Power Supply Rebuild - CNC Router
ฝัง
- เผยแพร่เมื่อ 27 ส.ค. 2024
- Recently burnt out the transformer on my CNC router. Decided to rebuild the drive/stepper power supply.
What you're about to see ... is a true story.
EDIT TO ADD: I use Gecko Stepper Drives. the formulas & power supply built in this video are done so in accordance with guidelines laid out by Gecko. These may or may not be applicable to your drives or your application. Do your homework & be safe!
------------------------
Mariachiando -- Doug Maxwell/Jimmy Fontanez
Quickest way to get an expert solution to a problem is to get on the internet and claim your solution works. Very clever This Old Tony, very clever indeed.
Good luck telling that one piece of sensible advice from the equally plausible-sounding 99 armchair opinions who will claim you're screwing up horribly for all the wrong reasons...
but.... it cut metal at the end.
nonetheless, stay tuned for my next video "DIY appendectomy"
The biggest difference aside of the voltage rating between the capacitors used before and the one now is probably the ESR (equivalent series resistance, a frequency dependent resistance which leads to heat in the capacitor with every change of charge = charge/discharge) based on the type of capacitor used those might vary quite a bit. I am not talking about grid frequencies or double grid frequencies, but the stepper drive load as well. If it works, there is no problem, even if it fails your stepper drives will be able to handle the full peak voltage - not work perfectly, but probably stay alive.
The worst effect of an unregulated power supply could be changes in torque depending on load, but the drivers actually compensate for that, them being variable input voltage capable. The driver datasheet or website probably contains enough information to answer those questions.
yes, the proof is in the pudding
@@ThisOldTony I saw that article. It was in "Suture Self: The Magazine of Home Surgery"...
I appreciate the civil engineer joke. Factor of safety and so on...
joke?
;)
To be discreetly precise as a computer;
Need Moar Practical Engineering! :D
I heard that joke and immediately went down here to see if you picked up on in XD
I can hardly wait for every This Old Tony's This New Videos. I'd also like to thank TOT and all the other TH-cam machinists for inspiring me to make a late-in-life career change to become a machinist. I've enrolled in our local comm. college and have recently begun making small bits of metal out of larger bits of metal for the first time.
Waco Glenn welcome to the secret club, dude. The rabbit hole never ends
as a Mexican, you got many smiles out of me on this video. as an engineer, you got many smiles out of me on this video
Just giving you a heads up, Tony. You might be reading a lot of anger from electrical engineers because of your calculations :-)
For example: Your stepper motors take 12 Amps in total, however the Gecko Drive (as does any sane Stepper Motor Driver) uses current chopping to generate the constant current (of 4 Amps per motor). Due to this fact, it effectively works as a buck or stepdown converter. And consequently, the Gecko is nowhere near 4 Amps at the (up to 80 V) input voltage.
If your motor has a DC resistance of 0.5 Ohms, then the idle power consumption per motor is I^2*Rdc which is 4^2 Amps^2 * 0.5 Ohms or 8 Watts. Thus in idle, your whole setup takes 8 Watts per motor or 24 Watts in total from the power supply. This relates to roughly 0.4 Amps at the 60 V input to the Gecko Drives.
Added to this is now the *mechanical* power that you request from your motors in your machine. And this is where it gets really tricky. Worst case, your Gecko Drive will not chop current anymore, because the demanded mechanical power is so high. In this case, your calculations are correct, the full 4 Amps times 60 V input voltage will be requested from the power supply. However this is only a worst case calculation and I wonder if your stepper motors will handle 240 Watts of mechanical power (with an ideal efficiency). I don't know the efficiency of those motors, I will guess it is somewhere in the 80% range. This means that the stepper motor has, beside its idle power consumption _additional_ heat losses of almost 50 Watts. I think the motor will not be able to handle that :-)
I guess what I wanted to say is, that your power supply is really overkill :-) Also the big cap might get you into trouble during powerup of the system, blowing the breaker all the time due to the startup current. By the way, the bridge rectifier definitely has to be put onto a heatsink! 12 Amps times roughly 2x0.8 Volts forward drop voltage is already 20 Watts, which is too much for the wimpy case without a heatsink. The hole in the middle is not there just for fun.
no power factor correction stage after the full bridge rectifier! IM PISSED
Given how crazy big that transformer is I wonder if that 1500W rating is 1500W dissipation in the transformer itself rating rather than a 1500W driven load rating....
the peak current through that rectifier is in the 100's of amps. It won't last long without any power factor correction. Neither will the wiring in his house.
not an unreasonable size for 1500VA, need about 15kg of steel and copper to do that, the dissipation on the transformer would be very little
Also the back EMF of the motor won't let him get all the amperage at high velocity. So yeah, waaaaay overkill.
