I might be missing something here, but is there any reason why (18:00) you don't have that 10K resistor for the test button connected to the left hand side of the capacitor?
Great project, excellent mods in this last version. This project needs a custom PCB designed so people could order the pcb online and easily assemble it on their own.
Thanks. I will be looking at designing a custom PCB when I have time. However, one of my followers Rob Bruno has already designed a new PCB using my design (up to Part 5) but we have not tested it yet. It should be OK and uses surface mount components. If you are interested it is available from: www.oshpark.com/shared_projects/tXUxCSVa You can also download the circuit from the link below: www.dropbox.com/s/orwivgbt23gl158/ElectronicDCLoad1.pdf?dl=0 Regards, Louis
Scullcom Hobby Electronics I was going to do a custom pcb for this as well but got side tracked on other projects. I may make one here soon. awesome project btw.
Thank you so much for an educational and comprehensive build. I am definately building this! The only thing I find missing is an updated parts list and a library of current (latest) schematics and software. It's quite time consuming to go through the files for each video plus having to catch vital information off the videos themselves. I would like to design a custom PCB for the project for myself and if it works well I would be willing to share that design to the community. Thank you once again and now if you'll excuse me I am going to start these videos all over again.
I thank "you" for the great project. One can not be disappointed when you release a new video. It is always fun and educational to watch your videos. Keep up the good work.
Hello, I know that a lot of time has passed, but the content you have prepared is still very useful. Thank you ! I wonder why BJT's emitter is connected to -5v? In this case it sounds like the BJT will always be transmitting? It seems like it would be more appropriate if we connected it to ground potential instead of -5v. I probably didn't fully understand the principle of its operation. Can you help me with some keywords for a deep research?
On the load balancing between 4 MOSFETs, why not use a quad opamp and control the current through each MOSFET individually? I mean, if you have to allow 1A load, each opamp will in turn control 0.25A, so that the sum is always 1A? That way each MOSFET is protected. We would need 4 shunt resistors, but that should be OK to increase the robustness. We may need the -5V bias to turn off the MOSFETs.
Hi, I like to modify my 1 mosfet DC Load (1x ir250 mosfet) to 4 or 6 Mosfets using the end stage of your design with the 2n2222 protection to avoid a run away. Would that work? what max A would you recommend for your protection circuit to work correctly ?
Hello Louis, In case i wish to change the mosfets for more power which you mentioned its possible. But to be precise or to doubly sure is there ANY LIMIT in the LCD meter reading or we need to change the firmware. kindly clarify it.
Hi , Tx for the mosfet protection circuit. I blew the irpn250 mosfets in my first DL project simulating a 7 amp 30V load for exact this reason. Great job.
Another fantastic video, Louis. I think I speak for all of your subscribers when I say that we are always excited when we see you've posted another one. You clearly know your stuff and I'm happy to learn from you.
This is really a professional project now :) Even the case looks like it's a commercial product, and the internals getting better and better every time :-D It's really fun and edifying to even just follow your work on this one.
Very nice project, i've been tinkering with the same thing for about a year now. Here is a link to my WIP 300W Dummy Load. forum.pjrc.com/threads/34863-Teensy-CC-Dummy-Load-300W?p=106516#post106516 I do want to consider improvements, such as revamping the input power, i'm considering switching to a Lipo/5V only approach and getting ride of the SEPIC design. I just recently replaced the LM4040 thanks to your videos on Reference's, I plan to use a MAX6071AAUT21+T. I'm also considering changing my LM324's but so far have not found anything that provides enough benefit to justify the premium price.
o my this is turning into a james bond 007 Q-Lab 🔬save the world project 😃it is wonderful to see all the improvements while learning along the way 🥰 thanks a lot 🥠☕
Better would be having each gate-driver NPN transistor having a seperate emittor resistance, so that they don't affect each other. Now, when a driver drains current away from a to much conducting MOSFet, it lets the other drivers throttle down the other MOSFets, just the opposite of what you want! Why not connect the testbutton to the base of the Opto Coupler transistor? That gives uniformity in the generated pulse an debouncing of the button!
I used 0.22 Ohm double resistors in one package to balance out the mosfets, works perfectly... got the resistors from an old audio amp, most audio amps have those particular resistors to balance out the PNP and NPN transistors so those resistors are already used for balancing, perfect to repurpose those for a project like this
This project is coming along quite nicely. I haven't given a lot of thought to this, but my gut feeling tells me that the 47k to -ve should not be shared amongst all the NPNs, as they each refer to that emitter rail, but the others can affect it's voltage. I know it's very low current, but that emitter rail must surely be bouncing around as the NPN's drift in and out of conduction. As I say, I haven't tried to analyze this in any depth, but 3 extra resistors cost almost nothing, and can't hurt operation...
