Replacing Old Resonant Transformer from an Industrial Battery Charging circuit

[Mithun_K_Das]

This is a very old fashioned industrial battery charger that uses 440V line converter into 220V DC via a resonant transformer. The problem is, the resonant transformer is not working and its saturated. Also they are not happy with it and don't want to use it any more.

So what I need to do is, making an alternative electronic circuit that can handle the load. Ratings are as follows:

Voltage: 300V AC.
Current: 0-10A.
Peak: 30A;1.5sec.
Run time: 24Hrs at 280V/5A.
Life time: must be over 2yrs


What I tried to do is making a TRIAC Based dummy AC load what can be connected in series with the line. Problem is the Amp is too high. I tried with BTA26, one pcs only. But when the load is connected circuit shuts down due to heavy load. And looses the control.

What I need to do is extending the current handling capacity of the TRIAC. But it is not wise to make parallel connection with TRIAC. So what is the way to control this heavy load in this voltage range?

 

[chuckey]

I think you will find that that "resonant inductor" is actually a saturable inductor. The problem is the difference with the peak and the mean of the rectified DC current. You should include an inductor in series with your new circuit to reduce the peak current and increase the mean.
The standard way would be to have two of the diodes in the bridge rectifier replaced with SCRs.
Frank

 

[Mithun_K_Das]

I was planning to use MOSFET to control the AC load with the help of Bridge rectifiers. Will it be good?

 

[chuckey]

If you just remove the reactor, can the existing diode bridge handle the voltage? If it can the easiest way is to use a 10A 400 V FET in series with the output of the bridge. When its off Iout = 0, when its on Iout = 20 A?. So you feed its gate with PWM to regulate the current (at a high speed Fc = 50 KHZ?). The existing arrangement, monitors the voltage drop across R32, so can actually track the current wave form to make a much smoother DC. use this resistor for your max current overload. Use a very small inductor to smooth the 50 KHZ pulses to pure DC. Like the top of a H bridge SMPS -loads of examples in this forum.
Frank

 

[Mithun_K_Das]

Yes, the Diodes are heavy duty and rated 60A/600V each. It will definitely can handle the current. But I'm afraid of the Heating problem of the FET. It is easy to use FETs in this case but only one problem if the Fets get hot at low duty cycle.

 

[Anna Conda]

You need a standard circuit with two controlled SCR's (>800V 35A) providing the +ve out and two 1kV (35A) diodes providing the -ve o/p in a bridge (rectifier) arrangement, standard control of the SCR's will give you regulated 220VDC out, best to have a large o/p choke if the leakage in the transformer is not high (e.g. 10mH min) and some decent sized 350V electro's on the o/p, e.g. 10,000uF - if you can't find on the net we have some std circuits...
All other approaches are bound to fail due to overload impulses and mains spikes
regards, Anna.

p.s. it is usually the resonant cap (on a winding of the Tx) that goes low uF over time and stops the ferro-resonant transformer action - replacing the cap (often >660Vac) with a new one often fixes the problem.

 

[chuckey]

Anna Conda The resonant stabilizers I have seen do not use a control board, just the transformer and the capacitor, that's why I thought it was a saturable reactor, the control board having a AC input for the DC saturation control.
Frank

 

[Orson Cart]

Correct, a ferro resonant transformer has a separate winding with a cap on it to run the steel into saturation and provide the voltage limiting, if you are doing away with this, as implied, you will need a control circuit driving SCR's as Anna suggests.
The saturable reactor control shown in your photo is a different thing again, if you remove it you again need a more modern SCR controller per Anna's suggestion.

 

[Mithun_K_Das]

Actually I'm thinking about using IGBT or MOSFETs. Because driving SCRs/TRIACs may not be so efficient in this case. The Batteries are connected to a bus bar that runs a VCB (Vacuum Circuit Breaker) that takes 280V/5A to run and 10~12A at start up. That means, the system will have a instant load when VCB is switched on of around 10~12A at 280V. If I use a TRIAC, it may fall as the positive/negative half cycle breaks down in this moment.

Also a IGBT may fall but IGBT will much more conductive in nature than TRIACs; No duty cycle dependance for IGBTs but TRIACs/SCRs. Only one thing is problem, supporting the instantaneous current while VCB is switched on.

- - - Updated - - -

Every thing will be solved only if we can get a device/system that can take the instant load of 12A.

Will a Capacitor work good for this?

 

[Orson Cart]

Generally SCR's are the more rugged devices, a 35 amp SCR is usually good for >500A half cycle pulse.

 

[Mithun_K_Das]

This is the name plate of that Resonant Transformer. But the language is not known to me. Can anybody help me what this Chainees language says?