LM1084 Output Current is not obtained.

Two points to consider:

1)     Is the output oscillating due to incorrect output capacitor?

2)     Is the LM1084 package getting hot, going into thermal shutdown?

Based on the information given, there’s a thermal problem, for sure.

The expected power dissipation would be ((Vin-Vout) x Iout) = ((12V-8V) x 3.5A) = 14W

You did not say which specific device you are using, LM1084IS-ADJ (TO-263) or LM1084IT-ADJ (TO-220), but it might not matter.

I'll guess this is TO-263 with RΘ(ja) of about 40°C/W when mounted to PCB per the JEDEC JESD51-7 standard.

So, the ~theoretical~ junction temperature would be = ((Pd x RΘ(ja) ) + Ta) = ((14W x 40°C/W)+25C) = 585°C

But that’s not going to happen as the thermal shutdown will activate at about 160°C and the output will shutdown.

Hi Jones,

Thank you so much for your replay.

1)     Is the output oscillating due to incorrect output capacitor?

Ans:  I am not sure.I have tested with 10 uF and 22 uF tantalum . When load is there  Vout is dropping from 8V to 4V and also it is varying. With Load Iout is only 0.2Amp.

2)     Is the LM1084 package getting hot, going into thermal shutdown?

Yes it is heating but it is touchable and going into thermal  shutdown.

I'll guess this is TO-263 with RΘ(ja) of about 40°C/W when mounted to PCB per the JEDEC JESD51-7 standard.

So, the ~theoretical~ junction temperature would be = ((Pd x RΘ(ja) ) + Ta) = ((14W x 40°C/W)+25C) = 585°C

But that’s not going to happen as the thermal shutdown will activate at about 160°C and the output will shutdown.

 I am using LM1084IS-ADJ (TO-263).I need maximum  Iout=3.5Amp. Load needs constant up to 2.5Amp.

Can you please suggest any solution for above  problem.

Awaiting for your reply..

Thanks & Regards 

Amaresh

"... Yes it is heating but it is touchable ..."

 At thermal shutdown the device case should be really hot.

No idea at all what might be happening here other than thermal shutdown. Did you double check Vin to make sure that it is not falling when loaded?

The only possible solutions to avoiding thermal shutdown are: 

   1) Reduce the dissipation in the device. This calls for reducing the Vin-Vout differential, and/or reducing the load current. In your situation reducing Vin from 12V to 10V cuts the dissipation in half, but that’s still not enough to help with device dissipation. LM1084 needs at least about 1.3V of Vin-Vout head-room to stay in linear operation.

   2) Reduce the RΘ(ja). This calls for adding heat-sink to the package. To dissipate 14W and keep the junction temperature out of thermal shutdown would require a RΘ(ja) of no more than 125°C/14W = 9°C/W. In this situation, using the TO-220 package with huge (RΘ(sa) < 8°C/W) bolt-on heat-sink . Increasing the copper area directly under the TO-263 DAP with a multi-layer board with lots of thermal vias could get RΘ(ja) down to approx 30°C/W.

   3) Reduce the ambient temperature. This usually calls for adding a fan, or improving ventilation. Probably not realistic solution in this situation.

The given RΘ(ja) rating of 40°C/W means that if Tambient is 25°C, you are going to be limited to about 2.5W of device dissipation, possibly 3W if you push it. With your 3.5A current requirement, that puts the Vin-Vout allowance at less than 1V, below what the LM1084 needs at high output current.

For this level of dissipation (14W) I always recommend using a buck SMPS. The poor efficiencey of almost any LDO limits device dissipation to the range of about 2 watts to 4 watts, depending on package type, mounting, air flow, etc.

WEBENCH® Designer at www.ti.com can help find a more appropriate solution.  

Hi Jones,

Thank you so much for your replay.

Yesterday I connected 0.1uF  capacitor at the output.Now Iout=0.35Amp but voltage is dropping when load is there.

Also now  no Thermal shutdown is there.If  you touch deeply then only  can feel the hit.

Please give me some suggestion to solve this problem.

Awaiting your reply..

Thanks &Regards

Amaresh