Thanks for your comment and question. You may have a point but I have found that as the actual current through the 47K is only about 100uA or less it has little effect. The Mosfet base current is negligible. You could also increase the resistor value to 100K if you wish and it would still work. As a mater of interest I have seen HP Agilent use a similar technique in some of their bench power supplies and they also use one return resistor feeding in their case 2 transistors. As you say adding 3 extra resistors costs nothing so if you wish you can do that. I will consider adding that as an option to the final design. Regards, Louis
Louis, great update. Awesome job.Wish I had just a smattering of your design knowledge. I will now get back to the PCB design. Just starting the transition form Eagle CAD to Altium Circuit Studio. Hope to have something in a week or so (using Eagle). Quick question would a 10uf smd tantalum cap be suitable for the bypass. Noticed you used one for the trigger ckt. That would save a lot of room on the design. Also they are easier to solder that the SMD electrolytic caps. Look for something soon.
Thanks Rob, Yes a 10uF SMD tantalum capacitor should be OK. Look forward to seeing the results of your work on the PCB. The problem is I keep adding things to this project which is a PCB designers nightmare :) Regards, Louis
I was thinking... is it possible to do any kind of AC test? If not, what kind of challenge we have? Thanks, I always excited when I see a new video release of yours here in my channel notifications.
Thanks for your comments. I suppose you could do an AC test if you added a diode series with the load. I was thinking of adding a diode in anyway to give some reverse voltage protection.
Great project Louis. I ordered a ton of components last week and most of them are here. I had to buy the enclosure and a few other parts from China so they won't be here for another week or so. I still have other parts to buy along the way but I'll buy them as and when I get to them. I think the easiest way for me to tackle the project is to just follow it through from Part 1 to 6, making the mods, as you did, along the way. I've just finished the PCBs so I'm about to start populating them. Keep up the good work. www.dolomitesprint.com/images/misc/DC_Load1.jpg www.dolomitesprint.com/images/misc/DC_Load2.jpg
Thanks for the comment and the links to your project images. Looks like you made a nice job of the PCB's. I am in the progress of just finishing an update to this project (Part 7) and should be able to upload it to my RU-vid channel soon. Regards, Louis
This looks like a very feasible design. I have only one problem with it, there is no cutoff relay. When you're doing battery testing, you want at least 2 to cutoff methods. If for any reason the MOSFET's fail to turn off there is no way to stop the test. Batteries don't care if the alarm has sounded. Invariably you'll get distracted and fail to stop the test before you leave. The result is fire or at least a ruined battery.
dear sir may i ask you please to create a pdf with all the pcb's in order to save photopaper or gels? i tried to convert to doc and combine all to one but the images loose quality a lot. or maybe to export the images? thank you.
Thanks for you comments. I will have a look and see if it is possible. I am also looking at redrawing the final full schematic for this project, in which case I could try and design one PCB with all the circuit on it.
Yes the PSMN2R5-60PLQ may be a good alternative. I had a look at the datasheet and it should be fine. Thanks for bringing this Mosfet to my attention. Regards, Louis
Thanks a lot for your work, Louis! Your style of presentation reminded me of Jim Williams (RIP). Can you tell me what dI/dt can provide this e-load in transient mode with actual mosfet drive circuit (2k2 gate resistors)? It's very important when testing SMPS and debugging FB loop compensation. Is it possiple to change I/R/P value under load without stop operation?
Hi Tor, Thanks for your comments.With regards the 2K2 gate resistors I use they do slow down the mosfet switching times in transient mode. Normally you would use a low value gate resistor of 100 ohm or less to ensure fast switching times. However, for high current mosfet's as in this DC Load the gate channel capacitance can be quite high and a rapidly changing drain voltage can produce several mA of transient gate current which may damage the mosfet. Under normal conditions the dI/dt would be around 100A/uSec. However, with a 2K2 gate resistor it is lower. When looking at the waveform of the current switching rise and fall slopes I am getting a period of about 100uSec in transient mode. Yes for normal operation of the DC Load in constant current, resistance or power it is possible to change the I/R/P value on the fly with the load switched on. Regards, Louis
Have you ever been interested in this INA219 module? www.aliexpress.com/item/Free-shipping-Zero-drift-CJMCU-219-INA219-I2C-interface-Bi-directional-current-power-monitoring-sensor-module/32813710560.html Interesting solution for simple programmable lab PSU as for me. What do you think? P.S: sorry for my bad english, I'm from Ukraine :)
Thank you for these series of videos on the DC Load and appreciate your way of explanation and pace too. I enjoyed and learned new things as well. May I ask couple of quick questions here please? 1) How effective is to use multiple MOSFETs in parallel? 2) What is the minimum Amp can this load be set to ? 3) Can load value be increased in 1mA to 2mA and so on incrementally?
