Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: [1] 2 Next   Go Down
  Print  
Author Topic: "Dummy Load" for Power Supply Testing  (Read 13125 times)
WB6DGN
Member

Posts: 617




Ignore
« on: October 17, 2011, 02:01:19 AM »

I am trying to design and build a load for adjusting and repairing power supplies.  In the past, I've used the usual collection of auto headlamps, combinations of resistors, etc. with varying results but with absolutely no convenience or precision.  My thought is to construct a bunch of relatively low current "modules" that, when used in varying combinations, would result  in load currents from 1 to 100 amps in 1 amp steps.  For example, 9 - 10 amp "modules", 1 - 5 amp "module" and 5 - 1 amp "modules" would accomplish this and wouldn't require expensive high current switches to switch them in and out of the circuit.  Using a load like this would allow me to adjust things like current foldback much more accurately than any method I've used in the past.  Calculating nominal resistances, effects of test lead size and length and so on is not my problem; that's all easily done.  My dilemma comes with trying to decide what to use to construct these "modules".  Nichrome is out because of its wide variations in resistance with temperature.  Two possible candidates that I've found are Constantan or Manganin wire.  I've spent several days searching the internet for information such as resistance per unit length, power dissipation (temperature), price, etc. with little to no result.
Does anyone out there have any advice/suggestions as to, (1) what type of material to use,  (2) specs. and/or pricing on my two above prospects or others, or (3) any other comments about my proposed project.  One final comment; cost IS a factor.  I really don't want this project to get out of hand if I can help it.  What that translates to in terms of dollars, though, I'm really not sure at this point.
Thanks for any and all help or advice you would care to offer.
Tom
Logged
KA4POL
Member

Posts: 1978




Ignore
« Reply #1 on: October 17, 2011, 04:36:44 AM »

For exactly the same purpose I recently built a 10A regulated load with one 2N3055. The GND output is connected to the emitter, the positive input goes via an instrument (10A) to the collector. Now you just need a 100 Ohm .5W resistor soldered to the collector and in series to that resistor a 10k 2 W wire potentiometer with the wiper and the other end connected to the base.
It should be possible to do a similar circuit with a Mosfet for higher currents or some more 2N3055s in parallel. For the latter you would need an additional small value resistor to equilibrate the transistor currents. You definitely need good cooling.
Cost should be fairly low, particularly in comparison to commercial units.
Logged
K8AXW
Member

Posts: 3827




Ignore
« Reply #2 on: October 17, 2011, 07:50:48 AM »

POL is very close to your answer and needs.  I have a construction article in my files for a "power waster."  As I recall (It's been filed for many years) it uses a series of 2N3055 transistors mounted on a large heatsink.  It has associated circuitry that allows a gradual increase in the amount of current being drawn.

If you're interested in this please respond.  I'll dig it out and see that you get a copy.

I personally use a very long (like 30") resistance tape wound on a ceramic form.  It's good for only 20A. 

It must be noted though that any resistor like this changes value as it heats up.  This can be minimized by a muffin fan to cool the element as it's being used.

As for 100A..... that would be interesting to build.
Logged
AD4U
Member

Posts: 2164




Ignore
« Reply #3 on: October 17, 2011, 08:33:21 AM »

At a recent hamfest (Shelby NC) I bought 10 each NOS 1 ohm 250 watt wire wound ceramic power resistors for $2 each.  They can be configured in series and / or parallel to give most any combination of ohms (10 ohms to 0.1 ohm) or wattage needed to load test any "low voltage" PS.  If all else fails, keep your eyes open for a similar deal at a hamfest.

Dick  AD4U
Logged
TANAKASAN
Member

Posts: 933




Ignore
« Reply #4 on: October 17, 2011, 08:36:56 AM »

I have a transistor based dummy load in a cupboard somewhere that can handle up to twenty amps at 13.8 volts. At full load the four fans are doing a reasonable impression of a hovercraft and the heatsink still runs warm to the touch. Build a one hundred amp load and you're going to be dissipating 1,380 W as heat, this is a major cooling problem.

Tanakasan

 
Logged
WB6BYU
Member

Posts: 13239




Ignore
« Reply #5 on: October 17, 2011, 09:17:19 AM »

I'd agree with the transistor version - basically you can configure each transistor as a constant current
load (at least over a reasonable range of voltages) and switch it on and off with a logic signal.
For higher current loads include the cooling fan current as part of the total.

The usual issues regarding pass transistors in power supplies apply:  provide plenty of heat sinking,
and don't count on the rated power dissipation or maximum current from the transistor (because
the thermal transfer to the heat sink is often the limiting factor.)  Also the gain of the device may
drop at higher currents, so you'll need more drive capability if you are switching it with a low current
control.  Over 5A you may need two (or more) pass transistors with a drive transistor that can
deliver up to an amp of base current.

