Overheating mSATA-IDE converter

[oussama]

Hi everyone!

The attached file is a picture of an mSATA to PATA converter that I bought from eBay. It is used for example to put mSATA SSD devices in older laptops that utilises PATA interface and not SATA. The converter is built around the chip JM20330 and uses a 3.3V voltage regulator SA1084D-3.3 (datasheet: http://www.alldatasheet.com/datasheet-pdf/pdf/301356/SILAN/SA1084.html)

mSATA devices works with 3.3V while PATA with 5V. For that reason, the converter contains a 3.3V voltage regulator.

When I used this converter both in combination of my USB-IDE converter and inside my laptop, the following becomes very hot:
-the mSATA SSD device
-The converter board (mainly the regulator)
-the host system (either the laptop or the USB-IDE converter)

I'm afraid if I use it for a long time, that I destroy my laptop.

I have looked in the regulator datasheet, and they suggest using two 100uF bipolar capacitors (page number 1) one for the input and the other for the input. Looking at the converter I only see surface mounted (ceramic?) capacitors with two different sizes, and it is obvious that these are non-bipolar.

Please look at the converter and try to find/guess the reason for overheating. Could it be a bad choise of capacitors?

Sorry I do currently not have a working oscilloscope for the moment; it would otherwise have been a good help to see if the voltage is stable or not.

I appreasiate any help or suggestion

Helpful resources:
==============
44 pin IDE pinouts: http://pinoutsguide.com/HD/Ata44Internal_pinout.shtml

44 pin IDE numbering: http://www.unitechelectronics.com/44IDEHARDDRIVE.gif

mSATA pinouts (page 8 for table, page 19 for numbering): http://www.google.com/url?sa=t&rct=...naaDXCLzFH6vhdl68IBe3lA&bvm=bv.51495398,d.bGE

 

[KrisBlueNZ]

Hi oussama and welcome to the Electronics Point forums

Thanks for the thorough problem description.

Re the capacitors. They're probably not the problem. The data sheet shows input and output decoupling capacitors on page 6; the input capacitor is 10 uF and the output capacitor is 22 uF. I don't know where you got 100 uF from.

Both capacitors are shown as polarised on the data sheet but they do not have to be polarised. The board uses multi-layer ceramic capacitors (MLCCs) which are not polarised and have a high capacitance/volume ratio. It's possible that they are 10 uF and 22 uF respectively but MLCCs are not marked so it's hard to be sure.

I would be worried about the laptop and the mini-SATA solid state drive getting hot, especially the SSD. That's definitely a bad sign! Can you measure the voltage on the 3.3V supply rail to the SSD to make sure it's 3.3V +/- 10%? Also, can you measure the voltages on the input and output pins of the regulator?

 

[oussama]

Thank you KrisBlueNZ for welcoming me and your answer.

OK good to hear that the caps are not the problem. You are actually correct, I was looking at a different datasheet for a similar component (AP1084D3.3)... sorry for confusing. Thank you for telling the type of these caps, I want to learn such things

I have measured the input for theregulator, and it is around 4.99V -4.98V depending on the voltmeter I'm using
The output of the regulator is 3.28V , which is 0.05V less than 3.3V - do you think that it looks good?

I have also measured the voltage exactly beside the mSATA connector and it is also 3.28V

could a high power consumption be the reason?
The SSD consumes 2W and the converter around 0.22W => 2.22W/3.3V= 0.67A
So the power that is drained from the PC is 0.67A*5V = 3.35W

The harddrive that was in the laptop before was marked as 5V and 0.55A, which didn't make the laptop very hot

Or could the problem be that the converter is only using the "+5 VDC (Motor)" and not the " +5 VDC (Logic)" line in the laptop IDE connector?

I have used another converter (from mSATA to SATA) which also includes a 5V to 3.3V regulator, and the SSD didn't become so hot. The converter was used in another laptop with SATA built in. See attached picture. The regulator was larger in size and marked as UZ1084L33

Thank you for your help

 

[KrisBlueNZ]

Yes, 3.28V is fine for the 3.3V rail.