Hi Pradip, Thanks for your comments. In answer to your questions: 1) In this project we use 4 power mosfets in parallel. As no two mosfets are identical you need some balancing to stop any one using more current than the other (current hogging) and so over heating and eventually going in to thermal runaway. To reduce/eliminate this possibility, in this project I use individual ballast resistors in the source return of each power mosfet. Also I use some active feedback as I explained in my last video (Part 6). 2) In practice the minimum Amp load can be about 2 mA. 3) The load can be incremented in 1mA steps. Regards, Louis
Hi Louis, thank you for this explanation. Since IGBT are meant for handling such a high load, would you please suggest to use a single IGBT instead? Any trial you did with IGBT as well before considering the multiple MOSFETs in parallel please?
Great project and excellent videos. I've been following this project and I've built sections of the circuit on a breadboard to help understand how it works. I now plan to build the complete project over the next 1-2months. However, looking around at commercial DC loads - I see most have a Constant Voltage (CV) mode. 1) Can you please explain what CV mode is and why you would require this? 2) How easy would it be to add to this project? i.e. Could this option be provided by software or additional hardware + software Thanks
Hi Kevin, Constant voltage (CV) mode allows you to set a fixed voltage. The electronic DC load will sink the necessary current to keep the voltage at the set level. This could be simply achieved by adding an additional option in software. No additional hardware is required. I will have a look at this as an option but it depends how much programme memory space we still have available in the Arduino (we have used most of it to date). Regards, Louis
Thanks Lewis for another great video... will definitely build it up. I would love to know what you are working on next.Keep up the great work! Big thumbs up.
Thanks Brett for your comments. I am working a a few new projects for future videos. I try to add something new and introduce some new aspects of electronics where I can. I am currently looking at some RF oscillator solutions for the hobbyist. May also look at power supplies. Regards, Louis
Louis, as always an excellent project. I built a portion of the circuit with two mosfets and two transistors, in my tests one of the mosfets got significantly hotter that the other. So to understand what was going on with the circuit I simulated it on LTSpice. To simulate the difference in conductivity I varied slightly the .1 resistors and the simulation showed significant differences as expected. I cut out the transistors from the circuit and there was no significant change on the previous result. What made a difference, at least on the simulator, was to have the base of the transistor on the source of the neighboring Mosfet. Using your schematic, the base of Q6 on the source of Q3, the base of Q7 on the source of Q2, etc. What do you think of this arrangement?
Hi Dan, Thanks for your comments. I checked my unit again and it seems to work OK. However, as you seem to be having some issues I suggest trying a few of things: 1. Increase the value of the 0.1ohm resistors (R22 & R23) to 0.5ohm. 2. Instead of using just the one 47K resistor (R30) supplying all the emitters of the transistors from the negative supply - instead use a 47K resistor in each of the emitter returns so as to separate the emitters. 3. You could also try increasing the 2.2K mosfet gate resistors to a higher value say 4.7K or even higher. Also in my circuit I used the 2N2222 transistor as I found that it worked better than some other general purpose NPN transistors in this application. As the 2N2222 is good a fast switching and responds very fast (much better than a 2N3904). In this application I would recommend using the 2N2222. With regards your suggestion of connecting the base of the transistor to a neighboring Mosfet seems strange as it would not reduce the gate/source voltage of the mosfet that was over heating. I will have another look at my circuit and do some more tests. Regards, Louis
This turned out to be a great project and a very fine tool for every electronics enthusiast's bench. I can't wait for the finalized version. I hope there will be a new PCB design when this is finalized, that includes all the add-ons and corrections. Thank you for doing this and sharing with us.
Thanks Bora for your comments. A number of my followers are already working on a new PCB design (double sided PCB). I may also design a new single sided PCB if the hobbyist wants to try make one themselves. Regards, Louis
Yes sir! The "hobbyist" do want to try :D I have made my own double sided PCB's before, some turned out ok, some not so much, but I've decided that I's better to have them made professionally. Chinese manufacturers do really great jobs, with very fair prices. Recently I ordered a 8cm x 30cm double sided PCB, with more than 300 drill points. The price was 50USD for 5 copies (that's the minimum), through hole plating, solder mask and silk screen on both sides, shipping included!
