Hi all.

In the local and national made APS/inverters they use power BJT transistors TIP35C (100V, 15A, 125W). They use about 8 transistors on each side for 1000VA, one transistor costs around 0.5$. They also use two 40A rectifier diodes for the charger, the diode costs costs about 5$.

I have a "made in china" APS that uses 3 'HEXFET' Mosfets on each side, the IRFP064N (55V, 110A, 200W, <8mΩ). Using these transistors they do not need rectifier diodes since these mosfets has internal diodes that are used... This mosfet costs about 1$ in my country. I gave the approx. prices for giving and idea for comparing, even if I think they are of good brands.

My question is: Isn't it better, cheaper, easier, more efficient (as less heat is produced from the pairs of the 3 mosfets as I sensed)?

Another question: I'm thinking to repair APSs with bad TIP35Cs by replacing them with IRFP064Ns after editing the circuit (I repair electronics). Any ideas about editing the circuit before I check the Chinese APS circuit board?

Regards,
Hazim

 

The diodes in the HEXFET are for protection purpose only. They cannot be used as charger rectifiers.

And you cannot switch BJT with FET, without changing the entire controller.

In short...you must be crazy to do that

 

Right.. BJT are current controlled while FETs are voltage controlled.... In my first question.. I see it cheaper and easier for the APS to be made with MOSFETs (considering the circuit changes...) and I'm wonder why they still use BJTs...
What you said about the rectifying diode is what I know, but in the Chinese APS there is no rectifying diodes, the connection done is as I said, the mosfets are used for rectifying too.

 

Right.. BJT are current controlled while FETs are voltage controlled.... In my first question.. I see it cheaper and easier for the APS to be made with MOSFETs (considering the circuit changes...) and I'm wonder why they still use BJTs...
What you said about the rectifying diode is what I know, but in the Chinese APS there is no rectifying diodes, the connection done is as I said, the mosfets are used for rectifying too.

Unless it's very specifically frequency controlled I see no way that a MOSFET could be used for rectification as it's basically the equivalent of a water tap you turn on or off. The included diode on some are more there for protection than anything else.

BJT's have advantages but it totally depends on the circuit.

Can you post a schematic of he board in question?

 

Hi again.

Sorry for being late to reply in this thread, I had much work to do.
The issue that was inquiry is about the rectifier of the charger. There is no diodes for rectification, from the connections it's like the MOSFETs used for the inverter are used for rectification when charging the battery.

Regards,
Hazim


The schematic:

 

Note: The two transformers in the circuits are not the ones used actually, also the feedback coil is in the same big transformer of the inverter&charger.

 

When considering saturated switches, BJTs (bipolar junction transistors) are somewhat similar to power MOSFETs in a few ways:
1) They are available in similar packages; ie: TO-220, TO-247, etc.
2) They usually have three terminals (four, if you count the tab; but that's connected to one of the other three)
3) One terminal (base, gate) controls current flow between the other two terminals (collector to emitter, drain to source)
4) They frequently have body diodes.

But, BJTs are current controlled, and MOSFETs are voltage controlled.

A BJTs base to emitter junction acts more or less like a forward-biased diode. When used as a saturated switch, you need 1/10 of the desired collector current flowing from the base to the emitter. When the current flow from the base to the emitter stops, so does the collector current.

A MOSFETs' gate acts more or less like a capacitor. When Vgs (the voltage on the gate, referenced to the source terminal) is below the threshold value (for simplicity let's just say < 1V), the drain terminal is electrically disconnected from the source terminal; there is no current flow. When Vgs = 10v (5v for logic-level N-ch MOSFETs), the MOSFET is saturated, and the resistance between the drain and source terminals is very low.

You must charge the gate until Vgs is ~10v to turn it on, and then discharge it until Vgs < 1v to turn it off.

As your circuit is now, the gates would charge, but there is no means provided for discharging them. You would wind up with all of the MOSFETs conducting simultaneously, which would result in loud noises and smoke.

The SGx524 is designed to drive transistors' bases, as is the rest of the circuitry.

