Author Topic: Why did my LT3042 die?  (Read 3274 times)

0 Members and 1 Guest are viewing this topic.

Offline giosifTopic starter

  • Frequent Contributor
  • **
  • Posts: 886
  • Country: gb
Why did my LT3042 die?
« on: April 03, 2021, 09:26:11 am »
Hi,

I bought one of those little boards found on eBay which use an LT3042 to provide low noise voltage supply.
I set the output voltage to 15V by changing Rset to 150 kOhms, fired it up and all looked good.
I then used it for about 1 hour or so, drawing around 50 mA from it.
During this time, I noticed the LT3042 IC was getting rather hot to the touch, but thought this might be normal.
After the 1 hour, though, the IC died and I am trying to understand the reason for that.
Looking at the datasheet for the IC (to the degree I can read & understand it), I don't think I've exceeded any of its operating parameters (except maybe for heat dissipation?).

Anyone has any ideas?

Of course, this being bought from eBay, that IC could be anything *but* an original LT part, so that is one possible explanation, but I'm trying to see if I might have overlooked something else.

Thanks!
 

Offline 2N3055

  • Super Contributor
  • ***
  • Posts: 7005
  • Country: hr
Re: Why did my LT3042 die?
« Reply #1 on: April 03, 2021, 09:36:24 am »
What was the input?
 

Offline giosifTopic starter

  • Frequent Contributor
  • **
  • Posts: 886
  • Country: gb
Re: Why did my LT3042 die?
« Reply #2 on: April 03, 2021, 10:41:52 am »
Sorry, forgot to mention that: input was 20V.
 
The following users thanked this post: AndersJ

Offline PeteH

  • Supporter
  • ****
  • Posts: 114
  • Country: ca
Re: Why did my LT3042 die?
« Reply #3 on: April 03, 2021, 11:35:21 am »
What did your wire harness look like?

Running the part at 20V when the abs max is 22V with default decoupling on the eval PCB is risky.

Does your load ever change in current?

The chip mounted on a fairly large PCB should be able to handle the ~300mW of loss but would get warm depending on the layout of the board.... Could get very hot if not done well.
 

Offline Jay_Diddy_B

  • Super Contributor
  • ***
  • Posts: 2741
  • Country: ca
Re: Why did my LT3042 die?
« Reply #4 on: April 03, 2021, 01:08:20 pm »
giosif and the group,

Are you talking about a board like this one?




It has ceramic input capacitors. It doesn't have an electrolytic capacitor that will provide damping. Have a look at this application note. The application note explains why you need damping.

Link: https://www.analog.com/media/en/technical-documentation/application-notes/an88f.pdf

This is probably why the LT3042 failed.

Jay_Diddy_B
 
The following users thanked this post: giosif

Offline giosifTopic starter

  • Frequent Contributor
  • **
  • Posts: 886
  • Country: gb
Re: Why did my LT3042 die?
« Reply #5 on: April 03, 2021, 04:08:28 pm »
Yep, that's the board.
Ok, I will read the application note.
Thank you!
 

Offline giosifTopic starter

  • Frequent Contributor
  • **
  • Posts: 886
  • Country: gb
Re: Why did my LT3042 die?
« Reply #6 on: April 03, 2021, 04:18:06 pm »
What did your wire harness look like?
Amm... Two simple wires going from by bench power supply into the input pins of the board?
I'm not sure what you mean with the question.

Running the part at 20V when the abs max is 22V with default decoupling on the eval PCB is risky.
I am aware of the 22V max limit but, for an output of 15V, I needed to go above that.
Also, one of the typical application examples is showing 20V at the input.

Does your load ever change in current?
Yes, but I don't think it would stray way too much from the 50mA.

The chip mounted on a fairly large PCB should be able to handle the ~300mW of loss but would get warm depending on the layout of the board.... Could get very hot if not done well.
This is the small board Jay_Diddy_B mentioned above.
 

Offline 2N3055

  • Super Contributor
  • ***
  • Posts: 7005
  • Country: hr
Re: Why did my LT3042 die?
« Reply #7 on: April 03, 2021, 06:41:52 pm »
What did your wire harness look like?
Amm... Two simple wires going from by bench power supply into the input pins of the board?
I'm not sure what you mean with the question.

Running the part at 20V when the abs max is 22V with default decoupling on the eval PCB is risky.
I am aware of the 22V max limit but, for an output of 15V, I needed to go above that.
Also, one of the typical application examples is showing 20V at the input.

Does your load ever change in current?
Yes, but I don't think it would stray way too much from the 50mA.

The chip mounted on a fairly large PCB should be able to handle the ~300mW of loss but would get warm depending on the layout of the board.... Could get very hot if not done well.
This is the small board Jay_Diddy_B mentioned above.

