inverter based on thyristor

Status
Not open for further replies.

oik

Newbie level 4
Joined
Jun 10, 2014
Messages
7
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Visit site
Activity points
49
hi
I am trying to built a twelve (or six) pulse inverter based on thyristor to use in High voltage applications
i use matlab but it doesn't work by universal bridge
i built it by GTO thyristor and it work correctly but i still need to built it by thyristor.

i really search hardly but i didn't find any help.
 

i use matlab but it doesn't work by universal bridge
i built it by GTO thyristor and it work correctly but i still need to built it by thyristor.
???
Isn't GTO one kind of Thyristors ?
 

Thyristor inverters can be rather simple implemented with current fed DC bus. Voltage fed bus doesn't work without a commutation means.
 

i use matlab but it doesn't work by universal bridge
i built it by GTO thyristor and it work correctly but i still need to built it by thyristor.
???
Isn't GTO one kind of Thyristors ?

yes but it work in opposite way than thyristor and have different features
i want it by normal thyristor type

Thyristor inverters can be rather simple implemented with current fed DC bus. Voltage fed bus doesn't work without a commutation means.

i use inductance after the DC source but it doesn't work
 
Last edited by a moderator:

I believe there are reasons why designers prefer turn-off capable switch elements for inverters. I must confess that I'm not particularly familiar with thyristor inverters. I've seen it in operation - far back in last millenium.

The thyristors are commutated with the help of other phases. So besides DC link characteristic, there's are also requirements for the load impedance.
 

A thyristor with forced commutation circuit can do the job of a GTO. But there is a minimum time requirement between ON and OFF state which should be considered also. See some pictures of forced commutation circuit below.



 

Is this an academic problem, or is it going to be built into an actual unit?

There are several reasons that forced commutation thyristor converters have been mostly superseded by GTOs or IGBTs: They were very bulky, expensive and extremely complex, the components themselves were special, and when failure did occur, the resulting event was absolutely catastrophic.

I worked for a company that designed and sold single phase forced commutation converters during the mid 1980s.
We were so glad when GTOs and IGBTs finally became mainstream, the circuit design was so much simpler.
 


it's an academic problem
and i chose the thyristor because we work with high voltage level about 500kv
 

Mitsubishi and others make extended gate IGBT's that can switch > 1kV and >1kA making >1GW switch which is not possible with Thyristors.

This academic problem will be old school with cheap Thyristors but complex pulse trigger and flyback snubbing with load SOA breakdown major issues due to slower switch off times with inductive loads that have large voltage during commutation at 0 current.

Get with the program on eIGBT's...
 

yes, actually i am trying to built an open loop hvdc system base on thyristor and i faced a problem with the inverter

Output of your inverter with conventional thyristors should be grid tied, otherwise it will not work without forced commutation.

The HVDC system in post #6 third link
https://www.mathworks.com/help/physmod/sps/powersys/ug/thyristor-based-hvdc-link.html
shows the example of like this


Here the output of the inverter is grid tied.

Try to upload the schematic/circuit diagram of your project.
 


what i find that i need a current regulator or any thing to keep the current constant and protect it from reach zero

is there any matlab tool can do this or how can i design such circuit
 

According to your system schematic, it seems that right side ac source (0.935pu) is a self generating three phase ac grid and receiving power from current source dc bus.

Firstly you need to understand natural commutation during power transfer from current source dc bus to single phase ac source using SCR. For an example, see the picture below.



If we trigger SCR1 & SCR4 at +30 degree and SCR2 & SCR3 at +210 degree (180+30) and so on, then at first half cycle power will be transferred from dc bus to ac source but after then the system will not work. SCR1 & SCR4 will remain ON all the time and SCR2 & SCR3 will remain OFF all the time.

But if we trigger SCR1 & SCR4 at -30 degree and SCR2 & SCR3 at +150 degree (180-30) and so on, then commutation will be done by ac source properly and the system will work. Power will be transferred from dc bus to ac source.

May be your triggering pulses are not like the second way.(-30 degree example)

So shift your all three phase triggering pulses backward (left side) and search for proper operation. (shifting x degree left is equivalent to shifting (360-x) degree right)

what i find that i need a current regulator or any thing to keep the current constant and protect it from reach zero

is there any matlab tool can do this or how can i design such circuit

Increasing the dc bus inductance will make the current more constant and protect it from reach zero. You should not need any current regulator.
 


i change the firing angle and it works correctly with DC voltage supply

but when i use the rectifier the current still reach zero and i try to increase the inductance but it still not working

i use parallel capacitor it prevent the current from reach zero but it not constant

any idea to help to keep it constant??
 

Perhaps you need feedback control on rectifier to maintain a constant DC voltage across DC bus capacitor.
 

I worked for a company that designed and sold single phase forced commutation converters during the mid 1980s.
We were so glad when GTOs and IGBTs finally became mainstream, the circuit design was so much simpler.

Schmitt trigger: Please tell me which inverter turned out to be more expensive, the IGBT version or the SCR version? How much should an inverter cost now (2014) when manufactured in bulk (say 1K units) from the point of view of business. M planning a small start-up, your help will be deeply appreciated.
 

Status
Not open for further replies.

Similar threads

Cookies are required to use this site. You must accept them to continue using the site. Learn more…