Capacitive discharge welder

Thread Starter

gopalyajur

Joined Jan 3, 2010
100
Hello All,

I am trying to build a simple capacitive builder. Part of the circuit is shown as below
upload_2016-8-2_16-3-45.png

The gate drive gets signal from a microcontroller. The pulse time can be varied between 10 - 200 ms.
Capacitor bank consists of four 100 F , 2.7 V ultra capacitors connected in series. They are charged with a seperate power supply and is disconnected during welding trials.

When the gates of MOSFETs (IRLB3813PB) are turned ON I am not able to do any welding. However, I measure around 10 V at the electrode outputs.
Looks like there is not enough current for welding.

I tried drving with pair of IGBTs as well, but faced the same problem.

Can someone help me where the problem could be?

Thanks
 

MaxHeadRoom

Joined Jul 18, 2013
28,619
The Cap value seems very low, what material are you welding? Also what is the storage voltage?
You can't regulate the welding time after a full discharge.
Max.
 

jpanhalt

Joined Jan 18, 2008
11,087
There is probably no need to control the "off" function to end a weld. It is very quick. Have you considered using a high-current SCR like most of the hobby units use? I use considerably less capacitance (ca. 0.8 F) in my unit but can charge to at least 30 V. I have found 16 to 18 V does what I need to do, which is battery tabs and light fixturing of thin-wall SS tubing.

A slightly higher charge voltage gives more "headroom" for losses.

John
 

AlbertHall

Joined Jun 4, 2014
12,345
On your diagram there is no connection between capacitor negative and circuit ground. If it is actually wired like that then it definitely won't work.
 

MaxHeadRoom

Joined Jul 18, 2013
28,619
The capacity and peak charge voltage is going to be determined by the nature of the application, the industrial ones I have worked on use SCR discharge and very large switch selectable capacitor banks and around 200vdc.
For small fasteners using a projection on the stud.
Max.
 

Thread Starter

gopalyajur

Joined Jan 3, 2010
100
Thanks everyone for fast response.
I am trying to weld nickel tabs.

I connected four 100 F capacitors in parallel , which gives a total capacitance of 25 F!!! Is that not enough.
The caps are charged to 10 volts.

I even try ed with 3000F maxwell caps at 2.7V. With mosfet it does not work as well.

The problem with fast discharge is, it just punctures the metal surfaces.
 

MaxHeadRoom

Joined Jul 18, 2013
28,619
4x100μf in parallel is 400μf.
You mean series?
I see one guy on utube used a 1farad from an auto radio system, welding battery tabs for a home made CD welder!
Max.
 

tcmtech

Joined Nov 4, 2013
2,867
I am trying to build a simple capacitive builder.
Obviously not.:rolleyes:
If you were you would only have one capacitor, 500K uF - 1F at most, and a large SCR. :oops:

SCR would dump all the energy at once. Cap ground goes through mosfets.
And what's wrong with that?
If you used the correct size of capacitors along with the proper circuit design and layout it's not an issue so unless you're welding 14 ga plating you don't need 25F.

Nevertheless the circuit does not work the way it should
Probably because you're overcomplicating what is a well known and simple process and using the wrong type of components and layout for the job. :rolleyes:
 
Last edited:

jpanhalt

Joined Jan 18, 2008
11,087
Here is a detailed description of a small, CD welder: http://www.philpem.me.uk/elec/welder/

I built a welder along very similar lines and have been using it for the past 8 years or so. As I mentioned earlier, I used an SCR and a much smaller capacitor bank. Mine was only about 0.4 F and was made from twelve, 33,000 uF capacitors in parallel. That allowed me to use higher voltages. Here View attachment Capacitor_Comparison[1].pdf is a comparison I did on capacitors.

The other attachment is the completed welder.

Completed_welder[1].jpg
John
 

Thread Starter

gopalyajur

Joined Jan 3, 2010
100
Here is a detailed description of a small, CD welder: http://www.philpem.me.uk/elec/welder/

I built a welder along very similar lines and have been using it for the past 8 years or so. As I mentioned earlier, I used an SCR and a much smaller capacitor bank. Mine was only about 0.4 F and was made from twelve, 33,000 uF capacitors in parallel. That allowed me to use higher voltages. Here View attachment 109982 is a comparison I did on capacitors.

The other attachment is the completed welder.

John
 
Hello All,

I am trying to build a simple capacitive builder. Part of the circuit is shown as below
View attachment 109871

The gate drive gets signal from a microcontroller. The pulse time can be varied between 10 - 200 ms.
Capacitor bank consists of four 100 F , 2.7 V ultra capacitors connected in series. They are charged with a seperate power supply and is disconnected during welding trials.

When the gates of MOSFETs (IRLB3813PB) are turned ON I am not able to do any welding. However, I measure around 10 V at the electrode outputs.
Looks like there is not enough current for welding.

I tried drving with pair of IGBTs as well, but faced the same problem.

Can someone help me where the problem could be?

Thanks
I know this thread is quite old now so this response is mainly for others who may come across it, but I considered the same design recently and discarded it when I realised the ESR of those caps is not low, and adds up to quite enough to prevent the current ever getting high enough to produce a weld.
 
Top