Like always, JDB is right on the money...
If you have a meter of cable and ceramic capacitor it can ring really bad. Also, one thing people forget, most of lab PSU have inductive output impedance(hence big elcos on the output, to compensate).
Also LT3042  is a low dropout regulator that can work with 350mV drop on regulator, i.e. for 15V you need 15.35V on input. 5V is way too much. 16V would have given you healthy reserve.
 

Offline coppercone2

  • Super Contributor
  • ***
  • Posts: 9981
  • Country: us
  • $
Re: Why did my LT3042 die?
« Reply #8 on: April 03, 2021, 06:44:18 pm »
I put a 3042 on a little home made PCB a while back and it also died, but I got mine from digikey.

These guys are sensitive, I think it was ESD, I was pretty careful soldering it but my ESD mat was not hooked up because I moved a buncha stuff and cleaned and forgot to plug it back in.

Anyway I cut the PCB in half and ran it on batteries, the 3042 can go on a power supply board if I ever finish that amplifier.

They are touchy. I recommend sticking with 3080 series unless absolutely necessary
« Last Edit: April 03, 2021, 06:48:33 pm by coppercone2 »
 

Offline giosifTopic starter

  • Frequent Contributor
  • **
  • Posts: 886
  • Country: gb
Re: Why did my LT3042 die?
« Reply #9 on: April 03, 2021, 09:22:28 pm »
Like always, JDB is right on the money...
If you have a meter of cable and ceramic capacitor it can ring really bad. Also, one thing people forget, most of lab PSU have inductive output impedance(hence big elcos on the output, to compensate).
Something I learned today.  :)

Also LT3042  is a low dropout regulator that can work with 350mV drop on regulator, i.e. for 15V you need 15.35V on input. 5V is way too much. 16V would have given you healthy reserve.
Indeed. I should have read the datasheet more carefully, but I was being impatient...
 

Offline giosifTopic starter

  • Frequent Contributor
  • **
  • Posts: 886
  • Country: gb
Re: Why did my LT3042 die?
« Reply #10 on: April 03, 2021, 09:29:03 pm »
[...]
They are touchy.
Yes, I learned this first hand.

I recommend sticking with 3080 series unless absolutely necessary
I will check the specs for the 3080, but I'm after something with very low noise.
 

Offline Jay_Diddy_B

  • Super Contributor
  • ***
  • Posts: 2741
  • Country: ca
Re: Why did my LT3042 die?
« Reply #11 on: April 04, 2021, 05:19:28 am »

Snip ...

Also LT3042  is a low dropout regulator that can work with 350mV drop on regulator, i.e. for 15V you need 15.35V on input. 5V is way too much. 16V would have given you healthy reserve.
Indeed. I should have read the datasheet more carefully, but I was being impatient...

giosif and the group,

Be careful, although the LT3042 can operate with only 350mV of input to output differential:




Other parameters will be compromised if operated with a low input to output voltage differential: This is PSRR:




Output noise spectral density:





I would choose a differential voltage 1.3V if I want the best PSRR and noise performance. Typical applications for the LT3042 are normally for lowest noise and highest possible PSRR. To allow for tolerances in a production environment you may need to allow more voltage. If the input supply is 5% and the LT3042 is 2% and the Vset resistor is 1%, a larger differential may be required. If is for home use, you can tweak the circuit.

Best Regards,
Jay_Diddy_B
 
The following users thanked this post: bingo600, giosif, 2N3055

Online bingo600

  • Super Contributor
  • ***
  • Posts: 2010
  • Country: dk
Re: Why did my LT3042 die?
« Reply #12 on: April 08, 2021, 04:33:59 pm »
For an Non EE person...

The headroom (Vin) should be around 1.3v above (Vout) ?
After reading the AN , am i correct in understanding that one should leave the ceramic input cap , and add an additional 47uF Elco in parallel ?

I have a few of these boards , that i expect to use for some OCXO regulation

/Bingo


 

Offline Jay_Diddy_B

  • Super Contributor
  • ***
  • Posts: 2741
  • Country: ca
Re: Why did my LT3042 die?
« Reply #13 on: April 09, 2021, 12:03:24 am »
Bingo600,

Choosing the headroom, is a complex question, the answer is (as always) it depends.

1) You have to consider power loss in the regulator. The power loss is:

(Vin - Vout) x Iout

The maximum dissipation depends on the PCB, but 1 to 1.5W would be a good starting point. This may limit Vin-Vout.


2) You have to decide if you *need* the highest possible noise performance and PSRR. You might be able to operate with a Vin-Vout < 1.3V

3) You may want to consider what rails you have available, You may also be able to choose the output voltage.

4) You may need to consider the tolerances of the LT3042 and the supply being used to feed it. This is to make sure that you get the performance you need under all conditions.


