Change nothing, it is beautiful and a child of necessity, it is quite simply the finest of fine 'Jank' and I for one would be proud to have it grace my lab.....cheers.
Some suggestions: add a fan (it increases massively the power handling capability), choose a MOSFET with a comfortable safe operating area SOA for your needs (the irl540 that was in that schematic is a cheap good choice, you DON'T want a MOSFET with a very low RDSon, it creates hot spots), add a 10~47nF between the opamp output and negative input (this limits the bandwith improving the stability, and prevents your mosfet from burning due to Mhz oscilations), replace the 1ohm resistor with a 0.1ohm and amplify the signal 10x or divide the pot output voltage by 10. Bonus tip, you can use a quad opamp 4 Mosfets and 4 resistors for a more powerful load, the control part remains the same, however you would need a bigger heatsink... And remember there no such thing as "too big heat sink" 😂👍
Amazing and quick build, i love it, just one thought, it would be recomendable to put a 104 (100nf) ceramic disc cap as close as possible to the lm385 p-supply pin, This will give smoother operation. Also perhaps you should hook it up to a scope to see if there is any osscliations, which can lead to the whole cct heating up more than it should. A very small o.1 ohm resistor with a 10 uf electrolitic cap just after the bread board supply will form an RC filter which will knock out any input ripple, because in general these cct's have low ripple rejection in general, but man that was fun to watch.
You should add a diode between the opamp and the mosfet gate in case you are going to test more high voltage and the mosfet my get shorted will protect the high voltage to dont affect the opamp!
I made a comment on the previous video about using the cheap power supply as a bench power supply with some added external fusing. I forgot to mention that I never use mine for more than 2 or 3 amps.
Change the resistor for a few cheap Ali 2 ohm 100W resistors in parallel, add a fan and perhaps use a dual op amp with the second one monitoring a thermister on the heatsink so that if it gets above a set temperature then it provides a bit of negative feedback to the load. Then it is a pretty capable little load.
Very nice video, thanks ! You can make a low ohm Resistor with just a piece of wire.. in a bucket of water for cooling .. MikesElectricStuff did that once.. long time ago
You can change the load current (power) of the DL24 "on the fly," while testing. You do not have to build your own, except as a project. I provide Windows PC (not Linux, sorry) for the DL24 on my web site. Free download.
Scope it...there's a good chance that circuit is oscillating! Also, that FET (most FETs) are not specified for linear operation, so don't be surprised if it goes up in smoke...
on the other hand, I have yet to see a mosfet going dead in linear operation if the thermal constraints are observed. keep the junction temperature within safe limits, and it should be fine.
Add a diode between the opamp and the MOSFET gate to keep the opamp safe in case you are testing some higher voltage like that in case if the MOSFET get shorted the higher voltage will be blocked by the diode
I'd drop the 1 ohm resistor value a bit. Would limit your current at low voltages. I.e at 5V the max you could possibly get is 5A, but you also need to drive the gate to at least 10V. I built something similar back in the '90s. I think I used a 0.1 ohm res. You would probably want to add a resistor in series with the pot to limit the max current. An alternative is instead of connecting the pot to a reference 12V supply, connect it to the output. Then the load changes from constant current to constant resistance.
nice one all you need to add would one of those cheep dc volt/amp meters and a wall wart PS be good as any test loads you could buy. I made one that has no active parts I just had a selector switch with real thick contacts to vary the load and a bunch of power resistors (I kind of have an obsession about them) in series / parallel in a box way bigger then that not cumbersome but closer then this one
I have some power resistors (more than will fit in the drawer I showed) that I sometimes use for loads, but this one should be easier to adjust on the fly.
@@pileofstuff it is the selector switch that made it possible I must have bought it at a swap meet do not remember and never seen one that heavy duty before it switches 8 positions and the contacts are 5mm 3/16 thick. a lucky junk find
@@pileofstuff I did finally come up with a situation where a much smaller form factor load like this electronic load would be just the thing. I had to try and figure out what current a transformer hidden in the wall some place could deliver, so I am going through my junk collection with the intention of making something similar then the passive one I have in an old cable set top box
A little pointer! If you expect the FET gate voltage to reach 10 V, please note that the 10 V for your choice FET is expected to be Gate to Source. Your source voltage climbs with the current increase. At 6 A, 1 ohm, you are increasing the source potential already to 6 V and that leaves 12 V - 6 V = 6 V instead of the data sheet 10 V….
