» Electronics » Power supplies »Electronic load with infinitely adjustable current

Continuously adjustable electronic load


Over time, I accumulated a certain number of different Chinese AC-DC converters for charging the batteries of mobile phones, lights, tablets, as well as small switching power supplies for electronic crafts and actually the batteries themselves. On the cases, the electrical parameters of the device are often indicated, but since most often it is necessary to deal with Chinese products, where it is sacred to overestimate performance, it would not be out of place to check the real parameters of the device before using it for crafts. In addition, it is possible to use power sources without a case, on which information about their parameters is not always available.

Many may say that it’s enough to use powerful variables or constant resistors, car lamps or simply nichrome spirals. Each method has its drawbacks and advantages, but the main thing is that using these methods of smooth adjustment of current is quite difficult to achieve.

Therefore, I collected for myself the electronic load on the operational amplifiers LM358 and the composite transistor KT827B with testing power supplies with voltage from 3 V to 35 V. In this device, the current through the load element is stabilized, so it is practically not subject to temperature drift and does not depend on the voltage of the source under test, which is very convenient when removing load characteristics and performing other tests, especially long ones.

Materials:
- chip LM358;
- transistor KT827B (composite NPN transistor);
- resistor 0.1 Ohm 5 W;
- 100 ohm resistor;
- 510 ohm resistor;
- 1 kΩ resistor;
- resistor 10 kOhm;
- variable resistor 220 kOhm;
- non-polar capacitor 0.1 μF;
- 2 pcs oxide capacitor 4.7 uF x 16V;
- oxide capacitor 10 uF x 50V;
- aluminum radiator;
- stable power supply 9-12 V.

Instruments:
- soldering iron, solder, flux;
- electric drill;
- jigsaw;
- drill;
- tap M3.

Assembly Instructions for the device:

Operating principle. The device by the principle of operation is a current source that is controlled by voltage. A powerful KT 827B composite bipolar transistor with a collector current of Ik = 20A, a gain of h21e of more than 750 and a maximum power dissipation of 125 W is the equivalent of a load. 5W resistor R1 - current sensor. Resistor R5 changes the current through resistor R2 or R3 depending on the position of the switch and, accordingly, the voltage on it. An amplifier with negative feedback from the emitter of the transistor to the inverting input of the operational amplifier is assembled on the LM358 operational amplifiers and the KT 827B transistor. The effect of the OOS is that the voltage at the output of the op-amp causes such a current through the transistor VT1 so that the voltage on the resistor R1 is equal to the voltage on the resistor R2 (R3). Therefore, the resistor R5 regulates the voltage across the resistor R2 (R3) and, accordingly, the current through the load (transistor VT1). While the op-amp is in linear mode, the indicated value of the current through the transistor VT1 does not depend either on the voltage on its collector or on the drift of the parameters of the transistor when it is heated. The R4C4 circuit suppresses the self-excitation of the transistor and ensures its stable operation in linear mode. To power the device, a voltage from 9 V to 12 V is required, which must be stable, since the stability of the load current depends on it. The device consumes no more than 10 mA.

Work sequence
The electrical circuit is simple and does not contain many components, so I did not bother with the printed circuit board and mounted it on the breadboard. Resistor R1 raised above the board, as it is very hot. It is advisable to take into account the location of the radio components and not to place electrolytic capacitors near R1. I didn’t quite manage to do this (I lost sight of it), which is not entirely good.

A powerful composite transistor KT 827B installed on an aluminum radiator. In the manufacture of a heat sink, its area should be at least 100-150 cm2 at 10 watts of power dissipation. I used an aluminum profile from some photo device with a total area of ​​about 1000 cm2. Before installing the transistor, VT1 cleaned the heat sink surface from the paint and applied the KPT-8 heat-conducting paste to the installation site.

You can use any other transistor of the KT 827 series with any letter designation.

Also, instead of a bipolar transistor, you can use an IRF3205 n-channel transistor or other analog of this transistor in this circuit, but you must change the value of the resistor R3 to 10 kOhm.

But there is a risk of thermal breakdown of the field effect transistor with a rapid change in the passing current from 1A to 10A. Most likely, the TO-220 case is not able to transfer such an amount of heat in such a short time and boils from the inside! To everything you can add that you can still run into a fake radio component and then the parameters of the transistor will be completely unpredictable! Either the aluminum housing of the KT-9 of the KT827 transistor!

Perhaps the problem can be solved by installing in parallel 1-2 of the same transistors, but I practically did not check - the very same number of IRF3205 transistors are not available.

