my-o-my this is just like the Heath-Kit project days 😃such a wonderful video for the folks just starting out in program electronics as well as the old timers 🥰 thanks a lot 🤩🍎🎂☕
Hi Great project ! I been following the whole project . One question . the MPC 3426 is it possible to have a software version with ADS 1115 ? MPC 3426 is kinda hard to find it , and it's too bad cause it's such a great project .
@@PauloSilva-ll4vs That's beautiful ! I did my own version of a DC load .... and I used the ADS as well . The thing is ....I don't have the kind of menu the project has .... Mine can deliver 200w .... and 15 amps in current mode .... I used a DC load kit and adapted it to my design .... I adjusted the DAC to give me a range fom 0 to 500mv .... its very stable considering I haven't used used a Vref . I also copied the fail safe . ....Shutdown at over 200watt ; over voltage 28 volt and over current 15.5 amps
Brilliant, thanks again, been waiting for part 4, my own has taken shape, used those keypads, but then they wouldn't fit on the front panel of the cases I had, so made my own from some salvaged switches.... Thanks to you for explaining it all simply and inspiring us all to go and build one! Learnt tons, looking forward to more, thanks Louis....
Good stuff again Louis! When I build my own electronic load, and went to parallel Mosfets to share the load, I quickly noticed a thermal imbalance between the Mosfets. One would get very hot, while the other stayed cool. In my opinion, there are two things you can do to prevent or counter that. One is to create a feedback/driver loop for each Mosfet. This will ensure a more balanced load. An other method, the one I adopted, is to measure the temp on each Mosfet and reduce the gate drive of that Mosfet if there is a thermal imbalance. I used a transistor bolted to the Mosfet fastener as the temp pick-up and some analog circuitry that keeps the relative Mosfet temps within a degree C or so from each other. Adding a transient capability is very usefull, especially for testing power supplies, etc. In that case, the pulse characteristics will dictate the configuration of the Mosfets and drivers, shunt and the wiring. Everything will most likely have to be physically closer to what you have now to get a reasonable rise/fall time for some serious testing purposes. Looking forward to see your solution!
Thanks Paul for your comments and ideas. I am continuing to test the Mosfets in parallel to find the best solution. So far it seems to work quite well, may be I am just lucky with the 4 mosfets. I am going to monitor the current flow through each one at the same time. Also looking at possibly altering the wiring arrangement as I discussed on the video. Regards, Louis
I like your teaching style. Its very simple to understand due to the they way you go step by step, even your videos are arranged by starting simple and getting it working then adding a few features/upgrades. I assume you have some form of teaching background ? If not you really missed your calling. This channel has become one of my favorite electronic channels. Great work.
Thanks William for your comments. I did teach electronics back in the late 1970's and early 1980's for about 8 years. If you are interested in my background you can find some information on my website at: www.scullcom.uk/about/ Regards, Louis
Just had a look at your CV on your website. What a background! It is now no surprise to me that you are as knowledgeable as you are and so good at putting information over. I am just grateful that you are so willing to share your knowledge. Thank you.
I know I'm little late, but just wanted to say that it is really worth the time to read about Louis on his homepage, what an amazing person, lots of people are teachers but a good teacher is rare, one for that feedback of his students is important, one that likes to be questioned, one that grows from criticism, this is the only way how a professional can learn to explain complex things to hobbyists and newbies, with such teachers only there would be only a few "bad" students, but most people would be scientists and we would travel through the universe already, I know that from myself and am pretty sure about that, always sucked in maths, but was always the best in physics which was mostly math, the difference was that the math teacher just waited for the end of the day while the physics teacher loved what he did, loved physics, was always excited about what he teached us and infected us with his excitement, I never had a lower grade than a 1- in physics, while in math I even had once a 6 end of the year, but nobody cared, they just say it is the kids fault, as a young person you can say whatever you want, they don't listen anyways, they even told my parents to visit a psychologist instead of testing the competence of the teachers they hire, years later all know I have an IQ of 138 but the lowest school grade that count in Germany, am I an idiot or does the school system suck? We all know about the answer, atleast about schools ... 😄
Nicely done Louis. At least one of our local hams is working on your GPS reference project and has designed a PCB for it that can be sent out to one of the PCB houses (he plans to send you the file) and he's then going to start working on this project. If I didn't already have a electronic load on the bench I'd be building one of these!
Hi Louis, absolutely excellent work ! Please let me put some future suggestions - ideas. 1. A DC Load in most cases is required for high current flow applications. So, an expansion to 30A even to 50...60 Amps could be a serious update. 2. Linear MOSFETs (like those from IXYS) have better performance and thermal response for this case of application. 3. Paralleling MOSFETs with its own driver - OpAmp and its own shunt resistor are very much better topology. This is a conclusion from tests and measurements that i have made (better MOSFETs matching, thermal activity, linearity, etc) and the results from market devices (if you can open and review one of these you will see this connection topology). 4. Op Amps to drive the Mosfets like those for audio applications are excellent choice because of high dynamic range, high slew rate, low noise and high current output to driving the gate of mosfet. A good example is LM833. 5. The differential amp needs some kind of protection at input stage to avoid malfunctions caused from the spikes of the under test power supply. 6. Why not and a Constant Voltage mode like the market devices ? You have already all the required stages. Is just programming code ... Keep the good work! Thanks for your attention. With kind regards Simon
Hi Simon, Thanks for all your suggestions. I have been playing with some IXYS Linear Mosfets (tested the IXTH110N10L2) but they did not work in my circuit as they require different drive condition. They are expensive to. Something to look at for the future. I will be looking further in to using power mosfets in parallel and may have a look at individual driver circuits or other solutions. Agree about protection for the OP Amp. I may add some diodes on the input to do that job. I have thought about a constant voltage mode and it could be easy to add in software - just need to ensure it can sink the required currents. Regards, Louis
I have been building Electronic Loads myself for a year or so, and it's very nice to watch some of the same evolutionary steps I've taken. But I would like to warn you about a mistake that I also made before. Regarding Mosfets in parallel: RdsON really has a positive temperature coefficient BUT the gate threshold voltage has a negative temperature coefficient, wich in such an application (linear mode) will probably be the most dominant effect of temperature drift. It's good to keep it in mind.
Brilliant work, Louis! Very nice upgrades. I'm always happy to see another video appear on your channel. You do a very nice job explaining everything; you should be *much* more prominent in the electronics maker community. I've been eyeing that enclosure myself, for an isolation transformer and power meter project. I'm happy to see that it appears to be of good quality. Cheers!
I remember years ago Dave Jones did a video or two on DIY loads and the community response to it spawned interpretations of all sorts of variations and capability. It seems history is repeating here and the project has certainly evolved from much humbler beginnings to what now looks a very capable load. Since the design has come so far I would add a second call for a unitised PCB in the future. For the record I no longer subscribe to Jones' channel which IMHO has devolved from it's beginnings whereas your approach better suits my requirements. Congratulations and gratitude for the effort and excellent explanations. One detail I noticed when you started the demo was you nulled the voltage reference before powering the source which introduced another offset when the load's input was energised but with the load off. I suspect the small errors reported in use would be further reduced if the nulling was done just prior to engaging the load.
Great Video as usual here. Thanks for the improvements, especially with the keypad input. Go on with the great work Louis. I am looking forward for your next episode. Greetings from Germany Michael
I have watch several of your video and they are very thorough and educative, and i have learn a great deal from you. just want to say thank you and keep bringing projects like this
I so do appreciate all your hard work and creativity in this project. I look forward to seeing the software update for the voltage correction and all the other updates. In a few weeks I am going to try to replicate your build. Thanks again and very awesome work.
Like any other of my projects,I want more out of it... I'm going to try incorporating a automatic pwm fan control and try it with a arduino capacitive touch panel. I will document it as i go and let you know how it goes.
Thanks for the great quality content! Your channel has quickly become one of my favourites on TH-cam, when it comes to electronics. I have a question: Is it really ok to parallel mosfets in linear applications, without adding source resistors? If I remember correctly Paul Horowitz and Winfield Hill in "The Art of Electronics" third edition (page 192) argue that it's OK to omit source resistors only in switching applications, and not in linear ones.
