Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Dc electronic load 6v 30amps

Status
Not open for further replies.

wizpic

Advanced Member level 3
Joined
May 23, 2004
Messages
787
Helped
108
Reputation
216
Reaction score
72
Trophy points
1,308
Location
London
Activity points
6,331
I'm looking building a constant load for a 6v system with 30amps maximum using 4 mosfets and 4 50w (to247 style package) all mounted on a heatsink.ive done some searching but not 100% sure if my heatsink is big enough to take all the heat.like this one in the link/photo
https://uk.rs-online.com/web/p/heatsinks/7742821/
What I don't want to do I see overkill or burn the components out
Hi how is the bet method for calculations for the best hetsink with 4 mosfets and power resistors all mounted on the same heatsink with a fan mounted on heatsink for cooling and may be another one in the case to help remove heat


Thanks
 

Attachments

  • IMG_0243.JPG
    IMG_0243.JPG
    21.9 KB · Views: 57

Hi,

6V x 30A = 180W.

If you want a temperature rise of 50°C...(example. Heatsink only)
Then you need a heatsink with 50°C / 180W = 0.78 K/W.

Not difficult.

Now you need to calculate in the different thermal resistances.
R_th_jc, R_th_ch.....

You find this informations at every heatsink manufacturer and at every power semiconductor manufacturer..and in the internet ... and here in the forum .... and...

Klaus
 

Just a rough (back-of-the-hand) calculation tells us:

say the power to be dissipated as 180W- keep some margin- make it 200W.

Heatsink thermal resistance is 0.4; that means you can have a max temp rise of the heatsink 200X0.4=80C

Say ambient temp is 30 (it can be more inside a box)= Actual temp of the heatsink is 80+30=110;

Say the temp drop at the pad and the transistor package is 20 (conservative guess) then the junction temp of the transistor is 130C

You are cutting it close; need to do the actual calculation with real values.
 

Thanks guys, I guess I should looked at or chosen the mosfets and then the mosfets and resistor data sheets, I will choose a mosfet with with a low RDS and capable of carrying the current and thermal resistance along with the working temperature.
 

Hi,

Low_R_DS is not necessary, because you are running a linear regulator.

In this case R_DS_ON is not valid and doesn´t not make a difference in power dissipation.

Klaus
 

I was considering using Pwm with a dedicated mosfet driver like in the first prototype-type. The only difference is the first design is that I used 2 mosfets and 4 x 0.47R free standing 90w each resistors. But to make assembley easier I thought about using a heatsink like above or similar where it all is mounted on 1 pcb and heatsink as I require 4 channels. This is why I thought low RDS Mosfets
 

Hi,

PWM in an electronic load? Are you sure? How does it work?
I can't see how you get around to dissipate 180W.

Klaus
 

PWM does not make the FETs any cooler/

PWM is used to make FETs cooler in SMPS when you are storing the energy in a choke to charge a cap then to be transferred to a load

Just use a power resistor to take most of the heat that has a heatsink clamp which may operate safely at a higher temp then use the FET's in linear more or use more FET's with high air velocity fan. Even a decent CPU heatsink might only be rated for 100W CPU's and very dependent on case surface area and the right amount of good thermal grease and surface coplanarity.
 

I was considering using Pwm with a dedicated mosfet driver like in the first prototype-type...

Let us be clear on the principle: if your load is taking 6V at 30A and dissipating all the energy as heat, then using any fancy MOSFET will not help; you have got a big heatsink and you can mount 200 resistors each 1W series-parallel to work 30A at 6V.

But it ALL depends on the load; if your load is going to be a motor, you need to consider adding inductance to the load and the actual dissipation may be lower. There are models to simulate large inductors. But you have not specified the load in the beginning in detail.
 

Hi,

you need to consider adding inductance to the load and the actual dissipation may be lower
It is DC, so even an inductance will not change power dissipation.

Klaus
 

It is DC, so even an inductance will not change power dissipation.

