3 Phase Rectifier

[IRQ57]

Hello,

I would like to clear a doubt: Which is the difference between the output voltage of a SCR 3 phase semi-controlled rectifier and a fully controlled SCR 3 phase rectifier thru the full range of variation of the firing angle.

This question is due to a problem that I am finding in a semi controlled rectifier ( 3 SCR's and 3 diodes), build by me, in which is not possible to regulate continuosly from 0 to 300 VDC ( 225 VAC 3 PH supply). It jumps from a few volts (20-30V) to 150V and from this point to 300VDC smoothly. This happens with resistive loads from 300 Ohms to the full load of 10 Ohms 30 A.

I could not find information about. Any help will be wellcomed.
Thanks in advance.

 

[zevon8]

If you do not provide good transient protection for the SCR's with snubbers ( series resistor and capacitor of say 100 ohm and 0.1 uF for 60Hz ) along with fast MOV's, you could exceed the dv/dt rating of the SCR and have all kinds of accidental firings. This will destroy the SCR's.

You may not have sufficient gate drive for the SCR. The gate current changes as the SCR fires, so be sure to provide a quick rising, solid signal. Verify the signl you provide is what the datasheet specifies for the firing conditions you need.

 

[IRQ57]

Thanks Zevon8,

All SCRs are protected with snubbers whose values have been taken from a table and are very far from the dv/dt that they can withstand.

I have seen the perfect rectagular pulses applied to the opto couplers and they start with at almost 0° and expand to 180°. I know that this is not necessary in a 3 Phase rectifier, but it reaches both extremes. The gate current is the maximun permitted in the datasheet.

What I don't know is if its a normal condition for this configuration and in a fully controlled is different.

 

[Nigel Goodwin]

Thanks Zevon8,

All SCRs are protected with snubbers whose values have been taken from a table and are very far from the dv/dt that they can withstand.

I have seen the perfect rectagular pulses applied to the opto couplers and they start with at almost 0° and expand to 180°. I know that this is not necessary in a 3 Phase rectifier, but it reaches both extremes. The gate current is the maximun permitted in the datasheet.

What I don't know is if its a normal condition for this configuration and in a fully controlled is different.

I presume you are using three totally seperate drive circuits?, with each one fed from it's own phase's zero crossing point?.

Perhaps you should try posting your circuit?.

 

[IRQ57]

Hi Nigel,

You are right, my circuit has 3 zero cross detectors, 3 ramp generators, and three comparators between each of the ramps and one reference DC voltage, variable, following the algebraic sum of two sensing inputs.

Each time that the ramp crosses the reference, a driver is activated and feed an MOC random phase optocoupler .This happens following the phase sequence of the three phases.

Due to the presence of the 2 other phases it is not necessary to start firing from 0 degrees , in my first attempt I used 3 transformers connected in star to generate pulses lagging 30° the correspondent phase and I experiment the jump from 20-30 VDC to 150 and then continuolsy up to 300 . Then I modified the connection to delta to get control from 0 to 180 degrees, obtaining the same result.

Before to start a fully controlled version, doubling the number of 50A 800V SCRs and other less expensive components I am trying to get an expert answer to my fundamental doubt about the difference in the voltage output between a semi and a full controled rectifier with respect to the variation of the firing angle. And I am making reference to an ideal circuit, with a pure resistive load, a perfect sine wave, as that only viewed in the text books, and a perfect range of the firing angle.

Thank you so much for your support.

 

[Phasor]

It is my general understanding that there is no difference in voltage between a semi-controlled and a fully-controlled rectifier. I don't know what the advantage of a fully controlled rectifier is - maybe someone else can enlighten us?

 

[IRQ57]

Hi Phasor,

Here there is a brief description of these two different configurations for a 3 phase rectifier and some of their characteristics:

**broken link removed**

**broken link removed**

I hope this will show the differences.

Regards

 

[IRQ57]

Hi,

I found the origin of my problem in an error in a land pattern. Now in the semi controlled configuration, three SCR's and three diodes, it is possible to raise the output voltage from 0 to 300VDC smoooooothly without any kind of jumps, beautiful!!.

Many thanks for your interest in this case.

Closing problem.

 

[john1]

land pattern ?

 

[IRQ57]

Hi John,

Yes, a land pattern in the printed circuit that I developed to contain about 12 ICs and a lot of resistances, capacitors, diodes etc.

Errare Humanum est.

 

[john1]

Land Pattern.

I didn't find a clear explanation of what a land pattern is.
From wading through reams of stuff i get the impression that
it refers to the layout of pads which are intended for
soldering to a surface mount device, and left clear by the
solder mask for that purpose.

I have not met that term before, and i may have interpreted it
incorrectly. It may refer to the shape of the solder mask
that surrounds an area of pads for a surface mount device.

I have given up looking now.

John

 

[Nigel Goodwin]

I don't make PCB's, but I've always understood a 'land' was any contact on a PCB where you connect a component to, so a 16 PIN chip would solder to 16 'lands'. There's no requirement for surface mounting, and the term predates surface mount by a great many years - presumably back to the first PCB's?.

 

[zevon8]

It seems to be a generally applied term to any suface that is raised above the surrounding area. Rifling in a gun barrel is also refered to as "grooves" and "lands." CD's have pits and lands, ..... you get the idea,

 

[IRQ57]

Hi,

I am very sorry for the confusing term that I used. During my 33 years working as a technician in the computer field, "Land pattern" was used to describe the copper lines that interconnect componentes in a card, a board or a mother board.

All the manuals that I used , written in english, use this term to refer to the copper lines, so, up to this moment, I was sure of the meaning of this expression and I don't know another one. As it is evident my native language is not english, and in that condition I ask you to help me to replace "Land pattern" for something more conventional.

Thanks in advance.

 

[john1]

Hi IRQ57,

Please do not apologise, it is a perfectly acceptable term.
It is probably the correct term, even if i haven't met it.

I am glad that it was a simple matter to put right.

There are so many terms these days that it is not possible
to avoid them in electronics.

Your english is excellent, your profile does not have your
country listed.

Regards, John