Maybe you can program the fan to be pwm controlled in relation to the heatsink? Then it will be less annoying:) And some vibration isolators if you've got them!:) (i hate loud fans) The project looks amazing! This is way better than looking at EEVBLOG! You have so many good ideas and go deep into the wiring and component setup! Even making proffesional looking enclosures. It's great to watch! Thank you!;) Keep up the good work!!
I really love this channel. Everything is very clear, with the right amount of theory, nicely prepared and presented. And it is classy! Just a remark: I am by no means a mosfet expert, but I am a bit dubious about the way you parallel your mosfets. No problem in switching application; it is true that the Rds has positive tempco - but only after the Vgs is higher than the tempco crossover point. Below this point the Rds has negative tempco, so the warmest of those mosfets will pass most current, possibly resulting in thermal runaway. It is phenomenon similar to hotspotting. If I were designing this kind of product, I would probably decide for an individual sensing/driving circuit for each mosfet unless some fit linear mosfets like the IXTK90N25 were available.
Thanks Medvídek for your kind comments. I will have a further look in to the issue with using power mosfets in parallel. Using a Linear Power Mosfet like the IXTK90N25 would be a good choice but I would have to alter the gate drive circuit. Also they are very expensive at around £19 (23US$) each. The IXTK90N25 is not easy to source in the UK. I have however purchased a IXTH110N10L2 and been test that one. Regards, Louis
You are very welcome. It is true that IXYS devices are a bit on the expensive side, but perhaps we both remeber times when >$10 was quite common price tag for almost any power mosfet. Maybe this app note could came somewhat handy: www.microsemi.com/document-portal/doc_view/14700-new-500v-linear-mosfets-for-a-120kw-active-load Best regards!
woaaah.. now that's what I was talking about in the last video.. thank you soo very much sir for the video. and thanks for adding the keypad input.. great video as always.. regards Parth
Nice job on the project and explanation of the design. One suggestion - when you are using the keypad entry, perhaps use the encoder for editing any mistakes during keypad entries.
beautifull project which deserves a BIG thumbs up! I made a very similar project some months ago, I think I will upgrade it / redoing from scratch with your great help! since I'm planning to use Eagle CAD to develop the schematic and the board layout, if someone is interested I can share the project files when finished, I think I'll end up with just some small modifications on your design to better match my needs, but if there's someone interested in modify the project it could help not to redraw all. Greetings from Italy!
Thanks Miky for your comments and offer of sharing your Cad file when completed. As I am planning on some additional features I will be having another look at my PCB design once I have finished all the upgrades. Regards, Louis
The "transient" feature you mentioned in the videos will be a great addition! In my project, to keep the PCB as small as possibile, I won't use an entire Arduino but just the standalone microcontroller (I have already made lot of other projects based on the ATMega 328), I'm planning to add some sort of logging / plotting feature on the pc, by using the hardware serial port of the microcontroller and a TTL/USB converter (like the normal FTDI232) + galvanic isolation (see for example the IL612 IC by NVE Corp) which is usually not considered so much. Do you know this project? www.instructables.com/id/Arduino-Programmable-Constant-Current-Power-Resist/ I followed this when I made mine (your videos was not released yet :p) and I like the way he uses many cheap shunt resistors instead of a single more expensive low-error one.
very nice piece of work indeed. only one thing I just want to add to your work is that the ground connection of that differential amplifier should be connected close to the ADC. in that way you can achieved better precision. however I am not sure if you could connect 4 transistors in parallel. for equilibrium power sharing you need to have 4 sense resistors each for one transistor. when a MOSFET works at linear area the Rds is not the parameter that should be accounted. so the negative temperature coefficient may not help. now everything works great so the only thing that you can invest your time in is enhancing its precision. we look forward to see the finished product comparable to products available in the market. thank you very much for all of your good work.
Out of these 4 MOSFET, one of the MOSFET will always carry more current than other 3 due to manufacturing tolerance. So its temperature will rise more than other three and hence its gate threshold will drop due to Negative tempco. This consequence will pump more current on this MOSFET and dissipate more heat. So there will never be balance and one will start moving ahead of all other 3. In fact there will be very poor load balancing. Yes, there will be somewhat current flowing through rest of the other 3 due to thermal conductivity, but literally its not proper way to do it. To share load among many transistors, one has to create separate negative feedback with individual sense resistor.
Thanks Satyajit for your comments. Good point about the ground connections to the ADC, but in practice I have found the current arrangement to work quite well. May be a better designed PCB would help. With regards paralleling up 4 Power Mosfets the datasheet shows that the Ron factor as a function of junction temperature has a positive temperature coefficient. So as the junction of a Power Mosfet heats up its ON resistance increases, which in turn will then reduce the current. As I said this provide some auto balancing, but it is only about 80% effective (which is good enough for our project). As you say, you could improve the balancing by adding an additional low value resistor in each of the source ground returns, but this would cause other issues with regards the way we sense the current in the feedback loop for the control Op Amp. Regards, Louis
The ON resistance not much accountable when you run the MOSFET into its linear region. May be its working in your case due to the MOSFETs are from same manufacturing batch and their actual parameters are quite close to each other. Just find one from different batch and see how it performs. When you share the loads among many transistors, then you need to have control loop for each MOSFET separately. Also reading back is quite easy. Just use a summing amp
If you are showing theory you should provide whole formula. Vout = V2 ( (R4 / (R2 + R4) ) * ( (R1 + R3) / R1) ) - V1 (R3/R1) I don't think you can assume V1 magically becomes zero and simplify your formula like that. Differential amplifier operates on difference of its inputs. Your formula is completely ignoring one of the inputs in circuit that is used to combat the fact that none of the inputs are truly zero. Especially in your application. And why the long version when you can use: Vout = (V2 - V1) * (R3 / R1) I like your projects and your videos. Keep them coming. Thumbs up.
Thanks for the comment and highlighting the full formula for Vout. As I mentioned in my video it was not my intention to go in to great detail in this project video with regards how the formula is calculated. I wanted to illustrate for the hobbyist the practical use of a Differential Amplifier in the context of this project where V1 connection goes to the bottom terminal (negative terminal) of the load which in this application is effectively at 0 volt with respect to V2. Therefore, using the formula as I illustrated works fine and is an easy way to quickly calculate the effect of the resistor values I used. If V1 was not 0 volt you would have to use the full formula as you correctly have highlighted. Without going in to to much detail, in summary the full formula is derived from the following: If V2 = 0 then Vout = -V1(R3/R1) If V1 = 0 then Vout = V2 ( (R4 / (R2 + R4) ) * ( (R1 + R3) / R1) ) If neither V1 or V2 is 0 volt then you add the two equations together to give: Vout = V2 ( (R4 / (R2 + R4) ) * ( (R1 + R3) / R1) ) - V1 (R3/R1) To cover the full subject of Differential Amplifiers and Instrumentation Amplifiers, I think that requires a tutorial video of its own. Thanks again for highlighting the point and am sure this will inspire others to read more on the subject which can only be of benefit. Thanks for the thumbs up. Regards, Louis
Don't get me wrong, your math is not wrong. I just think by ignoring that value you are oversimplifying explanation what we are trying to overcome by using differential measurements. By saying that V2 = 0 beginners might think that zero will be the same at both ends of the negative wire and only V1 matters. But voltage drop is divided between both leads and negative point will drift significantly with changes of the current. In my opinion it is important to mention this as often as possible. In more precise work it tends to be very critical where "zero" point is connected. Great project, I have to build one. Have you thought about ordering and selling PCBs?
Nice work. I would probably connect the sense connectors to the load connectors via -lets say- 1k Resistors permanently. So the switching between sensing and no sensing is not necessary anymore. But nevertheless a very nice project. I'll probably build this for me too.
You wouldn't want to do that because it makes the sense useless. The idea is that the 4 wire sense removes any voltage drop in the load wires form the calculation.
Torsten B it's a great idea and I will surely implement this technique onto my next electronic load project. the resistance of the leads can never exceed 1K so that when it's open it's always connected to local terminal. your idea is really awesome
Thanks Torsten, Its worth a try as that is what they do with sensing inputs on power supplies, they normally use 100ohm resistors or even less. It should work as the input load will usually be low impedance. It is also interesting to note, for some reason, that where power supplies use this technique they also provide shorting links which can be connected between the terminals, in effect doing the same as what my switch is doing. Regards, Louis
I wish I could give you 10 thumbs up :) You are too fast for me. I am still waiting for parts from China to build the last version. Got everything except the LCD. I love the look of that box. Must have one. Where can we buy, and how much for the box? One thing I've found very annoying on these Encoders, is they are hard to get right most of the time. Too sensitive. I will count 3 or 4 rotation clicks as one, in software. Love all your projects. None of the useless blinking LED stuff, you find on other channels. Thanks again.
I found these boxes on Banggood.com Very happy with them and used them on multiple projects already. I would suggest to replace those plastic feet with stick-on rubber ones.
Thanks. For the project case try searching on eBay for: Sourcingmap® 245mm X 190mm X 110mm Metal Enclosure Project Case You should find a number of suppliers who sell it. Regards, Louis
I've never bought anything from eBay. I think I'll 3D print mine from this design, and shield it where needed. www.thingiverse.com/thing:1264391 Don't want to pay $30+ for a box, where I can use that money to build something else.
in the video explain how to increase the load capacity by inserting 3 Mosfet. But in the scheme you exposed exactly at time 16.52, such a configuration is not functional. Because I am also making a dc load and I tried that pattern and from the measurements made the first mosfet has more current than the second and etc.. This phenomenon, in prolonged circumstances, breaks the MOSFET!. It would be better, as I did, to repeat the operational control stage and then make a parallel so as to divide their power x4, that is, divide the load x4. In English since I am Italian I am not a top I hope I explained the fact. Nice video and thank you.
Why is R11 changed to 10K? It originally formed part of the divide-by-50 voltage divider when it was 2K, but with the sense circuit in place, the voltage out of that is already divided by 50. The non-inverting input to the voltage follower (U7C) is connected to the sense circuit and I can't work out what R11 is actually doing? It seems to be part of forming a voltage divider between the low impedance output of the sense circuit and R11 connected to the non-inverting input.
Hi Louis! I successfully completed version #4, great project, and thanks!!! How do I get "I set" and the actual current to agree ? The actual current agrees with my programmable power supply.
Hi Phil, Glad to hear it all went well. If the "Actual Current" reading is slightly different to the "I set" reading. This can be affected by component tolerances. I am looking at possible way of improving the calibration of the unit. But in the mean time you could play around with adjusting the multiplier factor in the part of the code that calculates the actual current for the LCD. Look for the following lines in the code: //---------Calculate Actual Voltage and Current and display on LCD------------ void ActualReading(void) { ActualCurrent = ((((current - currentOffset)*2.048)/32767) * 2.5); You could try changing the 2.5 multiplier value at the end of the line either up or down very slightly for example: If your "actual current" reading is higher than "I set" then try 2.495 If your "actual current" reading is lower than "I set" then try 2.505 It will be a bit of trialing and testing after you change the code by adjusting the factor up or down very slightly. This is not a precise way as the error factor may be non-linear but it may be a quick way to improve things a bit. I intend to try and have a look at ways of calibrating. Regards, Louis
Hi Louis. I'm not understanding where the sense leads are connected. At 30:28 you say that both sets of leads are connected to the power supply. There are only small gauge wires connecting the sense inputs to the load inputs so I'm guessing you won't be passing much current through these wires? I only have a cheap bench power supply with a single output.
The sense leads do not carry any current so only need small gauge wire. Usually a power supply with only a single output has screw terminals also possible with a integral banana socket. So hopefully it should be possible to connect both sets of wires. Just make sure the LOAD wires are heavy gauge and securely connected.
A comment on the video itself: you might find using a lapel microphone rather than the one in the camera will improve the sound quality, which currently varies quite a lot depending on whether you're facing the camera or turned away to look at the whiteboard
Thanks Paul for the comment. I have tried a few different microphone's but will have a look at a lapel microphone hopefully I can find one for my video camera.
Hi Tony, Although the project has some similarities with a programmable power supply it does not supply any power out. It simply puts a constant current load across the device under test, be that a power supply your are testing to confirm its specification or testing of a batterys capacity. You could use some elements of this design in a programmable power supply and that may be an idea for another project. Regards, Louis
Louis, Are you still using D3, UF5408 ? I didn't see it in the video. That diode starts conducting at about .45 Volts. (although not much at this level) Does that mean after about 5 AMPS, the current sensing will be off?
Hi Phil, I did leave it off last time in part 4 but as the project was intended to handle up to 3 Amp it should be OK to add one. It just gives you another layer of protection if a Mosfet went short circuit. Regards, Louis
Louis, I see that the OPA277 comes in 2 different versions. OPA277PA, which is $3.03 (Digikey US) and the OPA277P which is $6.01. The only difference between the two is the "input voltage offset". The OPA277PA (used in the project) has a input voltage offset of 20 UV and the more expensive one is 10UV. Could you comment on "input voltage offset" ? Great work and thanks again for your time!
Wow, good job. How about supplying the electronic. load with a battery and a charging circuit. I suppose that the internal current of the circuit is minimal
the current sharing wont work as you thought. The on resistance is the resistance in the active region. The load will work in the ohmic region of the mosfet and the thermal behaviour there is the opposite. You need separate drives for each mosfet to ensure it is operating in the ohmic region.
Thanks Magnus for your comment. This DC load works both in the ohmic and linear region. For lower drain voltages of say lower than a few volts it is working in the ohmic region but for higher drain voltages it is in the linear region. If you look at the datasheet you will usually find a graph showing "Normalized drain-source on-state resistance factor as a function of junction temperature". This clearly shows on resistance increasing with temperature (positive temperature coefficient). I was using this effect in my application but I did mention it was not fully effective and the efficiency would be about 20% less. For this particular project it seems to work well enough at the power levels I am considering. As you say you could add separate drive circuits for each mosfets and that could be considered for future upgrades if required, but I am also trying to keep the cost down for the hobbyist. Thanks again for raising the point. Regards, Louis
Thanks for your reply and I am sorry for the late response. Looking at fig 8 in the spec for the mosfet you can see that for currents below 250A the current actually increases with temperature for a given Vgs. Since both mosfets will have the same Vgs there is a big risk for on of the mosfets heating up more than the other and therefore the current ratio between them will shift resulting in even more heating and current.
Thanks Magnus. The increase in current due to temperature for a given VGS is something I plan to cover in part 5 of this project. I have made some changes to reduce this effect in order to prevent thermal runaway. Regards, Louis
Hi, Louis Excellent project. Perhaps I will repeat it as soon as possible. It stops only the presence of rare we DAC / ADC ICs. But I'll find a shortage of components on Ebay. I am from Russia, but I think these chips are no sanctions )))) What are your plans for the development of this project? Can I ask you to introduce the function for automatic testing of power supplies? It works on the algorithm. 1. The user enters the value of the rated voltage. 2. The user enters the value of the voltage drop (in volts or%). 3. The user enters the current rise time. 4. The user sets the maximum power in watts (optional). 4. Turn the test. 6. Apparatus load increases. 5. The data obtained after the test. Example U nominal = 12 V P nominal = 20W (Value at which started the voltage drops below a user-specified.) U max = 10,5 V ( This value = U nominal - (U nominal - U value of the voltage drop in volts or%) P max = 25W (The maximum capacity at which the voltage has dropped to the level set by the user, or the protection tripped (shut down the source of all) Test time ))) Thank you and I wish you every success. Best ragards Roman
Hi Roman, Thanks for your comments. Auto testing may be an interesting option. At the moment I am concentrating on improving the current arrangement and then looking at adding a Transient Mode with the option of a trigger input. I also want to look at adding additional menu screens to add more functionality. Other things could include Excess Temperature Protection, Reverse Load Voltage Protection, Data Logging, Calibration settings in EEPROM. Regards, Louis
As a protection against reverse polarity, I recommend to pay attention to the scheme. img.mysku-st.ru/uploads/images/02/80/11/2015/06/03/1469e1.jpg Best regards Roman
Thanks Roman. Since we now have the Sense Terminals I am now looking at adding a Power Schottky Diode in series with the load. The one I am testing is the MBR41H100CT which is a dual diode in a TO-220 package. I will wire the diodes in parallel. Although we get a voltage drop the fact that we now have the Sense Terminals it does not matter. Regards, Louis
Hi, Louis I try to model your design in Proteus, but unfortunately not even have a picture on the display. Although I connected all the libraries and compile the file successfully. I suspect that the problem in Proteus. Can I ask you to put HEX file to test this. Many thanks Best regards Roman
I am currently test some alternative as I mention but did not have time to cover them in the video. One alternative is: STP80NF55L-06 which is available from either: www.newark.com Newark Part No.: 26M3714 or www2.mouser.com Mouser Part No: 511-STP80NF55L-06 It is manufactured by STMicroelectronics and it is a Logic Level Mosfet with a Vds of 55 volts drain current of 80 Amp. Also its safe operating area is similar to the BUK956R1-100E. Give it a try it should be OK. Regards, Louis
Added them to my mouser order. Thank you. I think this whole project costs about $100 to build. Which is excellent. I could've saved more by using some stuff I already had, but I don't want to say put another 16bit ADC and then have to tweak your code to make it work. Just plug and pray...