Housing for electronic load applied from a faulty car radio. A handle for carrying the device is present. Bottom mounted rubber feet to prevent slipping. As legs I used caps from bubbles for medical preparations.

On the front panel for connecting power supplies placed a two-pin acoustic clip. These are used on audio speakers.

There is also a knob for the current regulator, a power on / off button for the device, an electronic load operation mode switch, an ampervoltmeter for visual monitoring of the measurement process.

An ampervoltmeter was ordered on a Chinese site in the form of a ready-made embedded module.
Continuously adjustable electronic load

The electronic load operates in two test modes: the first from 70 mA to 1A and the second from 700 mA to 10A.
The device is powered by a stabilized switching power supply voltage of 9.5 V.

When connecting an electronic load, a 0.49V value is displayed on the ammeter (the value may vary).This is a feature of the operation of the LM358 operational amplifier and the KT827 composite transistor, but this does not affect the measurement accuracy in any way. If you want an aesthetic look, you can use a field effect transistor, then the readings will be 0 V. Once again I repeat - these values ​​do not affect the measurement accuracy!


Conclusion
With this electronic load, I was able to squeeze about 100 watts with a 12V power supply, maybe more, but there is nothing to check. Smooth adjustment of current, minimum temperature drift and independence from the voltage of the tested source allows you to more accurately determine the characteristics of the tested power source.

This device is suitable for testing single power sources, but if you approach the matter wisely, you can create on its basis a multi-channel device for checking, for example, a computer power supply.
7.3
6.8
6.2

Add a comment

    • smilesmilesxaxaokdontknowyahoonea
      bossscratchfoolyesyes-yesaggressivesecret
      sorrydancedance2dance3pardonhelpdrinks
      stopfriendsgoodgoodgoodwhistleswoontongue
      smokeclappingcraydeclarederisivedon-t_mentiondownload
      heatirefullaugh1mdameetingmoskingnegative
      not_ipopcornpunishreadscarescaressearch
      tauntthank_youthisto_clueumnikacuteagree
      badbeeeblack_eyeblum3blushboastboredom
      censoredpleasantrysecret2threatenvictoryyusun_bespectacled
      shokrespektlolprevedwelcomekrutoyya_za
      ya_dobryihelperne_huliganne_othodifludbanclose
52 commentary
EL
Related video. (Not mine) Why field workers are burning in this scheme.
El
Scheme I. Nechaev from the magazine Radio No. 1 2005 (p. 35) In general, as I did not torment her, and paralleled two and 4 pieces. up to 10 Amps never pulled out, transistors burn. At low currents up to 2 amps you can work. In general, I decided to now collect according to the aka kasyan stripped from the Chinese scheme on LM324.
In case write me
Ali Bastre
He has already burned several field workers. 2pcs (paired) IRL3705 at a current of 5 A burn out in a few seconds. The heat sink does not even have time to warm up slightly. Sorry, good transistors. I'll try to switch to KT827. I hope the Soviet transistors do not fail!
Quote: El
I would very much like the author to add a digital ammeter connection diagram to the article. Already burned two .. ((
Take a look heremaybe help. An external shunt can be excluded, it is only to expand the limits of measurement.
Author
Pierced IRF3205 with a little more than 4 Amps
So you need to parallelize a pair of IRF3205, above was already about that. It is better to connect a voltammeter, as Pokhmelev already said, according to the scheme from the page of the Chinese seller where they bought it.
. See the connection.
El
I put a radiator with cooling, as planned. I drove 1 hour at 25V 1A, everything works stably. I decided to check the maximum load, IRF3205 broke through at a little more than 4 Amps. Replaced it, smeared thermal grease "from the heart" the same result, slightly more than 4 amperes increased and immediately struck. And it goes deafly in the short circuit (between the drain and the source 16-18 Ohms), if the tested PSU without protection, it will burn out.
El
assembled according to the scheme on IRF 3205
El
Thanks, according to the scheme you proposed, I also connected it, it did not work. The SChs checked on a laboratory PSU from 0.2 to 32 volts that the circuit keeps the specified load stably, (it does not give out more PSUs) I set 250mA, 0.5A and 1A, did not go higher, because temporarily, while testing the circuit on the table, I put a small heatsink, at 1A it had already warmed up to 60C. He collected the circuit on a signet. Did under the comp. cooler with fan. Resistor 0.1 put 10W ceramics.
I connected according to the image:

