Limiting power inrush on spot welder

[Kric]

I have a spot welder that works well for some applications, but it's too much for some welds using lighter material. Manually triggering the welder on/of as quickly as I can is too much for the material I'm using. So I built a timer unit that uses an SSR connected to an Arduino unit. I added POTs and a display such that I can set a weld time from .05ms to 6 seconds, a foot pedal to trigger the cycle, fan for cooling and a fuse to protect everything. My idea is to plug the welder into the timer and force the welder toggle to the on position. The problem is that the inrush is so big, the power to the fan sags, and it it's more than 1 second it trips the breaker, but doesn't blow the fuse.

If bypass the timer and plug directly into the outlet, the welder works as expected. If I set the weld duration for 6 seconds, and keep the switch on the welder off, press the foot pedal to send power to the welder, then manually turn it on for a few seconds that seems to work OK.

It's almost like the toggle switch on the welder somehow limits the inrush as it's turning on. Locking it to the on position and controlling the power through the SSR seems to cause a huge inrush which isn't controlled. I added an Negative Temperature Coefficient (NTC) thermistor, but that didn't do much.

I'm thinking that I could program the Arduino to do coarse bursts to the SSR (i.e., have it turn on for 50ms, then off for 50ms) for the duration of the weld. Obviously that may reduce the effectiveness of the welder, but I can get around that by increasing the weld time. Am I on the right track, or did I have a fundamental design flaw?

Any advice or opinions are greatly appreciated.

 

[Harald Kapp]

One possible explanation:
The power supply of the welder very likely has some big capacitors on the input side (possibly some more on the output side) which are responsible for a high inrush current when primary (mains) power is turned on.
Try switching the output side while the primary side (mains) is permanently connected. This way the capacitors will supply a big portion of the weld current and the breaker on the mains side will not be triggered by the peak weld current on the output side of the welder.

 

[Minder]

If this is a simple transformer type resistance welder, then you will not be able to reduce the inrush, this is the whole concept of the type of welder, the secondary is usually many thousands of amps, even for fairly small one,

 

[Harald Kapp]

If this is a simple transformer type resistance welder

Probably not. It's a spot welder, according to the op. These are afaik typically DC and controlled by a bunch of MOSFETs on the output side to achieve the timing.

 

[Minder]

Most of the spot welders I have worked on have been simple mains-transformer constructed, with a single-turn secondary .
There has been the odd capacitive discharge type, I must admit.
The disadvantage with the latter is the weld length duration cannot be controlled.

 

[Kric]

I think you're right. There has to be some build up on the input side prior to activating it. By limiting the power input externally there is no chance for a capacitor to charge, hence the huge inrush when it is activated. I'm thinking about a slightly different approach which is to add a current sensor module. I'll leave the power at the SSR on by default. When the welder is triggered the current sensor should signal the start of the weld. Once time expires I can kill the power to the SSR. A bit of modified code to have the foot pedal reset the SSR back to the default ON position. Any thoughts on that approach?