I remember in the industry they use to use "current sharing resistors". I think they were something like .01 ohm in each of the MOSFET (source pin) out put devices. This is called local feedback because every time one MOSFET would draw a little more current the IR drop across the "current sharing resistor" would go up thus decreasing the VGS of the associated MOSFET. Here is an example image I found on the internet... www.diyaudio.com/forums/solid-state/281960-unbalance-sharing-output-mosfets.html ....Thumbs up on your video....PEACE
Thanks for the comment. I have been looking at making the wiring from the source to the current sense resistor all the same length in that way the wire resistance would act as a low equal value resistor doing what you suggest. I will give it a try. Regards, Louis
Could you show in your next video all the external connections to your electronic load. In particular I would like to see where you are connecting the current sense input line. Thanks for all your videos....PEACE
If you download the zip file I in my comments below the video and unzip it you will find a full parts list which includes details on the case. Below is also a direct link to the eBay seller I purchased it from: Sourcingmap® 245mm X 190mm X 110mm Metal Enclosure Project Case www.ebay.co.uk/itm/112303472593?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT
Hi Rob, The heatsink is 250x90x30 Large heatsink from the drop-down list on that eBay link in the parts list. This heatsink is just a little to long so I cut it down to fit the back of the project case. Regards, Louis
Hello Could you help me ? I want to do an electronic load on your new circuit But with such parameters 150 volts 500W 30 Amps I'm interested in software Because I'm not strong at this
I may have a look an an option using either a IXTN36N50 or a IXTN60N50L2 Linear Mosfets which will give you the power handling you require. We would need to alter the OP Amp drive circuit as they both need a higher gate drive voltage. But my software should still work OK. The two Linear Mosfets I mention are quite expensive but you will only need one. One of my viewers has already taken my design and software and made a DC Load with an IXTN60N50L2. He has posted his circuit changes and diagram on his web site, the direct link is below: www.lavrsen.dk/foswiki/bin/view/Kenneth/ElectronicLoad If you have a look at what he did it may help. Regards, Louis
Simply that I have acquired them over many years. I still like to use an analog oscilloscope on occasions. I now mainly use digital oscilloscopes for most of my project work.
why dont you make a few and sell it , i am a maker , i am struggle to find one i can afford , unfortunately have no time either to make one, too much prototyping work at this moment..
Sorry due to my time restraints I am unable to make any to sell at the moment. My You Tube channels is simply to help the hobbyist with there own projects.
Louis - Great Project really enjoying it and am in the process of collecting the parts to build one. Your Schematic V4 states that R6 and R16 are both precision 0.1% - Did you mean R6 and R14 the 9K and 1K as the voltage divider on the +ve input to the op amp? Question - Is there any reason why you couldn't use and 18k and 2K for the divider network? (I already have some 0.1% spare from another project) I also have an extra large 20x4 I2C LCD Display that will work well with it (130 x55mm) Sometimes you just have accept your eyes aren't as young as they once were!
Thanks Richard. Thanks for spotting my typo error it should have read R6 & R14 that is the 9K and 1K need to be ±0.1% tolerance. If you have 18k and 2K at ±0.1% they will do also. The resistors I used also had a Temp. Coefficient of ±15ppm/°C ensuring stability. All the resistors in the Remote Voltage Sense circuit need to be ±0.1% tolerance and if possible ±15ppm/°C. A extra large LCD would be good :) Regards, Louis
![ทุกครั้งที่หลับตา (Lucid Dream) - AYLA's [ Official MV ]](http://i.ytimg.com/vi/4i1OBAQyv58/mqdefault.jpg)
Excellent video
Excellent Project
Excellent Channel
Great project.
Thanks Cesar.
my-o-my this is just like the Heath-Kit project days 😃such a wonderful video for the folks just starting out in program electronics as well as the old timers 🥰 thanks a lot 🤩🍎🎂☕
Can't say anything but thanks, your projects are not a kit building process but a lesson itself.
Thank you.
I will add details for the alternative Mosfets and a wiring diagram for the case as soon as I have had time to complete them.
Hi Great project ! I been following the whole project . One question . the MPC 3426 is it possible to have a software version with ADS 1115 ? MPC 3426 is kinda hard to find it , and it's too bad cause it's such a great project .
@@lucianungurean Hi I am doing a version with ads 1115
@@PauloSilva-ll4vs That's beautiful ! I did my own version of a DC load .... and I used the ADS as well . The thing is ....I don't have the kind of menu the project has .... Mine can deliver 200w .... and 15 amps in current mode .... I used a DC load kit and adapted it to my design .... I adjusted the DAC to give me a range fom 0 to 500mv .... its very stable considering I haven't used used a Vref . I also copied the fail safe . ....Shutdown at over 200watt ; over voltage 28 volt and over current 15.5 amps
@@lucianungurean HI great to hear you solved your problem, the unique change i did wss the ads, this project is the state of art.
@@PauloSilva-ll4vs Indeed is . I'm very happy to hear you got the ADS . These days a lot of people use ADS 115 in their projects .
Very professional unit, excellent to see the real pride you take! - thank-you for sharing this freely with everyone.
Thanks Michael.
Brilliant, thanks again, been waiting for part 4, my own has taken shape, used those keypads, but then they wouldn't fit on the front panel of the cases I had, so made my own from some salvaged switches.... Thanks to you for explaining it all simply and inspiring us all to go and build one! Learnt tons, looking forward to more, thanks Louis....
Thanks Tony for your comments.
Regards,
Louis
Good stuff again Louis!
When I build my own electronic load, and went to parallel Mosfets to share the load, I quickly noticed a thermal imbalance between the Mosfets. One would get very hot, while the other stayed cool. In my opinion, there are two things you can do to prevent or counter that. One is to create a feedback/driver loop for each Mosfet. This will ensure a more balanced load. An other method, the one I adopted, is to measure the temp on each Mosfet and reduce the gate drive of that Mosfet if there is a thermal imbalance. I used a transistor bolted to the Mosfet fastener as the temp pick-up and some analog circuitry that keeps the relative Mosfet temps within a degree C or so from each other.
Adding a transient capability is very usefull, especially for testing power supplies, etc. In that case, the pulse characteristics will dictate the configuration of the Mosfets and drivers, shunt and the wiring. Everything will most likely have to be physically closer to what you have now to get a reasonable rise/fall time for some serious testing purposes. Looking forward to see your solution!
Thanks Paul for your comments and ideas. I am continuing to test the Mosfets in parallel to find the best solution. So far it seems to work quite well, may be I am just lucky with the 4 mosfets. I am going to monitor the current flow through each one at the same time. Also looking at possibly altering the wiring arrangement as I discussed on the video.
Regards,
Louis
I like your teaching style. Its very simple to understand due to the they way you go step by step, even your videos are arranged by starting simple and getting it working then adding a few features/upgrades. I assume you have some form of teaching background ? If not you really missed your calling. This channel has become one of my favorite electronic channels. Great work.
