I have a 12v 80mm fan, super chinesium, same design, i pushed it to about 55V, it didnt die, but it started overheating above 20-22V. I dont think it died from overvoltage as the chip seems quite resilent but im sure it died from being overheated for so long
This is not actually a bug, but a feature! The MCU detected a fan failure so it blew up the heater while you were in the vicinity so it wouldn't burn your house down in case you left it plugged in.
I still think that the lack of airflow meant that the thermocouple was measuring possibly 1 hundred degrees or so less than the actual temperature. The other factor might be the magnetic switch when you put the thing down - because if the fan failed, you might think it had switched off when seated back in the cradle. You wouldnt be aware that what had happened is the fan died, and the station thought the tool was still being used (ie. not disabled by the reed switch - I often find I have to reseat mine a few times to get it to register that I've finished using it). That coupled with the temperature measured as much less than actual temperature perhaps caused ramp up in current provided to the element in order to compensate as it thought it was too low... In any case a bad design, these things need to consider failure of each individual component and how it might be fatal.
This is exactly what happened, the thermocouple isn't close enough, the heating element get hot from it's center, you can see it by where the metal gets red first, the distance is big enough so the set temperature will never be reached.
I find DaveCAD works better with the graphite powered human interface device. The rolling-buffered Ink-HID often enables write-protection accidentally. [edit: and yes the last one is also write once, read only]
Guys... Everyone touts DaveCAD as some great, free, open source CAD tool, but don't y'all realize it's free because of _micro-transactions!_ It's totally true! The software might be free to play, but you have to keep buying the DaveCAD media! That's where the paper industry makes all it's DaveCAD cash! I've seen prices from $0.95 a piece, all the way up to big lootboxes (with random colored media) that cost $15 and up! Crazy, I tell ya! :P
The MOC opto-coupler gate driver fails as well. Just this week I replaced a TRIAC that was conducting without external signal and that fail to correct the problem, it was only when I replaced the MOC3041 that the behavior went back to normal. Testing the whole driver circuitry is a good idea...
This is my next suspect as well. Is there a signal on the isolator's output? Input? MCU? Does the signal from the thermocouple make it to the MCU's input? Are the thermocouple's wires shorted or open? Any of those things could cause the triac to get stuck always-on. Given that fans rarely quit out of the blue under normal circumstances, maybe an over-voltage killed the fan and other stuff on the control side.
If your main AC is 230V than it will happen again, in Germany standart is MOC3063, because MOC3041 can't handle the peak voltages, but many manufacturers still use MOC3041, don't know why, there is no difference in price.
But there is no need for all that, because the problem that caused this all is what he says in the first seconds of this video, but dismisses it, because he is over-qualified to know the heating element is heating from it's center and without airflow the heat on the thermocouple is ways lower, so the system will keep on heating but never can reach the set temperature. So instead to resolder the two wires of the fan, check it's transistor and replace the housing of the handle, he decided to screw the whole station completely.
@@Wilson84KS I will say it again, MOCs are fragile and can fail in a way that keeps the TRIAC on all the time. Without proper checking the whole driver circuitry you can't say that's the fan that caused all that mess.
I don't think it's a SW bug. The SW knows temperature of the air going out but it don't know about air flow, presure or fan speed. And the air don't conducts the heat very well. The SW thinks "OK, it have 200 degrees C, user wants 350 so I put it to the full power to heat up". No way to detect somethink is wrong. It is definitely HW bug - missing thermal fuse as a last safety components in case of the failure (how much it costs in mass production?), missing feedback from the fan (as used on PC CPU fans), use HAL sensor above specified voltage.
