UC3843: Heating observed at 24V

Hello,

Normally, more pins and more area to dissipate heat, and lower temperature such as 14 pins vs 8 pins. What is your switching frequency? 

Regards,

Sean

• The switching frequency is around 1Khz.
• In our application we don't use the switching mode continuously, we use UC3843D (SOIC-8) as PWM (0% to 95%) generator and it generates PWM only when analog input (Input voltage : 0.2V to 1.5V) is received.
• We observed heating even when the IC is not switching. 
• (Switching Voltage - 24V)
  (Biasing Voltage - 24V)

1k Fsw is very low. I don't see the reason why the temperature is hot.

Two suggestions:

1. Something wrong with this IC, or resolder it, or change a new IC;

2. Lower bias voltage to 10V, and check the temperature again.

Thank you for your apt reply,

- We have already tested this IC at 12V and it was working fine.

- What can be done to avoid overheating for 24V, since we had selected this IC for 24V application?

A safe way is to insert a zener (12V) with NPN transistor to reduce 24V to 12V before it is connected to VCC. 

The gate voltage is 24V, and gate loss is large. I guess this can contribute to heat IC.

Regards,

Sean

Thank you for your response, we also had similar solutions in our mind but;


Do you mean that UC3843D with SOIC-8 package does not support 24V ? Which is mentioned in data sheet.

So is it safe to assume that UC3843D with SOIC-8 package will not support 24V even if it is mentioned in the datasheet?

Any Datasheet Errata available to confirm that ?

There may be some other reasons to cause IC to get hot at 24V. It needs more work to narrow down the cause. If PWM output doesn't drive MOSFETs, is IC still hot? 

Is every IC hot at 24V? 

I don't have a formula to calculate power loss and temperature rise at 24V. From D/S, it seems IC is able to operate at 24V, but the temperature may go too high.

• Even when the PWM output isn't driving MOSFET, the IC is still hot.
• We tested 2 ICs both were hot. (Procured from Mouser) 
• Due to current semiconductor shortage there might be a possibility of getting faulty parts, Is it possible for you to send us few genuine samples ?
• It seems that due to SOIC-8 package the heating issue is present.
• Can you suggest any alternative for SOIC-8 package which will work on 24V without heating ?

I wish I could send you some samples.  I don't have any part for evaluation either.

I looked at other parts, such as UCC2803, UCC28C43, they couldn't operate at 24V. I think it is better way to modify your circuit to step down Vin from 24V to 12V for VDD bias. 

• Is there any document / errata / sheet which would mention that UC3843D supporting voltages?
• There is significant delay in our communication, can we have a Microsoft Teams call wherein we can discuss this issue in detail ?
• For Teams call, this is my email ID - manas.sable@konecranes.com

Do you accept webex (cisco) meeting invitation?

If you insert a resistor between 24V and VDD, can you measure the voltage drop, and calculate the current? 

Can you figure out how much power is dissipated by IC at 24V and 12V respectively?

Thank you for your suggestion, we have tried the following and here are the results;

• We have measure current flowing through UC3843D
• With 1K resistor in series with 24V VCC
• With Ammeter in series with 24V VCC
• With 1K resistor in series with 12V VCC
• With Ammeter in series with 12V VCC
• Accordingly we calculate power dissipation.

Webex should be fine.