In short, AN88, suggests placing an electrolytic capacitor that is 3 or more time the value of the ceramic capacitor in parallel. The ESR of the electrolytic will dampen any resonances. In most cases a 47uF electrolytic will be fine. It is if you 'hot plug' or hard switch the input supply.

Jay_Diddy_B
 

Offline giosifTopic starter

  • Frequent Contributor
  • **
  • Posts: 886
  • Country: gb
Re: Why did my LT3042 die?
« Reply #14 on: June 10, 2021, 06:43:56 pm »
Coming with a late update: I bought another module like the initial one, installed a 47uF electrolytic cap at the input and am trying to power it with 17V.
With no load, the output stays at around 15V (14.85V, to be exact).
However, as soon as I apply a load of 50 mA, the output goes up to 16.36V.  :-//
If I then increase the input voltage, the output tracks the input, until I get to about 18.9V on the input, when the output voltage (under the 50 mA load) becomes exactly 15.00V.
From 18.9V up to 20V on the input (I didn't have the courage to go above that), the output voltage stays the same (i.e. doesn't track the input voltage anymore).

Initially, I thought maybe this was due to the module coming off eBay and the possibility the LT3042 in it might not be genuine.
But (and this is why I took so long for the update), I finally managed to order an LT3042 IC off DigiKey and install it on the board of the initial module from eBay (the one which prompted this discussion) and the behaviour is the same as with my 2nd module from eBay.

I am bit confused at this stage and not sure what I'm doing wrong here.
 

Offline teddybear

  • Newbie
  • Posts: 1
  • Country: jp
Re: Why did my LT3042 die?
« Reply #15 on: September 09, 2021, 06:01:07 am »
I bought the same module, CJMCU-3042, from a Chinese seller very recently. 

I've tested the unit under a resistive load (~100 mA) with a spectrum analyzer equipped with a low noise preamp and found that the modules were unstable.  There is a noise bump in the power spectral density at around 100 kHz, which is about 50-100 times higher than those shown in the datasheet.  The instability is caused by the poor design of the PCB.  The noise bump was disappeared after the following MODs to reduce inductance of the PCB traces.

0) First of all, add an appropriate elecrolytic cap. in Vin.
1) Additional X7R cap. from the Vin pad of Cin to the GND pad of Cout.  Use this point as GND of Vin/Vout leads.
2) Add wires between the GND pads.  The wire from Cset to Cout/Cin is especially important. 


24 Sep. 2021
Additional notes:
I found that the Cset cap. in the module was not a low noise part, at frequency especially below 100 Hz.  After changing the Cset to Murata's X7R cap. all of my modified modules for 5V, 12V and 12V with PNP boost works as expected from the datasheet.
« Last Edit: September 24, 2021, 01:15:07 am by teddybear »
 
The following users thanked this post: thm_w

Online Gerhard_dk4xp

  • Frequent Contributor
  • **
  • Posts: 341
  • Country: de
Re: Why did my LT3042 die?
« Reply #16 on: September 09, 2021, 08:01:32 am »
I have used a lot of LT3042 but I have never seen signs of instability.
And nearly all of them had to live with ceramic SMD capacitors.
The noise peak was always close to the frequency and size given in the
data sheet.

Be sure that the exposed pad is soldered, positively.
You cannot expect 100 mA from that tiny chip with a drop of several volts.
Even if it has switch-off for over temperature, the thermal cycles will kill it.

Since someone mentioned feeding an OCXO: mine draw 1.5A
during warm-up. A clear NO for the 3042.
Exactly for this reason I have made a board with LT3042 + TO220-
transistor that has to take the heat. It is exactly the circuit of the
data sheet.  In the noise plot, 0dB is 1 nV/rt(Hz). The noise peak is
just 3 nV/rt(Hz).

The steep rise on the left side goes on the much undersized input
capacitor of the preamplifier. Note that is is much worse than 1/f.

Power supply is an R&S NGT20, all open on the table with no
shielding whatsoever. The board is home-etched, in the mean time
there is a tidied-up version to make use of the leftover space in
different JLCPCB projects.


cheers, Gerhard
« Last Edit: September 09, 2021, 08:22:24 am by Gerhard_dk4xp »
 

Online radar_macgyver

  • Frequent Contributor
  • **
  • Posts: 705
  • Country: us
Re: Why did my LT3042 die?
« Reply #17 on: September 09, 2021, 03:10:56 pm »
Damping is often required when one uses a series inductor for EMI filtering, here is a reference:

https://training.ti.com/power-tips-damping-input-filter

This is in the context of a switch-mode converter which has a negative impedance, which can cause oscillation in conjunction with an undamped filter. In the AN 88 case, the inductance comes from the long leads.

To estimate the optimum damping (when filter inductance is known), I use LTSpice, and step through different values of damping resistor while looking at the AC response, and pick the one with the smallest peak.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf