SparkFun Forums

[Dave Mueller]

If the load can't move without power applied to the actuator, then there will be no regenerative braking, so you don't need a big series diode.
You're measuring 36V without a load because this is the peak voltage from the rectifier. With a load, this will drop. This is because your power supply is not regulated. If you put the big cap at the output of the rectifier, then it will charge to 36V and when you start the actuator, current will flow from the cap to the actuator until the capacitor voltage drops to 24V, then it will smooth the pulses from the rectifier.

Putting a resistor across the big cap is just a safety measure to allow the capacitor to discharge to a safe level when the power supply is off and there is no load.

Putting the cap across the LA terminals will cause an enormous current surge through the controller when it turns on. This will stress the controller as well as the bridge rectifier, and series diode if you use it. If you put the cap at the output of the bridge, there will be a surge when the supply is turned on but not through the controller. Diodes are usually spec'd to handle this turn on surge.

Heat sinking the rectifier is the most likely answer to the problem, without buying or building a properly designed regulated 24VDC power supply.

[mikelaurey]

Thank you for your suggestions.

So, to sum up the situation, I have to:

1) heat sinking the rectifier
2) add a large electrolytic capacitor (50V, 22.000uF) after the bridge rectifier and before the driver
3) add a resistor in parallel to the capacitor (10K resistor is OK?)
4) No diode is needed

Is everything ok?

[Mee_n_Mac]

That's about right. Here's my take on what's happening. You have some transformer (part unknown to us) and you think it takes 220 in and spits out 24. But the 220 is AC and that's 220 V RMS. The peak voltage is 1.414 (sq root of 2) larger. If the 24 out is also AC and V RMS, you should get peaks of 34 V. I suspect your driver has some bulk capacitance in it and so w/o any load from the LA, you should measure about 34 VDC on the output of your bridge rectifier. I say about because there will be a voltage drop due to the diodes (a bit over a volt) and wiring. If you looked at the waveform with an o-scope you might see the solid line in the pic I linked to.

Now when you turn on the LA it will draw some amps. The output will look more like the dotted line in the pic as the caps in the driver are not large enough to smooth out the droop. If measured with a DVM under these circumstances you will likely see something a lot less than 34 VDC.

Add in 20,000uF of capacitance and that droop will be smoothed out somewhat. My loosey goosey simulation says about 2 V pp at 7-8 A with an average of perhaps 32 or so VDC.

Is that 30+ VDC OK ? It's higher than the 24 VDC you were aiming for. Is that amount of voltage ripple OK ?

[mikelaurey]

This is the datasheet for my transformer: http://www.deatech.net/files/2e_eur2.pdf
The code is EMS25012 and the power is 250VA.

My driver can work with a voltage input range of about 10 - 50VDC without problems and I can set an overvoltage alarm by software.
Don't you think that the 22,000uF capacitance would help me to smooth the waveform?

[Mee_n_Mac]

Don't you think that the 22,000uF capacitance would help me to smooth the waveform?

Yes it should. Using 2 would smooth it out more and reduce the effective ESR of the caps and ripple current through each one.

[mikelaurey]

Two capacitors in parallel between the rectifier output and the driver input?
It seems to be a great idea! Thank you!

[Mee_n_Mac]

Two capacitors in parallel between the rectifier output and the driver input?
It seems to be a great idea! Thank you!

Yes, 2 caps in parallel have the capacitance of their sum, in your case 40,000 uF.