Solution Finder Get Tech Support

TNY290PG - OUTPUT LOADING ISSUE

Posted by: anurag lokhande on

Comments

Submitted by PI-Wrench on 01/19/2022

PI Expert/PIXLS is a reliable tool for transformer design.

In order to properly test the transformer once it is constructed, you will need some sort of inductance meter. These are readily obtained from online vendors.

Also, in order for the flyback transformer to be able to store energy during the primary switch on-time without saturating, the core will need to have the proper air gap. The estimated gap required is shown in PI Expert and PIXLS. The gap is normally introduced by grinding away a small portion of the center leg of one of the transformer core halves. However, the gap can also be implemented by using a thin non-conducting shim in between the core halves. Some materials that can be used are thin mylar sheet, paper, or the same electrical tape used to construct the transformer. If you use a shim on the outer legs of the transformer to introduce a gap, keep in mind the the gap will be twice the shim thickness, as you are interrupting the magnetic path twice -  at the center leg as well as the outer legs of the E core. 

You can tune the transformer design to work with available shim materials by playing with the secondary turns in PIXLS until you get a gap that is twice the thickness of the available shim material. 

Submitted by anurag lokhande on 01/20/2022

Dear PI-Wrench, Thank you for your reply.

-------------------------------------------------------

As you suggested in the previous thread

Below is the suggested transformer design using PI XLS:

Primary Winding Turns: 61

Secondary Winding Turns: 7

Bias Winding Turns: 13

Lpri: 1187 uH

(I don't have this transformer)

-------------------------------------------------------

Primary Winding Turns: 120

Secondary Winding Turns: 12

Bias Winding Turns: NA

Lpri: 1200 uH

(This is tested one and working fine for 12V 2.4A with input 100VAC to 250VAC)

-------------------------------------------------------

Here I want to ask, both transformers should work for the same design as Lpri is same? Am I right?

-------------------------------------------------------

I did all calculations using good application notes. But it went wrong. 

Can you please give reference to good app notes that will really help?

-------------------------------------------------------

How to download file you had attached.

Attachment    Size

TinySwitch-4 Design1.pixls    58 KB

-------------------------------------------------------

Submitted by PI-Wrench on 01/20/2022

I tried rewinding the design from the start using PIXLS, with the understanding that you want a design with a universal input range. The first thing I saw was a warning  message that the TNY290 does not have sufficient capability for the job at hand. This limitation will likely show up at low input voltage, where the part is pushed hardest. The design does work with 20W output as was specified in the design example that you cited in the previous thread.. If you really need a design with 12V, 2.4A (29W) capability, I think you would need to choose a PI device with more power capability - my suggestion would be a device from the TOP-JX family. I am attaching a PDF of the spreadsheet I ran for this design using the TOP266E.

Submitted by PI-Wrench on 01/21/2022

I would suggest that you use PI Expert or PIXLS as a design aid, as all the corner conditions are taken care of. It is possible to do the design by hand calculation, but the characteristics of the part and variations in voltage, current limit, etc, will not be taken into account. Also, very few of our parts use conventional  fixed-frequency PWM over their entire operating range. As a particular example, the Tiny Switch uses pulse skipping rather than straight PWM to achieve regulation.  Using PI Expert or PIXLS takes into account the operating characteristics of the device in question, and the boundary conditions. Trying to achieve all of this by hand calculation is tedious and prone to error. 

Submitted by anurag lokhande on 01/22/2022

Hi PI, Thank you for inputs you have given. I have noted.

But why transformer I used, working with 1200uH and transformer designed using PIXLS, it gives 355uH and this is not working for me on board. I checked.

 

 

 

Submitted by PI-Wrench on 01/24/2022

The lower inductance transformer works with the TOP266 because the allowed primary peak current is higher than for the TNY290. If you try to extract excessive power from the TNY290, the spreadsheet tries to compensate by dialing in an excessively continuous design. Keep in mind also that the spreadsheet takes into account the spread of peak current limit values in a given population of parts, so that the transformer will work with all instances of a given device. You may be working with samples that have a current limit on the high range of the distribution.

Submitted by anurag lokhande on 02/06/2022

Hi PI-Wrench, 

Thank you for your continuous support here. I really glad to meet you.

I placed everything well now. 

Please see attached waveforms of converter, it shows some noise pockets. What can be the reason for this? 

My observation --> Transformer making very low audible noise. Can it cause this results? 

Can I get an evaluation board app note from POWER INTEGRATIONS, that already built and working well?

How to avail the evaluation board for testing from POWER INTEGRATIONS?

 

 

Attachment Size
POWER SUPPLY 12V 2A OUTPUT WAVEFORMS.pdf 529.05 KB
Submitted by PI-Wrench on 02/07/2022

If you are looking at the output ripple using a standard scope probe and ground clip, you will experience spuriously high output ripple due to the loop antenna effect of the probe ground lead. After all, the oscilloscope is a sensitive high impedance receiver of sorts. Please try using the setup detailed  in the attached files for ripple measurement. This is a standard setup used commonly in the power supply industry to reduce spurious pickup when performing output ripple measurements.

 

 

Attachment Size
Ripple_measure.pdf 275.67 KB