Where "LOAD" is the electronic load itself, and "load voltage" is the test power source. I did not connect the power supply of the multimeter, but I took the plus from the source to power the electronic load (I have 12 volts).
And do you manage to test power supplies with voltages above 20 volts?
El
This was done first, and googled the possible options, but something went wrong: with a minus of power hanging in the air, there was no indication at all. At the same time, the multimeter in the gap shows perfectly. I tried a bunch of options. The first burned out when it stuck into the gap (-), the second into the gap (+) of the tested PSU. I took the power of the ammeter from the power supply of the circuit. Therefore, the question arose, how did you connect?
If you bought online, see the diagram on the seller’s page.
El
I would very much like the author to add a digital ammeter connection diagram to the article. Already burned two .. ((
Check installation.
By the way, I looked at the circuits, the current limits were switched incorrectly: when switching the load current during the passage, a short-term surge of a very large current is possible, limited only by the collector resistor and the channel resistance of the fully open transistor. The switch should be redone: turn on 1 kOhm constantly, and connect a 110 Ohm resistor to it in parallel.
1. The shutdown time for currents from 1A to 9A is approximately the same 16-17.5V.
2. The successive shutdown of the field workers did not give an effect.
Yes, the input load from a primitive LBP, but for the first time I noticed this malfunction on two series-connected batteries. As if the load of defense is leaving, maybe something is not so connected. And at your load did not try to supply 24 volts (stable, with battery)?
Author
with a smooth increase in input voltage

Are you applying voltage to the load input from the LBP? In the process of adjustment, I had a case when I applied voltage from the LBP, but I could not regulate the load by throwing the voltage with a fixed output voltage from another PSU, the circuit worked. I did not understand, but the assumption that this is due to the current (voltage) stabilization circuit on the LBP itself.
Finally, from the third message, the situation began to clear up. It was impossible to answer the previous two, since one did not contain at all no information other than a silent photo of a ball of wires, and the other contained mutually exclusive statements and two spoilers that did not open.
By last post.
1. Compare the moment of shutdown at different currents.
2. Try to disconnect the drains of field workers one at a time.
I can’t download the video, but in general, with a smooth increase in the input voltage, the load is disconnected at a voltage above 16 V.
I ask to prompt in solving the problem. I can take measurements, if necessary.
Thank you for the clarification! He collected the load on two channels, according to the scheme, he took the transistors more powerful - irfp260n.


Everything works, loaded to the limit of the multimeter. But, when trying to connect to 24 V, the load does not work, i.e. not regulated.

What could be the mistake?
For complete happiness, put another 0.1 uF ceramic between 4 and 8 legs of the microcircuit.
Quote: ocherett
In addition, the quality of Chinese irf3205 can greatly affect
Chinese powerful transistors are often fake: under the marking of a decent transistor, a frail crystal like our KT817 is hidden.
Author
Yes, the second scheme should work.
Then so?
Thanks! Also look, very roughly threw the scheme, what needs to be changed? Or do you need to connect the gate of the second transistor with foot 7 (after 10k)?
Author
And even better, since the LM358 is a dual op-amp, make a circuit to the second channel (legs 5,6,7), repeating the chain R1, VT1, C4, R3.Combine non-inverting inputs for general adjustment
Author
I redid the circuit on irf3205, and it struck almost immediately (current less than 1A, 5 seconds in time) i.e. there was no heating

The TO-220 body, like the irf3205, is even less able to dissipate heat due to the smaller flange area. In addition, the quality of the Chinese irf3205 can greatly influence, perhaps the transistor crystal simply did not manage to transfer heat to the radiator and burned out.
And how to use the second channel of the amplifier, for another transistor, do the same strapping with resistances?

R1, VT1, C4 duplicate, R3, which goes to the 2 leg LM358, I think it will be common. For fieldwork 10kOhm
Quote: ocherett
And you already get 36 * 3 = 108 W, which is on the upper side of the transistor