Hi Dave, This testing circuit will not work properly ... You will feel it is going from ON to OFF suddenly .... The reason is A1 and A2 ... A1 should be down and A2 is up ... The gate is referenced to A1 not A2 so the voltage should be higher than A1 to trigger properly the Traic .... Good Luck
10:42 - _"We can control the turn on and turn off time of this TRIAC."_ You can only control the turn on time. The TRIAC will continue conducting until the "end" of the phase when the current becomes too low. Oops... ;-P
I love ya Dave! You always give me a chuckle. When I finally do make it to australia, (been wanting to go for 35 years) I definitely want to buy you a pint!
If I owned it, I’d definitely be adding a bi-metal thermal overload in series with the heating element... couple dollars and 5 minutes work might save your house or life...
I still dont think there was any failure other than the fan, the rest of it could still function correctly and overheat to the point of melting the plastic. I suspect that if you just put the mains power on the heating element without any control the element would burn out very quickly possibly before the thing got so hot as to melt the plastic that much.
Looks like some very loose windings on one of the coils in the fan, did one break away? I'm guessing the handset was thrown down into its rest, which dislodged one or more of the fan coils, and likely damaged the thermocouple. I believe if the thermocouple goes open circuit, the controller will just deliver 100% power to the heater constantly. The drive circuit is quite similar to the one in my cheap chinese desoldering gun, which was shipped to me with a broken heater, and the seller shipped me a replacement heater element (with built in thermocouple). On installing the new heater, i failed to notice that one of the precrimped thermocouple wires was not stripped properly, and crimped down on the insulation, making it open circuit. When i turned it on the thing glowed white hot. I yanked power and attached the thermocouple properly and it works great.
Should have had a three wire fan. The third fan wire for sense or feedback. The heat element would not have been allowed to come on or remain on without receiving the Fan sense or feed back signal from the fans third wire.
Great video! Interestingly, the vacuum tube (valve) equivilent of the SCR was called a thyratron which may be where the term thyristor was derived from. I worked on a radar system the used hydrogen thyratrons - yes a tube filled with hydrogen gas - to switch high voltage DC ( 15 Kv @ up to 600 amps). The tube was manufactured by EEV, an English valve manufacturer.
IF the TRIAC failed that requires more failures to also happen to cause this. The fan does *NOT* stop until the sensed temperature is below 100° (I believe. At least on my 858D+.) thus the fan would continue going full speed (whatever it was set to) until it has cooled down. A thermocouple error is a different story and could cause this. Note that even when set to the max temperature of 450° AND max air flow that the heater is not on continuously. Now consider if the flow was set low but the thermocouple is now reading low (and thus never getting to the set point) how hot it will get with 100% heater duty cycle. The situation gets even worse if a small nozzle is installed. I believe this is what may have happened.
Maybe. Could have been the (incorrectly specced) supply voltage, or maybe some form of SCR latchup (ironically) that could have killed the chip, or something else. The diode and RC filters would be there for a reason.
Dave, why didn't you just connect the whole station as it was to the heat gun, just with the heater replaced with a light bulb? Then you could check if it controls the fan and reacts on thermocouple temperature.
i'd replace it anyway, even if it was just a hand unit failure. most of the triacs in these aren't too expensive, so it just makes sense to make sure its going to be okay. with a new handheld bit, it should be right as rain.
It could have been the control loop. I have played with the firmware and replaced the micro on on my atten 858d+, and I remember even with full airspeed, full trottle on the heater ramps up the temperature really quick. If something goes wrong with the regulation it will go red hot even with the fan on, i think. I bought some thermal fuses, but found no good way to mount them in the heater module.
I have a version of this. The iron and rework station actually work quite well. But I had to fix things before use (like they fused the ground instead of mains) and I installed a quality mains switch from Mouser. Still, I never leave it plugged in, because I don't trust it 😆
@@ernieschatz3783 Watch the _other video_ before commenting, he said he "measured that fan and it showed open circuit", suggesting he used continuity mode on a multimeter. But that wouldn't be representative, since the fan is driven by an IC the multimeter's low voltage wouldn't suffice to make it start doing anything. So when testing that way it could very well be showing "open circuit" even if it worked fine.
