28->24 volt regulator keeps blowing my 24->10 regulator

[duhuhu]

Hey guys!

I have been having a lot of problems with a board I am working on. The board was originally designed for 24 volts, but my input is 28 (non negotiable), so to make things easier than redesigning the whole power circuit, I am just doing a step down to 24 right when it goes into my box. The problem that I am having is that when I hook up the whole circuit with 28, It blows my 24->10 volt regulator. But when I run the original circuit on 24, it works just fine. I have checked the entire rest of the board and there are no shorts. This board was given to me by the last guy to work on the project (who now cannot help me due to conflicts of interest), and it does not have a top coat, so there are exposed wires, but they don't seem to be causing any noticeable issues...

Unfortunately I don't have the specs on the 28->24 volt regulator. It is just a little black box circuit that my professor gave me. But I can say that it draws 10mA by itself (which seems right to me).

the attached image is the power circuit (excluding the black box regulator), the second chip (U2) is the one that keeps blowing out.

Do you guys have any ideas for what I could do to help keep this issue from happening?

I am not a wizard with electronics, in fact this is my first big project, It may be something stupid that I am doing, so if I say something that sounds really stupid to you guys, please just bear with me.

Thanks!

 

[Svein]

The schematic seems OK - have you checked the pinout against the datasheet (http://www.ti.com/product/UA78M10/datasheet/pin_configuration_and_functions#SLVS0595200)?

 

[duhuhu]

Yeah, I spent a couple hours with probes just checking the pinout to the schematic before I put power through it. And I double checked all the data sheets. but like I said, I am pretty new to this, so I don't understand a decent amount of what is on there.

 

[Svein]

I would try to disconnect everything connected to the output of U2. If it still blows, there is something wrong with your connections. If not, bring in one component at a time.

 

[duhuhu]

I'll try it out on a breadboard, I am tired of soldering and desoldering it all, and I will let you know what happens

 

[duhuhu]

It's the freaking black box... I hooked the ammeter to it when I turned the thing on this time, and as soon as I put the load on it, something clicked and it tried to draw 20 amps... Thanks for the help... I guess I will have to make the 28->24 volt myself.

 

[jim hardy]

The problem that I am having is that when I hook up the whole circuit with 28, It blows my 24->10 volt regulator. But when I run the original circuit on 24,

how much current does your circuit draw?
oopps i mis-read part number
78Lxx MxxURV series seems rated for 35 volts in but 0.75 watt
with 30 degC per watt, at 18 volts across the device you'd only handle about 160 milliamps.

watt max dissipation
http://www.ti.com/lit/ds/symlink/lm78l05.pdf
http://www.ti.com/lit/ds/symlink/ua78m10.pdf


it might be as simple as just adding a heatsink or changing regulator to a larger package.

sorry for mistake

 

[berkeman]

Does it blow U2 when you first turn the system on, or is it blown on the 2nd power up?

If the latter, then it could be a reverse current issue through U2 when the input power is turned off. A reverse diode from 10V to 24V at U2 can fix that problem if it is the cause of the damage.

 

[duhuhu]

how much current does your circuit draw?

My circuit should draw between .8 and 1.5 amps when it is fully assembled, but I am having troubles just with the voltage step down.

Does it blow U2 when you first turn the system on, or is it blown on the 2nd power up?

It is pretty much right away, The last time I tried it today was on a breadboard, and I had an ammeter attached to the black box. I found that the problem is with the black box because it was drawing 20 amps as soon as I put a load on it rather than just having it generate the potential.

 

[rbelli1]

My circuit should draw between .8 and 1.5 amps when it is fully assembled

I hope this is due to other devices drawing 24V current. If you are trying to draw 1.5A on the 10V rail you're going to have problems as the regulator is only rated for 1/2A. What else is on the 24V rail?

BoB

 

[duhuhu]

What else is on the 24V rail?

I am running a vacuum pump and some solenoids. The 10V rail is for some pressure transducers, lights and an accelerometer. The 5V rail is for some level sensors and thermocouple amplifiers.

