Boost Converter with Capacitor, DMM and Scope - Muppet 2
ฝัง
- เผยแพร่เมื่อ 7 ก.พ. 2025
- Sponsored by JLCPCB ($2 for 10 Boards): jlcpcb.com
Mustool MT8206 on Banggood:ban.ggood.vip/...
coupon：12MT8206
Banggood12th anniversary page: ban.ggood.vip/...
The boost converter certainly makes the bulb brighter, but the DMM voltage goes DOWN as the bulb brightness increases! Also, the top of the flyback spike is clipped (or crowbarred) by the MOSFET. The solution is a 1000F capacitor on the boost converter output.
www.jameco.com...
A man who understands the use of "less" and "fewer" in a sentence. From the bottom of my heart, I thank you. I wish the rest of TH-cam would use correct language and grammar.
The example you give - "less" and "fewer" is nothing compared to some TH-cam users. Iv'e seen "duh", "wa", "yuse", all sorts.
They seem to think it's clever to destroy the English language. Tie the devils over the mouth of a cannon and fire it. Indian mutiny. Bit 'orf' don't you think.
The 15Khz actually gets a bit louder at 50% duty cycle with the capacitor installed! It's a bit annoying but you get used to it... It's like hearing a CRT TV in the old days
Yeah, 15.625kHz on old PAL tellies if I remember rightly :)
Indeed! It was actually fairly audible (at least for me), even without the capacitor!
I remember those days, when I had to leave a store because of the squeel of those old Televisions. Can't happen any more as my hearing has been failing for some time now. If I hear such a noise now days, it is usually my hearing aid getting some feedback.
Jerry, if you think that was bad, some of the early "ultrasonic" alarms weren't quite "ultra" enough. I could hear them as a kid. When they were disabled, they only disabled the alarm part - the transmitter ran 24/7.
Could really hear it at about 3:09. Couldn't hear it other places.
I love this, it's a great example of how a capacitor effects a circuit. Great help for a buck converter knowledge..
Finally, the circuit makes sense to me. I was wondering what the diode was for without the capacitor.
A very helpful discussion to see what's happening at the component level!
Realy enjoying the muppet show, thanks Juian.
To clarify for the learners out there: It is a clamp not a crowbar. When triggered, crowbars latch fully on - usually until the power source is removed, with the intention of making the input protection fuse blow. Clamps prevent a voltage from going higher than a certain point, by applying a load, in this case, the zener built into the IRFZ44. Look at the schematic symbol at 6m50s, the body diode is presented as a zener.
OK, that's interesting - thanks :)
Good catch. Interesting that some schematic symbols show the body diode as a zener while others don't.
Julian, the crowbar effect your seeing is actually the body diode doing its job. If you look at the symbol for the MOSFET it shows a zener there. That is there to help protect the drain source junction. Hope this helps explain what you were seeing.
OK, that's interesting - thanks :)
The body diode is a parasitic structure in the mosfet (actually a whole bipolar transistor with B and E shorted). It's drawn as a zener to indicate drain to source breakdown. It's not there to help with anything, ideally it wouldn't even be there.
Crowbar circuits short points in a circuit together like a crowbar across a battery, waht's happening here is called clipping. Crowbarring would happen when the mosfet burns up and turns into a jumper.
Julian - Put some emi caps into the circuit. One next to the input cap, one across the output cap. Also for a super design you could add some common mode inductors and filter caps pre and post converter to really clean up the power and kick out noise.
Yess! Julian! I love this project!
Cheers Karl :)
The meter is reading RMS voltage so it's doing fine for what it's been asked to do. The circuit sort of emulates a PWM without the capacitor so it reads that but in (reverse?)? The lower the duty cycle, the lower the RMS reading.
Adding the cap after the diode is nice. It shows the power of what a good cap can do to a circuit and why we need them.
Two questions about this setup is (there may be others l8r, depending on the answers), what does the size of the capacity of the capacitor do to the ending result? And what would happen if you put a capacitor "before" the diode; will it change anything?
try a larger capacitor or two in parallel. the overshoot and ringing observed on the scope indicates underdamping. if you're looking for smoother output voltage.
There was some noise on the output with the electrolytic cap there... i wonder if you get rid of that noise if you also put a cheramic cap in paralell with the electrolytic.
Since the cheramic caps are better on high frequency stuff compared to the electrolytic.
I've been thinking about the Mustool meter/scope quite a bit. Too bad you will not get a good test of it before the sale ends. I'm looking forward to what you think of it.
