TNY290 going into fault mode well below rated current limit[SOLVED]

Thread Starter

jlawley1969

Joined Feb 22, 2021
97
Hi,
I am using a tny290 (as shown in schematic) I have it on protoboard. It works fine with no load as well as up 10Ohm load (about 0.5A) but above that it is going into fault mode as if I am short circuiting it or something. The tny290 is supposed to be rated up to 1360 (2550) mA so why am I getting this issue at such a low load current?
 

Attachments

jlawley97

Joined Oct 5, 2019
25
A lot depends on the transformer you are using.

Also which package are you using and how big a heatsink?
ok so it works with the 750871110(1600uH,48uH leakage, 14.4:1 turns) but not wit the 760871113(1300uH,40uH,~21:1 turns)
i figured they were similar enough to work
the 750871110 fits the bill electrically but is just too big.
do you reckon there is anything I can do to the 760871113 to make it more "accepting"

right now I just havea fan blowing on the board and also nothing is really getting how anyways.
 

ronsimpson

Joined Oct 7, 2019
2,988
There are many things that could be happening.
1) the duty cycle hit the limit. Can you test that.
2) current is at limit. (could be core saturation on the smaller core)
I need to run. More later.
 

Irving

Joined Jan 30, 2016
3,845
do you reckon there is anything I can do to the 760871113 to make it more "accepting"
Not sure, need to run the numbers, but as Ron said, core saturation coud be your enemy here.

right now I just havea fan blowing on the board and also nothing is really getting how anyways.
Might not seem that way, but internally could be going into current limit. Depends which package type and how mounted. Your fan might be having no useful effect.
 

Thread Starter

jlawley1969

Joined Feb 22, 2021
97
Not sure, need to run the numbers, but as Ron said, core saturation coud be your enemy here.


Might not seem that way, but internally could be going into current limit. Depends which package type and how mounted. Your fan might be having no useful effect.
Core saturation current is supposed to be 0.8A. for the 760871113

I am currently using the "P" package (DIP-8C) so no exposed bottom pad. Do you suppose it is reaching internal current limit due to insufficient grounding?
 

Thread Starter

jlawley1969

Joined Feb 22, 2021
97
There are many things that could be happening.
1) the duty cycle hit the limit. Can you test that.
2) current is at limit. (could be core saturation on the smaller core)
I need to run. More later.
The max duty cycle is about 62% and so would that mean that the turns ratio of like 23:1 for the 760871113 would be too much?
 

ronsimpson

Joined Oct 7, 2019
2,988
If the IC can not make more than 62%, what is the duty cycle you see? Yes I am thinking turn ratio problems. I did not do the math. Do you have a scope?
 

Thread Starter

jlawley1969

Joined Feb 22, 2021
97
If the IC can not make more than 62%, what is the duty cycle you see? Yes I am thinking turn ratio problems. I did not do the math. Do you have a scope?
i have a hand held scope where I can see the shape and voltage/time divisions. So I will freeze the signal and see do the calculation for duty cycle in a moment. What formulas would you use to calculate the proper turns ratio because 1) i dont want you doing the calculations when I can do it myself and 2) Ive tried to find a solid formula that explains all the variables but alot of application guides either don't say what all the variables mean explicitly or are formulated for a specific chip in a specific conduction mode. As well as the logic of the formula not being consistent across the guides
 

Irving

Joined Jan 30, 2016
3,845
I am currently using the "P" package (DIP-8C) so no exposed bottom pad. Do you suppose it is reaching internal current limit due to insufficient grounding?
Its possible.

Thermal resistance, Tja, for the P package is 70degC/W when all 4 S-pins are soldered to 232mm^2 of 2oz copper (70um thick). You have virtually no heatsink so easily going to be 3 times that, lets say Tja =200degC/W. From the datasheet, Rds(on) for the TNY290 is around 3ohm typically.. TNY290 will go into current limit at junction temperature of 142degC (with a 75degC hysteresis) so assuming ambient is 25C the max dissipation Pd is:

Pd = (Tjmax - Tamb)/Tja = (142-25)/200 = 0.59W​

but Pd is also:
Pd = Id^2 * Rds(on) * D, so re-arranging​
Id = √(Pd/(Rds(on) * D), Id = √(0.59/(3 * 0.62) = 0.56A which is much lower than output capability suggests.​
 

Thread Starter

jlawley1969

Joined Feb 22, 2021
97
Its possible.

Thermal resistance, Tja, for the P package is 70degC/W when all 4 S-pins are soldered to 232mm^2 of 2oz copper (70um thick). You have virtually no heatsink so easily going to be 3 times that, lets say Tja =200degC/W. From the datasheet, Rds(on) for the TNY290 is around 3ohm typically.. TNY290 will go into current limit at junction temperature of 142degC (with a 75degC hysteresis) so assuming ambient is 25C the max dissipation Pd is:

Pd = (Tjmax - Tamb)/Tja = (142-25)/200 = 0.59W​

but Pd is also:
Pd = Id^2 * Rds(on) * D, so re-arranging​
Id = √(Pd/(Rds(on) * D), Id = √(0.59/(3 * 0.62) = 0.56A which is much lower than output capability suggests.​
I finally got around to soldering to a vector board and it did not seem to help
 

Thread Starter

jlawley1969

Joined Feb 22, 2021
97
how big? what type of board?
63mmx93mm "mbss protoboard"
since I last posted here I was able to get up to 5V 0.7A and so figured the next step would be going to the protoboard
I bridged all the source rails together to give it a bigger sink
it didnt work any better or any worse
 

Thread Starter

jlawley1969

Joined Feb 22, 2021
97
...It was the TVS clamping diode
as labeled (poorly)in the original schematic, my TVS diode had a breakdown of 200v and a max clamping of 274V. This was a P6KE200.
I went down to a P6KE150 and I am able to charger at 1Amp even with the smaller transformer

The reason I was using the P6KE200 was an artifact from when I was working with the TOP258. I only now just received the 150 in the mail.

I will most definately be posting in the near future but as this thread, this problem has been solved. Thank you everyone and Ill remember to label my schematics more thoroughly in the future.
 
Top