Handling high currents (~75A)

Go To Last Post
37 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi - I'm working on an active load for testing high power batteries. My goal is to be able to load it with about 75A. The power dissipation is not a problem. However, I'm realizing that the current, well, is... I had originally planned on running the current through a PCB to a MOSFET mounted on a mega heatsink. But I think I'm going to melt the PCB unless I get a fancy PCB with really thick copper. So I think I should run all high current through discrete wires. At 75A, according to my reference materials, I'm going to see about a 110 degree C temperature rise in 10 AWG wires. Ouch! But maybe this is just fine? I mean, that isn't going to be damaging anything. Sure it it is power loss - but the thing is designed to be a load. So who cares.

I need to have a sort of H bridge carrying this current. I guess I can just solder big fat 8 or 10 AWG wires to gigantic MOSFETs (probably TO-247s), and though the MOSFETs will heat up just from the wires heating up, I think I should be fine, as they are designed to operate at up to 175 C. Wires will have rubber or Teflon insulation so that they can handle the heat. I should also note that this load will run for hours at a time - so it really needs to be able to take the heat! Oh, I should mention that there will be fans running on this system at all times.

Does this make sense at all? Anybody have any words of wisdom for dealing with this much current? Thanks!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

The power dissipation is not a problem. However, I'm realizing that the current,

Is power not directly related to current then? I though it was P=IV or P=I^2R or something?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

clawson wrote:
Quote:

The power dissipation is not a problem. However, I'm realizing that the current,

Is power not directly related to current then? I though it was P=IV or P=I^2R or something?

I was referring to the load's power, not the power dissipation in the wiring.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Don't use just one FET. Use several in paralel. TO-220 leads melts at @ 75A, so it's a good thing to keep the current away from that.

Felipe Maimon

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The length of the wire is important. The shorter the better. In 12 volt 110 watt radio transmitters with up to 35 amps draw we consider 8 AWG to be the minimum wire size. This should only be exceeded for really short wires a few inches long. If you use any connectors and want them to last for more then a few times, you will have a hard time finding 75+ amp connectors that can take 10 AWG wire (this tiny wire gage will be lost inside large gage connector pins). In any case wiring from 8 gage to 4 gage is commonly available. The high power stereo after market has lots of high current accessories for 12/24 volt applications. If you avoid the big name brands the costs are reasonable. Google "Anderson Connectors" for some other alternatives.

If you pull 75 amps though 10 AWG for a long length of time, you risk melting the insulation, especially any place the wire touches other structures. You might even manage to cause sparks or start a fire. If your batteries produce hydrogen like a lead acid battery can, you might blow up a battery if something else sparks or catches fire. You can forget having anything that even remotely resembles a safety certification unless you use the correct wire size and connectors for your application. If you sell something like this you are risking major lawsuits.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The starter motor current in an automobile is on this order of magnitude. Go open your car hood and take a look at the wires going to the battery. That is what you will need, if not bigger.

PCB - no way!

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Curtis golf cart controllers are rated up to several 100 amps. Fork lifts have some awesome fat cables in the controller.

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ka7ehk wrote:
PCB - no way!

Actually, the OSMC (Open Source Motor Control) boards can handle more than 100A continuous with a 4 Oz layer PCB.

Felipe Maimon

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Mike B wrote:
The length of the wire is important. The shorter the better. In 12 volt 110 watt radio transmitters with up to 35 amps draw we consider 8 AWG to be the minimum wire size. This should only be exceeded for really short wires a few inches long. If you use any connectors and want them to last for more then a few times, you will have a hard time finding 75+ amp connectors that can take 10 AWG wire (this tiny wire gage will be lost inside large gage connector pins). In any case wiring from 8 gage to 4 gage is commonly available. The high power stereo after market has lots of high current accessories for 12/24 volt applications. If you avoid the big name brands the costs are reasonable. Google "Anderson Connectors" for some other alternatives.

If you pull 75 amps though 10 AWG for a long length of time, you risk melting the insulation, especially any place the wire touches other structures. You might even manage to cause sparks or start a fire. If your batteries produce hydrogen like a lead acid battery can, you might blow up a battery if something else sparks or catches fire. You can forget having anything that even remotely resembles a safety certification unless you use the correct wire size and connectors for your application. If you sell something like this you are risking major lawsuits.