Please forgive my ignorance. I continually read that buck converters are noisy. As they are quite capable of driving a PC processor I can only assume that the 'post processing' of the output must be very good. You used a capacitor, is there any way an 'add on' can reduce the noise on the many Chinese ones available.
I noticed that after you added the capacitor you switched the input meter from voltage to power. what was the input to output voltage with the capacitor?? Thank you Tom
You shouldn't put the scope ground level at the very bottom of the screen here. 4:56 It is an analogue signal with an inductive load and you *can't assume it isn't going negative.* Move it up at least one division so it shows any negative voltages and you can adjust it up more if required.
It would be alright to do this on an analog scope because the trace can still be seen below here but on a DSO it is a hard line.
Julian, would have been interesting to see what a moving coil analogue meter made of the output voltage.
Do you still own any (I’m presuming that you had one in the past... mainly because my first multimeter was a moving coil analogue type...).
I believe that's called an output integrator when you use a capacitor to average out the PWM. it's how they "recover" a sine wave from PWM in pure sine wave inverters and the audio output in D class amplifiers.
Are you sure it isn't the scope cutting the top of those traces? The volts/div selection determines the maximum voltage it can measure (because it determines to which multiplier/divider the amplifiers/attenuators inside the scope are set). It doesn't matter where you move the zero volt line to.
The MOSFET might be getting hot, because the inductor is driven into saturation at which point it is like a short.
An inductor that size will be far from saturated with this amount of load.
Also, no scope will display a cutoff like that ever.
There´s a good reason that Mosfet claims only 55V VDSS. You can´t just use any mosfet for any purpose. At some point it just becomes conductive when off.
I believe that symbol for the MosFET body diode is a zener, which explains your clamping.
Yeah, I'm reading that it's like a PiN diode, with zener or avalanche characteristics - interesting stuff :)
Hi Julian,
One micro comment, in relation to the capacitor.
Excellent presentation and theory.
I looked at the Mustool DMM on Ali. The one you have should go to 5GHz, not 200Khz. So why didn't it show the range properly? Or am I looking at the wrong thing?
MT8205 MT8206
Bandwidth AC 10KHZ ● ACA/ACV/DCA/DCV 20KHZ
Frequency 2KHZ-200KHZ ●5HZ-5MHZ
Are we looking at the "bandwidth" instead? If so, what's up with the frequency?
Love my Keysight scope. Think it's time I played with a similar circuit.
The ringing you saw after installing the cap is probably not there. You are using a ground wire for the probe which creates an antenna. If you measured with a shorter wire (or none ideally) the ringing would would likely diminish or out-right disappear.
It´s unlikely that this comes down to his probes. The frequency isn´t very high here.
It´s probably due to the diode he´s using. Diodes take a bit of time to turn on.
@@Basement-Science hmm, good point. I have seen scope ring in 200khz applications before but not in anything this low. It is an easy thing to rule out though.
"Less voltage, fewer volts" :) I'm getting that warm fuzzy feeling that only correct grammar can bring.
I’m late for my comment, but here comes anyway: Your volt meters are measuring the average which is dominated by the FET conduction pulse width. If you had a true RMS meter, it would weigh in (like as a square) the transient portion and would increase just like the light bulb intensity. In that respect, the light bulb also responds to the true RMS.
What is the influence of the inductance of the coil to this whole thing? I simply can't imagine, that it doesn't matter.
It kinda matters - big inductor, lower switching frequency :)
Is that clamping because of the MOSFET body diode reverse breakdown voltage?
Re: 7:38 I hope you know that light bulbs are NOT linear devices, a matter of fact there are a bunch of circuits that use light bulbs because of this facts. They do change their resistance with the temp of the filament but it's not a straight line correlation. I am sorry I don't have a link right now on the subject, but I remember this from a project I did back in the 70's and from popular electronics article.
typo in the description; it says 1000F capacitor instead of 1000uF.
According with inductor value i think it's better increase the switching frequency. So the average output voltage will be higher and output voltage and current ripple in the capacitor will be smaller. Do you think so too?
we learned two things in todays lab, signal aliasing and low pass filtering
You could try the multimeter on AC to measure the spiky DC accurately. True RMS multimeters don't care whether it's AC or DC input, and it should show the voltage rising with the brightness of the bulb if the RMS integration is working correctly. (I'm not sure why they bother with DC settings at all on true RMS multimeters!)
a great proof of concept
Did you ever explain why the dvm's showed lower voltage as the pwm increased ? I heard a lot about crowbarring, and capacitor, but no explanation.
