Author Topic: Measuring amps  (Read 5103 times)

0 Members and 1 Guest are viewing this topic.

Offline HalvmandTopic starter

  • Regular Contributor
  • *
  • Posts: 111
  • Country: dk
Measuring amps
« on: March 24, 2015, 09:31:12 pm »
Hi all.

I have been in to electronics for about a year now, and the obligatory power sypply have been built and put to use.

So in the weekend i decided that i need a load.
It's based on the one from daves video, but suitet to what i have in my junk boxes. (Amazing how much you can accumulate in just over a year  )
 
Late last night i got it assembled for a test before i commit to build a chassis ( in leftover plywood of course  ;) ), but it didn't work. It was bed time any way ..

Today i got it to work though. somebody switched two wires  ::)
Allthough it is working, the current measuring display does not follow the current on my multimeter. i've checked with two and they agree. Over a range of 4 A the meter is out a whole amp.
You can see in the scamatic that op amp c is amplifying the voltage from the load resistor by a factor of 10. I need this because i only have a 0-30 v led meter. The meter works fine. I've checked it beforehand with my brand new AM520 DMM  :D
Is there anything inherently wrong with what I am doing here?

Thanks in advance. I love the forum and the videos  :-+
 

Offline HalvmandTopic starter

  • Regular Contributor
  • *
  • Posts: 111
  • Country: dk
Re: Measuring amps
« Reply #1 on: March 24, 2015, 09:39:22 pm »
I would post pictures. but can't seem to reply with attachments attached.

Forgot to mention pad 1,2 and 3 connects to a 50k ohm pot.
« Last Edit: March 24, 2015, 09:44:05 pm by Halvmand »
 

Offline Mr Smiley

  • Frequent Contributor
  • **
  • Posts: 324
  • Country: gb
Re: Measuring amps
« Reply #2 on: March 24, 2015, 10:48:11 pm »
Mmmm,

Your load goes from Ground, through the mosfet, through RL then back to ground  :palm:

 :)
There is enough on this planet to sustain mans needs. There will never be enough on this planet to sustain mans greed.
 

Offline HalvmandTopic starter

  • Regular Contributor
  • *
  • Posts: 111
  • Country: dk
Re: Measuring amps
« Reply #3 on: March 24, 2015, 10:53:47 pm »
To be clear, that load is the power supply beeing tested. my bad.
 

Offline Seekonk

  • Super Contributor
  • ***
  • Posts: 1938
  • Country: us
Re: Measuring amps
« Reply #4 on: March 24, 2015, 10:55:24 pm »
What is the value of RL so we can determine what voltages you are working with?  Is it off in a linear way or does it happen at some point. The current meter and RL connections can add an offset. These problems always have a very logical reason.
 

Offline mrkev

  • Regular Contributor
  • *
  • Posts: 225
  • Country: cz
Re: Measuring amps
« Reply #5 on: March 25, 2015, 02:27:32 am »
Ok so:
1. Is it following the  voltage at PAD1 ok or is there any other problem
2. Do you have the amplification at the IC1C right?
3. Knowing  RL, range of V-meter and current range would probably help....
 

Offline HalvmandTopic starter

  • Regular Contributor
  • *
  • Posts: 111
  • Country: dk
Re: Measuring amps
« Reply #6 on: March 25, 2015, 08:25:12 am »
RL= 0.1 ohm
I've designed for 10 A, although It Will rarely see more than 5.
Meter reads 0-30 v, 0.01 v res, 4-30v vcc.
The amplification at op amp c should be about correct. I've included the pot for calibration if it wasn't, but that doesn't make it linear.
For eksampel. I could adjust to 1 A and cal the led meter to one amp. When I then increase the current to say 4 A the led meter will read 5...

Could it be the resistor devider values thats too high? Would 9k and 1k be more sensible? I'm going for a gain of 10.

 

Offline rs20

  • Super Contributor
  • ***
  • Posts: 2317
  • Country: au
Re: Measuring amps
« Reply #7 on: March 25, 2015, 11:04:35 am »
So, you've got a known signal here and a wrong signal there. Divide and conquer, debug the problem. If I were you, I would connect my multimeter so that it was measuring the voltage across RL (if you only have one multimeter, take that one that's measuring current at the moment*). Then you can verify that your gain stage is working correctly. If you're finding that it's not providing a consistent gain of 10 across different RL voltages, then you know your gain stage is suspect. If the gain is a consistent 10x, you know the load resistor has a terrible tempco or something. Let us know what you find.
 