Mike - do remember that my application is significantly different than that of the ham radio enthusiast's. Heat is expected. Welcomed, as it'll take some of the load off of my MOSFET active load. I just have to keep it under control. The 100 degree C temperature rise figure I gave is for a wire in still air. With a fan blasting on it, it will be significantly lower, I suspect. I think 10 AWG will be fine, especially if I have a Teflon insulation on it (melting point is over 300 C).

You are right that I will have to be careful about connectors. Anderson connectors seem like a decent choice. I'm also thinking about using some bullet style connectors from the RC industry. Some of them are rated for over 200A continuous. That's pretty serious stuff.

This is for a one-off project for myself - so I'm not worried about getting it certified by anybody but myself. And I'm not a big fan of lawyers - so it's unlikely that I'll be suing myself either.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ka7ehk wrote:
The starter motor current in an automobile is on this order of magnitude. Go open your car hood and take a look at the wires going to the battery. That is what you will need, if not bigger.

PCB - no way!

Jim


Automobiles are designed for a significantly different environment than what I'm designing for... I can't imagine I need to go that extreme.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

nleahcim wrote:
Mike - do remember that my application is significantly different than that of the ham radio enthusiast's.
Wrong, that is a commercial radio standard practiced by people that are more serious than enthusiastic :). My reply was based on the little information in your original post.
nleahcim wrote:
so it's unlikely that I'll be suing myself either.
Don't miss out on that, I expect it will be the next new fad in designer lawsuits :lol:.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

May want to look into crimping all connections, or exactly how to solder for these conditions.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

This raises memories: When I was around 12 year old my father charged a car battery in the basement. He used a jumper cable to connect the battery to the charger, who had a connector/contraption to go into the cigarette lighter receptacle. While rummaging around I caused a short circuit and the battery discharged across the jumper cables. The cables burnt out completely, only some melted plastic remained.

My father told be, that in case of problems with electricity you just had to disconnect the faulty equipment and all would be well, so I disconnected the charger from the mains.

However, high currents should be worked with correctly. In your case, I understand, you want to test battery performance under high load (70A). This will necessarily generate a considerable amount of heat, who will have to go somewhere. At 12V you'll generate close to a kW, this has to be designed properly to dispose of the heat.

There are special 'resistance wire' with defined resistance and able to survive high temperatures. The interconnect cables and switches/mosfets should be properly designed and of sufficient caliber to be handle the current with not much heating.

Markus

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

At such currents, it seems viable to investigate moving away from traditional wires and look at alternatives. Using solid aluminium bars to carry the current, for example. Not square profile, L profile or flat will give you a way to mount it, and will have the added benefit of a larger surface area, which means built-in cooling too.
Yes, aluminium has higher resistance than copper, so if you can spring for copper bars, that'll be better, but it's not needed. Just make 'em a bit bigger. I'm freelancing on a semi-regular basis as a lighting engineer, working with 63A (or 125A), 400V cables for the massive power consumption. In those cables, it's not copper anymore, due to both weight (they're heavy enough as-is!) and cost.

Just don't forget your screwdriver across the terminals, or it'll be welded stuck... quite scary stuff!

Also, allowing a 110° temperature rise seems very high, in either case. Keep it cool or you'll accidentally melt the floor or something.

Come to think of it, copper pipes with circulating water inside is awesome for keeping away heat. Works best with non-DC power though (due to skin effect). That's what that insane overpowered AVR-driven induction heater used... it's lurking somewhere on these forums, can't remember where atm.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
active load for testing high power batteries

Is this your hobby or a commercial project? All you want is to make sure the battery is not damaged or you want to measure its performance under 75A? Or perhaps there is a ISO directive you want to comply to?

With 12V Lead-acid at 75A it would give ~900 watts of heat dissipation outside and depending on the internal resistance, some energy from inside. I suppose you do not want to measure voltage drop (you can check it in one second, dissipating 1kJ), but the energy taken outside (requires discharging and dissipating several MJ).