I enjoyed this series very much. Thank you very much Would it be possible to use a ATmega328P chip instead of a Nano? However. It must be able to run the software you created but would fit nicer into a homemade pcb. I am not fond of the surface mount components (just because I'm a hobbyist and not that good with it) It would be great if you do a redesign of the PCB for the completed project. Kind regards Stephan
Thanks Stephan. Yes you could use the ATmega328P chip and this is something I have considered. If you install the Arduino boot loader then the software can be loaded just like the Nano. As I have made a number of changes as the project has progress I will probably redraw the schematic and maybe redesign the PCB to match as a final end to this project. I hope to get back to this project soon with an update but as I have to go in to hospital for some treatment it is likely to be a few weeks before I am able to do it. Regards, Louis
Dear Louis Thank you for your response. I hope and trust that your visit to the hospital will turn out well. Kind regards Stephan PS: I must say. It looks as if you can buy a Nano for less that the ATmega chip. PPS: I enjoy the videos that you have created
Hi Stephan, Thanks had to rest for 5 days after my hospital treatment. All went well. I am now getting back to the DC Load project to try and bring it to a conclusion. I have now redrawn the schematic with all the previous modules included. Have made a few changes and some additions. I am working on a one board PCB now which will replace all the previous modules. I just need to test some of the additions and make some changes to the software. I hope to be able to film everything soon and then upload Part 7 of this project. Yes I agree the Arduino Nano can be cheaper than buying the ATmega chip. So I find it easier to just use the Nano. But if you want a neater arrangement once a project is finalised then building the ATmega chip in to the PCB does work well. Regards, Louis
Thank you very much for the video. As a further upgrade it would be excellent to add a USB port to send the data to a computer via the Arduino (for the purpose of data storage, display and analysis) both in this and the miliohm meter project. All the best.
Thanks for your comments. I will have a look at adding a USB output. This project seems to just keep developing, its knowing when to stop :) I am currently also looking at a completely new project as well. Regards, Louis
I've finished the project but have a problem. I'm using software version 24. When I switch on the 12V power supply to the load terminals the display immediately shows 4.3A even though the load is off. If I try to limit the current, nothing happens. Toggling the load on & off makes no difference. I've spent hours trying to work this out but failed. Can anyone help me out? www.dolomitesprint.com/images/misc/EDCL1.jpg
Hi, Sorry to hear you are having some problems. You need to do some basic checks: 1. First check that you do not have and shorts across PCB print tracks, especially around the small SMD IC's. (particularly around U7, U5, U6). 2. What voltage do you read at pin 7 of U6 (MCP3426)? 3. With no load set what voltage do you have on pins 12 and 13 of U7D (AD8630)? 4. Even with this error on the display (actual current) can you still set a current with the rotary encoder and does that reading show on the "Set I = " reading on the LCD. 5. if you press the rotary encoder are you able to move the cursor on the "Set I = !" reading of the LCD display. Check the above first and let me know what you find. Regards, Louis
1. I checked U5, U6 & U7 many times using my 60x magnifier and everything seems fine, the tracks also. 2. I'm seeing 0.45V on pin 7 with the 12V load connected but load set to off and no load set. 3. The same 0.45V on both pins with the 12V load connected but load set to off and no load set. 4. Yes. 5. Yes. I haven't temporarily disconnected R9 yet.
It might be a silly question but can you double check that all the surface mount IC's are mounted in the correct orientation on the PCB with pin 1 positioned near the dot on the print! Its easy to mount them in reverse at times. Before going through the details below check R17 is 0.1ohm and that the earth side is firmly connected to ground. If OK continue below: If the load is set to off you should have zero volts on pin 7 of U6 and also on pins 12 and 13 of U7D. The 0.45V reading you are getting is causing the actual current reading to show 4.371A. If the power Mosfets are getting hot then this would suggest that something is causing maximum drive voltage to the Mosfet gates - if so disconnect R15 (100 ohm) and see if reading on the LCD actual current goes to nearly zero amp. If that what happens let me know. If there is now change to the actual current reading on the display carry on with the details below: So we need to establish why you are getting 0.45V on pin 7 of U6 and on pins 12 and 13 of U7D. First disconnect the R9 (100 ohm) resistor to isolate the output of U7D from U7 and check again the voltage on at pin 7 of U6 (MCP3426) - if the voltage is still 0.45V then either U6 is faulty or their is an issue with pin 8 of U6 not connected to ground. If after disconnecting R9, pin 7 of U6 is now zero or near zero volts then the fault is either with U7D or the output stage. To check if U7D is the fault temporary disconnect R16 - if the voltage on pins 12 and 13 of U7D is still 0.45V then U7D is likely faulty. If the voltage on pins 12 and 13 of U7D is now 0V then the fault is in the output stage - first temporary disconnect R15 to remove the drive voltage from the Mosfets if you are still getting 0.45V on 12 and 13 of U7D then check the output Mosfet's - if the voltage on 12 and 13 of U7D is zeros then the fault is likely U7. Finally, sometimes when soldering surface mount IC's solder my run under the IC case and cause issues. Let me know how you get on with the above suggestions. Regards, Louis