Nothing is plugged into the board yet. I want to get the voltage regulators figured out before I have the possibility of ruining my expensive hardware

 

[rbelli1]

have you checked the pumps and solenoids? Can they work on 28V? Can you tell us the part numbers? The voltage difference is small so it would be best if you could get away with just plugging into the higher voltage.

BoB

 

[duhuhu]

have you checked the pumps and solenoids? Can they work on 28V?

These parts were given to us by the company who is sponsoring this thing. I don't have the part numbers on hand, but I can get them tomorrow. They were designed to work on 24 volts, so we don't want to mess with that if we can help it (These parts are too expensive to replace). 12.5% might be too big a jump in voltage.

 

[davenn]

I hope this is due to other devices drawing 24V current. If you are trying to draw 1.5A on the 10V rail you're going to have problems as the regulator is only rated for 1/2A. What else is on the 24V rail?

BoB

I agree with Bob

you are asking way too much from the reg both in input voltage and output current

tho the max input voltage for the 78M10 is 28V you can't expect it to be able to handle the max current of 500mA at that input voltage
even at 24V the limits are still being pushed
consider voltage drop from 24V to 10V across the reg = 14V drop x current you are trying to get 1A ( mid range of your 800mA to 1.5A)
that 14W of heat the poor reg has to try and dissipate ... no wonder it's dying

you always need to consider that datasheet ratings are for close to ideal conditions, any variation from that and you will run into problems

get the input voltage down to 20V or less and make sure that the reg is very well heatsunk and you may get 0.8 - 1A out of it

Now finally. DISCONNECT the -10V reg and the 5V reg and anything else connected to the 10V reg output and establish that on it's own the 10V reg is working as expected. then step by step add the other sections of circuit onto the 10V reg output and monitor what happens with each step
This will confirm whether or not there is any other wiring/design problems elsewherecheers
Dave

 

[jim hardy]

I am running a vacuum pump and some solenoids.

Inductors...

 

[duhuhu]

guys, most of the current is on the 24V line (in the *inductors*). On the 10 and 5V lines my best estimate would be that there is about 300mA of current being used at a constant rate. THE PROBLEM I AM HAVING THOUGH has been before I draw any current for those components on any line. Right now they are just empty plugs on my power distribution board that are not attached to the components. If you read my post above:

Nothing is plugged into the board yet. I want to get the voltage regulators figured out before I have the possibility of ruining my expensive hardware

I have not been putting a load on the 24->10 regulator. The problem was with the black box, and I am going to work on that today, and try to figure out a way to make a high current regulator that won't blow out my other regulators.

 

[jim hardy]

The problem was with the black box,

some regulators need a minimum load.
some can't handle inductive load.

we don't even know if the black box is an inductor based buck switcher or linear.

 

[duhuhu]

some need a minimum load.
some can't handle inductive load.

we don't even know if the black box is an inductor based buck switcher or linear.

That is why I am going to try to build one from scratch... This thing clearly can't do what I need it to without destroying my equipment. Would you reccomend that I do one that runs both the inductors and the power board, or would it make more sense to make two so that I can set up each one in a more optimized way and just manually jump them into the board?

 

[jim hardy]

i'd separate the inductors from the electronics with separate regulators.

Your 10 volt regulator spec sheet shows 28 volts as top of "recommended operating range"
section 7.3 at http://www.ti.com/lit/ds/symlink/ua78m10.pdf

so if 28 is really 28 and power dissipation is okay you might get away with just a diode to keep inductive "kick" away .

 

[duhuhu]

if 28 is really 28 and power dissipation is okay you might get away with just a diode to keep inductive "kick" away .

Well that is the kicker... We are expecting some unstable power with fluctuations up to 2 volts at an unknown frequency (They are refusing to tell us much about the system that we are mounting on). I was just going to put some big capacitors on it to help deal with that, but I am not too happy with how little they are telling us. Hell, we don't even know what side of the box they want our plug to be in. Or what kind of plug they want.