Impressive how many parts one can salvage from an old MP3 player.
And still have it working after salvaging said parts!
They are there just to justify the price.
iwtommo I read your comment right as he stopped the music
For 30 years I designed power supplies; switchers, linears, battery chargers, and some jet engine starter/ generator stuff.
I like to have some inductance between the full wave rectifier and the capacitors. That makes a real averaging filter and cuts down on the peak currents [both start up inrush and steady state 120 Hz]. I like snubbers across the rectifiers so I don't ruin AM radio reception. I look at capacitors in terms of ripple current rating, effective series resistance, effective series inductance, and if electrolytic, how many hours until the electrolyte has out gassed. If you pick up the capacitor with tweezers so you are not weighing the fingerprint, you can measure the rate of electrolyte loss. Small wires do not hurt anything, but they can get hot. Then it becomes a matter of the insulation on the wire's temperature rating. High temp insulation is hard to strip.
Tony, thanks for this. Most of us are tooling guys, or CAD guys, programmers, or electronics guys. Not always a lot of overlap but you bring all of it to the table.
Keep introducing the math! It's what keeps the electronics making sense!
I really enjoyed the show, a pleasure to watch as always.
I am a civil engineer........
I wish I understood something of electronics, though I understood the unscrewing of those two bolts
I find it so difficult when you are talking to decide on the fly if you are being serious or not.....love it!! Keeps me on my toes!!! Keep up the amazing work!!!
I’ve never heard of the bucket analogy, that’s a great way to put it! 2 electrical degrees, 21 yrs in electrical maintenance, never heard the bucket... Thanks
A man of my own heart! One thing goes wrong and now you own a transformer the size of a puppy
FULLL BRIDGEEEE RECTIFIERRRR!
rectum frier
electro boom
Not a puny one diode rectifier!
I like how both AvE and ElectroBoom have inside jokes about full bridge rectifiers
yea its like they are almost catering to two different crowds of people
Your humor..... Love it
when you took the mp3 player apart an the huge cap fell out i'll admit i laughed way to hard at that.
chuckiebthatsme dude me too, mad chuckles
That wasn't a cap
I'm pretty sure it's a huge capacitor
It was the rectifier, not the cap.
It was both. One capacitor and one brigde rectifier.
just replaced the speed control board on my mini mill and then got watch this. its been a great day. thanks Farthead.
As someone who did a lots of electronic building in the past you made a very nice heat sink
I don't know why I'm watching this again.. but I'm sure happy I did.👍
carbide [carbide]
Not sure why but this cracked me up.
A couple things come to mind when I see your design for this simple un-regulated power supply: 1) The larger the capacitance bank, the larger your in-rush current will be. Therefore, make sure to size your fuses to the proper value for safe operation and use double-time delay fuses to prevent them from blowing every time you turn it on. 2) Un-regulated supplies do not maintain their output voltage when faced with large current draws. This means hard acceleration will pull down the DC bus voltage. Make sure your accel ramps are defined such that you aren't asking for too much motor speed at peak acceleration ('cause you won't get it). 3) Conversely, when decelerating the motor driver will "pump up" the DC bus voltage due to the regenerated voltage coming back from the motor. Make sure you have enough headroom to accept this voltage or things may go boom. Especially electrolytic capacitors (100V cap should be ok). 4) In servo and stepper systems, voltage equates to motor speed (because you have to overcome the motor's back-EMF to go faster) and current equates to torque. Your stepper drives are probably ok if they are rated for the nominal DC voltage of your supply, but remember the regeneration voltage that will be back-rectified. You might want to stick an o-scope on the DC bus and watch it as you acceleration and decelerate all axes simultaneously to make sure everything looks good under peak load conditions. Whew. That was a lot. Sorry for the book I wrote you!
Your sense of humor gets me every time Tony! I think I watch you videos just to laugh at those moments! And occasionally, I learn stuff :P
The little puns and goofy stuff (like the power joke and the Mp3 player) is only getting better. Makes these even more fun to watch. :)
It's amazing how these videos keep getting better and better, they were already top notch!
the john walsh ad was the best part of the video.
When a TOT video pops up, man you know it's going to be good. Thanks for getting the music and funnies back in, now I really have to concentrate again to see all the 'hidden' stiff. Kindest regards. Joe.