Make sure to only download original DaveCad™ software from the vendor website. You can check the authenticity of your product by examining the logo, it should have a letter D with a small face drawn inside of it.
Should have tested the hall sensor to see if that was the fail point. I might have to change the fan in mine to avoid a similar failure even though it is a later model it won't hurt to check.
Why not to make some dummy heat gun to test if there main unit has something wrong? I mean with high resistant you should be able to demitre if triac works. I also think that you have such nice adjustable voltage source that you could simulate thermal voltage of thermal element.
It is likely that without the resistors and capacitors installed as specified in the data sheet, transient voltage spikes hit the driver chip and popped it.
Am I the only one who learns something new about electronics, immediately feels like I'm a genius cause its just crazy amazing, then immediately feels like a moron cause this is the BASIC, elementary level stuff AND this is stuff that's been around for about 250years already?
Dave, I downloaded and installed DaveCAD, but Im having trouble getting my Papermate graphics and line tracing hardware to work. Any ideas? Are there additional plug-ins I need? Also I updated to Inkjoy, if that helps.
i recently decided to buy a "my first" rework station and glad i decided against this one. i noticed about 5 different "brands" with this same form factor and thought twice about it.
It could still be a high voltage break down, I have seen high voltage transistors fail only under high voltage and tested OK at low voltage, drop 240V onto it and see how it goes, this is exactly why I built my high voltage leakage tester project: th-cam.com/video/_N-gh16ADH0/w-d-xo.html
Did you check out the reed switch to see if it was damaged? If the fan failed, the heat could have risen enough to thermally cause the reed to open, maybe? If so, then that switch opening would signal the unit to turn on the heating element
My money is on just a fan failure not providing enough heat over the sensor to control the heater , if you hook up the heater wrapped inside something to simulate the original wrapping and check the sensor I would bet you`ll find the sensor does NOT operate like it would with air flowing over the unit Sometimes you over complicate the simplest things
hey dave, yesterday i watched ur tutorial playlist, and i found this video unlisted for some reason in the playlist tutorial, just want to let you know
I've had a TRIAC fault where there was a partial short from gate to anode which was enough to keep it switched on regardless of the gate signal. There was no short from Anode to Anode.
YT had a video replay from a security PVR in someones house showing one of these cheap tools catching the bench on fire. I tried searching for it but couldn't find it... could someone post the title, if they've seen it? The owner just made it to put-out the fire.
Can ya test the coils? I watched till the end to see if the coils were actually bad, or a transient took out the hall effect chip. Wonder if the filtering circuitry that was left out was responsible for filtering out any reverse transients from the coils somehow.
Fan fail - I wonder if the fan wires are cheesy aluminium and one has gone o/c maybe - you could stick 12v directly on the fan board and just check it still doesn't spin up Dave (I noted the comments above about fan chip voltage range also)......But it seems odd that we have two failures happening at the same moment here, so I wonder if something in the power supply has gone intermittent and is knackering multiple things, like fan and triac drive/control loop at the same time ?
I have a cheapest&noname 858D station and there're not 24V on the fan. It regulates within about 4-17V through linear regulator. Btw, the fan is marked as 24V. I believe the IC inside is the same as your one, with 3-16V power supply. So, probably you killed it instantly☻ I wonder why they mark it as 24V? Probably because the entire thing is made of cheapest crap, so they're using some standart rotor, that's just cheaper I suppose and no one cares about its marking.
