find it but I think I must have misunderstood you or something.

/Henrik.

[Mariss Freimanis]

Henrik,

First let me say you have a well worked out circuit so my comments were not intended as any kind of negative criticism.

Somehow I missed your post; I would have responded otherwise.

Dump1 circuit: You can replace the TIP147 with a small-signal PNP like an MPSA92 which has a Vco of 350VDC. Install a collector resistor sufficient to limit collector current to 1mA or 2mA at 170VDC. Have the end of this resistor drive the gate of a suitable n-channel MOSFET whose drain circuit contains the dump resistor. MPSA92 collector -> 100K -> MOSFET gate, MOSFET gate -> 10K -> GND. Place a safety 12V zener gate -> cathode, anode -> GND.

Dump2 circuit: The main advantage over dump1 is the wasteful rectifier in dump1 is eliminated. It will work with 3-phase as well as with single-phase.

Returned energy: Never post until the brain is fully saturated with caffeine; it wasn't and I wasn't thinking. The theoretical maximum returned voltage is twice the supply voltage, i.e. motor back EMF equal to supply voltage, '-' side of motor to 'on' top bridge MOSFET.

I have attached a diode decoupled supply voltage scope trace going to a G320. Supply voltage is 48VDC, NEMA34 motor is decelerating from 3,000 RPM to zero in 100 milliseconds. Note the voltage going to near 80VDC, 30 volts higher than supply.

Mariss

[H.O]

Hi,

First let me say you have a well worked out circuit so my comments were not intended as any kind of negative criticism.

Not taken as such - just explaining the reasons the circuit is what it is. It works and it has served me well but there's always room for improvement... Good to know that the Dump2 circuit does indeed work with a three-phase rectifier, thanks!

/Henrik.

[Mariss Freimanis]

Let me hasten to add that another 1N4004 going to the bridge must be added in the circuit. A single-phase full bridge rectifier has 4 diodes, a 3-phase bridge has 6 of course.

Mariss

[Xerxes]

Originally Posted by Mariss Freimanis Returned energy: Never post until the brain is fully saturated with caffeine; it wasn't and I wasn't thinking. The theoretical maximum returned voltage is twice the supply voltage, i.e. motor back EMF equal to supply voltage, '-' side of motor to 'on' top bridge MOSFET.

I think it's not that way. Drive acts as step-up converter during braking where motor acts as inductor. So voltage can be pumped up almost infinitely. During braking current direcition just reverses and motor EMF doesn't play any role.

Attach more mass to the 48VDC motor and try running the test again

[ftkalcevic]

Sorry for waking such an old thread.

I was wondering if anyone makes a board, or kit, or module that provides the power dumping features?

On my servo power supply, I built a board based on Mariss's design (from the gecko yahoo group) with a few different parts. It work fine for a couple of years, then one day, I turned the power on, and after about a minute, there was a big bang, and lots of smoke coming out of the enclosure. The power resistor and the pcb around the resistor were toasted.

If I can't find a ready made one, I'll try building another.