If you need a cell energy meter used under ~75A load, it is time do decide what accuracy of current/voltage/energy you need and must it be exactly 75.000000A? If all that is required is a 75A +-3% load with lead-acid battery and 2% energy measurement accuracy, then you must know lead-acid voltage is quite stable and with 0.1C current will drop much less than 10% (check the datasheet) through 50% working cycle.

You can easily buy a precision resistor or make a heater of adequate known calibrated resistance (for example 1%) using Wheatstone. With a ~75A relay and integrating the voltage (with 10 bit AVR AC) you will be able to measure energy dissipated at the accuracy of 2-3% with current kept much below +-5% level. The relay will not influence the measurements, because the voltage drop across it will be insignificant and known.

The relay and heater can be mounted on the cell with short 2x15cm wires, while the interface (LCD display with uC and control panel) can be placed on any length of wire externally. Remote voltage measurement is very easy. Same as relay control.

My advice is to use a small ~1.5L water container and mount your heater inside. This will let you perform discharge test without any maintenance well above 1 hour, all without fans, 110*C MOSFETS, noise and dust cleaning.

No RSTDISBL, no fun!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Brutte wrote:
Quote:
active load for testing high power batteries

Is this your hobby or a commercial project? All you want is to make sure the battery is not damaged or you want to measure its performance under 75A? Or perhaps there is a ISO directive you want to comply to?

With 12V Lead-acid at 75A it would give ~900 watts of heat dissipation outside and depending on the internal resistance, some energy from inside. I suppose you do not want to measure voltage drop (you can check it in one second, dissipating 1kJ), but the energy taken outside (requires discharging and dissipating several MJ).

If you need a cell energy meter used under ~75A load, it is time do decide what accuracy of current/voltage/energy you need and must it be exactly 75.000000A? If all that is required is a 75A +-3% load with lead-acid battery and 2% energy measurement accuracy, then you must know lead-acid voltage is quite stable and with 0.1C current will drop much less than 10% (check the datasheet) through 50% working cycle.

You can easily buy a precision resistor or make a heater of adequate known calibrated resistance (for example 1%) using Wheatstone. With a ~75A relay and integrating the voltage (with 10 bit AVR AC) you will be able to measure energy dissipated at the accuracy of 2-3% with current kept much below +-5% level. The relay will not influence the measurements, because the voltage drop across it will be insignificant and known.

The relay and heater can be mounted on the cell with short 2x15cm wires, while the interface (LCD display with uC and control panel) can be placed on any length of wire externally. Remote voltage measurement is very easy. Same as relay control.

My advice is to use a small ~1.5L water container and mount your heater inside. This will let you perform discharge test without any maintenance well above 1 hour, all without fans, 110*C MOSFETS, noise and dust cleaning.


This is for a hobby project. I'm testing lithium-ion and lithium-polymer cells, not lead acids. My goal is to get some pretty precise measurements from this thing. Currently, I have a 16b ADC doing measurements, and a 16b DAC controlling the active load. I don't have a specific goal in mind - more just that I want to maximize whatever I can get. I am currently planning on measuring the current through the cell (with kelvin shunt resistor), voltage of the cell, temperature of the cell, and temperatures of various other components to make sure they don't die. I am also interested in measuring the volume change of the cell, but I haven't figured out a nice way of doing this, and it's also not a priority.

I quite like your idea of heating up water. By my calculations, discharging my largest cell would increase the temperature of 1.5L of water by 16 C. But getting everything in the water will be tricky. It may make sense to look at a non-conductive liquid. But then it'll be icky to work on. I think I may try to have my heatsink dipped into some water, but have all the electronics dry. This'll probably still require a fan for the wires, but not as big of one as the ambient temperature will stay cooler.

Also - a simple relay for switching off the load won't work here. I'm planning on putting the cell in the middle of an H-bridge. I want to test the cell's charging characteristics, as well as discharging.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Not sure if it would help in the charging side, except for an intitial current limit, but many years ago I developed a battery capacity tester for 12V SLA batteries, which would be disccharged at a constant current to a cutoff voltage of 10.8V over a 90 minute plus time period. <90 minutes was a failure.

With 80-200AH batteries this was quite a load.

What I did was a constant current source based load, comprising 8 TO3 transistors on a huge 1 degree C/W fann cooled heatsink loaded by banks of 12V 50W Halogen Lamps that were bank switched into operation depending on the required dishcharge current.