One item missing is a snubber circuit on the transformer primary side. When the transistors turn off, there is no place for the transformer primary current to go, so there will be a very high voltage spike, which will tend to break down the collector-emitter junctions of the transistors.

 

Thank you for the information even if I know most of them. I forgot to add two (100 ohm 5W resistor in series with a 100uF capacitor) between both common drains and ground. I don't know if this solves the problem you mentioned. Maybe I have something else wrong with the circuit when I "reverse-engineered" it.
The main point is about the rectification of the charger... Are the Mosfets really used as the rectifier? the charging current reaches 40 or 50A....

Regards,
Hazim

 

Thank you for the information even if I know most of them.

I kind of figured that you knew the difference between BJTs and MOSFETs; the basic info was as much for others that have an interest as it was for you.

It needed to be said, however, to explain the reasoning why you would experience problems if you attempted to power up the proposed circuit.

I forgot to add two (100 ohm 5W resistor in series with a 100uF capacitor) between both common drains and ground. I don't know if this solves the problem you mentioned. Maybe I have something else wrong with the circuit when I "reverse-engineered" it.

Yes, those are snubbers. You should incorporate them into your schematic, as they are a critical detail. If the snubbers fail, the transistors will quickly be destroyed due to the very high peak voltages from the transformer primary; there will be no other path for the current.

The main point is about the rectification of the charger... Are the Mosfets really used as the rectifier? the charging current reaches 40 or 50A....

The circuit you posted does not appear to be a charger; it has a 230V output with no apparent means of rectification - if the output is supposed to be DC, you left something out.

 

Thanks again.
This is an "APS" or automatic power supply that has a big 220V relay that switches the power supplied to the AC outlet from the inverter to the AC mains input. The AC mains input is also connected to the relay and then to the big transformer to charge the battery... offcourse there is a circuit board for the charger but I didn't mentioned that before because the rectifier diodes are usually around 40A diodes mounted to the same heatsink where the inverter's power transistors are mounted, but there isn't any. Also, whatever it is the circuit of the charger, there is no high current diodes appearing in the hole case and the MOSFETS and their heatsinks become hot during charging process.

If you want I can record a short video for the "APS", this would clarify more for you.

 

Hi.
This is an old thread but I still want to understand how these UPS's charges the battery at such high current (up to 50A) without a single high current rectifier diode. I'm not sure but this is how I see the rectification process is:

First of all, the connections are as follow: the center tapped terminal of the transformer's 12V side is connected to the positive terminal of the battery. Each of the other two terminals (+/-12V) is connected to one side of the MOSFETs to the drain pins. Then all of the sources of the MOSFETs are connected to the negative terminal of the battery.
I see that during the charging process the gates are high, i.e. the MOSFETs are ON, and so only positive voltage passes through the MOSFETs.

I don't know if this is right, but if not how this UPS charges the battery?

Regards,
Hazim

 

I believe that the part of the inverter is used as a charger.

check with a scope to see if the power switches receive any signals when is charger mode

 

MOSFETs are being used as "ideal diodes" nowadays. That's why it's getting difficult to find high-current diodes anymore, as engineers are using MOSFETs as synchronous rectifiers instead.

Have a look at this Wikipedia entry: http://en.wikipedia.org/wiki/Active_rectification

 

That's it SgtWookie. "The control circuitry for active rectification usually uses comparators to sense the voltage of the input AC and open the transistors at the correct times to allow current to flow in the correct direction."
Thank you and thanks R!f@@.

 

I did not know what exactly happens but it is always been my assumption that the FET are used as the charger too.

 

MOSFETs are being used as "ideal diodes" nowadays. That's why it's getting difficult to find high-current diodes anymore, as engineers are using MOSFETs as synchronous rectifiers instead.

I never knew that... thanks

 

You're welcome. They're also using that technique in various other types of circuits. Googling "synchronous buck" will give you lots of interesting results.
Have a look here:
http://en.wikipedia.org/wiki/Buck_converter#Synchronous_rectification

If you scroll to the top of the article, you'll see representations of a simplified buck regulator using a diode; the diode is replaced by a MOSFET that is switched synchronously.