Yes, I think it was, but after that, I redid the circuit on irf3205, and it struck almost immediately (current less than 1A, time 5 seconds) i.e. there was no heating, so I think it’s not. At the moment there is IRFP260N, I want to try on it, but doubts remain. And how to use the second channel of the amplifier, for another transistor, do the same strapping with resistances? Here, with a potentiometer it comes to 3 pins, and if there are two channels, then where?
KT827 went beyond the border OBR (area of ​​safe operation). With IRF3205 - it is not clear. We can assume a poor filtering of the output voltage of the PSU, the presence of "needles" on it. If you have an oscilloscope, you need to look at the shape of its output voltage.
Author
Up to 3A smoothly, in a couple of minutes (this is kt827), and irf3205 struck almost immediately, the current was less than 1A
. Well, the dissipated power of the KT827 transistor is -125 W and this is at + 25 ° C, and if the temperature is higher, the power will be even less. At 50C, there is already about 100W of dissipation. The heat sink is very important (radiator, CBT, etc.) And you already get 36 * 3 = 108 W, which is on the upper side of the transistor
Up to 3A smoothly, in a couple of minutes (this is kt827), and irf3205 struck almost immediately, the current was less than 1A.
The current was immediately given 3 A or did they gradually raise to this value?
Please tell me, I took off the power of about 100W (short-term) and 60W (long-time) discharging the battery, everything is fine. But, when I tried to test the Chinese ac / dc 220v / 36v block, a transistor (kt827) struck, at a current of about 3A, after I redid the assembly to irf3205, and the situation is the same - everything works correctly for 12 volts, but it struck instantly on this block. What could be the reason? At the output, the ac / dc block produces 35.8 volts.
Quote: kouroff
What to do to increase the limit of the tested power, say up to 200 watts
Include several modules in parallel, as done, for example, here. Naturally, you need to choose the right transistors for current, voltage and power.
Quote: kouroff
What does the voltage limit depend on?
Forget the filthy word "voltage"! ireful
Maximum voltage depends on the maximum allowable Uke power transistor, it must have a margin of at least 20%.
What can be done to increase the limit of the tested power, say up to 200 watts, and on what does the voltage limit depend?
Quote: Wonderful Fox
I’m used to power everything from lithium and I want to make the device stand-alone, and sticking there 2-3 batteries 18650 is not very hunting
If 2 pieces is a lot, put a Step-Up converter.
I repeated the project, spread everything on the printed circuit board, it worked perfectly, did it on the field effect transistor, I haven’t tested it yet, I’ll come up with a case so I will immediately start testing this device, but with a rather small heatsink and transistor in the TO-220 case 100+ watts I shot for a short time (until the radiator starts to overheat) I think it’s a suitable device, the only power supply is not convenient ... I’m used to power everything from lithium and I want to make the device stand-alone, and sticking there 2-3 batteries 18650 is not very hunting
Sorry, I disappeared the circuit - everything works correctly.
There are no clairvoyants.;) Describe the situation.
Good afternoon. I assembled this circuit, the power supply loads, and there is no separate battery (no reaction). What could be the reason?
Quote: ocherett
There is no fundamental difference. Maybe it’s not established by standards
You are mistaken. There is a fundamental difference. When the power is on, the marking indicates the status is “off”.
The metal-glass case is indicated in the datasheet and the word "metal" just means aluminum! Do not believe me, submit for analysis!
Try to prove that it is aluminum. Just do not work.))
KT-9 (aka TO-3) has a steel flange and cover, and a copper gasket between the crystal and the flange for heat distribution.
Here I forgot to write that I adjusted the output power to 9.5V
If it really stabilize, then it was necessary to put at least the simplest linear stabilizer on a zener diode or TL431.
Here is the usual divider! What to write here?
"10 A" and "1 A" - informatively, "1: 1" and "1:10" - slurred.
Author
The power key is turned upside down.

There is no fundamental difference. Maybe it’s not established by standards
Really - aluminum?

The metal-glass case is indicated in the datasheet and the word "metal" just means aluminum! Do not believe me, submit for analysis!
What is this "stabilized" source with an output voltage of 4.5 ... 9.5 V?

Here I forgot to write that I adjusted the output power to 9.5V
On the front panel you need to write not the mysterious “1: 1” and “1:10”, but the understandable “10 A” and “1 A”.

Here is the usual divider! What to write here?
Why are there two R3 resistors in the circuit?

Here's a typo. I agree
The device is powered by a stabilized switching power supply voltage of 9.5 V.
What is this "stabilized" source with an output voltage of 4.5 ... 9.5 V?
Either the aluminum housing of the KT-9 of the KT827 transistor!
Really - aluminum?
The power key is turned upside down.
Power R1 should be at least 10 watts, and better - more. And on the lid of the device you need to bring the graphs of the OBR of the transistor, since there is no protection against excess power. There are two schedules - standard (for short-term operation) and for long-term operation (taking into account the radiator area).
On the front panel you need to write not the mysterious “1: 1” and “1:10”, but the understandable “10 A” and “1 A”.
Quote: LeoBrynn
And where is the link?
In this case, the absence of a link is not critical, since the device is quite standard and mass-produced. But the connection diagram had to be given, and at the same time tell whether the option of using an ammeter shunt as a current collector was considered.
The product is just normal. And "sucks", in your opinion, is your unfounded proposal. How do you imagine that?
And the controller on the "face" of the device is drawn the other way around.
Yes, a sloppy topic. sorry. You take the unit from the computer and the controller to it. all.

We advise you to read:

Hand it for the smartphone ...