Inverter building using Wiseguys Power board and the Nano drive board


Author Message
KeepIS

Guru

Joined: 13/10/2014
Location: Australia
Posts: 1768
Posted: 08:58am 10 Jun 2024      

I decided to go back and check the Ground plane on my Big Inverter, this has been running The Nano controller for a few weeks 24-7, we are only on mains when I do some Modification to the Inverter.

About a week ago, one of those modifications was to add Ferrite beads to the Power board and fit a 15V isolated module to enable -3v gate bias, and also halve the value of the Gate resistors.    

Today I revisited the Ground plane installed with the Previous controller board.

Checking the interference on the board was pretty good, but what I hadn't had a chance to do was check the level of noise on the incoming leads, well this time that was a bit eye watering.

This has Nothing to do with the Nano Controller or the Power board, it has been a "sometimes" overlooked problem with previous inverter builds going back to day one.

The way the Controller board is designed with a ground plane setup in mind is new to most DIY inverter controllers.

The noise on the connectors CON6 Pins 2,3,4,5 (LEDS) and J1 pin 1 Inverter Off, was a few volts of fast transients. Basically any wire running through the inverter and coming back to the controller can bring some noise with them.

I tested a 0.1uf across the offending pins and all responded. Some pins should have a much smaller value around 0.01uf for discharge / timing reasons.

Note Do not bypass Sensor leads (AC VFB or Current Sensor) with the exception of the Thermostats, where I used a 0.01uf.

I removed the controller board, fitted suitably voltage rated bypass caps from the Sockets Pins under the board to ground and retested.

This was easy to do and worth the effort, noise came down into the 50mv P-P region.

My 48V supply input is quite clean and did not need a bypass, a couple of connections already a 0.1uf near the pin.

Another result was a slight Idle offset that I had on AC current dropped by half.

All readings were taken with respect to the Controller board ground, I have a pin that I fitted to controller board for testing.

Do not connect anything (I mean test lead ground clips) to the ground plane!  

I hope the above makes sense.
.
Edited 2024-06-10 19:23 by KeepIS