TL494 DCDC converter wide range input issues

Zac1

Member level 1
Joined
Mar 17, 2024
Messages
36
Helped
0
Reputation
0
Reaction score
1
Trophy points
8
Activity points
380
With a help of few people I managed to design fairly wide range DCDC converter using TL494

I confirmed it works in range of 30-190V DC
Has fairly stable 12V output with ~40mV peak to peak noise
Can deliver up to 3A.

However it has one fatal flaw that haunts me in my dreams. It kills mosfet Q5 and diode D2 if you suddenly apply 160V+ (high dV/dt)
It does work properly if slowly raising voltage across 40-190V. Also works up until 140V (sudden turn on).

I have managed to capture the problematic moment on thermal camera

(http://imgur.com/CHKR4Ew)
Basically there is a full short on Q5 and D2 for a moment, then fuse pops and both components have dead short. I have tried to remedy this situation with help of PNP clamp (Q1) but it doesn't do much. I am out of ideas how to fix this... perhaps the driver is at fault? Or maybe diode is too slow?


 

Solution
it needs to be speeded up / higher gain - until it becomes unstable - then back off a bit
I think this can be done after I order PCB, right now i need to order asap to have something to play with.

both soft start and/or dc gain can be adjusted with just cap/resistor values.


Why is the switcher driving high current into D3 even in steady state although the nominal output voltage is said to be 12V? There seems to be a problem with voltage feedback operation point.
I am sorry, i did not understand question? D3 after startup should see a max of 52micro amps.


As already stated, soft start time constant should be much higher, I'd test with 5 or 10 uF C5.
Will do, right now i think i have complete PCB that should...
In your sim - look at I_L1 or I_fet, M1 - this tells you real information about the start up !

I imagine your choke is saturating at start up at the higher Vin, ( > 36A into zener ) and this is what is killing the fet etc,

time to make the soft start softer and longer.
 
Last edited:

    Zac1

    Points: 2
    Helpful Answer Positive Rating
In your sim - look at I_L1 or I_fet, M1 - this tells you real information about the start up !
This is at 100V


And at 180V


If i increase dead time to like 1000 cycles by replacing cap C5 (100n -> 1u) i get swomething like this at 180V:


Should i still increase max dead time?
 

For any one trying to run the TL494 model in LT spice - there are a host of issues with the model and LT spice's changing of internal workings

covered in good detail here: ( read to the end )

--- Updated ---

What is the max current your inductor and diode can handle at start up ? is it ok for the rest of the ckt if the start up is slow ?

if yes - make it slow !

The main issue is - the model is not accurate for the buck inductor - above a certain current - it saturates and the current really peaks up ( 100 amp easily ) causing issues for the fet and diode

The Duty cycle for 12V out and 190V in is 12/ 190 = 6.3 % allowing the Duty cycle to go way over this at start up from 190V is asking for trouble - hence a slow soft start that allows the control loop to " catch " the duty cycle and Vout before it overshoots too much is kind of a big deal !
 
Last edited:

    Zac1

    Points: 2
    Helpful Answer Positive Rating
For any one trying to run the TL494 model in LT spice - there are a host of issues with the model and LT spice's changing of internal workings
Thanks for the link, i did replace my sub model. Didn't change much probably because i am using mosfet driver which doesn't really care about voltage on the output.

What is the max current your inductor and diode can handle at start up ? is it ok for the rest of the ckt if the start up is slow ?

if yes - make it slow !
There is no minimum startup time requirement so i can make it as slow as I like. Even if it starts up after like 1 second.

Yes i understand that. Right now i have in sim like 1000 cycles do you suggest to make it even slower? (ie increase C5) or max dead time?
 

another part of the problem is that the response of the error amp is very slow - so after the Vout crosses 12V it takes a while for the error amp to respond - this results in some over shoot - it needs to be speeded up / higher gain - until it becomes unstable - then back off a bit

at least root cause of the problems has likely been found - the excessive current in the ckt at start up causing inductor saturation.

p.s. I have found the LT spice model for the TL494 ( the improved one ) very poor indeed when it comes to simulating the error amps !
 
Last edited:

Why is the switcher driving high current into D3 even in steady state although the nominal output voltage is said to be 12V? There seems to be a problem with voltage feedback operation point.

As already stated, soft start time constant should be much higher, I'd test with 5 or 10 uF C5.
 

it needs to be speeded up / higher gain - until it becomes unstable - then back off a bit
I think this can be done after I order PCB, right now i need to order asap to have something to play with.

both soft start and/or dc gain can be adjusted with just cap/resistor values.


Why is the switcher driving high current into D3 even in steady state although the nominal output voltage is said to be 12V? There seems to be a problem with voltage feedback operation point.
I am sorry, i did not understand question? D3 after startup should see a max of 52micro amps.


As already stated, soft start time constant should be much higher, I'd test with 5 or 10 uF C5.
Will do, right now i think i have complete PCB that should be enough for testing.

You can zip the .asc file and attach next time.
Which ASC file? of entire SIM?

EDIT:

Just a quick question, does it make sense to add fuse between VBAT+ and entire DCDC + side?
 
Last edited:

Solution
Save schematic as ASC file. then zip

Yes a fuse or Polyfuse.
 

I'm sure you can decide based on the thermal time constant.
Unfortunately i cannot decide because the moment we are talking about high voltages as well as (kinda) unknown current spikes I simply don't know which fuse to use.
 


    Zac1

    Points: 2
    Helpful Answer Positive Rating
mount near Q5

hold 1A trip 2A if that is enough
Thanks, but that is what i was worried about... they are all THT. Either i will have to get really creative to mount them or completely redo the layout.

EDIT:
Ok its doable, may need some leg bending but should fit with some small changes
 
Last edited:

the PPTC won't trip or protect FET from inrush anyways, an NTC ICL helps with that but dissipates heat running.
Yes I am aware of that, I was just wandering it it makes sense to add any kind of fuse in case some short happens or something.
 

Last edited:

perhaps OTP is more prudent. this can be a thermistor on a wire adhered to FET or SMD IC with latency.
I am bad with "three word shortcuts" so I'll assume you mean over temp protection but that is not going to ever be a problem in long run. According to my tests it never gets hotter than 32 deg in 23 ambient with no cooling whatsoever.
Short run, it will be too slow to react to 0.1s internal combustion.

Overvoltage is going to cause significant damage to rest of the circuit so PSU is last thing to worry about.
Overcurrent is one of the things i may be a little concerned about because in one of my tests i managed to "half burn" MCU and that resulted in PCB getting very hot, drawing more than 2A. So I do hope that fuse may prevent some catastrophic failure in case of dead short somewhere on the logic board.

Cost is irrelevant in this case, all those protection doesn't even add to 1% of total cost but size is. I am nearing the limit of what i can fit on this PCB.
GaN Mosfets for ESC alone cost 126$
 

Also a 2 or 3 A fuse will protect your circuit is it plays up some more under test ( or you build it wrong before test )
 

Similar threads

Cookies are required to use this site. You must accept them to continue using the site. Learn more…