The advantage of the Lamps over resistive dummy loads is that at least some of their energy (dissipation) is converted to light instead of heat.

You do need this light to escape, else it's the same as a resistor.

It worked well and the current range was 125A or so.

Food for thought.

Ron.

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
measuring the current through the cell (with kelvin shunt resistor)

Good choice, did you try www.lem.com ? No voltage drop, no heat.

Quote:
I am also interested in measuring the volume change of the cell

Could you write something more about it?

How did you figure out this:

Quote:
But getting everything in the water will be tricky

from this sentence?

Quote:
1.5L water container and mount your heater inside

I did not suggest to put anything else but a heater inside. Just like an electric kettle works. I never suggested to put hands or electronics down there.. Just a single, small resistor - nothing else.

Quote:
increase the temperature of 1.5L of water by 16 C

Well, the subject suggested discharging high power and high capacity (Pb-Acid, 12V ) cell with 75A, that is why i suggested 900W for an hour(0.9kWh of energy). How much energy can your largest cell hold?

Quote:
It may make sense to look at a non-conductive liquid

What is the highest voltage of your cell? I have a 230VAC in my kitchen 2400W kettle and I cannot see a problem.

Quote:
fan for the wires

???

Quote:
I want to test the cell's charging characteristics, as well as discharging

Sorry, I did not notice it in your post. I thought you were only thinking about discharging it at 75A.

Quote:

I don't have a specific goal in mind - more just that I want to maximize whatever I can get

Well, usually when you measure something, you use that information and make a decisions - controls. If you do not have a specific goal to maximize, then you actually do not know which information is valuable. So you simply cycle the cells and what next?

No RSTDISBL, no fun!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Brutte wrote:
Quote:
measuring the current through the cell (with kelvin shunt resistor)

Good choice, did you try www.lem.com ? No voltage drop, no heat.

I'll take a look. I tend not to trust magnetic sensing though - too prone to noise. I'm going to have a big fat switching power supply on this thing too (for charging the cell) so there will be plenty of noise.
Quote:

Quote:
I am also interested in measuring the volume change of the cell

Could you write something more about it?

Lithium-polymer cells expand when warm. (and maybe other times too - I haven't seen much good literature about this). I've heard they can expand by as much as 10% during normal usage.

Quote:

How did you figure out this:
Quote:
But getting everything in the water will be tricky

from this sentence?

Quote:
1.5L water container and mount your heater inside

I did not suggest to put anything else but a heater inside. Just like an electric kettle works. I never suggested to put hands or electronics down there.. Just a single, small resistor - nothing else.


My heater is a MOSFET. Maybe a bank of them. My heat, however, will be spread out among the wiring, H bridge FETs, and load FETs. So all of that needs to be cooled. The control electronics, of course, do not. I did not mean that they would be cooled.
Quote:

Quote:
increase the temperature of 1.5L of water by 16 C

Well, the subject suggested discharging high power and high capacity (Pb-Acid, 12V ) cell with 75A, that is why i suggested 900W for an hour(0.9kWh of energy). How much energy can your largest cell hold?


Other people assumed lead acid batteries. I have zero interest in lead acids. They are an old technology and have terrible energy density. Not that they don't have their place. I'm only interested in li-ions and li-polys. My cells hold about 30WH.

Quote:

Quote:
It may make sense to look at a non-conductive liquid

What is the highest voltage of your cell? I have a 230VAC in my kitchen 2400W kettle and I cannot see a problem.

4.3V. I was referring to using a non conductive liquid if all my power electronics were in a liquid. But if I have them mounted to a heatsink that is in water, no problem there.
Quote:

Quote:
fan for the wires

???

If I mount a big heatsink such that it is sitting in water, the things directly connected to the heatsink will be cooled a lot better than things connected to things connected to the heatsink. In other words - I expect my wires to get toasty but my FETs to be better off. Perhaps they will share the thermal load better than that though - not sure. Without simulation/a bunch of numbers, I'm just guessing.
Quote:

Quote:

I don't have a specific goal in mind - more just that I want to maximize whatever I can get

Well, usually when you measure something, you use that information and make a decisions - controls. If you do not have a specific goal to maximize, then you actually do not know which information is valuable. So you simply cycle the cells and what next?