The waveform was spiky, the meter can only read a sinewave accurately.
PWM is not a sinewave.
@@youpattube1 the higher the voltage got the narrower and sharper the pulse became and the more the meter misread it as a lower voltage.
youpattube1 - as the duty cycle was changed, so the voltage spike got taller, but the on time of the transistor was longer. During this time energy is going into the inductor. So no energy is going to the lamp. Normal digital meters have both relatively slow sample rates and only appropriate waveforms that are not pure smooth DC. They use an averaging system (designed to show the correct AC value when measuring a pure sine wave). Because of this, they don’t respond to the short voltage spike, but do notice the loss of voltage during the time the transistor is on. Hence as the transistor on time is increased, they show a lower voltage.
Now we have an R-C-L circuit. Can you calculate the resonance frequency? And what would happen with the output? Or input? Or polyfuses :-)
Next, I want to charge my supercapacitor bank using the Muppet :)
Now to be fairly honest: is that wise? Supercapacitors are sensitive to 'high' voltage (a few volts above their maximum voltage is sufficient), it will shorten their live considerably. Now I may be wrong, because I saw on the scope the peak voltage is no longer there when you connect a (small, 0.0001 F) capacitor. Seen from a circuit perspective, the supercaps are just a (huge) resistance. What I am trying to say? Please do the charging and show us what you find out (also use the scope, check if there is any ripple left). I really don't know if you could damage them or not.
Great practical example! Now how would you do current limiting??
Thanks. Measure current with the Arduino - then use that measurement to adjust the constraint on PWM value.
@@JulianIlett I know voltage and current are interrelated but how would you for example boost the voltage to double input but limit the current, you would need a high duty cycle to boost voltage but then a low duty cycle to limit the current, it seems contradictory, or is there another factor at play
Andrew Waters what you´re asking for is *_regulation._* So far this thing is completely unregulated. If you were to remove the lamp and run it, the voltage would just rise until something dies. In the same way, if you increased the load, the voltage would drop.
What you need to regulate the voltage *or* current is circuitry that measures it and adjusts the duty cycle.
By the way, measuring that with the arduino or another ADC will lead to very slow regulation since an ADC takes time to do its thing. Normally OpAmp circuits are used for this, which are normally integrated into SMPS controller ICs.
It would be interesting to get a "true RMS" meter on the boost circuit without the cap. I suspect it would show a voltage rise. The bulb is behaving like a "true RMS" measurement, while the DC voltmeters are filtering the waveform.
Bingo
I don't believe that clamping and crow-baring are interchangeable terms.
Clamping constrains to some non-zero voltage (60v here), whereas crow-baring is a special case where the clamping is to zero volts.
It comes from the notion of throwing a crowbar across the output of a power source - an extremely low resistance that won't burn out, which absolutely forces the power source output to zero volts. Many power supplies include "crowbar overvoltage protection", which triggers an SCR across the output should the output voltage rise too high. The heavy short circuit thus caused blows the power supply's fuse, protecting the load from the overvoltage.
Quite right - I should have called it clamping :)
I approve of your shorting out fuses. Annoying troublemakers.
Heiko van der Laar - he should of used a six inch nail instead though 😂
crowbar! I've heard that term for years and sort of had an idea what it meant. So what does it actually mean and where did the name come from? is it from the pattern you are seeing there on the scope?
emaglott - no. The term was misused here. He most likely mean clamped. A crowbar in electronics actually refers to a protection circuit that once triggered by a higher than normal voltage, it conducts until the circuit voltage drops to near zero. The idea is that by shorting the supply, it will rupture the fuse. Occasionally a breakdown in an electronic component will have a similar affect. But this is not what was happening in this video.
Oh. Thanks for the clarification. Why is that called crowbar?
emaglott - I can’t for certain tell you why, as I don’t know how the name came about. It may be because the circuit has the same effect as dropping a metal bar (crowbar) across the supply terminals (the voltage drops to near zero, a high current flows and blows the fuse). Or it may be named due to the sudden and rapid removal of a fixing when a much larger crowbar is used to lever it off of a structure - in the circuit, the voltage is suddenly and rapidly removed as first the crowbar circuit shorts it out, and then the fuse blows.
Search using “electronics crowbar protection” for more information.
@@Mark1024MAK- I was looking up some info on ignitron rectifiers and found crowbar information! It is a reference to dropping a crowbar across the output terminals of a power supply according to: en.wikipedia.org/wiki/Crowbar_(circuit)
So the temperature did drop on the mosfet with the capacitor installed?
Yep :)
Next goal: Regulating the voltage output so the voltage is adjusted for the load.
Up to about 20v with the capacitor it is just as loud as before.
Is Vdss not drain source saturation voltage?
Bet you can't hear Radio 4 on LW with that on?
If your load is always a bulb use a smaller capacitor maybe only 10u to take the spike only and leave the square
Yes, good idea. My next load will be a supercapacitor bank - I want to charge it :)
Julian, you probably improve then efficiency by adding a cap on the input also. I used to fix smps's in the telecom's sector and every single one had come capacitance on the input.. Just a thought. :)
It wouldn´t change much in this case. The power source (his other converter) is pretty close to it.
Yes that does have caps.
How about a video getting started with a stm32 board?
won't the ESR of that cap be high? won't it be better to have a low ESR cap in parallel?
This converter only runs on 15kHz, so it´s not too critical. Of course lower ESR would decrease the output ripple a bit more.
Useful video
Now compute the value of capacitor that will form a resonant circuit with the inductor a5 the 15 kHz frequency, and see what happens!
Enjoyed this, thank you. I saw that meter on banggood last week, I’ll be interested in your review. I think I must have lost 15kHz too :-(
Here's hoping we get a full review, not just a hand wavy sort of thing. Perhaps actually test its accuracy on all ranges, check the resistance threshold of the continuity, look at auto ranging times compared to other devices, look inside to assess the construction & safety, check the current consumption. And on the 'scope look at bandwidth (comparing it with your proper cal'd 'scope). Examine and demonstrate the history functions (such as they are).
Where did you get those nice wires with banana plugs on both ends?
I made them :)
They look like they were made in factories. Nice job!
Is it just me who can hear the inductor squealing on the video... It's horrid!
And the capacitor doesn't help much!
i can hear too, and i paused the video and see if the noise is coming from the video
Yeah when the light is on I can hear it but barely, almost too high frequency for my ears I guess.
Hmm, looks like the capacitor mob got their way.
Only briefly :)
I could still hear the high pitch ring either way. Though it did change pitch slightly when the cap was installed, and it went a little higher. I think you had your audio set a little differently, or maybe the mic was in a different spot, something. It wasn't "quite" as annoying as the first video 😉
I do not hear any difference with or without the capacitor.
A capacitor...finally.
..how novel
Maybe after like 5 more videos there might be some actual regulation too lol
He's the kind of person that wants to go slowly...
As always a 👍 from me.
An "ever so slightly voltage pulse" of 6 volts. Sure!
11.34 volts spells out HELL.
very good video
Try to pop a led on the boost converter 🤔🤷♂️
I love it
The problem with Bangood is that they dont do PayPal :(
Yeah they do, at least on what I've ordered so far...
Can you run it with a 555 ???
Yes, probably, in a PWM configuration (with diodes). But the Arduino is needed for synchronous buck where two non-overlapping waveforms are required :)
you make me subscribe
It is silly they still use "7-segment" display font for digits on graphical LCD. So silly.
learnful
Real men don't use polyswitches...
I thought they were called capacitators lol
Stop cheating and build an oscillating board
hi，my friend
Can you help me promote the shop?
huh, thats quite an interesting multimeter for $50. i wonder if that "oscilloscope" mode can be used for anything useful or is just a gimick.
I saw it on banggood on sale and almost pulled the trigger, but figured the scope would be fairly useless.. They show it displaying a 60hz sine wave, I wonder if it has the resolution to accurately display a modified sine inverter output waveform? Or some dc graphing? I'm going to guess it's only useful for around power line frequency or below, and has trouble auto ranging and triggering on anything but perfect signals. Can't wait to see Julian put it through a few tests.
I'm thinking I'll feed it with 1kHz sine from the Wein Bridge oscillator, several kHz from a Royer oscillator (also sine) and the output from a modified sine mains inverter. Maybe also a 555 square wave.
Perhaps the sources to use (for 'scope testing) are your DDS and waveform generator modules. The AD9833 and AD9850 if I remember.
i'll be very interested to see that.
Julian, maybe you could also compare the JYE TECH DSO138 Oscilloscope (kit) that you did a while back. What was its bandwidth?
Can you help me promote the shop?