Home
JAQForum Ver 20.06
Log In or Join  
Active Topics
Local Time 17:21 29 Mar 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 : Transformer choice for 300V Warpverter

     Page 2 of 2    
Author Message
genetrySid
Newbie

Joined: 05/09/2020
Location: United States
Posts: 8
Posted: 09:09pm 29 Dec 2020
Copy link to clipboard 
Print this post

So I have to say that all 3 inverters that burned up were stock PJ inverters (with my upgraded CPU firmware)...but their FET drivers are lousy at best.  Mismatched and uneven at the root (how about 15v for both H FETs, 12v for one L FET and 10v for the other L FET...not to mention using 2 completely different driver topologies between the H FETs and L FETs).  

On the 2 inverters that blew up at 68v, the FETs were (6x/channel) HY3810s, which are 100v FETs...right about at that 150% mark.  However, when I've checked the waveform at the FET outputs (with respect to battery negative), I have not observed any splashback--so to me, this would indicate that the FET voltage limit is not being exceeded.  But to be fair, I haven't had the patience to watch this waveform when the FETs blow...
The other inverter that blew up at 40v had (6x/channel) 60v FETs (again, 150%).  Guess about the only point I am making for myself is that the FETs voltage rating must be over 1.5 times the battery input voltage??

Yes, an inductor/choke is always between the FETs and tranny--boy, talk about making a world of a difference to the no-load current ;-).  I've noticed that 2 turns around a stock PJ choke is usually about the best that can be attained--using more chokes and/or more turns does not help the no-load current at all.  The test bench inverter (with my new control board) has a total of 4 donut chokes on the tranny (1 turn each), and the no-load current remains the same from 20v to 40v.  Haven't taken a 'scope to it, though you might be onto something there.

I have not yet done an abusive test of this sort with my completely new designed control board (using a completely balanced 14.7v across all 4 channels, and fully identical 4.0A drivers), so I guess I can't say I have had a blowup of that fashion on anything except a stock PJ inverter...which really is not a point at all   .
 
Warpspeed
Guru

Joined: 09/08/2007
Location: Australia
Posts: 4406
Posted: 09:15pm 29 Dec 2020
Copy link to clipboard 
Print this post

Quite right Nick.

Lots of possible contributing factors to higher idling power with higher dc voltage.
If the transformer is being fed with a pure 50Hz sine wave of constant amplitude on the primary, the magnetising current should remain the same as dc input voltage changes.

But... The mosfets themselves are (or should be) switching into a series choke, and may be having to charge and discharge significant capacitance at 23 Khz.  That is reactive power and does not generate heat in itself, but that power sure must come from the dc supply. Energy stored in a capacitor goes up square law with voltage, so more volts means a lot more power required to charge and discharge stray capacitance at 23 Khz.

Toroids are particularly bad for capacitance down to the core.  Something to think about...

As Nick says, mosfet switching losses will increase with increasing voltage too.

There may also be some linear voltage regulators driving fans or LED displays. The current will not vary, but the regulators themselves may have to dissipate more power as the dc input voltage rises.

All kinds of things are possible and can be speculated about.
Only way to know for sure is an in depth complete power audit.
Find out where all the extra power is actually going.
Cheers,  Tony.
 
Warpspeed
Guru

Joined: 09/08/2007
Location: Australia
Posts: 4406
Posted: 09:25pm 29 Dec 2020
Copy link to clipboard 
Print this post

  genetrySid said  
Yes, an inductor/choke is always between the FETs and tranny--boy, talk about making a world of a difference to the no-load current ;-).


The main effect of the choke is isolating the very fast high frequency square waves generated by the mosfets from the transformer. And the reduction in idling power is a direct result of putting some series impedance between the very high stray capacitance of the transformer, and the mosfet switching bridge.

It took me a very long time to convince people that a choke was necessary, but I think that battle is now won.
Cheers,  Tony.
 
genetrySid
Newbie

Joined: 05/09/2020
Location: United States
Posts: 8
Posted: 09:26pm 29 Dec 2020
Copy link to clipboard 
Print this post

Thanks for the suggestions; I'll have to dig into this with the info.  Mostly wanted to make sure there wasn't a glaring black hole that power was going into--because when 8kw of power gets mis-directed, smoke is usually very quick to follow.
 
     Page 2 of 2    
Print this page


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

© JAQ Software 2024