TNY285K Efficiency
Hello PI team,
I have design a product with flyback AC/DC switching power supply based on TNY285K device, universal input 85Vac to 265Vac, output 5V.
Transformer is a 093.830 from KASCHKE, and BP/M capacitor value is 1µF, so the device is used as current limit smaller device.
Let see the schematic in attachement.
During the validation tests I have drawn the efficiency curve @ 250Vac input by increasing the power load from 50mW to 7W and decreasing from 7W to 50mW.
In attachement, let see the result.
As you can see, their a 'step' of efficiency (48% to 68%) between 1,5W and 2,5W when the power load is increased and between 1,5W and 0.5W when the power load is decreased like an hysteresis effect.
What do you think about this result ?
I think is a strange effect, that bothers me becaus I use this power supply under 1,5W at 99% of time.
I have also try with BP/M capacitor value 100nF and 10µF and also without drain snubber C102, R100, D101 => I obtain the same strange effect.
Input Precision Power Meter : ZIMMER LMG95 (calibrated on 2018-04-24)
Output ampermeter : FLUKE 179
Output voltmeter : FLUKE 179
Output Electronic Load : TTi LD300
Thanks by advance for your answer.
Mickael.
Comments
Hi Jedidiah,
Thanks for your fast answer.
If I well understand, the bad efficiency (~55%) between 0,5W to 2,5W is normal for you ?
It bothers me because I expected to reach around 70% of efficiency after 0,5W...
Would it be possible to improve the efficiency in this area (upper 0,25W) to reach faster 65-75% efficiency ?
Regards,
Mickael
Hi Mickael,
Can you send to me the specs of the transformer you used in your design? Also, do you have PCB layout of your circuit? Let me check how we can improve the efficiency.
Regards,
Jedidiah
Hi Jedidiah,
Many thanks for your answer and your help.
In attachement you can find :
- Transformer datasheet
- PCB layout
- BOM
Regards,
Mickael
| Attachment | Size |
|---|---|
| Transformer Kaschke 093830 datasheet (45.75 KB) | 45.75 KB |
| PCB Buttom (6.28 MB) | 6.28 MB |
| PCB Top (8.39 MB) | 8.39 MB |
| Board Top (1.86 MB) | 1.86 MB |
| BOM (356.75 KB) | 356.75 KB |
Hi Mickael,
Here are several ways we can try to improve the efficiency of your circuit:
1) With your current schematic, remove C102, R100, and D101. Change your snubber circuit in drain to source from DT1 and US1M combination to an RCD clamp circuit.
2) Another way to improve the light load efficiency is to lower down the switching frequency but you might need to change the transformer to a bigger core to avoid saturation.
Regards,
Jedidiah
Hi Jedidiah,
.
Thanks for your last answer.
So I've done the folowing tests (please let see the efficiency curves in attachement)
.
- In red : the curve of the original efficiency (Vin = 250Vac, Transformer = Kaschke 093.830 and drain RC snubber according to the schematic)
- In blue : the curve with RC snubber (C102, R100, D101 removed) replaced by an RCD snubber instead of DT100.
===> Efficiency is improved but not so much. Please note that the maximum output power is decreased...
- In orange : the curve with another transformer Kaschke 094.930 (datasheet in attachement) and RCD snubber.
===> Efficiency is clearly improved at low power but also degraded between 0,5W and 2W. Please note that the maximum output power is dramaticaly decreased...
.
What do you think about this result ?
Do you think that the solution coul'd be to contact the transformer supplyer and find a best transformer ? (No more schematic effect or chip TNY effect)
.
Regards,
Mickael
| Attachment | Size |
|---|---|
| Efficiency comparison at 250Vac (33.05 KB) | 33.05 KB |
| Transformer Kascke 094930 datasheet (46.1 KB) | 46.1 KB |
Hi Mickael,
Yes. We could iterate different transformer with bigger Core Area for efficiency.
Regards,
Jedidiah
Hi Jedidiah,
.
Thanks a lot.
Yes could you please let me know if you know some transformer standard references of any manufacturers to improve the efficiency on my project ? :)
.
Best regards,
Mickael
Hi Mickael,
We usually construct our own transformer based on available core and bobbin in the public market. Some customers are able to customize their own core or bobbin because of their mode of business is on power supply industry.
You may try to use our website to optimize transformer design. You may find in the options core and bobbin to use.
https://piexpertonline.power.com/site/login
Regards,
Jedidiah
Hi Mickael,
Whenever you get the efficiency from low to high then high to low, the current limit (Drain Current) at light load will be different. TinySwitch varies at four current limit.1,2, 3, and 4 current limits. When you start from light load to full load, it uses the current limit gear 1,2, and 4. While if you start from full load, your current limit will be from 4,3, and 1. Note: 1 as the lowest current limit and 4 as the highest.
Regards,
Jedidiah