Thanks William for your comments. I did teach electronics back in the late 1970's and early 1980's for about 8 years. If you are interested in my background you can find some information on my website at:
www.scullcom.uk/about/
Regards,
Louis
Just had a look at your CV on your website. What a background! It is now no surprise to me that you are as knowledgeable as you are and so good at putting information over. I am just grateful that you are so willing to share your knowledge. Thank you.
I know I'm little late, but just wanted to say that it is really worth the time to read about Louis on his homepage, what an amazing person, lots of people are teachers but a good teacher is rare, one for that feedback of his students is important, one that likes to be questioned, one that grows from criticism, this is the only way how a professional can learn to explain complex things to hobbyists and newbies, with such teachers only there would be only a few "bad" students, but most people would be scientists and we would travel through the universe already, I know that from myself and am pretty sure about that, always sucked in maths, but was always the best in physics which was mostly math, the difference was that the math teacher just waited for the end of the day while the physics teacher loved what he did, loved physics, was always excited about what he teached us and infected us with his excitement, I never had a lower grade than a 1- in physics, while in math I even had once a 6 end of the year, but nobody cared, they just say it is the kids fault, as a young person you can say whatever you want, they don't listen anyways, they even told my parents to visit a psychologist instead of testing the competence of the teachers they hire, years later all know I have an IQ of 138 but the lowest school grade that count in Germany, am I an idiot or does the school system suck? We all know about the answer, atleast about schools ... 😄
Nicely done Louis. At least one of our local hams is working on your GPS reference project and has designed a PCB for it that can be sent out to one of the PCB houses (he plans to send you the file) and he's then going to start working on this project. If I didn't already have a electronic load on the bench I'd be building one of these!
Thanks Dino. look forward to seeing the PCB when its finished.
Regards,
Louis
Wish I could give a 2 thumbs up, this project shows a lot of effort and success!
Thanks for the thumbs up :)
Hi Louis,
absolutely excellent work !
Please let me put some future suggestions - ideas.
1. A DC Load in most cases is required for high current flow applications. So, an expansion to 30A even to 50...60 Amps could be a serious update.
2. Linear MOSFETs (like those from IXYS) have better performance and thermal response for this case of application.
3. Paralleling MOSFETs with its own driver - OpAmp and its own shunt resistor are very much better topology. This is a conclusion from tests and measurements that i have made (better MOSFETs matching, thermal activity, linearity, etc) and the results from market devices (if you can open and review one of these you will see this connection topology).
4. Op Amps to drive the Mosfets like those for audio applications are excellent choice because of high dynamic range, high slew rate, low noise and high current output to driving the gate of mosfet. A good example is LM833.
5. The differential amp needs some kind of protection at input stage to avoid malfunctions caused from the spikes of the under test power supply.
6. Why not and a Constant Voltage mode like the market devices ? You have already all the required stages. Is just programming code ...
Keep the good work! Thanks for your attention.
With kind regards
Simon
Hi Simon,
Thanks for all your suggestions.
I have been playing with some IXYS Linear Mosfets (tested the IXTH110N10L2) but they did not work in my circuit as they require different drive condition. They are expensive to. Something to look at for the future.
I will be looking further in to using power mosfets in parallel and may have a look at individual driver circuits or other solutions.
Agree about protection for the OP Amp. I may add some diodes on the input to do that job.
I have thought about a constant voltage mode and it could be easy to add in software - just need to ensure it can sink the required currents.
Regards,
Louis
Great job, a lot of people will find this project very interesting. Especially the use of the nano and the code libraries.
Thanks Rich.
I have been building Electronic Loads myself for a year or so, and it's very nice to watch some of the same evolutionary steps I've taken. But I would like to warn you about a mistake that I also made before.
Regarding Mosfets in parallel: RdsON really has a positive temperature coefficient BUT the gate threshold voltage has a negative temperature coefficient, wich in such an application (linear mode) will probably be the most dominant effect of temperature drift. It's good to keep it in mind.
Thanks Sergio for your comments. I am currently taking a closer look at the Mosfets in parallel and testing some options.
Regards,
Louis
Brilliant work, Louis! Very nice upgrades. I'm always happy to see another video appear on your channel. You do a very nice job explaining everything; you should be *much* more prominent in the electronics maker community.
I've been eyeing that enclosure myself, for an isolation transformer and power meter project. I'm happy to see that it appears to be of good quality. Cheers!
Thanks for your kind comments, much appreciated.
Regards,
Louis
I remember years ago Dave Jones did a video or two on DIY loads and the community response to it spawned interpretations of all sorts of variations and capability. It seems history is repeating here and the project has certainly evolved from much humbler beginnings to what now looks a very capable load. Since the design has come so far I would add a second call for a unitised PCB in the future. For the record I no longer subscribe to Jones' channel which IMHO has devolved from it's beginnings whereas your approach better suits my requirements. Congratulations and gratitude for the effort and excellent explanations.
One detail I noticed when you started the demo was you nulled the voltage reference before powering the source which introduced another offset when the load's input was energised but with the load off. I suspect the small errors reported in use would be further reduced if the nulling was done just prior to engaging the load.
Thanks for your comments and for highlighting the nulling procedure.
Regards,
Louis
Great Video as usual here. Thanks for the improvements, especially with the keypad input. Go on with the great work Louis. I am looking forward for your next episode.
Greetings from Germany
Michael
Thanks Michael for your comments.
freundliche Grüße
Louis
AWESOME AWESOME AWESOME! Excellent project and video. Thank you Sir!
Thanks.
I have watch several of your video and they are very thorough and educative, and i have learn a great deal from you. just want to say thank you and keep bringing projects like this
Thanks Melvin for your comments. Much appreciated.
regards,
Louis
I so do appreciate all your hard work and creativity in this project. I look forward to seeing the software update for the voltage correction and all the other updates. In a few weeks I am going to try to replicate your build. Thanks again and very awesome work.
Thanks Gene. Hope your own build of this project goes well.
Regards,
Louis
Like any other of my projects,I want more out of it... I'm going to try incorporating a automatic pwm fan control and try it with a arduino capacitive touch panel. I will document it as i go and let you know how it goes.
Look forward to hearing your result.
Regards,
Louis
WIRING DIAGRAM LAYOUT CAN BE DOWNLOAD FROM LINK BELOW:
www.scullcom.com/DC_Load_wiring_layout.pdf
Thanks for the great quality content!
Your channel has quickly become one of my favourites on TH-cam, when it comes to electronics.
I have a question:
Is it really ok to parallel mosfets in linear applications, without adding source resistors?
If I remember correctly Paul Horowitz and Winfield Hill in "The Art of Electronics" third edition (page 192) argue that it's OK to omit source resistors only in switching applications, and not in linear ones.
Very nice. Looks like I have to get back to work on the PCB. Thanks much. Really enjoying this project.
Thanks Rob. The PCB may change again if I continuing adding more functions :)
Regards,
Louis
Maybe you can program the fan to be pwm controlled in relation to the heatsink? Then it will be less annoying:) And some vibration isolators if you've got them!:) (i hate loud fans)
The project looks amazing! This is way better than looking at EEVBLOG! You have so many good ideas and go deep into the wiring and component setup! Even making proffesional looking enclosures. It's great to watch! Thank you!;) Keep up the good work!!
Thanks Simon for your comments.
Regards,
Louis
Well done. This turned out really well, and great looking frontpanel also. I'm planning to build one myself too.
Thanks.
really great! I am learning a lot from this series!
Thanks David, glad to hear its helpful.
Regards,
Louis
excellent comprehensive well explained project, taught me a lot, thanks again
Nice to hear you are benefiting from my videos.
I really love this channel. Everything is very clear, with the right amount of theory, nicely prepared and presented. And it is classy!
Just a remark: I am by no means a mosfet expert, but I am a bit dubious about the way you parallel your mosfets. No problem in switching application; it is true that the Rds has positive tempco - but only after the Vgs is higher than the tempco crossover point. Below this point the Rds has negative tempco, so the warmest of those mosfets will pass most current, possibly resulting in thermal runaway. It is phenomenon similar to hotspotting.
If I were designing this kind of product, I would probably decide for an individual sensing/driving circuit for each mosfet unless some fit linear mosfets like the IXTK90N25 were available.
Thanks Medvídek for your kind comments. I will have a further look in to the issue with using power mosfets in parallel. Using a Linear Power Mosfet like the IXTK90N25 would be a good choice but I would have to alter the gate drive circuit. Also they are very expensive at around £19 (23US$) each. The IXTK90N25 is not easy to source in the UK. I have however purchased a IXTH110N10L2 and been test that one.
Regards,
Louis
You are very welcome. It is true that IXYS devices are a bit on the expensive side, but perhaps we both remeber times when >$10 was quite common price tag for almost any power mosfet.
Maybe this app note could came somewhat handy:
www.microsemi.com/document-portal/doc_view/14700-new-500v-linear-mosfets-for-a-120kw-active-load
Best regards!
Thanks for the link.
Regards,
Louis
You Rock dude! great project, and you are doing a great job!
Thanks.
woaaah.. now that's what I was talking about in the last video.. thank you soo very much sir for the video. and thanks for adding the keypad input..
great video as always..
regards
Parth
Thanks Parth, glad you enjoyed it :)
Regards,
Louis
Your channel is the best..
Thank you.
Nice job on the project and explanation of the design.
One suggestion - when you are using the keypad entry, perhaps use the encoder for editing any mistakes during keypad entries.
Thanks. Not sure about using the encoder to correct keypad entries. I may go to another menu screen as an option.
Regards,
Louis
It becomes better and better.
Thanks.
extremely well made , both the device and the video , congrats.
Thanks.
beautifull project which deserves a BIG thumbs up! I made a very similar project some months ago, I think I will upgrade it / redoing from scratch with your great help! since I'm planning to use Eagle CAD to develop the schematic and the board layout, if someone is interested I can share the project files when finished, I think I'll end up with just some small modifications on your design to better match my needs, but if there's someone interested in modify the project it could help not to redraw all. Greetings from Italy!
Thanks Miky for your comments and offer of sharing your Cad file when completed. As I am planning on some additional features I will be having another look at my PCB design once I have finished all the upgrades.
Regards,
Louis
The "transient" feature you mentioned in the videos will be a great addition!
In my project, to keep the PCB as small as possibile, I won't use an entire Arduino but just the standalone microcontroller (I have already made lot of other projects based on the ATMega 328), I'm planning to add some sort of logging / plotting feature on the pc, by using the hardware serial port of the microcontroller and a TTL/USB converter (like the normal FTDI232) + galvanic isolation (see for example the IL612 IC by NVE Corp) which is usually not considered so much. Do you know this project? www.instructables.com/id/Arduino-Programmable-Constant-Current-Power-Resist/ I followed this when I made mine (your videos was not released yet :p) and I like the way he uses many cheap shunt resistors instead of a single more expensive low-error one.
Excellent video Louis, thanks!
Thanks Javier.
Excellent
very nice piece of work indeed. only one thing I just want to add to your work is that the ground connection of that differential amplifier should be connected close to the ADC. in that way you can achieved better precision. however I am not sure if you could connect 4 transistors in parallel. for equilibrium power sharing you need to have 4 sense resistors each for one transistor. when a MOSFET works at linear area the Rds is not the parameter that should be accounted. so the negative temperature coefficient may not help. now everything works great so the only thing that you can invest your time in is enhancing its precision. we look forward to see the finished product comparable to products available in the market. thank you very much for all of your good work.
That is true for LEDs. Each LED would need its own dropout resistor. LEDs do not posses the inherent PTC characteristics of MOSFETS.
Out of these 4 MOSFET, one of the MOSFET will always carry more current than other 3 due to manufacturing tolerance. So its temperature will rise more than other three and hence its gate threshold will drop due to Negative tempco. This consequence will pump more current on this MOSFET and dissipate more heat. So there will never be balance and one will start moving ahead of all other 3. In fact there will be very poor load balancing. Yes, there will be somewhat current flowing through rest of the other 3 due to thermal conductivity, but literally its not proper way to do it. To share load among many transistors, one has to create separate negative feedback with individual sense resistor.
Thanks Satyajit for your comments. Good point about the ground connections to the ADC, but in practice I have found the current arrangement to work quite well. May be a better designed PCB would help.
With regards paralleling up 4 Power Mosfets the datasheet shows that the Ron factor as a function of junction temperature has a positive temperature coefficient. So as the junction of a Power Mosfet heats up its ON resistance increases, which in turn will then reduce the current. As I said this provide some auto balancing, but it is only about 80% effective (which is good enough for our project). As you say, you could improve the balancing by adding an additional low value resistor in each of the source ground returns, but this would cause other issues with regards the way we sense the current in the feedback loop for the control Op Amp.
Regards,
Louis
The ON resistance not much accountable when you run the MOSFET into its linear region. May be its working in your case due to the MOSFETs are from same manufacturing batch and their actual parameters are quite close to each other. Just find one from different batch and see how it performs. When you share the loads among many transistors, then you need to have control loop for each MOSFET separately. Also reading back is quite easy. Just use a summing amp
Look at Fig5 on page 4 here
www.ixys.com/Documents/Articles/Article_Linear_Power_MOSFETs.pdf
Is that what you are looking for?
well done its one awesome project and the final product is awesome
Thanks.
If you are showing theory you should provide whole formula.
Vout = V2 ( (R4 / (R2 + R4) ) * ( (R1 + R3) / R1) ) - V1 (R3/R1)
I don't think you can assume V1 magically becomes zero and simplify your formula like that. Differential amplifier operates on difference of its inputs. Your formula is completely ignoring one of the inputs in circuit that is used to combat the fact that none of the inputs are truly zero. Especially in your application. And why the long version when you can use:
Vout = (V2 - V1) * (R3 / R1)
I like your projects and your videos. Keep them coming. Thumbs up.
Thanks for the comment and highlighting the full formula for Vout.
As I mentioned in my video it was not my intention to go in to great detail in this project video with regards how the formula is calculated. I wanted to illustrate for the hobbyist the practical use of a Differential Amplifier in the context of this project where V1 connection goes to the bottom terminal (negative terminal) of the load which in this application is effectively at 0 volt with respect to V2. Therefore, using the formula as I illustrated works fine and is an easy way to quickly calculate the effect of the resistor values I used.
If V1 was not 0 volt you would have to use the full formula as you correctly have highlighted.
Without going in to to much detail, in summary the full formula is derived from the following:
If V2 = 0 then Vout = -V1(R3/R1)
If V1 = 0 then Vout = V2 ( (R4 / (R2 + R4) ) * ( (R1 + R3) / R1) )
If neither V1 or V2 is 0 volt then you add the two equations together to give:
Vout = V2 ( (R4 / (R2 + R4) ) * ( (R1 + R3) / R1) ) - V1 (R3/R1)
To cover the full subject of Differential Amplifiers and Instrumentation Amplifiers, I think that requires a tutorial video of its own.
Thanks again for highlighting the point and am sure this will inspire others to read more on the subject which can only be of benefit.
Thanks for the thumbs up.
Regards,
Louis
Don't get me wrong, your math is not wrong. I just think by ignoring that value you are oversimplifying explanation what we are trying to overcome by using differential measurements.
By saying that V2 = 0 beginners might think that zero will be the same at both ends of the negative wire and only V1 matters. But voltage drop is divided between both leads and negative point will drift significantly with changes of the current.
In my opinion it is important to mention this as often as possible. In more precise work it tends to be very critical where "zero" point is connected.
Great project, I have to build one. Have you thought about ordering and selling PCBs?
very nice, well done sir!
Thanks.
Good job. Looks awesome. Really great channel.
Thank you.
Thanks for this. Great project.
Thanks.
Excellent video. Thanks!
Thanks.
you are doing a great work👍
Thanks for the big thumbs up :)
Great master good job. Nice project great result...
Thanks.
Nice work. I would probably connect the sense connectors to the load connectors via -lets say- 1k Resistors permanently. So the switching between sensing and no sensing is not necessary anymore. But nevertheless a very nice project. I'll probably build this for me too.
You wouldn't want to do that because it makes the sense useless. The idea is that the 4 wire sense removes any voltage drop in the load wires form the calculation.
Torsten B it's a great idea and I will surely implement this technique onto my next electronic load project. the resistance of the leads can never exceed 1K so that when it's open it's always connected to local terminal. your idea is really awesome
Thanks Torsten, Its worth a try as that is what they do with sensing inputs on power supplies, they normally use 100ohm resistors or even less. It should work as the input load will usually be low impedance. It is also interesting to note, for some reason, that where power supplies use this technique they also provide shorting links which can be connected between the terminals, in effect doing the same as what my switch is doing.
Regards,
Louis
Credit goes to Torsten B and not me. Thanks
Thanks a lot. But I didn't invent this. Biggest credit goes to Louis who puts a lot of time and effort in his videos.
nice project 👍
Thanks.
You may use a heat sink with vertical fins because of free convection/thermodynamics etc..
very good project, thank you
Thanks.
Why don't you use the arduino eeprom to store the last used settings, so it turns on in the last state and settings used?
Thanks for your comment. I am considering storing some setting and may cover that in later upgrades to this project.
Excellent job!
Thanks.
I wish I could give you 10 thumbs up :)
You are too fast for me. I am still waiting for parts from China to build the last version. Got everything except the LCD.
I love the look of that box. Must have one. Where can we buy, and how much for the box?
One thing I've found very annoying on these Encoders, is they are hard to get right most of the time. Too sensitive. I will count 3 or 4 rotation clicks as one, in software.
Love all your projects. None of the useless blinking LED stuff, you find on other channels. Thanks again.
I found these boxes on Banggood.com Very happy with them and used them on multiple projects already. I would suggest to replace those plastic feet with stick-on rubber ones.
Do you have the link to share with us?
Thanks.
For the project case try searching on eBay for:
Sourcingmap® 245mm X 190mm X 110mm Metal Enclosure Project Case
You should find a number of suppliers who sell it.
Regards,
Louis
I've never bought anything from eBay.
I think I'll 3D print mine from this design, and shield it where needed.
www.thingiverse.com/thing:1264391
Don't want to pay $30+ for a box, where I can use that money to build something else.
in the video explain how to increase the load capacity by inserting 3 Mosfet. But in the scheme you exposed exactly at time 16.52, such a configuration is not functional. Because I am also making a dc load and I tried that pattern and from the measurements made the first mosfet has more current than the second and etc.. This phenomenon, in prolonged circumstances, breaks the MOSFET!. It would be better, as I did, to repeat the operational control stage and then make a parallel so as to divide their power x4, that is, divide the load x4. In English since I am Italian I am not a top I hope I explained the fact. Nice video and thank you.
Hi Louis!
I have the Sense PCB built. To "null" the input, what is the proper procedure? Short the input leads and adjust for zero ?
Hi Phil,
Yes short the sense input terminals and adjust the Null trimpot for a reading of zero.
Regards,
Louis
Why is R11 changed to 10K? It originally formed part of the divide-by-50 voltage divider when it was 2K, but with the sense circuit in place, the voltage out of that is already divided by 50. The non-inverting input to the voltage follower (U7C) is connected to the sense circuit and I can't work out what R11 is actually doing? It seems to be part of forming a voltage divider between the low impedance output of the sense circuit and R11 connected to the non-inverting input.
Hi Louis!
I successfully completed version #4, great project, and thanks!!!
How do I get "I set" and the actual current to agree ? The actual current agrees with my programmable power supply.
Hi Phil,
Glad to hear it all went well.
If the "Actual Current" reading is slightly different to the "I set" reading. This can be affected by component tolerances. I am looking at possible way of improving the calibration of the unit. But in the mean time you could play around with adjusting the multiplier factor in the part of the code that calculates the actual current for the LCD. Look for the following lines in the code:
//---------Calculate Actual Voltage and Current and display on LCD------------
void ActualReading(void) {
ActualCurrent = ((((current - currentOffset)*2.048)/32767) * 2.5);
You could try changing the 2.5 multiplier value at the end of the line either up or down very slightly
for example:
If your "actual current" reading is higher than "I set" then try 2.495
If your "actual current" reading is lower than "I set" then try 2.505
It will be a bit of trialing and testing after you change the code by adjusting the factor up or down very slightly.
This is not a precise way as the error factor may be non-linear but it may be a quick way to improve things a bit.
I intend to try and have a look at ways of calibrating.
Regards,
Louis
More excellence!
keep up the good work sir !!
Thanks.
Thanks! Are you going to make (in addition) schematic and PCB design to upgraded version? I mean one PCB with all "circuits"?
Thanks Pawel. When I have eventually stopped adding additional functions I may try and design a new PCB if I have time.
Regards,
Louis
Scullcom Hobby Electronics Ofcourse it is not your duty but it would be nice for people who haven't started making their load yet. Great work anyway!
Hi Louis. I'm not understanding where the sense leads are connected. At 30:28 you say that both sets of leads are connected to the power supply. There are only small gauge wires connecting the sense inputs to the load inputs so I'm guessing you won't be passing much current through these wires? I only have a cheap bench power supply with a single output.
The sense leads do not carry any current so only need small gauge wire. Usually a power supply with only a single output has screw terminals also possible with a integral banana socket. So hopefully it should be possible to connect both sets of wires. Just make sure the LOAD wires are heavy gauge and securely connected.
A comment on the video itself: you might find using a lapel microphone rather than the one in the camera will improve the sound quality, which currently varies quite a lot depending on whether you're facing the camera or turned away to look at the whiteboard
Thanks Paul for the comment. I have tried a few different microphone's but will have a look at a lapel microphone hopefully I can find one for my video camera.
I have a question, please excuse my ignorance. Your project seems kind of like a programmable power supply, is that what it is?
Hi Tony,
Although the project has some similarities with a programmable power supply it does not supply any power out. It simply puts a constant current load across the device under test, be that a power supply your are testing to confirm its specification or testing of a batterys capacity.
You could use some elements of this design in a programmable power supply and that may be an idea for another project.
Regards,
Louis
Louis,
Are you still using D3, UF5408 ? I didn't see it in the video.
That diode starts conducting at about .45 Volts. (although not much at this level) Does that mean after about 5 AMPS, the current sensing will be off?
Hi Phil,
I did leave it off last time in part 4 but as the project was intended to handle up to 3 Amp it should be OK to add one. It just gives you another layer of protection if a Mosfet went short circuit.
Regards,
Louis
How should we handle this when we increase the power handling capabilities ?
Hi Phil,
You could wire 2 of the UF5408 diodes in series.
Regards,
Louis
Louis,
I see that the OPA277 comes in 2 different versions. OPA277PA, which is $3.03 (Digikey US) and the OPA277P which is $6.01. The only difference between the two is the "input voltage offset". The OPA277PA (used in the project) has a input voltage offset of 20 UV and the more expensive one is 10UV. Could you comment on "input voltage offset" ?
Great work and thanks again for your time!
Phil,
You may have noticed I added a 20K trim pot to correct for any input offset voltage so that the cheaper OPA277PA should be OK.
Regards,
Louis
Wow, good job. How about supplying the electronic. load with a battery and a charging circuit. I suppose that the internal current of the circuit is minimal
Thanks. I am sure you could keep adding things to this project.
the current sharing wont work as you thought. The on resistance is the resistance in the active region. The load will work in the ohmic region of the mosfet and the thermal behaviour there is the opposite. You need separate drives for each mosfet to ensure it is operating in the ohmic region.
Thanks Magnus for your comment. This DC load works both in the ohmic and linear region. For lower drain voltages of say lower than a few volts it is working in the ohmic region but for higher drain voltages it is in the linear region. If you look at the datasheet you will usually find a graph showing "Normalized drain-source on-state resistance factor as a function of junction temperature". This clearly shows on resistance increasing with temperature (positive temperature coefficient).
I was using this effect in my application but I did mention it was not fully effective and the efficiency would be about 20% less. For this particular project it seems to work well enough at the power levels I am considering.
As you say you could add separate drive circuits for each mosfets and that could be considered for future upgrades if required, but I am also trying to keep the cost down for the hobbyist.