I've ordered some additional FETs,for stock including the specific one called out in the design I was basing my build on. That said, this "made from what I have on hand" version serves my current (pun not intended) need.
Hi I really enjoyed your video, I wonder if you could 3D print a case for it and add one of those small fans to blow over the MOSFET and resistor. I would like to see you make that video.
@@pileofstuff It's almost certainly a flyback topology power supply. Energy is stored in the transformer when the primary side MOSFET is on and delivered to the load when the MOSFET switches off. With 1/2 the designed input voltage across the inductance of the transformer, current in the transformer primary can only rise to half of the 240V case in a given on time the and energy stored will be about 1/4 . At low loads there is duty cycle to spare any everything is OK, as the load increases the duty cycle hits a maximum and no more energy can be stored and delivered . 16.7/5 is about √(220/120) so the output current is about what you would expect.
@@pileofstuff Lower output current on lower input voltage is quite common on small form factor high power power supplies like for example in server Power supplies.
Nice home-made load! I've never seen a bench power supply like yours (the one with the 7 segment leds). What is the brand / model of the tiny power supply?
If u want to build a "good" load, u have to do more than using a cooler for your mosfet. i build a load with 300W and one transistor (IXFK94N50P2) - and its horrible, to cool it down to the maximum temperature allowed - so its not burning up. as a measuring resistor, im using the Riedon PCS-0R1D1. Its a liitle bit "more" testing needed to "waste" hundreads of watts ;)
make it more nicer add 100w resistor at list and add also a voltamp meter like that you can see how many volts you are pushing inn and how many amps you are drawing!
I can scroll thru the menu and set it to s specific current, then turn it on, so it can provide a load. What I don't think it can do is vary it's loading on the fly.
the resistor is 1Ohm. 6A * 1 Ohm = 6V across the resistor which is 6*6 = 36W. too much for the resistor. If you put second 1 Ohm in parallel it'll be 6A * 0.5 Ohm = 3V * 6A = 18W (9 W each resistor) The rest will go to the transistor which has more heat transfer capability - metal to metal contact. But that heat sink looks like it's good for 15-20 W at most. Maybe even less without fan. But if it's cold you can do much over that for short time.
Yes great design. In addition to finding a higher wattage load resistor, i believe it would be prudent to install a 5v or 12vdc fan with perhaps a thermistor. I just might have to build one of these for rooting out those pesky Chinese failed gadgets.
I thought that PSU was not rated for 110V input? if so that will seriously reduce it's output capability. It may run but won't provide the 16A it states.
@@pileofstuff yeah that. It would increase potential power draw. Both by its own use and allowing the the resistor to take more current. Like I said, crazy idea.
That was fun and interesting. The build has a real salvaged appeal, like something you grabbed from a pile of stuff. Thanks Mr. Stuff. Take care.
Change nothing, it is beautiful and a child of necessity, it is quite simply the finest of fine 'Jank' and I for one would be proud to have it grace my lab.....cheers.
I appreciate that you have embraced my "style" in the appropriate manner.
@@pileofstuff :)
wow didn't expect you to breadboard that power supply. It helps answer questions and possibly bring more. good show
I like looking at the bottom regardless of how it looks to see how you did what you did.
Very nice. I need to go to my salvaged parts and try to duplicate your success. Thanks for a great video!
Good build. I might build something similar now because I probably have about the same parts laying around.
Some suggestions: add a fan (it increases massively the power handling capability), choose a MOSFET with a comfortable safe operating area SOA for your needs (the irl540 that was in that schematic is a cheap good choice, you DON'T want a MOSFET with a very low RDSon, it creates hot spots), add a 10~47nF between the opamp output and negative input (this limits the bandwith improving the stability, and prevents your mosfet from burning due to Mhz oscilations), replace the 1ohm resistor with a 0.1ohm and amplify the signal 10x or divide the pot output voltage by 10. Bonus tip, you can use a quad opamp 4 Mosfets and 4 resistors for a more powerful load, the control part remains the same, however you would need a bigger heatsink... And remember there no such thing as "too big heat sink" 😂👍
yeah, that would be the "proper" way to do it.
What you saw was the "quick and dirty using what I have on hand" approach.
Amazing and quick build, i love it, just one thought, it would be recomendable to put a 104 (100nf) ceramic disc cap as close as possible to the lm385 p-supply pin,
This will give smoother operation.
Also perhaps you should hook it up to a scope to see if there is any osscliations, which can lead to the whole cct heating up more than it should. A very small o.1 ohm resistor with a 10 uf electrolitic cap just after the bread board supply will form an RC filter which will knock out any input ripple, because in general these cct's have low ripple rejection in general, but man that was fun to watch.
You should add a diode between the opamp and the mosfet gate in case you are going to test more high voltage and the mosfet my get shorted will protect the high voltage to dont affect the opamp!
4:13 - Looks like me working out a patch on some of my modular synths!
Also 15:18 - "Give us a reading on the 1202 Program Alarm …”
I made a comment on the previous video about using the cheap power supply as a bench power supply with some added external fusing. I forgot to mention that I never use mine for more than 2 or 3 amps.
Change the resistor for a few cheap Ali 2 ohm 100W resistors in parallel, add a fan and perhaps use a dual op amp with the second one monitoring a thermister on the heatsink so that if it gets above a set temperature then it provides a bit of negative feedback to the load. Then it is a pretty capable little load.
Very nice video, thanks !
You can make a low ohm Resistor with just a piece of wire.. in a bucket of water for cooling .. MikesElectricStuff did that once.. long time ago
You can change the load current (power) of the DL24 "on the fly," while testing. You do not have to build your own, except as a project. I provide Windows PC (not Linux, sorry) for the DL24 on my web site. Free download.
Scope it...there's a good chance that circuit is oscillating! Also, that FET (most FETs) are not specified for linear operation, so don't be surprised if it goes up in smoke...
That is possible. I was just being lazy with the capacitors.
on the other hand, I have yet to see a mosfet going dead in linear operation if the thermal constraints are observed.
keep the junction temperature within safe limits, and it should be fine.
@@s_s-g4d Actually not. Look into thermal runaway, SOA, etc... Most "digital" MOSFETs don't even show SOA to DC.
Excellent video! I need this myself. I have all the parts.
Add a diode between the opamp and the MOSFET gate to keep the opamp safe in case you are testing some higher voltage like that in case if the MOSFET get shorted the higher voltage will be blocked by the diode
I'd drop the 1 ohm resistor value a bit. Would limit your current at low voltages. I.e at 5V the max you could possibly get is 5A, but you also need to drive the gate to at least 10V.
I built something similar back in the '90s. I think I used a 0.1 ohm res. You would probably want to add a resistor in series with the pot to limit the max current.
An alternative is instead of connecting the pot to a reference 12V supply, connect it to the output. Then the load changes from constant current to constant resistance.
nice one all you need to add would one of those cheep dc volt/amp meters and a wall wart PS be good as any test loads you could buy.
I made one that has no active parts I just had a selector switch with real thick contacts to vary the load and a bunch of power resistors (I kind of have an obsession about them) in series / parallel in a box way bigger then that not cumbersome but closer then this one
I have some power resistors (more than will fit in the drawer I showed) that I sometimes use for loads, but this one should be easier to adjust on the fly.
@@pileofstuff it is the selector switch that made it possible I must have bought it at a swap meet do not remember and never seen one that heavy duty before it switches 8 positions and the contacts are 5mm 3/16 thick. a lucky junk find
Building useful things from assorted salvaged (or randomly acquired) parts does leave a guy with a good feeling.
@@pileofstuff I did finally come up with a situation where a much smaller form factor load like this electronic load would be just the thing. I had to try and figure out what current a transformer hidden in the wall some place could deliver, so I am going through my junk collection with the intention of making something similar then the passive one I have in an old cable set top box
A little pointer! If you expect the FET gate voltage to reach 10 V, please note that the 10 V for your choice FET is expected to be Gate to Source. Your source voltage climbs with the current increase. At 6 A, 1 ohm, you are increasing the source potential already to 6 V and that leaves 12 V - 6 V = 6 V instead of the data sheet 10 V….
I've ordered some additional FETs,for stock including the specific one called out in the design I was basing my build on.
That said, this "made from what I have on hand" version serves my current (pun not intended) need.
Hi, Great video. I notice on your bench a small electric screwdriver. what is the make and model?
It's a Kaiweets ES20
pileofstuff.ca/r/qwslg
@@pileofstuff cheers
Hi I really enjoyed your video, I wonder if you could 3D print a case for it and add one of those small fans to blow over the MOSFET and resistor. I would like to see you make that video.
A fan is a good idea.
But with as much heat as thin thing dissipates, I think a 3D printed case would get kinda melty.
What AC voltage are you running the power supply on? If only 110v, that might explain the reduced current
That is a possibility.
I was expecting it to not work at all if it was unhappy about the input voltage.
Agreed.
@@pileofstuff It's almost certainly a flyback topology power supply. Energy is stored in the transformer when the primary side MOSFET is on and delivered to the load when the MOSFET switches off. With 1/2 the designed input voltage across the inductance of the transformer, current in the transformer primary can only rise to half of the 240V case in a given on time the and energy stored will be about 1/4 . At low loads there is duty cycle to spare any everything is OK, as the load increases the duty cycle hits a maximum and no more energy can be stored and delivered . 16.7/5 is about √(220/120) so the output current is about what you would expect.
@@pileofstuff Lower output current on lower input voltage is quite common on small form factor high power power supplies like for example in server Power supplies.
@@AlpineTheHusky on the other hand, it wouldn't be surprising for the PSU to just suck.
Most interesting video let it live in its Frakenstien format
use some 55w car headlamps was the input 220v?
Car headlamp/side light 55W/23W, or household halogen bulbs at 10W/20W are my goto for a simple 12V load. Not adjustable, but very easy to set up.
Nice home-made load! I've never seen a bench power supply like yours (the one with the 7 segment leds). What is the brand / model of the tiny power supply?
It's this one: ua-cam.com/video/1Q3-kk_9o4s/v-deo.html
If u want to build a "good" load, u have to do more than using a cooler for your mosfet.
i build a load with 300W and one transistor (IXFK94N50P2) - and its horrible, to cool it down to the maximum temperature allowed - so its not burning up. as a measuring resistor, im using the Riedon PCS-0R1D1.
Its a liitle bit "more" testing needed to "waste" hundreads of watts ;)
True, but I didn't really expect this to be "professional grade".
This was a "use what's on hand for a quick easy result" type build.
It would be nice if u could make a video about how you made ur screen power supply you used in this video :)
This power supply? ua-cam.com/video/1Q3-kk_9o4s/v-deo.html
@@pileofstuff Oh it's a module, I thought it's a DIY ... Thank you very much .. Keep up ur videos
This was interesting.
make it more nicer add 100w resistor at list and add also a voltamp meter like that you can see how many volts you are pushing inn and how many amps you are drawing!
Did you test it at it's rated input of 180-240 volts?
make a full PCB out of the test rig you made and post it to a shematice ? (sorry for the bad spelling)
Nice little circuit. One question though, what beer were you drinking ?
Stir Stick Stout, from Half Pints brewery.
@@pileofstuff So it's hard to order another one when you've had a few then. LOL.
I was thinking about your first idea... How to fool the battery discharger to do the job?... Error 3 on ZB2L3 gadget .
I can scroll thru the menu and set it to s specific current, then turn it on, so it can provide a load.
What I don't think it can do is vary it's loading on the fly.
the resistor is 1Ohm. 6A * 1 Ohm = 6V across the resistor which is 6*6 = 36W. too much for the resistor. If you put second 1 Ohm in parallel it'll be 6A * 0.5 Ohm = 3V * 6A = 18W (9 W each resistor) The rest will go to the transistor which has more heat transfer capability - metal to metal contact. But that heat sink looks like it's good for 15-20 W at most. Maybe even less without fan. But if it's cold you can do much over that for short time.
Yes, this was just cobbled together with parts I happened to have on hand and a random salvaged heatsink.
@@pileofstuff ..and that is exactly what I enjoyed most about this. The essence of fine/fun engineering is building what you need with what you have.
Your LED matrix is broken?
If it gets whacked the last segment sometimes loses its mind. A quick power cycle restores it.
Yes great design. In addition to finding a higher wattage load resistor, i believe it would be prudent to install a 5v or 12vdc fan with perhaps a thermistor. I just might have to build one of these for rooting out those pesky Chinese failed gadgets.
Yes, it would definitely benefit from a fan.
I looked at the bottom too ;-)
I thought that PSU was not rated for 110V input? if so that will seriously reduce it's output capability.
It may run but won't provide the 16A it states.
good point
I'm still surprised that it works at all.
pushing 16A on that resistor is painful
That's the risks of using whatever parts I have on hand.
Just a crazy idea.
Add a pizoelectric heat transfer plate to aid in thermal management of the resistor.
Can't remember what the device is called.
A peltier cooler?
@@pileofstuff yeah that.
It would increase potential power draw. Both by its own use and allowing the the resistor to take more current.
Like I said, crazy idea.
@@kyleallred984 Crazy ideas don't offend me (as witnessed by this janky build...)
You can't run 6 amps through that breadboard.
That's why it's so badly melted.