Problem with LT3092 current source

Hi everyone,
I am trying to get stabille em field in inductor so what i need is inductor and some current source to drive that inductor.The inductor is wounded and have cca 2000 turns, which in this case isn`t so important.
Problem i have in this case is with current source.I need stabile and constant current source, so i used lt3092-programmabile cs which is capable to produce up to 200mA. Several of them connected parallel will produce, or should produce larger current equal to summary of current each of them produce.And that is what is this "thing" doing. But during in time, after several moments current which LT3092 is producing suddenly starts to drop.It is not constant any more and also magnetic field in inductor also drops.I have tried to substitute inductive resistance with equivalent resistor and the same thing is happening.
I have several of lt3092 and tried them on external board.On the start current they produce is stabile and gets rapidly low after several moments.I used two resistor configuration to set desirable current, which is available in datasheet.Current each of them produce is 180mA.
Does anyone have some idea about this or from yours experience with this ic`s?
Thanks

Maybe the chip is heating up?
Typical for current sources is that they can dissipate a lot of power, especially if the supply voltage is high.
Best is to make sure there is a groundplane with large surface on either PCB side underneath the chip (connected between them with via's for heat transport) an that this groundplane is connected to the heat sink of the chip (the tab or pin 9 in the package drawings)

Yes it is heating up and it has no copper area or any heat sink. They are supplied with externall adjustable supply from 12V-36V, but they should work with permanent 36V. When 12 V is input Vout is 11.2v, Vin=36V Vout=35.something V. They should be able to generate 200mA but they have 185mA out current so it is something about 10% bellow specification in datasheet...

Then the heating probably causes the problem (maybe it is a safety mechanism to avoid blowing up the chip)
You can check this by cooling it down temporarily (there are special sprays for that, maybe blowing do the trick)

you can do the following:
- buy the DD or ST package version with heat sink, and solder it to a large surface for better cooling
- use a supply voltage that is as low as possible
- put more chips in parallel.

There are 10 Lt3092's in group of two connected parallel so there are 5 separete output. Each group of them is acting almost similar.
Definitely i should cool them more because one output dissipate about 5,6watts and they have thermal coefficient about 20 Cell/Watts so they get more then 100 cell.
Another problem is that they are on bottom side of pcb so practicaly there is no room for any heat sink...
The last thing i like to do is creating new pcb..but if is there no other possibilities i will have to do that...

Try reducing the supply voltage, They probably run so hot now because they get a much too high voltage.
You can probably do this easily by changing a regulator or a zener value

vthielen said:

Try reducing the supply voltage, They probably run so hot now because they get a much too high voltage.
You can probably do this easily by changing a regulator or a zener value

Click to expand...

I have decreased supply to 15V and when supply them from external power supply they dont stop working but when i connect them on transformer via voltage regulator on the same voltage (15.5V) current is decreasing, as it was before with greater voltage.
Circuit is behaving different on 15V when are connected on power supply and on transformer.
Power of transf. is 50 VA, while total consumption of power is about 10W of LT3092, plus the power disipation on voltage regulator.
Transformator is 230V/30V, and after capacitor there is 42V; so the disipation on voltage regulator is (42-15)*I=25*(~1A)=20+W
Power of transformer is sufficient for all power losses in circuit..
Now i am confused why is there difference between supplying from external power supply and from transformer?