It's actually +8.5 and -8.5 and I get even more on my board (+-9.5)
For example the Velleman K8042 Kit seems to work for my applications
What I did....Still works except for a minor flaw where the plastic melted a little. 
Just some added data pins....Still is backwards compatible with normal SD cards IIRC.
Yes, try googling "boost converter".
1 Like
You also need a 24V center tapped power transformer to connect to the K8042
For the moment I'm using a 0-30V 2A power supply to generate my 5V. How about using two of them, one generating +9V referenced to the arduino ground and the other -9V referenced to the arduino ground ? That would make everything more stable I bet
That will be a much better solution
1 Like
Yeah I just wanted a final solution that only needed one micro usb cable or something very simple like one power supply but this is gonna be much easier
I think all the unstability I have is caused by the potentiometer. The voltage between my ground and the non inverting input of my second op amp is more or less put to 1.1V but in reality, it fluctuates a lot, very fast, between 0.8 and 1.4V. It could explain why the AI2 moves so much but doesn't explain why I have saturations like +7V sometimes...
It's a bit better now with another pot. So I can pretty much say the voltage is managed for my current-voltage characteristic.
Now I have to find a solution to measure the current I have in my SD pins with the analog inputs of the arduino.
Is that spec or did you use a meter with possibly no load between VCC/VDD and GND?
Hi, @rayou23
Can you please post some images of your project?
So we can see you component layout.
Have you tried a 10uF capacitor and a 0.1uF capacitor in parallel from the pot wiper to gnd?
Thanks.. Tom... 
1 Like
You're right I used a meter with no load btw Vcc and GND but I don't think it has any importance here for two reasons :
- I can power my op amps with +/- 8.5 or 9.5, it doesn't matter as long as I am superior to the output I wanna have
- I am not using the MAX232 for my project. I am now using : one power supply for +9, one for -9, and the 3.3V of the Arduino Due
I will post some images in a few hours, sorry for the late response.
The pot isn't causing any problems now. My only problem now is to measure the current that flows in my SD pin now that I have my voltage. So I can draw the characteristic!
But for sure a capacitor could have helped a lot with these voltage unstability!
Just to be sure I'm not overcomplicating : it wouldn't work simply doing I=(Vin-Vout)/R4 to get my current flowing in my SD pin right?
EDIT : Since Vout goes "backwards" in the 3.3-0V range I would of course re-adapt the voltage to the -3.3-3.3V range before doing the operation to get I
The reasons that you don't know are how an unterminated INPUT_PULLUP pin will show 5V but if you draw even a tiny current from it, you get < 1V.
More extreme: Take an led, put the short leg (normally ground) and put it in A0 and pit the long leg in GND. Write a sketch that reads the led once a second for 10 seconds to see how the V rises up to the led forward V.
You point the led above at a house light and run the sketch.
As long as I don't let the current flow, I get a rise to 1.8V just from a led pounted at a CFL!
Input pins only interact with the circuit connected to them when they are read, they sample (datasheet says) 1 uA otherwise, electrically neutral.
Digital circuits can do this in their electrically noisy way.
One read per second wont drain the lead but see what V the light that drives the led current can maintain at 100 reads per second, 1/10th of a milliamp per second drain.
AFAIK opamps need current from their inputs and amplify/compare that which is a load condition, not a static one. The current from the led is called ghost current for good reason yet in seconds it can fill a piece of wire to Vf, with a blue led maybe over 3 volts at almost 0 amps.
1 Like
Using drivers like those in the MAX is really neat. Serial through cables up in the ceiling between offices does take some juice to drive, the driven voltage has to have current flow to read correctly.
Looking through a TI MAX232 datasheet at the Electrical Characteristics -- Driver table I see that TEST CONDITIONS are specified to get true ratings.
The conditions are not as simple as just the 3K Ohm resistor to ground.
What does the sheet for your MAX232 say? This says high level TYP 7V amd low level TYP -7V.
With TTL serial, 0 bits are HIGH and 1's are LOW. When there's no signal the line is held high, less susceptable to induced noise. Cable=Antenna.
This is no total solution for you but maybe now you get better numbers and less wild swing responses to pot turns? A weak power source is like a straw in the wind.
Power is Watts is Volts x Amos. Volts alone may be a tiny spark and gone.
1 Like
I'm not sure I understand everything you're suggesting/asking so I'm going to take it point by point:
- You're talking about that max232 again when I'm not even using it in my circuit anymore. So you're suggesting it would be a much better power supply ? I don't question your call but I just don't see how. Of course the long serial cables would require current to work correctly because the longer the cable the higher the resistance. But why are you talking about these ?
- I am using the TI Max232 so we saw the same test conditions. I cabled everything like it is suggested in the datasheet. It's strange you saw +/-7V because I am seeing +/-8.5V
I am not seeing the use of the max232 anymore since I'm not using it for its original purpose or anything anymore (converting voltage levels TTL to RS32). I now use two power supplies.
But if you're talking about all this for my general knowledge in electronics and to make me understand why it didn't work with the 232, thank you very much for all the infos!
You read 8.5V with no load, just voltmeter says by your words and say that you met test conditions.
I didn't read any 7V. That is the TYP-ical Voltage TI got from a correctly wired DRIVER, the part that outputs the signal you were going to tap before you abandoned what didn't work for you.
1 Like
I thought we solved the current measuring issue in your other topic
Could be "X-Y".
2 Likes