Love your humor laden educational videos with GREAT background music. I am no thermodynamic engineer but would have drilled and tapped the other end of the heatsink for mounting to the enclosure. This would help keep the heat away from the capacitor. A small valued/high Wattage surge resistor in series with the cap should reduce initial surge current and a high value bleeder resistor to ground will help prevent a shocking experience when poking around after being powered down. As you know, "The Experts" come out of the woodwork after you have created something that works for you. Keep up the good work and never let the smoke out of electronic components as you will never get it all back in again.
Your sense of humor is unmatched. Just very enjoyable to watch! :)
Tony, when you added the Teflon washers for the cap terminals, it looks like you put just a washer on top leaving a gap between the cap posts and those washers. If this is correct, all the clamping pressure you are applying with the screw is acting directly on trying to pull the terminals out of the capacitor. Add a small spacer or make a "t" profile custom insert to replace the plain washer.
Good point, thanks DAS!
And maybe a beefy descharge resistor on that cap, so it can bleed of the power stored in the cap when you power off your machine .
So no nasty suprises when you want to poke arround in it and make some changes ,But I'm sure you knew that already .
Nice built ,Looks like something *S*S could use ,i hope they don't recrute you. :-)
You betcha.
Jacksat the discharge resistor is very good practice. If you want to test the power stored in a cap acito just discharge a smaller cap at a lower voltage with a screw driver and see how much of the screw driver melts
Good idea using bridge to warm up capacitor. It will be easier to start this thing on cold winter morning.
Wow, it's been 10 years since I've seen someone with a Sandisk Sansa.. Amazing mp3 player. That one went waaaaaay over the "use by" date for me too. Respect!
man i really like your channel. i've watched a video or 5 everyday. some in the morning, some at night, some both like today. your voice calms me. world is hectic right now but i get a few laughs in and learn about machining watching your channel. thanks for the relief .. i also enjoy you know some electrical engineering concepts D
the close-up slicing death of "Thanks f watching, / Fartheasd," with that musical accompaniment, is modern art
Well, I didn't really understand all of that, but, I sure did enjoy it. Keep making 'em, Tony, I enjoy them....
Just a tip for you on your DIY heatsinks. If you cut your fins to sit vertically in operation you will pick up quite a bit of passive cooling in the form of "chimney effect", as opposed to stagnant horizontal laminar flow. Why heat sinks on the back of old high power audio amplifiers nearly always ran vertically.
As to question of more small caps being better than one big one, the answer is yes. Due to many smaller caps having a lower combined ESR (Effective Series Resistance). Basically lowers the impedance of your filtering. (Pushes back harder, faster to smooth out bumps).
On figuring safety overhead values for components, industry rule of thumb is 20%.
Doc
Hi Tony. Another great video! Always get excited when I see you've posted a new one.
I used to work as an electronic engineer for a company which designed custom packaging equipment for the food industry. We used to drive big stepper motors from a pair of 24 volt regulated switch mode power supplies, wired in series to produce 48V. We never had any issues with back EMF damaging the PSUs or the stepper drivers. The advantage was they were cheap (24vdc is really common in industrial control), and available in compact din rail mounting cases with all the international certifications for sale all over the world.
You'll never need to replace that beast though! That transformer could happily power 10 CNC routers 😉
NO! Not the Farthead block!!!!! I really like the bridge rectifier/Filter cap mount. Excellent job. Looks like you'll have more power to tackle even more larger projects.
Miracle of medicine... This Old Tony discovers new cure for urinary incontinence.
Love the music selection for this episode.
That's what i did for my wood lathe earlier this year. 90vdc treadmill motor. 90vdc? I don't have that kind of power supply! Screw it, we'll take 120vac, rectify it, put a huge cap on the output. Boom. Wait, that's way too much voltage. Ah yes, variac to the rescue. Talk about precise speed control. Beauty.
Hi ToT. You make really great videos. I like this one about the power supply. As a former electrical/electronics guy, I have a recommendment. When you make stuff with ginormous™ capacitors, it doesn't hurt to put a high value resistor across the capacitor terminals, like a 100k or 1M ohms or so. A low wattage little guy that won't substantially load the supply or waste power, but during periods of inactivity, will slowly draw the voltage down to zero-ish. This is particularly nice protection if you happen to power the supply up while it's not connected to the motor drives. After you turn it off, the voltage will bleed back down to zero. Anyway, just saw that monstah capacitor and immediately though a nice axial resistor would be right at home across the terminals.
Pretty result. The bucket analogy was good also.