Offline HalvmandTopic starter

  • Regular Contributor
  • *
  • Posts: 111
  • Country: dk
Re: Measuring amps
« Reply #8 on: March 25, 2015, 02:18:02 pm »
Thanks for the replys.

I have taken the measurments.
Im = Current measured by the led meter
Iam = current measured with multimeter in series.
Uin = Voltage at op amp c non inverting input
Uout = Voltage at op amp c output

Im = 1 A
Iam = 1.12 A
Uin = 120 mV
Uout = 0,93 V

Im = 2 A
Iam = 1.60 A
Uin = 170 mV
Uout = 1.47 V


Im = 3 A
Iam = 2.20 A
Uin = 238 mV
Uout = 2.51V

Im = 4A
Iam = 2.82 A
Uin = 314 mV
Uout = 3.69 mV

I noticed that the led meter doesn’t correspond to the Uout measured with the multimeter. That is a bit wierd since it followed the dmm to the last digit when i tested it. Hmmm.
Otherwise the whole system seems to be off, and each step add to the final measurement on the led meter.

The temp coefficient does not seem to be a problem below 3A but above it does.
Am i right on this? It does not heat up much. Maybe 10 deg. C. at 4 A

I hope you guys see something I dont.
 

Offline HalvmandTopic starter

  • Regular Contributor
  • *
  • Posts: 111
  • Country: dk
Re: Measuring amps
« Reply #9 on: March 25, 2015, 02:48:47 pm »
I might have found something.

When the power supply under load reaches 1.5 - 2 A it goes in to crazy mode.
See the picture. thats 5 v / div. It will continue to do this until it's down to 0.5 A. Surely this has something to do with the readings at higher amperage..

 

Offline HalvmandTopic starter

  • Regular Contributor
  • *
  • Posts: 111
  • Country: dk
Re: Measuring amps
« Reply #10 on: March 25, 2015, 02:50:22 pm »
Btw it is a 4.5 A IBM lab top supply i am using. An old one...
 

Offline HalvmandTopic starter

  • Regular Contributor
  • *
  • Posts: 111
  • Country: dk
Re: Measuring amps
« Reply #11 on: March 25, 2015, 03:13:29 pm »
It's the power supply.

Now it works perfect  :D Display is stable and reads just what the multimeter does. Just as long the supply is stable.

Good investment, that old scope.
 

Offline macboy

  • Super Contributor
  • ***
  • Posts: 2252
  • Country: ca
Re: Measuring amps
« Reply #12 on: March 25, 2015, 05:10:35 pm »
It's the power supply.

Now it works perfect  :D Display is stable and reads just what the multimeter does. Just as long the supply is stable.

Good investment, that old scope.
No bench should be without a scope. Even a simple old 10 MHz analog one is a valuable debug tool, especially with circuits like power supplies and loads where the control loop might not be stable.
 

Offline kezat

  • Contributor
  • Posts: 29
  • Country: us
Re: Measuring amps
« Reply #13 on: March 25, 2015, 05:36:31 pm »
In my experience you need a bit more resistance in the system then 0.1 OHM or you get osculations. Do you have a 3-10 OHM power resistor you can add in series to the power supply under test. If not just a meter or 2 of relatively thin wire might do it.  Might be a better way to do it then adding a resistor but that is what worked for me.

The deference in the two readings you are getting at the same point with the two meters may be how each handles the osculating voltage your trying to measure.
 

Offline HalvmandTopic starter

  • Regular Contributor
  • *
  • Posts: 111
  • Country: dk
Re: Measuring amps
« Reply #14 on: March 25, 2015, 06:45:48 pm »
macboy: I can commit to that. Also the mac part..

kezat: The 0.1 ohm resistor was chosen because I figured it would dissipate less heat and thus be more stable. I measured the resistance of the 0.1 ohm and a Chinese 50 w 8 ohm  when fresh out the freezer  and on a hotplate. That chines one was terrible. Not to mention that at 5 A it would dissipate 200W. 3 ohm at 5 a is still 75 W. That resistor alone would break the whole budget.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf