Hello,
I designed a PCB with 3pcs TC7660 charge pump ICs. They are going to be connected paralely to reduce output resistance. Each of them can be disconnected at the output side via jumper.
I have a problem to make them working even when they are separated from another components. It seems like there is a bug in the schematic/PCB design, because I already burned many of them. But I did not find a bug and I have no Idea what could be wrong. When I solder the TC7660 chip to the PCB and power it with +5V, there is not expected -5V at the output and the power consumption is rapidly changing a lot (up to 200mA) when the charge pump output is disconnected from the rest of PCB. Sometimes I can also hear acoustic noise from the failing chip (>10 kHz).
PCB is designed and soldered according to datasheet, only pin 6 is not connected (should be grounded only for Vcc < 3.5 V) and I am using 10 uF ceramic capacitors instead of electrolytic.
Of course I have carefully checked there are no shorts at the PCB.
Does anybody have idea what could be wrong and why these charge pumps are burning at my PCB?
I marked 3 positions in attached picture. It fails ommediatelly at position '3', a bit later at position '2' and currently it works at position '1'(but also failed once there).

 

Is that the complete schematic diagram? EDIT: no just the basic unit replicated three times
Where did you get the curious notion that you could connect them in parallel? EDIT: from the datasheet - wow
Things burn when they draw excessive current or there is an unanticipated low impedance path to Ground.
A 10uF ceramic capacitor sounds dubious, are you sure that is what you have?
Did you use low ESR capacitors? This is something you have to specifically look for and ask for.

 

Is that the complete schematic diagram? EDIT: no just the basic unit replicated three times
Where did you get the curious notion that you could connect them in parallel? EDIT: from the datasheet - wow
Things burn when they draw excessive current or there is an unanticipated low impedance path to Ground.
A 10uF ceramic capacitor sounds dubious, are you sure that is what you have?
Did you use low ESR capacitors? This is something you have to specifically look for and ask for.

It is just part of the diagram, but the rest is now disconnected.
I was looking for short at the PCB, but I did not find a problem.
Yes, i am using 10 uF ceramic capacitors (1206 package, 10V) but no extra low ESR. But even that I expect the ESR to be much lower in comparation with electrolytic capacitors (datasheet expects using of electrolytics).

 

I don't see any bypass caps for each IC, and I agree with @Papabravo that the caps on the board are not likely to be 10µF.

Didn't see this before posting.

Yes, i am using 10 uF ceramic capacitors (1206 package, 10V)

 

What is the tolerance and the tempco on those 1206 capacitors?
Can you give us a link to the datasheet for the capacitors?
Depending on the frequency a high ESR capacitor can get quite warm.
Your expectations on ESR might or might not be valid. Short of looking at a datasheet it is impossible to tell.

 

Do you have boards without ICs soldered in yet? You might check to confirm they aren't drawing current on their own. Then add just one of the 3 circuits and see what happens.

 

Do you have boards without ICs soldered in yet? You might check to confirm they aren't drawing current on their own. Then add just one of the 3 circuits and see what happens.

As my daughter used to say when she was six: "That would be a great idea -- Daddy!"

 

Note the comment about oscillation:

https://product.tdk.com/en/techlibrary/solutionguide/mlcc_replace-guide.html

 

What is the tolerance and the tempco on those 1206 capacitors?
Can you give us a link to the datasheet for the capacitors?
Depending on the frequency a high ESR capacitor can get quite warm.
Your expectations on ESR might or might not be valid. Short of looking at a datasheet it is impossible to tell.

The tolerance of capacitance should be 10%.
Here is the datasheet
https://www.gme.cz/data/attachments/dsh.905-147.1.pdf

 

As my daughter used to say when she was six: "That would be a great idea -- Daddy!"

Currently I don't have. But i checked it befefore soldering with multimeter.
(I have 2 pcs of the PCB. At first I soldered all components to 1st PCB and after dtat I found this problem. After that I started to assembly 2nd PCB only with charge pumps and capacitors.)

 

Note the comment about oscillation:

https://product.tdk.com/en/techlibrary/solutionguide/mlcc_replace-guide.html

Thank you for mentioning this, I will analyze this idea more in details.