Transistor Or Resistor
in [Basics]
|
From: Chris W on 28 Apr 2010 18:36 I want to make a load center to test power supplies and batteries. I was thinking of using 50 Watt 4 ohm resistors for 12V loads but I will need 15 of them to get the current drain I want. I would also like to load 5V and 3.3V lines and of course that would require different resistors. I was wondering if this wouldn't be a lot easier with a power transistor. The 50 Watt resistors are going to cost a little over $3 each and I will probably need 30 of them to get the loads I want. The goal is to have a variable load of about 3 to 50 amps on as much as 14V and from about 1 to 25 amps on 5V and 3.3V. Can someone recommend a specific transistor that would work good? I am hoping I can do it with fewer transistors. I do plan on using a large heat sink and fan to keep this cool. Thanks, Chris W
From: Tim Wescott on 28 Apr 2010 18:44 Chris W wrote: > I want to make a load center to test power supplies and batteries. I > was thinking of using 50 Watt 4 ohm resistors for 12V loads but I will > need 15 of them to get the current drain I want. I would also like to > load 5V and 3.3V lines and of course that would require different > resistors. > > I was wondering if this wouldn't be a lot easier with a power > transistor. The 50 Watt resistors are going to cost a little over $3 > each and I will probably need 30 of them to get the loads I want. > > The goal is to have a variable load of about 3 to 50 amps on as much as > 14V and from about 1 to 25 amps on 5V and 3.3V. Can someone recommend a > specific transistor that would work good? I am hoping I can do it with > fewer transistors. I do plan on using a large heat sink and fan to keep > this cool. 2N3055, if they're still around. You'll spend lots on the transistors and heatsinks, though. Car headlights work well for this, at least for 12 volts, with car tail lights to trim the current. They're bulkier than resistors, but loads cheaper -- and they'll brighten your day :-). Or get a space heater that uses resistance wire, chop it up into ten equal sections -- viola! ten power resistors. -- Tim Wescott Control system and signal processing consulting www.wescottdesign.com
From: Chris W on 28 Apr 2010 19:14 Tim Wescott wrote: > Chris W wrote: >> I want to make a load center to test power supplies and batteries. I >> was thinking of using 50 Watt 4 ohm resistors for 12V loads but I will >> need 15 of them to get the current drain I want. I would also like to >> load 5V and 3.3V lines and of course that would require different >> resistors. >> >> I was wondering if this wouldn't be a lot easier with a power >> transistor. The 50 Watt resistors are going to cost a little over $3 >> each and I will probably need 30 of them to get the loads I want. >> >> The goal is to have a variable load of about 3 to 50 amps on as much >> as 14V and from about 1 to 25 amps on 5V and 3.3V. Can someone >> recommend a specific transistor that would work good? I am hoping I >> can do it with fewer transistors. I do plan on using a large heat >> sink and fan to keep this cool. > > 2N3055, if they're still around. You'll spend lots on the transistors > and heatsinks, though. > > Car headlights work well for this, at least for 12 volts, with car tail > lights to trim the current. They're bulkier than resistors, but loads > cheaper -- and they'll brighten your day :-). > > Or get a space heater that uses resistance wire, chop it up into ten > equal sections -- viola! ten power resistors. > The data sheet says it will dissipate 115 watts. And they cost $1.35 at mouser. That will total to a lot less than the 50 watt resistors. The cost isn't as big of an issue as the complexity of wiring up a bunch of resistors and having to switch them in one at a time so I have to use lots of switches too. On a side note can this thing. http://mouser.com/ProductDetail/Fairchild-Semiconductor/FDL100N50F/?qs=GBxGW0xXju923CYRyhG5QQ%3d%3d really disapate 2,500 watts? If so I assume you need a very good heat sink. Is really possible to realistically have enough heat sink to dissipate enough heat to put anywhere near that much wattage through that thing for more than a very short time? Chris W
From: John Larkin on 28 Apr 2010 19:20 On Wed, 28 Apr 2010 17:36:26 -0500, Chris W <1qazse4(a)cox.net> wrote: >I want to make a load center to test power supplies and batteries. I >was thinking of using 50 Watt 4 ohm resistors for 12V loads but I will >need 15 of them to get the current drain I want. I would also like to >load 5V and 3.3V lines and of course that would require different >resistors. > >I was wondering if this wouldn't be a lot easier with a power >transistor. The 50 Watt resistors are going to cost a little over $3 >each and I will probably need 30 of them to get the loads I want. > >The goal is to have a variable load of about 3 to 50 amps on as much as >14V Which is 700 watts. No single transistor is going to do that. A reasonable number would be 10 big power transistors on a huge fan-cooled heat sink. It becomes non-trivial to drive them properly. There are commercial transistor-based load boxes http://www.teknetelectronics.com/Search.asp?p_ID=18842&pDo=DETAIL&Kikusui_PLZ700W but as you can see they are a pretty big deal. Resistors are a lot simpler. These are neat: ftp://jjlarkin.lmi.net/Welwyn.JPG but the best heat sink is no heat sink, like just letting some nichrome wire get red hot. John
From: default on 28 Apr 2010 20:02
On Wed, 28 Apr 2010 18:14:12 -0500, Chris W <1qazse4(a)cox.net> wrote: >Tim Wescott wrote: >> Chris W wrote: >>> I want to make a load center to test power supplies and batteries. I >>> was thinking of using 50 Watt 4 ohm resistors for 12V loads but I will >>> need 15 of them to get the current drain I want. I would also like to >>> load 5V and 3.3V lines and of course that would require different >>> resistors. >>> >>> I was wondering if this wouldn't be a lot easier with a power >>> transistor. The 50 Watt resistors are going to cost a little over $3 >>> each and I will probably need 30 of them to get the loads I want. >>> >>> The goal is to have a variable load of about 3 to 50 amps on as much >>> as 14V and from about 1 to 25 amps on 5V and 3.3V. Can someone >>> recommend a specific transistor that would work good? I am hoping I >>> can do it with fewer transistors. I do plan on using a large heat >>> sink and fan to keep this cool. >> >> 2N3055, if they're still around. You'll spend lots on the transistors >> and heatsinks, though. >> >> Car headlights work well for this, at least for 12 volts, with car tail >> lights to trim the current. They're bulkier than resistors, but loads >> cheaper -- and they'll brighten your day :-). >> >> Or get a space heater that uses resistance wire, chop it up into ten >> equal sections -- viola! ten power resistors. >> > >The data sheet says it will dissipate 115 watts. And they cost $1.35 at >mouser. That will total to a lot less than the 50 watt resistors. The >cost isn't as big of an issue as the complexity of wiring up a bunch of >resistors and having to switch them in one at a time so I have to use >lots of switches too. > >On a side note can this thing. >http://mouser.com/ProductDetail/Fairchild-Semiconductor/FDL100N50F/?qs=GBxGW0xXju923CYRyhG5QQ%3d%3d > >really disapate 2,500 watts? If so I assume you need a very good heat >sink. Is really possible to realistically have enough heat sink to >dissipate enough heat to put anywhere near that much wattage through >that thing for more than a very short time? > >Chris W You can also use a combination of transistor and resistor(s) to give you control and limit the maximum current (most of the heat in the resistor not the transistor) In a simple circuit with a transistor and potentiometer to set the current - you will quickly learn all about thermal runaway. As the transistor heats, the gain goes up, so the current goes up. Not a problem if you are sitting there watching and correcting for it, but a problem for long term unsupervised testing. The converse also works - you get the current where you want it after multiple adjustments and the ambient temperature drops a little and the current drops too. An adjustable constant current circuit is a better safer bet. -- |