Max6675

bhaan

I am working on a project that deals with reading thermocouples. ·I believe the problem lies within my power supply somehow.·

I made a PCB with 20 MAX6675's on it that all read an individual thermocouple.· The MAX6675's are selected by shift registers and they share data and clock lines.· For initial testing purposes I have made a 5 volt power supply on a breadboard and connected that to my pcb.· My 5 volt power supply is a 7805 voltage regulator which is supplied by an "SR Components" class 2, 12 volt adaptor that I plug into the wall.· I have a capacitor before the regulator on the power supply and after it.· The problem I am having is my 5 volt regulator will get pretty hot at times.· If I plug in my AC adaptor, wait a few seconds, then connect my pcb to the 5 volts my regulator does not get hot.· However if I have my pcb connected to the 5 volt power strip and then plug in my AC adaptor, my 5 volt regulator will get pretty hot.· Do you know why my 5 volt regulator would be getting hot?· I don't believe I have a short anywhere, are my MAX6675's drawing too much?

Thanks,
Brian

Tracy Allen

Hi Brian,

It sounds like a problem of latchup, which can occur when there are two separate power supplies in the system. For example, I'm thinking the MAX6675s have one power supply and the BASIC Stamp and/or shift registers have a second separate power supply. Could there be anything like that going on? Do you have a multimeter to make some measurement?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

bhaan

I have one wall adaptor that comes in, then I have 3 total regulators(15 volt, 6 volt, and 5 volt). Nothing draws much current on any of the circuits. I measured the circuit and they all measure about right except the 15 volt one was a little low(~11.5 volts). I'll try to post a simple schematic later on today.

During my testing though, I unhooked the 15 volt and 6 volt circuits and regulators. My 5 volt regulator would still get pretty hot depending on which order I powered things up by(plug the regulated 5volt line into the pcb before or after plugging the adaptor into the wall). On the times it got hot, if I remember correctly, the board was drawing about 83mA, and on the times it didn't get hot it was around 2 or 3mA. I believe my 7805 should be able to handle about 1000mA, and all that I have powered on the board was the MAX6675's.

Tracy Allen

Does your system work fine as far as reading the thermocouples and everything else, once the power is hooked up in the good order? I just wonder about the printed circuit board, if there might be a wiring error or short circuit somewhere. Do any of the MAX6675s get hot too, when the power goes high? All that current has to be going somewhere.

Is there a common ground among all three of those power supplies, and are they all powered by the same wallwort?

Is there a BASIC Stamp in this circuit and where does it get its power? What else is in the circuit that requires so many different power supplies?

Sorry for so many leading questions.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

bhaan

Here is the schematic of what I have right now.· There are other components that will go on the board, but as of right now this is all that is connected on the board.· My board is a 4 layer board, so the middle two layers are power(5Vdc) and ground.· One thing I forgot to mention earlier, is that I have not added my 0.1 micro farad capacitors to the MAX6675's yet.· I know the bypass capacitors are recommended and I plan on adding them, but I did not think they would cause this problem.

I have not tested reading the thermocouples yet, but I will soon.· I have felt around on the MAX6675's when the regulator gets hot and I have not been able to find any of them feeling warm.· The three regulators all share a common ground and are all powered by the same wallwort, but as of now, only the 5 volt regulator is hooked up.·

I have not added my basic stamp to the circuit yet, there is only sockets where my other components will go.

Eventually the 5 volt regulator supplies 24 MAX6675's and 3 shift registers, the 7 volt regulator is used to power the basic stamp(BS2p40), and the 15 volt regulator is used for a 4-20mA control loop I have.· As you might be wondering, my projects goal is to control an actuator based on temperature data.· I have a full schematic I can post if needed, but as of right now my board only powers the 5 volt circuit with only sockets where the shift registers go like the attached·pdf schematic shows.··Everything on the 7 volt and 15 volt circuit have been disconnected.·

bhaan

I put my shift registers on my board and tried reading the thermocouples with my basic stamp, and it appears to be reading pretty good.· I am confused because after adding my shift registers the regulator did not seem to get hot no matter how I hooked up the power supply.· Why would adding my shift registers not make the regulator get hot?

Tracy Allen

The chip select and clock inputs on the MAX6675s are supposed to be digital inputs that are at or very near either zero volts or 5 volts. If they are at some intermediate voltage, (especially around 1.4 volts for chips with "TTL" input levels) the internal transistors can drop into an analog linear mode and the chip draws excess current. If the input is floating, not connected at all, as was the case before you added the shift registers, they can float to that unfortunate level. That evidently happened if you switched on the power supply in one way, but this sort of thing can depend on how you wave your hand near the circuit board or a whole bunch of factors. The bottom line is that you want the CS\ and CLK inputs to the chip to be set at definite levels. As the shift register. enforces definite voltages that seems to have taken care of the problem. And the Stamp enforces the level on CLK.

I would add that for the same reason, you should always set unused BASIC Stamp pins to be outputs (either high or low), or use a pullup or pulldown resistor. The whole point of that is to bring the pin to a definite low or high level. Otherwise the chip will consume excess power in a manner that depends maddeningly on the cosmic flow.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

bhaan

Thank you for your help and advice. I'll do further testing this week, but what you said makes sense now. Your help is very much appreciated.

Thanks,
Brian

bhaan

One other thing, since I have 24 MAX6675's, do I need a bypass capacitor beside each one?