I understand your point and I do not know the answer. If you have a DC motor that has a boilerplate power label of 180W (running at 6V and taking 30A current) as the REAL load, it certainly dissipates less than 180W of power (it converts electrical energy to mechanical work with rather high efficiency) as heat. What kind of electronic circuits can simulate such a load?

(My personal preference is that all motors be treated as AC and then many problems do disappear)
 

Hi,

From electrical view a DC motor which draws 30A at 6V motor consumes 180W.

A part of this energy is dissipated as heat in the motor itself and a part is transferred mechanically via the rotation and torque.
But this has nothing to do with inductance (of the motor).

Klaus
 

I once saw a hybrid design which switched power resistors in parallel with the Mosfets to dissipate some part of the load. At the highest power levels.

Since the resistors are well, resistors, its current changes with changing voltage, the Mosfet would be driven dynamically such that the end result would still resemble a CC load.

If the load drops below what the resistors will dissipate on their own, they are switched off and the Mosfets themselves take all the brunt.

A clever design, but cannot find it anymore.
 

This is not a motor control, It places a 20-30amp load on a set of batteries for testing, I know it's not good for batteries but the unit that works as saved many of batteries and still in service today. I just wanted to revamp the design and make better. The way it works at the moment that it measures the current and increases the PWM signal until the 20-30amps is reached and then it continuously monitors it so when the voltage falls it maintains the 20-30amps to which what the current is set for at the beginning of the test. I use a heatsink of 200x100x40mm thick with 8 mosfets on there (one for each channel) then 16 90W free standing power resistors 4 for each channel. The heatsink does not really get that hot you still can touch with hand.

Is it better to use a Dac then instead of PWM ?
 
Last edited:

Hi,

DAC...
The only solution I see is a linear/analog regulation loop. For sure you may use a DAC for the setup value.

Klaus
 

Are you try to use pulse method to reduce battery sulphation? There are many approaches to this . Some are blow torch sub millohm pulse injectors capable of large currents and others are xx ns 20A low energy pulses at ultrasonic rate that breakdown the lead sulphate crystals while a DC charge is being applied.
 

How good does this load have to be? Like, maybe a couple
of automotive relays (you can probably find them for old
6V systems, still) and a couple of light bulbs, and you've
a load that will just sit and take it indefinitely. Not great
for accuracy. But nobody has said anything about that.
 

Thanks Guys
But I've placed it on the back burner for now. There was this company that is trying to sell us this super charger that helps recover batteries so we all ready have that special charger that creates those pulses to breakdown the lead sulphate crystals. The idea was that we load test the batteries first that had been in service sometime see how good they where then place them on this charger and re-test them to see if the runtime goes longer, I already have designed a 30amp load tester which recorded the runtime, This was based on 4 90W free standing resistors and 2 mosfets for each pair of resistors using PWM seems to work quite well but the draw back is there all common grounded but applies the 30amp load individually to each battery, This other company said they could design a newer all singing all dancing but they used 4 power transistors and 4 50W power resistors mounted on a larger heatsink like I posted above, but the trouble was it kept over heating and blowing up where mine did not overheat you could always touch and hold the heatsink. The only difference was I used PWM to control the load and they used DAC'S in linear mode. They also tried to add reverse polarity protection so no matter which way you connected the battery it still ran the test, But this had a major flaw and draw back once you connected the batteries and turned the unit on before you started the test and realised you had one battery wrong and tried to connect it the other way without turning the unit off it would go bang and blow the power transistors up. I know it did not matter really as you could just start the test and it would work but to me this was a design fault of there unit.
So I wanted to design another mark 2 version and make it better and have all batteries isolated from each other and wanted to make sure what the best method or way was to go about it, the reason I asked here, I'd have one controller and 4 slaves that sent all data to the master.
Early test's show that this charger does help improve the runtime which means that it is breaking down the lead sulphate crystals in the battery. I wanted to make a start so if this did take off I could build more of them for our company as an in house tester but the project has been put on hold for the time been as our company saying it's to time consuming and they need to look into it more before they decided

Shame as it would have been an interesting project to work on again, If it was not for that other company trying to come up with an all singing all dancing version my other design would of been made but again because of all the faults and time they stopped it.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top