Wow! you explained electronics like my teachers never could!
You are the absolute best this old Tony !
That knowledge is power joke was amazing!
More smaller capacitors are better for many reasons.
Smaller caps are cheaper.
If one cap fails you only have to replace one small cheap one.
In the event of a capacitor faliure you get a soft failiure. This assumes the cap fails open circuit obviously if it fails short you will blow your low voltage AC fuse.
The ripple voltage will cause the capacitors to heat up, several smaller caps will have a greater surface area to dissipate heat.....
Mechanicaly I like the idea of using the heatsink to mount the smoothing capapacitor but heat is the enemy of electrolitic capacitors so I hope the heatsink does'nt get too hot.
Love your videos the perfect mix of real knowledge, great thequnique and humor.
Thanks, Tot, for making so many different topics understandable and of interest.
Even when various comments may already mentioned some of these points (or then not), I decided to add my 2 cents worth.
The single big versus several small capacitors is mostly determined by the ripple current rating. The ripple current is not straight forward to calculate, but you can start wit a guess that it is same as the load current. So, whether it in this case is 12 A or something else, let's take that just as an example. If you look at the data sheet of the candidate capacitors, you often find that 4 smaller capacitors have more ripple current capability than a single capacitor of the combined capacitance. There are two factors to this - the combined surface area of the four capacitor is larger and thereby provides better cooling. The other part is that the parallel connection reduces the ESR (equivalent series resistance) to one quarter and likely the individual ESR numbers for the small capacitors are not 4 * the ESR of the big one.
Then an ISO safety rule suggests that a bleeder resistor is dimensioned so that the capacitor voltage drops to a safe level (which varies by the environment) in 20 seconds.
A lifetime expectation is tied to temperature. I have chosen to follow the Military Handbook rules and limit the component voltage to 0.7 * specified maximum. Similarly the current to 0.7 * specified maximum. That way the POWER is limited to 1/2.
And finally, there are two parts to inrush current, the effect of the capacitor and the effect of the transformer. The rectifier bridge and wiring resistances can mostly be ignored as a first approximation. If the capacitor ESR is 0.1 ohm, at 64 V peak, you would in theory reach 640 A worst case. But then there is the inductance and resistance of the transformer. Those are normally expressed as a combined "regulation" factor which means the percentage drop of voltage from zero to full nominal load. Some tiny transformers may have as bad as 15 %, while really big ones (power distribution types) may have 2.5 %. The 1500 VA transformer likely has about 4 %. That translates to some 12 or 13 A at 115 V primary full load and a huge 1/0.04 * 12 A = 25 * 12 A or 300 A on the primary. Have your fuses or breaker selected accordingly!
Hoooooooooooooly crap... I thought my prototypes sucked... but you made that actually work for as long as it did is amazing.
I greatly enjoy your videos. I don't know anything about anything you talk about, but you make it interesting and entertaining. Thanks!
I strongly recommend a 6.8k 3 watt resistor across the capacitor terminals, to discharge the cap when the power is turned off. Also, you might want to look at AjaxCNC, they sell pre-made rectifier and capacitor combos, the rectifier board mounts to the top of the capacitor and is quite tidy. Just in case you need one in the future.
Tony, looks like you have used a crimp tool for uninsulated terminals on the yellow faston/spade connectors. There is a specific crimp tool for these insulated terminals.
If the wires are not subject to vibration It's better to remove the yellow insulation and solder these connectors. You can use some heat shrink tubing for insulation if necessary.
If you think there is a risk of vibration then use uninsulated connectors with the correct crimping tool. It is much easier to get a good crimp (cold weld) on this type of terminal.
With the oversized transformer, one needs to consider the amount of current it can deliver with its secondary shorted. This is the current the bridge rectifier will be asked to pass to the discharged filter capacitor on startup. If the filter capacitor is too large, the inrush current could be several hundred amps for a short duration on each power application. The bridge rectifier may take this abuse for a while, but it will likely fail far sooner than a proper design. In a past job I maintained several high voltage power supplies that supplied 10KV at up to 6A. These supplies had a circuit with six large SCRs between the mains and the three phase transformer primary to limit the inrush current, which provided a soft-start to the supply which kept the instantaneous current the rectifiers passed within their specifications.
I don't know what I enjoy more the video or your sense of humor with the music I love it😂😂😂😂😂😂😂😂 well your sense of humor...period
THANK FOR TO LOOK!!
Also: 1 cap > 4, no loss in the leads, no solder junctions to crack, no tracking down which cap blew.
I was so happy when I saw you put down thermal paste on the heat sink. I know it seems common sense, but you would be surprised how many wouldn't.
Yay! New ToT video!!!
(from 2 years ago... new to me though.)
My favorite video you have done so far lol love the Mexican music your the best, "diresta eat your heart out!"
Clear and lucid explanation - thanks! And also thank you for using the correct terms - current and power. I immediately kill any video where 'amperage' and 'wattage' are used.
OMG, This Old Tony, that poor Sansa Fuse... How could you?! Senseless desecration of one of the best MP3 players ever to exist aside, you created another great video. Thank you for the edutainment.
Informal advice:
Mount diode bridge above cap to avoid heating it which could shorten it's lifespan. Or better still side by side ensuring heatsink fins are vertical for slightly improved cooling.
Multiple caps are better for lower ESR (= improved filtering) but not essential in this application. Can be added alongside for same benefit.
Another commenter suggested inrush current on big cap could cause problems, which is plausible.
A separate breaker or fuse in the PSU inlet would be desirable if don't have already.
That's my unasked for tuppence worth! Keep up the slick humourous videos!
This is my all time favorite channel
Thanks for the info Tony. I have been contemplating upgrading my power supply for my small CNC router as well. Now to find a new spindle motor so I can cut aluminum
Watching that list clip, I forgot for a minute that you built a router, not a CNC VMC. Nice.
That music! Oh, the stuff with the numbers and metal chips flying everywhere was pretty ok too...
Wow, a DIY CNC that actually sounds like a real CNC rather than an RC car rolling through gravel.
Large linear supplies like this typically use inrush current limiting in the form of a large resistor between the xformer and caps. Once voltage rises above a certain point, a relay will bypass the resistor. Likely won't see many issues except potentially prematurely wearing your circuit breaker if you turn this on and off a lot.
This Old Tony: At 9:00 on the bottom of the capacitor, that white dot. That is probably a vent hole, it is there to release the magic smoke in a controlled way (without explosion where metal bits fly around and stuff like that).
When you mounted it at 11:20 I couldn't see a matching hole in your mounting plate. I would have drilled that third hole into it.
Love your videos and editing.
A friend of mine with more experience in such things commented to me that you should never mount your capacitor on the same heat with the rectifier. In the event of the rectifier partially shorting , the excessive heat turns the cap into a potential bomb.
Keep up the awesome videos.
Love the vid. This is the type of thorough info helps me. One note - I have trouble knowing when the camera is sped up during machining. It may be obvious to seasoned machinist, but a 2x or 4x label might help us newbies. Thanks for the videos. Waiting for a patron page.
Think you made Forest Mimms Proud, Tony...even the screen writing/font was similar. Two minor things I noted were a vent hole for the Cap in your heatsink (Yustin case) and a Bleeder resistor. Keep'n us grinnin while schooling us is Art! Thanks for all your woik and dilitentness! ~PJ I been cravin some good Tamales....good ideer and music...must be time.
Thanks PJ! 1) I think (hope) the cap has a vent in the back. I didn't want to coat my electronics in cap guts. 2) I didn't think about a bleed resistor but I was worried about how long that thing would store power on shut-down, but I noticed the LEDs on the drivers didn't stay lit for very long... so I assume they're draining the cap on power down?
Thanks for watching!
Typically the vents are on the terminal end and saw it there in the video. One that big would spew the blue goo into all the wrong places but more important is one that size are like horseshoes and hand grenades. -Close is not pretty if it can't vent properly....and a bleeder will help to keep it cool/last longer. One of those tubular ceramics with the heat sink around it would be good and probably bolt to your heat sink, IMHO. LED's are only about 13 Ohms each...probably need about 2k. Thank You Sir!
Hey Tony, that driver is a current chopper so I'd use the highest voltage you can pump into it up to its max. The drive will limit the current to the motor depending on your settings. Also, you can use a relay type pre-charge circuit to limit the current with a power resistor until a specific voltage threshold has been reached. Then the relay will switch to the main line in.
Your videos are incredibly informative to us plebes/wannabes.
Thanks Joe!
Would love to see you break apart a tormach mill and tell us what you think. They are an employee owned company that has a little bit of Chinese help, but my mill has been stellar so far for what it can do per dollar.
I didn't read all the comments, but I would have added a hole through the aluminum heatsink/mount positioned over the capacitor vent valve. Otherwise, that right there is a compact rectifier/filter assembly, really clever design.
About big caps (capacitance not physical size) ..
bigger results in lower ripple but thats is a double edged sword, also results in the charge pulses beinng extremely short. Like you dropped a spanned accross the transformer terminals. The mean current draw for the power you are using is a given, but the shorter the duty cycle, the higher the charging peak current, this can easily be 10x the mean, is brutal on the diodes. The copper in the traffo has a finite resistance, and V=IR so the higher the charge current te bigger the voltage drop due to that.
These brutal ripple currents are also brutal on the capacitor. may resut in overheating, and for sure will reduce its operational life. resistores and transformers should last forever if operated in spec, but electrolytic cas have a very finite life and die of old age evenn if not used, and death is accelerated by high temperatures and big ripple currents. Shoulnd be an issue for occasional hobby use but is a concern in 50 year old antique equipment or anything used industrially so getting a hard life
Not an issue for your application relly but in a idealised world of perfection (ie the extreme end of audiopphile ampifier design) the ideal is thus to have a quite small cap after the diode bridge to get a longer charge angle amd lower charge peak ... then have an inductor from that to a bigger reservoir cap. A so called pi filter. This gives a calmed down load on the mains and bridge (lower switching noise) simultainious with very low ripple. But inductors cost money
Parralleling caps will lower the effective series resistance in the cap everything else being equal, but everything else is never equal and a single cap for times the value will have a different ESR anyway even if the same series , due to the different geometry. In reality in applicatios where we care about low ESR, these ays there are other types, of cap (notably organic dialectrics such as OSCON) that give massively better performance making parralleling moot as a way to get low ESR
Love the humour and also beautiful camera work / editting. Thanks BR D
I'm "civil engineer" myself (doing a PhD in signal processing systems), and I did not know about the regulated power supplies not being able to handle these back surges. I feel ashamed of myself.... Thanks for the nice insight
Well, the energy has to go somewhere. On drives with large loads, you will usually see a honking big braking resistor. It's either that, or you have to limit your deceleration to what the drive can handle on its own.
On a mill, I don't see a huge problem with shedding inertia. But on a lift that's holding up several thousand pounds things can get out of hand pretty quickly.
Just to note, the rectifier and capacitors do not add any voltage, it's just measured differently for AC. AC voltage it is measured in RMS Voltage. 120V RMS at the wall outlet actually has a peak voltage of 170. So wall AC varies between -170 and 170, rectified it would vary between 0 and 170. With No load attached, that would let the capacitors charge up to 170 Volts DC.
As every other idiot out there, instead of thanking you for one of the best material on TH-cam, I will point out that at the end of the video it says: "gracia por mirar" when it actually should have said: "gracias por mirar ( or ver)”.
I feel better now. Or worse.
@ThisOldTony
I would add a discharge resistor across the capacitor so it does not retain its charge - for safety.
Neither the classic PSU that you have nor the generic switching power supplies can't give excess energy (when you decelerate the steppers) back to the grid. Both of this power supplies have a capacitor at the output, tho the switching PSU cap is smaller because the frequency is higher - the pulses are more together. With no way to put the energy back into the grid when the current flows form the steppers to the PSU ("regenerative/generator braking") then the only thing that can happen is the voltage on the capacitor goes up. For the switcher (smaller caps) this means that the voltage overshoots will be larger. Meaning that the classic PSU is less likely to be damaged in this scenario. Also for this reason you did good to go with a 100V rated cap.
As always enjoying your videos !!! Can't wait for the next one:D
You are a very good teacher. Maybe keep around a little thermal grease?
10 thumbs down= AvE logging in under his TH-cam burner accounts to show his disapproval of you working with Diresta.
Moneypit330 ^^^ Totally plausible.
Nah, he's too busy starting his TH-cam war with Wranglerstar. HAHAHA!
LT1FirebirdSLP
Maybe they'll cancel each other out!
I gotta say, it was satisfying seeing the call out of that fawkin wranglestar video. That dude reminds me of that picture perfect TV preacher - the one that has 4 mansions, 2 private jets, and a yacht that he snorts coke out of hooker's ass cracks when he's on. I've been impressed with the quality strides that HF has made in the past year or two(those Vulkan welders are tits!), but that pile of video had some steam coming off of it...
That it's now 666/10 dislikes for some reason makes this feel more plausible
A huge improvement over the old power supply. It is a bit overkill for the realistic loads those motors will take but better to be over than under. Yes the bridge does NEED heatsinking. I would add a discharge resistor (say 4k7 1W) across the capacitor. 67v won't normally hurt but the impressive sparks if you short it will make you jump. A couple of smaller capacitors to help prevent conducted noise from the stepper drivers (they chop the current to the motors at several kHz). Say a 100uF and 1uF capacitor across the bigun will help prevent some of that electrical noise getting out to the big wide world. A common mode choke in the mains supply along with an X rated capacitor (across line to neutral) and two Y rated (from line and neutral to earth) capacitors across the mains will also help protect all the sensitive micro electronics in the other parts of the system from the high current switching noise. Oh and you need T rated (thermal) fuses for both the primary and secondary side of the transformer to survive the inrush currents. Strictly you could do a proper EMC design but just these simple measures will make a big difference. Keep up the great videos!
Fantastic video buddy!!! A Electrical engineer that is a machinist and designer too!!! Who of thunk it!!!!
Ay, ay ay!!!!!!!
Who knew a teeny little MP3 player has such a honkerin' big capacitor built in...great find!
Pretty cool you hog dawgin' that cold rolled steel....very nice!
Great video! Keep em coming!
regarding that capacitor bank question:
You should check the ESR of those 4 caps vs. the single large one.
The lower the ESR the better the current delivery capability.
Also maybe add a large Ceramic or Foil Capacitor.. you know these blocky ones.
They have different current capability as well in terms of lower resistance to filter out sharp quick spikes if your gecko drivers suddenly demand more current.
Makes for a smoother ride on your cnc router
Personally I'd keep the large one for some fun projects...
I think it is a bit of a waste to put it in a power supply...
Try charging it with your arc welder and zap some stuff
It's a lot more fun and a lot more dingus :D
A little advice on capacitor selection, go for the higher temperature range as those will have a lower ESR (Equivalent Series Resistance). Also, more smaller capacitors in parallel increases the surface total surface area and decreases the total ESR. Low ESR is a good thing for electronics.
I do like how you mounted that capacitor though. I kinda want one of those brackets now. XD
Also, them having different capacities would make them filter out different frequency ranges, for better overall filtering
My only thought would be to have a beefy 100k-1meg ohm resistor across the cap (if it isn't already) to drain the cap when the system is powered down just for a tidbit of extra safety. Brilliant work as always.
Dave jones/eevblog did a great episode about why a person would use 3 of a smaller cap compared to 1 bigger one; it comes down to BoM simplification/quantity discount efficiency as well as form factor awesomeness.
good beefy future proof linear psu good job. 2 small concerns; accidental zaps and corrosive leakage.
add a bleeder resistor to cap for future "mishaps". that size of cap will zap you twice or thrice AFTER u short the terminals!
and check the possible electrolyte leak path in your build. maybe unlikely but prepare for the worst and expect that big boy to leak serious amount of corrosive liquid. it will eat away stuff on its path incase of a failure. as a bonus u can also add fuse mov ntc ptc thermostat something something as u see fit.
Multiple smaller capacitors are sometimes preferred for reasons of cooling and inductance, but it might not matter in this application. The small transformer probably survived until enough shmoo built up on it to be a cooling issue. Transformers can handle up to about 300% rated power if they are kept cool.
I can't believe he didn't add a smirk when he said "I don't know if one big one is better than a bunch of smaller ones" with a reference to Ms Tony.
great video Tony......keep up the great work.
When I made my plasma table i had bought a 24v switching power supply so i had thought i needed to run 24v steppers however I picked up some slo-syn branded 6v steppers and they have been working with no problem. As long as your amps are covered you shouldn't have a problem
I think it's better to have multiple capacitors in parallel rather than one big one. Lower ESR that way.
Does ESR matter in this instance?
Only if you really care about efficiency. But it is true that several Cs in parallel can be better than just one large capacitor. And if you want to filter out some of the nasty frequencies it is a better idea because capacitance and frequency are related. Current controlled drivers can produce some nasty transients of high frequency (in this case anything in the range of kHz is HF) and having several caps of various capacitances can help solve this. Mainly at the fundamental frequencies of the drivers and some other harmonics.
Higher ESR = more heat generated. Possibly you could get more life out of your caps.
@ Euro Bum
Yes, you are right on the heat generation issue, but this is by far a high performance high end application. That capacitor will last a very long time before is gone, probably more than his milling machine. The EMI issues are to me far more concerning in this case. Someone bellow gave very good advice on what will make a better design, including the use of capacitors close to the drivers and moving the big "tanky" boy away from heat sources.
And if he ever finds that there are lighting glitches in the home or a wall clock gets crazy when the mill is on, I would add some ferrite beads to the power wires on the mill and the drivers. Like those that can be seen on power cables for printers. If everything is OK, no worries I wouldn't bother to improve anything and will just move to make cool stuff on that mill.
