Home
JAQForum Ver 24.01
Log In or Join  
Active Topics
Local Time 07:37 28 Nov 2024 Privacy Policy
Jump to

Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.

Forum Index : Electronics : Inverter # 4

     Page 6 of 6    
Author Message
Tinker

Guru

Joined: 07/11/2007
Location: Australia
Posts: 1904
Posted: 09:50am 30 Oct 2018
Copy link to clipboard 
Print this post

Some more tinkering with this inverter.

Gaspo was at my place a while ago testing his neat little EG8010 Adapter board with the interlocking transistors on it - it worked very well BTW.

He brought his digital scope to check things and showed me that there were big voltage spikes on the 12 & 5V rails on my control board. I had not noticed these on my ancient 35MHz analog CRO before as they were running at 23KHz and extremely narrow but peaking at the rail voltage.
Thinking that these spikes might stress components unnecessarily I started to investigate how to reduce them.

First trying the DC filters that were inside the Aerosharp box (shiny metal case with blue & red wires coming out and two others that can be cut out of the large PCB).
These did not show much difference when spliced into the 12V power leads, perhaps my wires were too long.

Next I increased the 12V rail filter capacitor which did the trick.

My control board is different from the OZ version as it uses a DC/DC converter to get the 12V from the battery voltage and then a 5V regulator to get the 5V.
I had a 1000uF cap at the 12V rail and quite a few strategically located 10uF caps and 0.1uf cer disks on the 5V rail.

That 1000uF was increased to 4700uF, the biggest size I could fit on my board. For good measure I also soldered a 0.1uF cer disc cap directly across the 4700uF cap's terminals on the other side of the PCB.

Now those spikes are only about 1V peak to peak, good enough for me.

For those wanting to investigate that, the little kit digital CRO has trouble showing those spikes. On my analog CRO I had to crank up the brightness and set the time base to 20us to see them - just. Its a very narrow pulse with some ringing after it.

I notice the Ozinverter (as per tinyt's drawing) has only a 100uF filter cap on the incoming power rail. Curious to know if these spikes are on those boards too. The board obviously works, as does mine with the spikes on the low voltage rails.
But prevention is always a good idea.
Klaus
 
noneyabussiness
Guru

Joined: 31/07/2017
Location: Australia
Posts: 513
Posted: 10:48am 30 Oct 2018
Copy link to clipboard 
Print this post

Klaus, different power supply, linear power supply's (which the ozinverter use) although being " inefficient " are notoriously quite electrically. .so 100uf would be plenty, however switching supply's have a tendency to be " noisy ".
I think it works !!
 
Tinker

Guru

Joined: 07/11/2007
Location: Australia
Posts: 1904
Posted: 03:47pm 30 Oct 2018
Copy link to clipboard 
Print this post

  noneyabussiness said   Klaus, different power supply, linear power supply's (which the ozinverter use) although being " inefficient " are notoriously quite electrically. .so 100uf would be plenty, however switching supply's have a tendency to be " noisy ".

Yes, but the noise is at the EG8010's high side frequency, not from the the DC/DC converter unless that one runs at exactly the same frequency - I will double check that tomorrow with my spare converter.
If you have a suitable CRO do take a look to confirm the "quietness" .

Edit: Yes, this noise spike originates from the Mosfet drive - not the DC/DC converter.Edited by Tinker 2018-11-01
Klaus
 
Tinker

Guru

Joined: 07/11/2007
Location: Australia
Posts: 1904
Posted: 09:30am 31 Oct 2018
Copy link to clipboard 
Print this post

A comment by Poida had me wondering, can a single HY4008 handle switching 50Amps in my inverter?
Easily checked, set the thing up with just 4 HY4008's and 4 x 10000uF capacitors.

2350W load - 51.5A at the input cable - and it ran that no trouble at all. Equating that to a set of 24 HY4008's would mean they could run over 14Kw of load - way more than the dual stack toroid core is designed for. Perhaps just 20 HY4008's would be plenty for such an inverter?

While I had all this set up I did some heating measurements with my infrared thermometer, a type which uses two red laser points converging into one for the correct distance.

I let the inverter run until the EG8010 controlled fan (TFB) turned on.
The hottest mosfet leg would be the source as the drain current goes via the heat sink.
Source leg, wide part: 66 degrees
Source leg narrow part: 76 degrees - the conductor area difference is 1mm sq to 0.8mm sq.
The screw terminal was also at 76 degrees, I'm sure they can handle at least twice that by which time the mosfet probably has died.

The heatsink temperature was 48 degrees at the top.

Next some testing to check if the reported TFB and IFB related blowouts had disappeared with the new 8010 Adapter board with interlock transistors on it.

Removed the NTC sensor from the heat sink and heated it by soldering iron.
Fan turned on at about 50 degrees, further heating shut the EG8010 down - 5 flashes on the status LED - (could not measure the exact temp. as that required 3 hands ).

The fan kept on running during the shut down. The EG8010 soft restarted after removing the heat source and the NTC had cooled a little.
No drama with that re starting, this could can happen fully automatically.

Then some IFB testing with a shorted turn of a beefy transformer secondary through the current sensor.
Over current shuts the inverter down by tripping the SCR. Fault light flashes then (on my control board). Two flashes on the status LED.

Reset button must be pressed (after the over current is reduced) to soft re start the 8010.
It did that many times with no drama.


Klaus
 
     Page 6 of 6    
Print this page


To reply to this topic, you need to log in.

© JAQ Software 2024