EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: soligen on March 22, 2016, 11:19:21 pm
-
I built an electronic load project similar to what Dave and others have posted (schematic attached). My additions are an ADC and Arduino nano to display amps/volt/watts, and also control turning the load on/off. I thought I had it wrapped up, but I have one additional issue that I could use some help on.
If the electronic load is powered on, then when I connect the DUT, my LCD display (generic 16x2 display) gets garbled. If the load device and DUT are both mains earth referenced, or are both floating, then I don't have the issue. It seems to only happen when one is earth referenced and the other isn't.
Putting the o-scope on the 5v line, I see some big spikes when I connect the DUT and the diaply garbles. The worst of the spiking only lasts 100 - 200 ns, but sometimes swings all the way below ground with 10ish volts peak to peak. The 5v line was measured directly on the nano board, and I disconnected the fan completely to eliminate any inductive load noise issues.
I added a 10uf ceramic cap directly to the nano board (C12), and I added a 220uf electrolytic between D5 and the nano VIN pin (not on schematic). Still no help.
I am guessing that this 5v noise is causing my issue. Any ideas on how to clean it up? Does it require some massive amount of capacitance to clean up?
I could work around this issue by always connecting grounds together before powering on, but seems like it would be best to suppress the spikes.
On a related note, would it be best to have a test device like this always mains earth referenced, or leave it floating? I initially left it floating so I would have no concerns about connecting it to anything, but I'm a noob - there may be other considerations I don't know about. Since the o-scope is mains earth referenced for safety, it makes me wonder if this should be too.
Thanks
-
Forgot the schematic. Here it is:
-
It seems to only happen when one is earth referenced and the other isn't.
How exactly are you powering both devices when the issue occurs ?
If the issue happens when only the DUT is earth-referenced, can you try earth-referenced DUT + battery-powered load ?
I'm thinking that the power supply you're using to power the load is not completely isolated and you're charging/discharging a capacitor when you connect the grounds.
-
Excellent suggestion, and in fact I think you have it. the device works fine on powered by battery, however I don't want to battery power it because I am relying on a fan for cooling, and I think that will suck down the battery too quick.
What I was using was an old Nikon battery charger whose main DC rail is 8.7V. I thought it was isolated - sure seemed to be based on my multi meter readings and examining the board. There is a small transformer and obvious spacing between the high and low sides with an optocoupler for feedback.
Any suggestions on where to get a small 120VAC to 9ish VDC power supply that is 100% isolated?
-
Pictures of the charger you're using may help. I'm thinking it has some sneaky capacitors. Does it have any mains ground connection at all or is it 2-wire only on the mains side ?
-
Hi,
The grounding should be done carefully.
I would make the bottom of the current sense resistor a 'star point'. The negative input should be connected here. The connection from Arduino Nano ground should go there. This will make sure that the transient that are generated will not upset your circuit.
Can you post a picture of your construction?
Regards,
Jay_Diddy_B
-
Pictures of the supply will take some dis-assembly, so I cant get them quickly. It is a 2 wire connection to the charger, but I run all three to my enclosure (metal) so the enclosure is grounded.
Here is a picture on my board layout. This is V2 of the board, and I am using V1, however the biggest difference is how the fan connects, and i am testing for now without the fan. My board is single sided, and the top side connections are done with wire jumpers. Note that 3 of the top side wires are just stitching ground together as I thought the ground plane a bit spread out otherwise. The load's main current path (right side) is also reinforced with 14ga wire soldered on the bottom of the board.
Some stuff on the board is extra - extra pads and test points. Ignore the TO-220 device with air wires. This is off board connected to the center screw terminal with 16ga wire. Also the shunt resistors shown on the board are really off-board connected using the top screw terminal.
-
Here is where I am at. Everything seems to work fine when I power the unit with a battery. When I power through the little hacked battery charger, then the display gets garbled (the black plastic unit clamped inside the enclosure). For another test I used a regular transformer I have in my salvage bin, rectified it, and passed it through a dc-dc converter to get voltage down to under 9V. The transformer I used is in the larger view picture, and you can see the little board with the rectifier and a cap, and the DC-DC converter. Seems like using a transformer like this that the circuit should be isolated, however the display garbles running from this power source too.
just to take the DUT out of the equation, the display garbles even with no DUT connected. All I have to do is touch a lead from my negative terminal to my earth grounded enclosure, and garbage appears.
I don't know what to try next, or if I should live with it as the work-around is to connect grounds before powering on.
-
Where are you powering the LCD from ?
Have you tried 6v zenner's across the Arduino power supply ? That might fix the issue.
-
Hi,
I would try shortening the leads on the LCD display.
What do other members of the forum think?
Regards,
Jay_Diddy_B
-
USB? Is it connected to the PC when you test?
-
If you are using the USB to develop the code, can you try and run it stand alone?
LCD wiring does appear fairly long. No shield. No core. Do you know if the unit, besides the LCD is working and that it really is just the LCD that is having problems?
-
The LCD is powered from the arduino 5V regulator. The USB is only for programming - it is disconnected during operation. I just left the cable in place because it is hard to re-connect in the tight quarters.
I don't have a 6V zener handy, so I cant try that for a while - until I can source some. However, the spikes pull the supply to ground too sometimes (20 - 50mHz of spikes). It might trim the highs, but what about the lows?
-
Try to post a screenshot of a trace of the Arduino 5V and the 9V input when you connect the terminal to the ground.
Also measure the voltage with a voltmeter in between the probe and the ground and see what you get. If you place a 1M or 10M resistor in between the terminal and the earth before connecting them, do you still get the issue ?
BTW Where are you getting your 7.5v and -2.5v from ?
-
7.5V comes from VR1 (TL431) . -2.5 comes from the riser board attached at the lower right. It's just a totem pole into a charge pump, regulated by another TL431. Its on a separate board simply because I had soldered it up while I was still breadboarding.
Tonight I'll see if I can figure out how to get screenshots from the scope
-
Thanks everyone for the help. I just thought I would post where I ended up. I tried several other things, including removing parts to simplify the circuit and trying a 3rd power supply. Nothing changed. All I can think is that my supplies are not as isolated as I thought.
Since the work around is pretty simple (attach the ground to the DUT before power on). I decided to let this go and move on since the device does what I need it to as long I I obey the work around, which is probably good practice anyway.
Just one last question. For a device like this, is it considered best to have my circuit ground tied to mains ground, or left floating? Is there a safety issue one way or other?
Thanks