All things being equal...
(... things were not equal)
A few thoughts and answers to your questions: Using multiple small capacitors will result in a better filter because they will have lower ESR (equivalent series resistance) - less ripple. The difference is probably not worth worrying about when driving motors. This is unregulated, after all.
The enemy of capacitors is heat. It is the primary driver of service life. Mounting one on a heat sink is not good engineering practice. A high value resistor across the cap to keep you from getting bit when you work on it is always a good idea for large caps.
When you increase your motor drive voltage by 25%, you're going to increase the motor current draw by 25%. So instead of designing for 12A and 880W, you've got 15A and 1005W, though your 67V will sag some under heavy load, so it won't be quite that bad. That extra current (and the heat it causes) is the main hazard to running the motors at a higher voltage. The fact that the motors aren't used at 100% duty cycle reduces the impact of this a bit.
EET here with some real world experience, what you did is not totally wrong but here is some information. More small capacitors is always better. the ESR (Equivalent series resistance) rating of them is much lower that way using multiple smaller capacitors. they filter the ripple out much better. and a larger rated diode will not create less heat. there will always be a 1.4V drop across them (0.7V per diode)
now I've worked on large machines that used very powerful stepper motors for positioning. they used the same power supply you've just made. Almost exactly. the power supplies never turn off and they were made in the 70s. most have never needed parts replaced until recently.
The updated power supplies we are switching to slowly just use more but smaller capacitors.
heres what you should have done. instead of using a $60 transformer and a $80 capacitor you should have just gotten a switch mode power supply. they are much more efficient, dont hammer your electrical in your house when you turn them on, and they are voltage adjustable. on top of all this they are dirt cheap. word of caution. use an appropriate fuse on the AC going in if you buy cheap. also if you buy cheap do not use them on hyper sensitive electronics. they have a ripple in the thousands of Hz(s) they must be made higher quality or use a supply like yours. that said 3d printers use them with no problem. anyways this could have been done for $20-30 or around 100 if you get the better ones
after finishing the video the stepper motor drives almost always have snubbed diodes to handle the spikes coming off the motors. If you are still worried about it you can get a beefy schottky diode and put it in reverse polarity on the power supply they are crazy fast switching. if for whatever reason the reverse polarity voltage spike is fead back to the power supply it will snub it out. the switching power supplies should have them already but sometimes they are under rated for massive spikes. an added beefy one will help them out in such an event. still the stepper drivers should have already accounted for this. usually its the reverse biased diode built into the switching mosfets. normally you will not need to do this unless you are trying to quiet down the noise created by a brushed motor or are using an inductor to help filter power supplies
Where are you getting an 800W SMPS for $20????? $100 would be a good price unless you find used or surplus. Few caps work better as a cap but in this case that coupled with the big transformer could cause other problems with surge current.
if you check around you will find them. usual suspects like ebay
Always amazing videos, thanks again for taking us along for the ride!
Good job, and very festive music! First video I watched after breakfast of huevos rancheros, which I'm sure will keep reminding me of your transformer all day, as I digest this...cheers!
The transformer is way oversized. I would need to do some calculations, but I highly doubt that you will ever get close to the 800W, not speaking about 1500W. The transformer will be able to handle momentary overload, I guess that's why you could get away with the small one in the first place. But bigger is better!
The very large capacitor will create a very high inrush current, when you plug this thing in. However, again the transformer will be able to handle this, as the very high current will only be present for a couple of ms. Generally speaking having multiple capacitors in parallel helps with the parasitic inductance and resistance. You should add a few smaller caps close to the stepper drivers. Exact value doesn't matter as long as they are as close as possible to the stepper drives.
Mounting the large capacitor on the same heatsink and above the rectifier isn't a good idea. Electrolytic capacitors will see accelerated "wear" with increased temperature. For this reason you should mount your capacitors as low as possible in the enclosure and if they are in the airflow you use for cooling, make sure that they are the first component in that airflow, so that they don't get warmed up by warm air.
The voltage rating is probably ok, for industrial gear you often shoot for twice the actual voltage to get a proper lifetime out of these.
"The very large capacitor will create a very high inrush current, when you plug this thing in. However, again the transformer will be able to handle this"
But the rectifier might have a problem with that. Overkill is not always good...
But moar bigger equals moar better, right?
Me encanta este video con una pasion tan fuerte! Craig
I guess your the only person who makes his own heatsink :)
Skaza, sorry there are quite a few of us out here who make our own.