The immediate goal is to take a big pile of LiPo cells that I have and make a matched pack out of a bunch of them (I have 50+ of these cells, and need a pack of 10). The better my resolution is in sensing and control, the better I can match my cells. 16b is a nice first pass. If that ends up not being enough resolution - I can modify it.

The long term goal is to just support my battery interest. That means running whatever the hell tests I feel like. No specific goal. Hence no specific specifications. I know lack of specifications drive some people crazy. My standard practice, when building one-offs, is to overbuild. 16b is almost guaranteed to be more than I need.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

rberger wrote:
Not sure if it would help in the charging side, except for an intitial current limit, but many years ago I developed a battery capacity tester for 12V SLA batteries, which would be disccharged at a constant current to a cutoff voltage of 10.8V over a 90 minute plus time period. <90 minutes was a failure.

With 80-200AH batteries this was quite a load.

What I did was a constant current source based load, comprising 8 TO3 transistors on a huge 1 degree C/W fann cooled heatsink loaded by banks of 12V 50W Halogen Lamps that were bank switched into operation depending on the required dishcharge current.

The advantage of the Lamps over resistive dummy loads is that at least some of their energy (dissipation) is converted to light instead of heat.

You do need this light to escape, else it's the same as a resistor.

It worked well and the current range was 125A or so.

Food for thought.

Ron.


Ron - that's a good point about using lights to lessen the load on your heatsink. I think if I didn't want as much flexibility as I want out of this system I would pursue that - but I want absolute control over the load - and lights (even with a FET current sink in front of them) are a constant minimum load.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
I am also interested in measuring the volume change of the cell, but I haven't figured out a nice way of doing this,

Well, one method is to immerse the Device Under Test into a liquid bath within a rigid walled container.
Measure the delta-heigth change in the water, (knowing the length and width of the container), and you know the volume change.

Put a hole in the side of the container and connect a vertically oriented, calibrated, capillary tube, (looks like a thermometer, but has an open top), and you can easily read the delta heigth.

Use a liquid with low thermal expansion caracteristics, or measure the liquid's temp for an additional term in the volume calculations.

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

30WH at 4.3V gives about 0.1MJ.. With 75A you will discharge them in 5 minutes completely. Are you sure they are designed to work under this load?

If you do not have a specific goal in mind, why don't you test them individually? You must agree the control is much easier when your cells are independent and also you can infer much more then, since you have more information.

No RSTDISBL, no fun!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi nleahcim,

Quote:
but I want absolute control over the load - and lights (even with a FET current sink in front of them) are a constant minimum load.
Using an OP-Amp based current sink driving transistors or mosfets, the load current is defined as Vin/RSC. The fact that the load resistor is a Lamp, bunch of Lamps or a Resistor does not alter this relationship.

Either way the power must be dissipated. What I tried to achieve was minimal dissipation in the semiconductors by switching the value of the load bank as required to minimise the voltage across the semicondctor part of the load. Lamps were used only to convert some of the total heat dissipation to light.

Ron.

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Brutte wrote:
30WH at 4.3V gives about 0.1MJ.. With 75A you will discharge them in 5 minutes completely. Are you sure they are designed to work under this load?

If you do not have a specific goal in mind, why don't you test them individually? You must agree the control is much easier when your cells are independent and also you can infer much more then, since you have more information.


Yes. They are designed for high discharge rates. They are 6.6AH cells rated for 10C, which means they're rated for 66A. I'm designing for 75A to give the system a bit of headroom.

I am testing them individually, not as a pack. The idea is to eventually build a pack.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

rberger wrote:
Hi nleahcim,
Quote:
but I want absolute control over the load - and lights (even with a FET current sink in front of them) are a constant minimum load.
Using an OP-Amp based current sink driving transistors or mosfets, the load current is defined as Vin/RSC. The fact that the load resistor is a Lamp, bunch of Lamps or a Resistor does not alter this relationship.

Either way the power must be dissipated. What I tried to achieve was minimal dissipation in the semiconductors by switching the value of the load bank as required to minimise the voltage across the semicondctor part of the load. Lamps were used only to convert some of the total heat dissipation to light.

Ron.


Ron - the problem is that if I want to make my load less than what the lamps are, I'm out of luck. Let's say I have a current sink in series with a 1 ohm lamp. No matter what, the load that my cell would see would always be at least 1 ohm. It is a loss of flexibility that I would like to avoid. My planned load is not a resistor. My planned load is a MOSFET. That way it can be whatever I feel like!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

DocJC wrote:
Quote:
I am also interested in measuring the volume change of the cell, but I haven't figured out a nice way of doing this,

Well, one method is to immerse the Device Under Test into a liquid bath within a rigid walled container.
Measure the delta-heigth change in the water, (knowing the length and width of the container), and you know the volume change.

Put a hole in the side of the container and connect a vertically oriented, calibrated, capillary tube, (looks like a thermometer, but has an open top), and you can easily read the delta heigth.

Use a liquid with low thermal expansion caracteristics, or measure the liquid's temp for an additional term in the volume calculations.

JC


Only worry there is that the liquid bath would interfere with the cell (cool it down and potentially short it). If I could find a liquid that didn't conduct heat or electricity, that might work...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Put the cell in a baggie, (like a plastic sandwich bag), then put it in another one for safety.
As you immerse it the air wil be squished out of the baggie(s).

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

hobbyking has some cheap high capacity wires/connectors.
For example: 8AWG silicone wire.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi nleahcim,

Quote:
Ron - the problem is that if I want to make my load less than what the lamps are, I'm out of luck. Let's say I have a current sink in series with a 1 ohm lamp. No matter what, the load that my cell would see would always be at least 1 ohm. It is a loss of flexibility that I would like to avoid. My planned load is not a resistor. My planned load is a MOSFET. That way it can be whatever I feel like!

The concept of a constanr current sink is that you can set the current to any value you like.

If you choose an RSC value of .0025 Ohms, then the load current will be 400 Amps per Volt of input to the Op-Amp based Cuerrent sink. At 75 Amps 14W will be dissipated in this resistor. The balance of the power will have to be dissipated somwhere else. e.g. Mosfet.

With a single 4.3V cell, a 75 amp load will drop about 187.5mV across the 0.0025 Ohm sense resistor. Leaving (4.3V-.1875V)*75A to be dissipated. That is 308.44W!

The use of Resistors/Lamps in the constant current sink diverts part/most of this power away from the Mosfet.

Using a constant current sink with a 0.0025 ohm sense resistor gives a sensitivity of 400A/V, or 2.5mV/Amp.

Using a resistive divider on the input to scale that to make 5V=75A would give you 15A/V. Even using an 8-bit D:A (PWM) you would have a resolution of around 300mA per PWM count.

The Idea of using resistors/Lamps is to divert power from the Mosfet, and to set a Maximum load current when the Mosfet is saturated.

Ron.

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

rberger wrote:
Hi nleahcim,
Quote:
Ron - the problem is that if I want to make my load less than what the lamps are, I'm out of luck. Let's say I have a current sink in series with a 1 ohm lamp. No matter what, the load that my cell would see would always be at least 1 ohm. It is a loss of flexibility that I would like to avoid. My planned load is not a resistor. My planned load is a MOSFET. That way it can be whatever I feel like!

The concept of a constanr current sink is that you can set the current to any value you like.

If you choose an RSC value of .0025 Ohms, then the load current will be 400 Amps per Volt of input to the Op-Amp based Cuerrent sink. At 75 Amps 14W will be dissipated in this resistor. The balance of the power will have to be dissipated somwhere else. e.g. Mosfet.

With a single 4.3V cell, a 75 amp load will drop about 187.5mV across the 0.0025 Ohm sense resistor. Leaving (4.3V-.1875V)*75A to be dissipated. That is 308.44W!

The use of Resistors/Lamps in the constant current sink diverts part/most of this power away from the Mosfet.

Using a constant current sink with a 0.0025 ohm sense resistor gives a sensitivity of 400A/V, or 2.5mV/Amp.

Using a resistive divider on the input to scale that to make 5V=75A would give you 15A/V. Even using an 8-bit D:A (PWM) you would have a resolution of around 300mA per PWM count.