The SSD consumes 2W. Is that from its specifications? That would be 0.6A drain on the 3.3V regulator. If the converter is specified as drawing 0.22W that would be specified at the 5V rail voltage so it corresponds to 44 mA for a total regulator load of about 0.65A. That's well within its specification but it will make it dissipate significant power.

Power dissipation in the regulator will be P = V I where I=0.65 and V is the voltage DROPPED by the regulator, which is 5-3.3 which is 1.7V. So from that formula, power is 1.1W.

Heating (above ambient temperature) can be calculated by multiplying the power dissipated (1.1W) by the thermal resistance to ambient, which is specified in degrees Celsius per watt. It depends on the area of the copper that the regulator is soldered to, the thickness of that copper, and whether there is matching copper on the other side of the board. Can you tell me any of those numbers?

Can you also give me some kind of description of how hot the regulator gets? How long can you hold your finger on the copper for, before it gets so hot you have to let go? What if you push firmly? Leave it running for at least five minutes before you test.

Also you could try with the SSD disconnected and see how hot the regulator gets then.

The heating in the SSD is a worry. As long as the SSD is receiving the correct voltage, it should not dissipate any more power, or generate any more heat, than it's designed to. Do SSDs generally get noticeably warm? Have you used this SSD before? If so, did it get warm at all? How much hotter is it now than before?

I doubt that using the motor supply instead of the logic supply would explain the problem. They probably come from the same point in the laptop, just through thicker or thinner wires. You shouldn't draw 0.65A from the logic supply.

The heating of the laptop is also a concern. Can you tell where the heat is coming from? Can you describe how hot it actually gets?

Can you find some kind of temperature measurement device and measure the actual temperature of the hottest component in the laptop, the adapter, and the SSD?

Can you measure current with your multimeter? It would be useful to know the current going into the regulator, or into the SSD. To measure current, you have to break the circuit and connect the multimeter across the break. Is there any suitable place where you could do this? For example could you remove the pin of a connector to break the circuit?

If you can, you need to keep the multimeter connected at all times while the power is ON. Set it up with the power OFF, make sure it's properly connected and will not come disconnected, and doesn't require you to hold anything in place, before you switch ON. You may be able to solder small bits of wire to the adapter circuit board and use alligator clip leads to hook the multimeter in circuit.

Your multimeter will probably have a 10A current range, that uses a separate socket on the front of the multimeter. It's best to use this one. It won't resolve to better than about 0.01 amps, but that's fine.

 

[KrisBlueNZ]

I've just looked at the SSD data sheet that you linked to. It says that the current consumption, while idle, is only 170 mA. This shouldn't cause significant heating in the regulator on the adapter board. Or is that data sheet for a different SSD? If so, can you link to the data sheet for the one you have?

I think you need to measure some currents.

Why are you so interested in pinouts? Have you changed any of the pin connections or made any other changes to the IDE or SATA connections?

 

[oussama]

It appears to be normal to become hot

Thank you KrisBlueNZ for the time you put in helping me.

I did not make any changes to the pins in neither the converter nor the SSD. The SSD PDF I refered to in my first post is there just to give a referens of the mSATA pinouts.

I have tested the converter with two newely bought mSATA SSD devices:
* Kingspec KSM-mSATA.5-064MJ (http://www.kingspec.com/en/product_xx24.html) which consumes 2W when actively used and 0.5W when idle
* Kingston SMS200S3 (http://www.kingston.com/en/support/technical/products?model=sms200s3) which consumes 0.4W idle / 1.2 (TYP) read / 1.8W (TYP) write

I have followed your advice and measured the temperature of the converter and the SSDs using a kitchen temperature meter. To my surprise I discovered that the SSD got very hot even in the laptop with builtin SATA connection (no mSATA-IDE converter used), up to 52C ! The regulator in the mSATA-SATA converter (the last picture that I uploaded) does also get hot, up to 41C.

So apperantly, it is designed to become so hot, which I think is a bit strange... however the SSDs that I use are designed to operate in temperatures up to 70C, so it should be safe according to the manufacturers.

I think I will trust the converters and the SSDs and I hope they do not destroy my father's laptop

Thank you very much for your help