ISL55110 output burning !

Thread Starter

captoro

Joined Jun 21, 2009
207
Hello,

I am wandering why my ISL55110 burns when I mix two signals together.
I am using input A on both ISL55110 and mixing them in the J1 jumper before going to a MOSFET.
Individually they work fine and I get a waveform on the header J1. But when I combine both, one of the ISL U10 , goes up in smoke !
signal 1 is 3Mhz
signal 2 is 50 hz
Why and how do I avoid this ?

ken
 

Attachments

Last edited:

ericgibbs

Joined Jan 29, 2010
18,766
Hi captoro,
Please post a sketch showing how the signal lines are 'combined' in the J1 header.

Also check phase sense of the two IN-A input signals.

Added the d/s for the devices.
E
 

Attachments

Wolframore

Joined Jan 21, 2019
2,609
U10 is exceeding current by going straight to ground when U13 is high at the same time..

its up to you to correct this, without seeing the system, it makes little sense to me as to even why you chose these components.
 
Last edited:

Thread Starter

captoro

Joined Jun 21, 2009
207
U10 is exceeding current by going straight to ground when U13 is high at the same time..

its up to you to correct this, without seeing the system, it makes little sense to me as to even why you chose these components.
why you chose these components ??
I am driving a MOSFET . So yes I am using MOSFET drivers, I thought there would be a diode on the output of the ISL55110.. so you are saying when its LOW, its sucking all the current in from the other ISL. Got that.
I choose ISL55110 because they have fast rise time.
How would you parallel multiple signals and amplify the signal on a MOSFET ?

ken
 

Thread Starter

captoro

Joined Jun 21, 2009
207
I see what you mean. I cant put two output together. The lower frequency output will suck in the current from the High frequency.The Q4 is only there in case i want to break the output from J1. We dont need to look at that as its not a concern.
The only was I see it now is for each ISL output to direct it to a MOSFET, the parallel the waveform. After that amplify the combined resulting waveform.
or Put a zener diod on each ISL output, then I probably wont have a true ground when the signal is low.
K
 

Wolframore

Joined Jan 21, 2019
2,609
Why not sum the control signals with a high impedance buffer and use one gate driver? Again, unless I know what the desired goal is, I’m at a loss as to the correct solution. Q4 is not decoupling U10, it’s a low impedance path to ground, it’s permanently disabling U10. Think of a better way to do that.
 

Thread Starter

captoro

Joined Jun 21, 2009
207
Why not sum the control signals with a high impedance buffer and use one gate driver? Again, unless I know what the desired goal is, I’m at a loss as to the correct solution. Q4 is not decoupling U10, it’s a low impedance path to ground, it’s permanently disabling U10. Think of a better way to do that.
Microcontroller communicate to LTC6904, LTC6903 and AD9850 to generate signals. Some signals maybe routed to AD835 to be multiplexed. (signal out of the 835 is low, so it is amplified in the LM7171) Signals is then transfered to MOSFEt driver ISL55110. to drive a MOSFET.

Do you have a high impedance buffer to suggest me ?

K
 
Last edited:
Top