The Idea of using resistors/Lamps is to divert power from the Mosfet, and to set a Maximum load current when the Mosfet is saturated.

Ron.


Ron - the only way this would work was if the lamps were parallel with the FET load. This then establishes a maximum load that can be placed on the packs. It is a loss of flexibility. The only way to leave me complete flexibility would be to have an individual switch on the lamp load, separate from the active load.

Understand that though my immediate goal is to text my 6.6AH LiPo cells - I suspect I will be using this for a lot of other testing in the future so I don't want to close any doors...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

thmjpr wrote:
hobbyking has some cheap high capacity wires/connectors.
For example: 8AWG silicone wire.

Sweet - I had been a bit worried I might have to buy a big roll of the stuff. Thanks!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi nleahcim,

Quote:
Ron - the only way this would work was if the lamps were parallel with the FET load. This then establishes a maximum load that can be placed on the packs. It is a loss of flexibility. The only way to leave me complete flexibility would be to have an individual switch on the lamp load, separate from the active load.

I think have failed to understand what a constant current sink is and what it will allow you to do in regards to loading your cell.

The purpose of a series resistor (lamps in my case) between the current sink and the battery is to shift most of the power dissipation into pasive components rather than your Mosfets.

The series resistor in your case is likely to be less than 50 miliOhms, but at 75A, a 0.05 Ohm resistor will move 281 Watts of heat away from your Mosfets.

Even 0.03 Ohms at 75A shifts 168W of heat away from your Mosfets.

Ron.

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

75A and PCB:
Just remember that most PC CPU's use that amount. And the PCB don't care if it's 1V or 12V.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

But perhaps you have to look into using CT's

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

rberger wrote:
Hi nleahcim,
Quote:
Ron - the only way this would work was if the lamps were parallel with the FET load. This then establishes a maximum load that can be placed on the packs. It is a loss of flexibility. The only way to leave me complete flexibility would be to have an individual switch on the lamp load, separate from the active load.

I think have failed to understand what a constant current sink is and what it will allow you to do in regards to loading your cell.

The purpose of a series resistor (lamps in my case) between the current sink and the battery is to shift most of the power dissipation into pasive components rather than your Mosfets.

The series resistor in your case is likely to be less than 50 miliOhms, but at 75A, a 0.05 Ohm resistor will move 281 Watts of heat away from your Mosfets.

Even 0.03 Ohms at 75A shifts 168W of heat away from your Mosfets.

Ron.


Ron - take a look at the simulations. They show a similar setup to what I'm going to be running (mine will be more complicated since I'll be doing differential current measurement across a kelvin resistor - but oh well). The component values will obviously be significantly different in my application.

The 0.1 resistor is my shunt resistor, and the 0.5 resistor is the lamp. Note that when the lamp is connected in series, the max current I can draw is limited. When I put the lamp in parallel, the minimum current I can draw is more than zero.

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

sparrow2 wrote:
But perhaps you have to look into using CT's

I think current transformers are only useful for AC currents. Unless I'm missing something about the technology or CT stands for something else?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

nleahcim wrote:

Quote:
Ron - take a look at the simulations. They show a similar setup to what I'm going to be running (mine will be more complicated since I'll be doing differential current measurement across a kelvin resistor - but oh well). The component values will obviously be significantly different in my application.

The 0.1 resistor is my shunt resistor, and the 0.5 resistor is the lamp. Note that when the lamp is connected in series, the max current I can draw is limited. When I put the lamp in parallel, the minimum current I can draw is more than zero.


I agree with your simulations but as you have chossen non practial values for
your original spec of 4.3V & 75A, the 0.5 ohm load and 0.1 Ohm shunt will limit
the maximum current. Total Resistance needs to be less than 4.3/75 = 0.0573 Ohms.

The concept in using a series resistor, in the Drain - Not Source!, is to shift
80% or so of the dissipated power away from the Mosfet.

Take a look at my values, Shunt=0.0025 Ohms, Series Resistor 0.05 Ohms, Mosfet
must be Less than .0048 Ohms. You can juggle the values of Mosfet & Resistor
so long as their total is the same or less. Just try and keep the Resistor
around 10 times the Mosfet RDS_ON to move 80% or more of the power dissipation
away from the Mosfet.
Ron.