@@EEVblog Anyway, there're no fan fault detection.The MCU can only turn it off and on. The fan does not have hall output for rpm detection, only power rails and nothing else. And there're no thermal fuse near heater. So I guess in normal condition closest side of the heater is cooling by airflow and only because of that the plastic around is not melting. If the fan fails, the whole heater became hot and plastic starts to melt. That's your case. And you're lucky! In case of TRIAC fail, the heater will became red in seconds and the plastic around will start burning☻ (say hello to Lukey solder stations!☻) Please look at this video and you will understand what I'm talking about: v=m5kzyUFb2Sg
its not uncommon in 3D printers to have the thermister fail open and the MCU attempt to heat the heater back up and cause thermal runaway thats why under temp protection is so important
On a 3d printer with the proper firmware, the printer will refuse to heat if it can't sense the temperature, if the temperature is too high or too low.
unfortunately this feature tends to be disabled to prevent false triggers another even lesser used feature is the hotend not heating up when it should be to indicate a thermistor or heater cartridge that has fallen off all fairly simple to implement in anything like this
Anyone know why the first thing checked was not the thermocouple and its wiring back to the processor? If the thermocouple actually failed first and it thought it was cold, would that not have shut the fan off? I know the fan is dead now but i would imagine a vast proportion of the heat generated after the initial fault happened would have exhausted out of the fan side as it is the pass of least resistance as it has bigger holes than the nozzle.
Test the thermocouple. I had a weller that wound up turning the power on to the iron full time and it turned out the thermocouple had failed open circuit.
The visible damage is a result of the failure. It seems that you have not yet found the "why" pointing to the root cause. Several more "why" cycles are probably needed...
He mentions it right at the beginning but is obviously too smart to realise that without airflow the gap between set and is temperature id high enough so the set temperature can never be reached, he just say the thermocouple is close enough and dismiss this possibility, because also too much engineer to know the heating element is heating from it's center, right there where the metal pipe begins to glow.
Waiting for Triac-testing part 3 since there was some issues with the testing procedure. Also want to see measuring the coils of the fan. There are so many things that you exclude in the videos that at least i want to see when diagnosing. Still i have no evidence what went wrong. Just some guessing.
Let's replace the fan and see if the unit works, if it doesn't work then my first target would be the heating element power drivers, after that the thermistor and finally the micrcontroller .
Yes....also my first reaction....measure the Fan coils. 😏. By the way..i have the same only Atten heatgun station for years....still working fine....i would like to buy an expensive Weller....but then i have to put and build an extra fuse on it.... 😏
Can't you get spare guns? You could have fixed it. They should only cut the heating element and it should not be via reed switch it should be an arm on the side of device with sprung microswitch. Or at least put the reed switches in several positions or put them in the station and magnet on the gun. if fan stops working there should be a detection and standby with warning lamp for fan fault. Just bought a most likely fake quick ali express special €135 857DW
So, does this mean the driver failed, but the motor is fine? I know you hooked it up the power supply, but if you hooked up the motor bypassing the driver would it work? Curious, not sure how these work.
Here's an off-topic comment: Wow, TH-cam gets really annoying! With uBlock Origin set as a "general cross site request blocker" and then allowing all the different youtube and googlevideo domains for the site to actually work, it really pushes in unwanted video content a lot! I don't want to be innundated with crap (every advertisement is crap, that's a general rule which never gets broken), and yet I have to reload the page 3 to 4 times and jump back to the time code where the video that I actually wanted to watch was interrupted by unwanted bullshit. I wish it were possible any other way (but then, paying isn't possible either because you'd lose your anonymity towards Google, which is an even bigger no-go). Some aspects of this modern world really are to be hated whole-heartedly.
I do wonder if there may be something wrong with the position of the thermocouple. Because the way it is mounted it may just look at the temperature of the air around it, and not the temperature of the heating element. So if tried to keep the temperature of the air but if there is no airflow (if the fan breaks or ...), then it would overheat the element and still think that the temperature of the element is ok.
Yes, that's it, the difference between set and is temperatures is without airflow big enough so the set temperature will never be reached, the heating element heats from it's center when there is no airflow, clearly visible by the glowing on the metal pipe, so the difference is huge.
don"t assume the fan was fed with 24v !, also are the coils in fan have continuity to "ic" driver?. If coils ok measured at driver ok then sus ic driver failure (knock on effect)
The heater assembly should have a one-shot or self resetting mechanical thermal cut-off placed near the middle of the heating coil body where it would do the most good. Pretty much any 'approved' heater would have this design. They can be as small as the reed switch they used. Even bargain basement hair driers have one. As it was built to a super low price, I get leaving it out, but would have saved a lot of unnecessary excitement. But then we don't get a laugh and a Dave video.
The failure mode with these is quite simple. The fan is individually controlled and there is no feedback. The thermal regulation is done by a microcontroller using the couple and triac. When airflow stops, the airflow from the heating element to the thermocouple also stops, making the thermocouple read low less, causing the heater to switch fully on. Since the heat will never quite reach the thermocouple to get it to the set temperature, this thing will go full meltdown. All it would have needed is fan rpm sensing and maybe even a thermal fuse in the heater block.
I would guess that device is reliant on the fan to reach thermal equilibrium and then when the fan fails or is "stuck off" it has no working feedback to know the fan has failed. I noticed that the fan only has 2 wires so likely no tachometer feedback. I also suspect that the safety feedback devices do not actually work or it is "pot luck" if they are actually connected to anything on the circuit board.
Are these devices not designed to antisipate situations like a fan stopping. Maybe there should be a tiny tachometer attached to the fan to give feedback to the microcontroler.
Inside the motor driver there’s a diode in series with the VCC. You checked it with an ohmmeter. The diode might not conduct with the ohmmeter. Did you test it with power?
At which temperature? In normal use they get pretty hot to get the airflow at 500 degrees even with an environment of less then 10 degrees C. Glowing-hot I guess on the element, actually you see it glowing quite clear when they are in normal use. A hairdryer is just getting 'warm' compared to this. If you had one, where to install? Maybe at 2000 degrees C, close to the core but in which direction? Before it reacts when it is on the bottom of the heater, much plastic will have been melted already, probably shorting out the contacts or making a connection with the mains to the metal (death trap for electrocution) or you have to get a real expensive one. And that will be the main reason... heating handle inclusive transport over the whole world is about 5 -10 Euro's -not much room for expensive safety-things...
Sometimes triacs/SCRs and cmos logic can have junction latchup that will clear itself once it’s fully powered down. It’s rare, but they can fail to fully shut off after the zero crossing occurs... I’m not too sure how to test for that fault though... seems to be a scary horrible design if it burns your house down if the cheap ass fan isn’t flowing enough air... it would be interesting to test just where the thermocouple actually shut the thing down... he said it was glowing red to white hot so surly the thermocouple was pretty damn hot...
Hi Dave! I did some work on TRIACs. One very common reason why they stop turning off is if they are on square wave power supply like old style inverters or UPS. They need time to turn off which they get on sign wave as it goes through zero crossing. There is almost 1ms time during 50Hz power line. While 50Hz square wave do not give this time for TRIAC to turn off! It do not damage the TRIAC but give 100% duty cycle to the load! Can you verify the state of power supply at the time it went kaboom!
QUESTION ! ! ! The bearing in the center of the motor at the end, was that "wallered" out by wear or did it get warp by the teardown? Worn bearing in motor cause it to hang up on something prevent it from turning? My guess actual cause was motor but it failed mechanically not electrically. Set it down but instead of sense it off it actually locked the motor. Sorry I know bad wording.
:) I have on my table right now a fan control unit that uses triac to control fan. And its weirdest possible malfunction. It works fine up to certain fan speeds. After that it goes crazy. :) Atmel mcu controls triac. I have no idea what the ef.... is happening.
I have some Stagelighting-Dimmerpacks that utilize the BTA16 and I must say everytime a Dimmerchannel blew up on me it resulted in the Triac staying on, dead short leaving the lights at full intensity. It’s an easy fix though by simply changing the damn thing but still every single one of the countless times it has happened failed in its on-state
![How A Turn Signal Flasher Blinker Works - Complete Redesign With Schematic [Tung-Sol P229D]](http://i.ytimg.com/vi/3ADoODcApNE/mqdefault.jpg)
3.5-16v supply voltage ic , on a 24v fan together with the missing RC filter and diode, i wonder that not more burned up like that
LOL, didn't notice that!