Or you could build two such loads with adjustable current draw - use one for the static load on
the supply and switch the second to adjust the dynamic characteristics.  But you'll need enough
heat sinking to dissipate several hundred watts, and the thermal design will be the most critical
aspect if you are going to run them for a while.
Logged
WB6MMV
Member

Posts: 97




Ignore
« Reply #6 on: October 17, 2011, 09:47:38 AM »

Hi Tom:  I have a simple electronic load circuit that I use to test power supplies (both linear and switching).  It has forced air cooling with temp control of the fan so that when the load heats up the fan speeds up.  It can handle around 50 amps on a continuous basis with the forced air cooling.  It has overtemp and a few other minor safety items to protect it, but it can easily be built without using the digital meters and other fancy stuff I added just because it was fun to do so.  You can take this basic circuit and get fancy with pulsing, and using digital control to do step testing and dynamic reactions if you like. 

You can make many variations of this supply by adjusting the number of pass transistors used.  I had a heat sink that would handle six TO-3 transistors which is what I use.  I also rate the transistors at 5 amps each to be conservative as I often leave the power supply under test for hours to check thermal issues.

The schematic is in the format used by express pcb for their board design service.  I can email you a copy of the supply if this doesn't work for you.  Or I can send you a print copy via snail mail also.

Let me know your preference. 

Ken WB6MMV
Logged
G3RZP
Member

Posts: 4557




Ignore
« Reply #7 on: October 17, 2011, 10:41:35 AM »

A few metal can 2N3055s, a small piece of aluminium, and a bucket of water.
Logged
K5LXP
Member

Posts: 4480


WWW

Ignore
« Reply #8 on: October 17, 2011, 11:51:34 AM »

I use two ways to put the hurt to supplies under test.

One is steel wire.  To get around the varying resistance over temperature I coil it up and put it in a bucket of water.  A standard shunt ammeter tells me the exact current.  Good enough for go/no go testing or serve as a load for discharging large batteries.

Another way is to connect a beefy DC-AC inverter, and use an appropriate 120VAC device as a load - anything from lamps to room heaters.  A combination of different lamps on a dimmer makes for a simple adjustable load.


Mark K5LXP
Albuquerque, NM
Logged
W0NFU
Member

Posts: 46




Ignore
« Reply #9 on: October 17, 2011, 11:46:31 PM »

Simple load is to use a headlight bulb.

73 - Larry W├śNFU
Lake Forest Park, WA
larry_w@comcast.net
Logged
KA4POL
Member

Posts: 1978




Ignore
« Reply #10 on: October 18, 2011, 12:02:03 AM »

In the past, I've used the usual collection of auto headlamps, combinations of resistors, etc. with varying results but with absolutely no convenience or precision. 
Simple load is to use a headlight bulb.
Obviously Tom is already one step further  Wink
Logged
WB6DGN
Member

Posts: 617




Ignore
« Reply #11 on: October 18, 2011, 12:45:33 AM »

Wow!  Thanks for all the good ideas and suggestions.  I guess my age is showing!  I should have thought of using a transistor based load instead of brute-forcing it.  A lot easier to "touch up" the load current too.  Yeah!  Disposing of nearly 1.5kW of heat efficiently is going to take a bit of thought but, at least, my work room is going to be nice and toasty this winter!  Hi! 

To WB6MMV and K8AXW I would very much appreciate copies of your designs to give me some additional ideas.  My email is <my callsign @ att.net> and my snail mail address is good on QRZ.

Once again, thanks very much to all who took the time to reply.  Your ideas are very much appreciated.  Also, nothing is set in stone as yet so I'm still open to further suggestions.

Tom DGN

Logged
WB6MMV
Member

Posts: 97




Ignore
« Reply #12 on: October 18, 2011, 12:23:34 PM »

Tom:  I just sent you an email with a bitmap of my schematic.  If it doesn't work, I will drop a hard copy in the mail.  Let me know if its readable for you to use.

Ken WB6MMV
Logged
K8AXW
Member

Posts: 3827




Ignore
« Reply #13 on: October 19, 2011, 10:25:04 AM »


DGN:  I finally found the schematic of the Power Waster in my files.  I'll get a copy made today and snail mail it to you.  I can scan it and email it but the quality suffers terribly when I email a scan as a .jpg.

Al
Logged
K8AXW
Member

Posts: 3827




Ignore
« Reply #14 on: October 19, 2011, 10:48:30 AM »

MMV:  I'd like a copy of that schematic also OM.  Please send it to:

aibling328@yahoo.com

'ppreciate it!

73

Al
Logged
Pages: [1] 2 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!