| Author |
 Topic  |
|
wasssup1990
Nobel Prize Winner
A Land Down Under
2261 Posts |
 Posted - Jun 05 2010 : 9:36:24 PM
|
| Yes that will work too, but of course you need to take in 16 samples before you can give an output. |
When one person suffers from a delusion it is called insanity.
When many people suffer from a delusion it is called religion. |
 |
|
|
codingplanet
Mad Scientist
United Kingdom
195 Posts |
Posted - Jun 06 2010 : 07:17:35 AM
|
| Yes, but that's not a problem. The CPU has nothing else to do in the meantime. I might increase the sample size actually, there's still a bit of noise and response it still excellent. |
|
|
|
wasssup1990
Nobel Prize Winner
A Land Down Under
2261 Posts |
Posted - Jun 06 2010 : 07:33:58 AM
|
| That's good. Is this your own CPU temperature and voltage meter? My i7 motherboard has a few pads where I can directly measure voltages like the VCORE. |
When one person suffers from a delusion it is called insanity.
When many people suffer from a delusion it is called religion. |
|
|
|
codingplanet
Mad Scientist
United Kingdom
195 Posts |
Posted - Jun 06 2010 : 07:50:57 AM
|
Yep, exactly right. The plan is to incorporate it into a homemade benchtable :)
For reading negative voltages, how about this:
When the lower voltage is 0v, the ADC input will be 1.25v. When the lower voltage is 0.5v, the ADC input will be 1v.
I could do with adjusting the resistor ratio so I get a better scaling but that's the general idea.
Edit: The diagram is wrong. It should show 2.25v right?
Edit2: No, that doesn't make sense, must be 0.25v?? I've confused myself, lol.
Edit3: No no, edit 1 is correct LMAO. |
Edited by - codingplanet on Jun 06 2010 07:56:45 AM |
|
|
|
wasssup1990
Nobel Prize Winner
A Land Down Under
2261 Posts |
Posted - Jun 06 2010 : 08:26:13 AM
|
Assuming there is negligable current flowing on the ADC input line, then the diagram is correct.
In order to read negative and positive voltages while utilizing maximal dynamic range you can simply set -Vref to the absolute minimum voltage and +Vref to the absolute maximum voltage you expect on the ADC. Your voltage reference in that case would be from -Vref. If you have no -Vref line on your PIC then it is probably connected internally to GND. In that case you will need a voltage level shifter circuit that uses an op-amp. Look it up. |
When one person suffers from a delusion it is called insanity.
When many people suffer from a delusion it is called religion. |
|
|
|
codingplanet
Mad Scientist
United Kingdom
195 Posts |
Posted - Jun 06 2010 : 08:44:46 AM
|
Well I cant use different vrefs cos then I'd mess up my other voltage readings. I think I need an op amp setup with unity gain and +2v offset, seen as I'm translating -2 to +0.5 to 0 to +2.5. DOing some research now.
Thanks again :) |
|
|
|
codingplanet
Mad Scientist
United Kingdom
195 Posts |
Posted - Jun 06 2010 : 10:05:30 AM
|
Putting 0.5v in gives 0v out. The way I've designed this circuit means that I don't need a negative voltage reference for the op-amp. However this means that the output voltage is inversely proportional to the input voltage, but this can be easily compensated for in software.
I guess I can connect this directly to my ADC? I probably need some kind of filtering still though. |
|
|
|
wasssup1990
Nobel Prize Winner
A Land Down Under
2261 Posts |
Posted - Jun 06 2010 : 11:35:00 AM
|
Umm have you measured the output impedance where you are taking the reading? I think the motherboard manufacturers would put a resistor in series with the pad you'll be measuring from. If that pad was directly connected to the VCORE and accidentally grounded then you might wreck your motherboard. You can do level shifting with resistors too and have high input impedance but your ADC input might have an internal resistance that is lower than the two resistors being used in your level shifter and would therefore sink the signal slightly. So it would be better to use an Op-Amp since you can design for high input impedance (from your VCORE) and low output impedance (to your ADC input). I'll leave the Op-Amp with you, I sense the force is strong with you. Hehe.
I mention impedance but what I really mean is resistance since we are only talking DC here. I'll tell you all this impedance stuff just FYI, in case you run into strange hair tearing out kind of problems.
See the new profile picture? Different eh? Hehe. I wonder what everyone else thinks. I might change it later. |
When one person suffers from a delusion it is called insanity.
When many people suffer from a delusion it is called religion. |
Edited by - wasssup1990 on Jun 06 2010 11:40:14 AM |
|
|
|
codingplanet
Mad Scientist
United Kingdom
195 Posts |
Posted - Jun 06 2010 : 11:45:18 AM
|
I will take the vcore point directly from the CPU side of the inductors, so my circuit will be connected directly to vcore. So I guess you're saying my circuit isn't quite ready yet? I'll have a ponder about it.
You've changed quite a bit since your last profile pic :D |
|
|
|
wasssup1990
Nobel Prize Winner
A Land Down Under
2261 Posts |
Posted - Jun 06 2010 : 12:06:46 PM
|
Yeah, now I'm a blond bombshell performer. LOL
Anyway I say just experiment and you'll learn from doing that. Try the two different methods I've mentioned, but also be aware of what I have cautioned you of. I know how expensive these i7 motherboards are, I've got one. The VCORE operates at low voltages but high currents, so a direct short will frizzle your motherboard tracks but hopefully a safety mechanism will kick in to prevent damage if it ever does happen. Good luck. I'm gonna hit the sack.
|
When one person suffers from a delusion it is called insanity.
When many people suffer from a delusion it is called religion. |
|
|
|
codingplanet
Mad Scientist
United Kingdom
195 Posts |
Posted - Jun 06 2010 : 12:16:19 PM
|
| I'll test it with an old S775 board. I'm thinking I'll need a voltage follower for my vcore. I'll do some experimenting tomorrow. |
Edited by - codingplanet on Jun 06 2010 12:16:56 PM |
|
|
|
codingplanet
Mad Scientist
United Kingdom
195 Posts |
Posted - Jun 08 2010 : 05:39:56 AM
|
| So neither my voltage follower nor the other circuit I designed work. But you probably already knew that. So what am I meant to do now? |
|
|
|
wasssup1990
Nobel Prize Winner
A Land Down Under
2261 Posts |
Posted - Jun 08 2010 : 05:49:01 AM
|
| Check your ground connections and ALL reference voltage lines. Studying for exams, can't help too much. |
When one person suffers from a delusion it is called insanity.
When many people suffer from a delusion it is called religion. |
|
|
|
codingplanet
Mad Scientist
United Kingdom
195 Posts |
Posted - Jun 08 2010 : 05:55:30 AM
|
| Already have done. I'm using an LM3900, could that be the problem? |
|
|
|
codingplanet
Mad Scientist
United Kingdom
195 Posts |
Posted - Jun 08 2010 : 06:15:44 AM
|
Ok so after 50 times checking it I still didn't find the problem. But I have now. I'm not sure who put this circuit together (ok, so it was me), but almost everything was in the wrong place. +12v was not connected to the amp & the ground wire had come out.
LMAO, maybe my voltage follower will work now :) |
|
|
|
Topic |
|
PIC Voltmeter