@@EEVblog no wonder it failed lol
I have a 12v 80mm fan, super chinesium, same design, i pushed it to about 55V, it didnt die, but it started overheating above 20-22V. I dont think it died from overvoltage as the chip seems quite resilent but im sure it died from being overheated for so long
I was wondering the same thing
@@EEVblogIs that white thing (in the fan) not a fuse? If you trace it out that's pin 1 (+) of the chip, but prolly the chip is gonesky
This is not actually a bug, but a feature!
The MCU detected a fan failure so it blew up the heater while you were in the vicinity
so it wouldn't burn your house down in case you left it plugged in.
Being white hot was the safety feature so anyone could see it was failing.
make video about triac and teach us young players!!!
I still think that the lack of airflow meant that the thermocouple was measuring possibly 1 hundred degrees or so less than the actual temperature.
The other factor might be the magnetic switch when you put the thing down - because if the fan failed, you might think it had switched off when seated back in the cradle. You wouldnt be aware that what had happened is the fan died, and the station thought the tool was still being used (ie. not disabled by the reed switch - I often find I have to reseat mine a few times to get it to register that I've finished using it). That coupled with the temperature measured as much less than actual temperature perhaps caused ramp up in current provided to the element in order to compensate as it thought it was too low...
In any case a bad design, these things need to consider failure of each individual component and how it might be fatal.
This is exactly what happened, the thermocouple isn't close enough, the heating element get hot from it's center, you can see it by where the metal gets red first, the distance is big enough so the set temperature will never be reached.
I'm using DaveCAD and so far I love the software! Simple, clean! No whackers!
Simple yes. but sadly nu updates.......
I find DaveCAD works better with the graphite powered human interface device. The rolling-buffered Ink-HID often enables write-protection accidentally. [edit: and yes the last one is also write once, read only]
@@muppetpaster Don't need em, no bugs, every needed feature included...
@@juliandahl1999 Oh beware....it might become obsolete! hahaha
It requires a high quality perefiral, mine one is glitching all the time.
Guys... Everyone touts DaveCAD as some great, free, open source CAD tool, but don't y'all realize it's free because of _micro-transactions!_ It's totally true!
The software might be free to play, but you have to keep buying the DaveCAD media! That's where the paper industry makes all it's DaveCAD cash!
I've seen prices from $0.95 a piece, all the way up to big lootboxes (with random colored media) that cost $15 and up! Crazy, I tell ya! :P
You wouldn't believe my commission
*Sent from my private jet flying to my private island*
The fan chips data sheet said max 16v and somebody gave it 24v - Ooops! That YHFS fan was NOT labeled correctly...
Yep, bad labling. Not sure what voltage they actually drive it at though.
The MOC opto-coupler gate driver fails as well. Just this week I replaced a TRIAC that was conducting without external signal and that fail to correct the problem, it was only when I replaced the MOC3041 that the behavior went back to normal. Testing the whole driver circuitry is a good idea...
This is my next suspect as well. Is there a signal on the isolator's output? Input? MCU? Does the signal from the thermocouple make it to the MCU's input? Are the thermocouple's wires shorted or open? Any of those things could cause the triac to get stuck always-on.
Given that fans rarely quit out of the blue under normal circumstances, maybe an over-voltage killed the fan and other stuff on the control side.
If your main AC is 230V than it will happen again, in Germany standart is MOC3063, because MOC3041 can't handle the peak voltages, but many manufacturers still use MOC3041, don't know why, there is no difference in price.