Thanks again for raising the point.
Regards,
Louis
Thanks for your reply and I am sorry for the late response. Looking at fig 8 in the spec for the mosfet you can see that for currents below 250A the current actually increases with temperature for a given Vgs. Since both mosfets will have the same Vgs there is a big risk for on of the mosfets heating up more than the other and therefore the current ratio between them will shift resulting in even more heating and current.
Thanks Magnus. The increase in current due to temperature for a given VGS is something I plan to cover in part 5 of this project. I have made some changes to reduce this effect in order to prevent thermal runaway.
Regards,
Louis
Hi, Louis
Excellent project. Perhaps I will repeat it as soon as possible. It stops only the presence of rare we DAC / ADC ICs. But I'll find a shortage of components on Ebay. I am from Russia, but I think these chips are no sanctions ))))
What are your plans for the development of this project? Can I ask you to introduce the function for automatic testing of power supplies? It works on the algorithm.
1. The user enters the value of the rated voltage.
2. The user enters the value of the voltage drop (in volts or%).
3. The user enters the current rise time.
4. The user sets the maximum power in watts (optional).
4. Turn the test.
6. Apparatus load increases.
5. The data obtained after the test.
Example
U nominal = 12 V
P nominal = 20W (Value at which started the voltage drops below a user-specified.)
U max = 10,5 V ( This value = U nominal - (U nominal - U value of the voltage drop in volts or%)
P max = 25W (The maximum capacity at which the voltage has dropped to the level set by the user, or the protection tripped (shut down the source of all)
Test time )))
Thank you and I wish you every success.
Best ragards
Roman
Hi Roman,
Thanks for your comments. Auto testing may be an interesting option. At the moment I am concentrating on improving the current arrangement and then looking at adding a Transient Mode with the option of a trigger input. I also want to look at adding additional menu screens to add more functionality. Other things could include Excess Temperature Protection, Reverse Load Voltage Protection, Data Logging, Calibration settings in EEPROM.
Regards,
Louis
As a protection against reverse polarity, I recommend to pay attention to the scheme.
img.mysku-st.ru/uploads/images/02/80/11/2015/06/03/1469e1.jpg
Best regards
Roman
Thanks Roman. Since we now have the Sense Terminals I am now looking at adding a Power Schottky Diode in series with the load. The one I am testing is the MBR41H100CT which is a dual diode in a TO-220 package. I will wire the diodes in parallel. Although we get a voltage drop the fact that we now have the Sense Terminals it does not matter.
Regards,
Louis
Hi, Louis
I try to model your design in Proteus, but unfortunately not even have a picture on the display. Although I connected all the libraries and compile the file successfully. I suspect that the problem in Proteus.
Can I ask you to put HEX file to test this.
Many thanks
Best regards
Roman
Hi Roman,
Below is a link for you to download the HEX file:
www.scullcom.com/Electronic_Load_software_V12.hex
Regards,
Louis
I can't get these MOSFETS anymore. They are obsoleted. Can you provide another in North America like mouser or digikey part numbers?
I am currently test some alternative as I mention but did not have time to cover them in the video.
One alternative is:
STP80NF55L-06
which is available from either:
www.newark.com
Newark Part No.: 26M3714
or
www2.mouser.com
Mouser Part No: 511-STP80NF55L-06
It is manufactured by STMicroelectronics and it is a Logic Level Mosfet with a Vds of 55 volts drain current of 80 Amp. Also its safe operating area is similar to the BUK956R1-100E.
Give it a try it should be OK.
Regards,
Louis
Added them to my mouser order. Thank you. I think this whole project costs about $100 to build. Which is excellent. I could've saved more by using some stuff I already had, but I don't want to say put another 16bit ADC and then have to tweak your code to make it work. Just plug and pray...
It seems that the Vout equation R1 in the denominator should be R3. Just a thought...Thumbs up
Thanks. Just to be sure I check and can confirm my equation is correct.
I remember in the industry they use to use "current sharing resistors". I think they were something like .01 ohm in each of the MOSFET (source pin) out put devices. This is called local feedback because every time one MOSFET would draw a little more current the IR drop across the "current sharing resistor" would go up thus decreasing the VGS of the associated MOSFET. Here is an example image I found on the internet... www.diyaudio.com/forums/solid-state/281960-unbalance-sharing-output-mosfets.html ....Thumbs up on your video....PEACE
Thanks for the comment. I have been looking at making the wiring from the source to the current sense resistor all the same length in that way the wire resistance would act as a low equal value resistor doing what you suggest. I will give it a try.
Regards,
Louis
Could you show in your next video all the external connections to your electronic load. In particular I would like to see where you are connecting the current sense input line. Thanks for all your videos....PEACE
OK will do.
I'd kill for a case like that! Where did you find it?
If you download the zip file I in my comments below the video and unzip it you will find a full parts list which includes details on the case. Below is also a direct link to the eBay seller I purchased it from:
Sourcingmap® 245mm X 190mm X 110mm Metal Enclosure Project Case
www.ebay.co.uk/itm/112303472593?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT
Thank you very much.
What is the size of the heat sink. The link to the ebay page has a size pull down? Thanks
Hi Rob,
The heatsink is 250x90x30 Large heatsink from the drop-down list on that eBay link in the parts list. This heatsink is just a little to long so I cut it down to fit the back of the project case.
Regards,
Louis
Buenas noches como esta tengo una pregunta cómo podría hacer una fuente línea de 0-100 volt thanks
what if I use 1% tolerance resistor instead of .1% tolerance ?
Will the voltage output will be unreadable or A little off that can be trim out?
You should be OK. It may be just out of calibration slightly. You could add a variable trimpot in series to adjust to correct level.
The biggest problem I see with that heatsink is that the fins aren't vertical, which discourages natural convection...
Simply suberb!!
Thanks.
Is there an option to clear the keypad input if you screw up?
I will be adding a clear option in Part 5 of this project. Also an option to input your own Battery Cut-Off voltage for the battery capacity mode.
Hello
Could you help me ?
I want to do an electronic load on your new circuit
But with such parameters
150 volts
500W
30 Amps
I'm interested in software
Because I'm not strong at this
I may have a look an an option using either a IXTN36N50 or a IXTN60N50L2 Linear Mosfets which will give you the power handling you require. We would need to alter the OP Amp drive circuit as they both need a higher gate drive voltage. But my software should still work OK. The two Linear Mosfets I mention are quite expensive but you will only need one. One of my viewers has already taken my design and software and made a DC Load with an IXTN60N50L2. He has posted his circuit changes and diagram on his web site, the direct link is below:
www.lavrsen.dk/foswiki/bin/view/Kenneth/ElectronicLoad
If you have a look at what he did it may help.
Regards,
Louis
Why do you have so many oscilloscopes?
Simply that I have acquired them over many years. I still like to use an analog oscilloscope on occasions. I now mainly use digital oscilloscopes for most of my project work.
why dont you make a few and sell it , i am a maker , i am struggle to find one i can afford , unfortunately have no time either to make one, too much prototyping work at this moment..
Sorry due to my time restraints I am unable to make any to sell at the moment. My You Tube channels is simply to help the hobbyist with there own projects.
big master pls mppt battery charger 50a or 100a ı waiting
Louis - Great Project really enjoying it and am in the process of collecting the parts to build one.
Your Schematic V4 states that R6 and R16 are both precision 0.1% - Did you mean R6 and R14 the 9K and 1K as the voltage divider on the +ve input to the op amp?
Question - Is there any reason why you couldn't use and 18k and 2K for the divider network? (I already have some 0.1% spare from another project)
I also have an extra large 20x4 I2C LCD Display that will work well with it (130 x55mm) Sometimes you just have accept your eyes aren't as young as they once were!
Thanks Richard.
Thanks for spotting my typo error it should have read R6 & R14 that is the 9K and 1K need to be ±0.1% tolerance. If you have 18k and 2K at ±0.1% they will do also. The resistors I used also had a Temp. Coefficient of ±15ppm/°C ensuring stability.
All the resistors in the Remote Voltage Sense circuit need to be ±0.1% tolerance and if possible ±15ppm/°C.
A extra large LCD would be good :)
Regards,
Louis