@@Wilson84KS My main AC is 127V
But there is no need for all that, because the problem that caused this all is what he says in the first seconds of this video, but dismisses it, because he is over-qualified to know the heating element is heating from it's center and without airflow the heat on the thermocouple is ways lower, so the system will keep on heating but never can reach the set temperature. So instead to resolder the two wires of the fan, check it's transistor and replace the housing of the handle, he decided to screw the whole station completely.
@@Wilson84KS I will say it again, MOCs are fragile and can fail in a way that keeps the TRIAC on all the time. Without proper checking the whole driver circuitry you can't say that's the fan that caused all that mess.
*hehe* Hardware guys are always blaming the software ;)
I don't think it's a SW bug. The SW knows temperature of the air going out but it don't know about air flow, presure or fan speed. And the air don't conducts the heat very well. The SW thinks "OK, it have 200 degrees C, user wants 350 so I put it to the full power to heat up". No way to detect somethink is wrong.
It is definitely HW bug - missing thermal fuse as a last safety components in case of the failure (how much it costs in mass production?), missing feedback from the fan (as used on PC CPU fans), use HAL sensor above specified voltage.
Hi Dave, This testing circuit will not work properly ... You will feel it is going from ON to OFF suddenly .... The reason is A1 and A2 ... A1 should be down and A2 is up ... The gate is referenced to A1 not A2 so the voltage should be higher than A1 to trigger properly the Traic .... Good Luck
Why ? Have you tried that ? I tried that many times and this is my conclusion
@Marcin Berman very funny
ما شاء الله ربنا يبارك فيك معلمي الفاضل
@Marcin Berman wut?
10:42 - _"We can control the turn on and turn off time of this TRIAC."_ You can only control the turn on time. The TRIAC will continue conducting until the "end" of the phase when the current becomes too low. Oops... ;-P
I love ya Dave! You always give me a chuckle. When I finally do make it to australia, (been wanting to go for 35 years) I definitely want to buy you a pint!
If I owned it, I’d definitely be adding a bi-metal thermal overload in series with the heating element... couple dollars and 5 minutes work might save your house or life...
I still dont think there was any failure other than the fan, the rest of it could still function correctly and overheat to the point of melting the plastic. I suspect that if you just put the mains power on the heating element without any control the element would burn out very quickly possibly before the thing got so hot as to melt the plastic that much.
Looks like some very loose windings on one of the coils in the fan, did one break away?
I'm guessing the handset was thrown down into its rest, which dislodged one or more of the fan coils, and likely damaged the thermocouple. I believe if the thermocouple goes open circuit, the controller will just deliver 100% power to the heater constantly.
The drive circuit is quite similar to the one in my cheap chinese desoldering gun, which was shipped to me with a broken heater, and the seller shipped me a replacement heater element (with built in thermocouple). On installing the new heater, i failed to notice that one of the precrimped thermocouple wires was not stripped properly, and crimped down on the insulation, making it open circuit. When i turned it on the thing glowed white hot. I yanked power and attached the thermocouple properly and it works great.
Should have had a three wire fan. The third fan wire for sense or feedback. The heat element would not have been allowed to come on or remain on without receiving the Fan sense or feed back signal from the fans third wire.
A silly question:did you check the optocoupler?
Great video! Interestingly, the vacuum tube (valve) equivilent of the SCR was called a thyratron which may be where the term thyristor was derived from. I worked on a radar system the used hydrogen thyratrons - yes a tube filled with hydrogen gas - to switch high voltage DC ( 15 Kv @ up to 600 amps). The tube was manufactured by EEV, an English valve manufacturer.
IF the TRIAC failed that requires more failures to also happen to cause this. The fan does *NOT* stop until the sensed temperature is below 100° (I believe. At least on my 858D+.) thus the fan would continue going full speed (whatever it was set to) until it has cooled down.
A thermocouple error is a different story and could cause this. Note that even when set to the max temperature of 450° AND max air flow that the heater is not on continuously. Now consider if the flow was set low but the thermocouple is now reading low (and thus never getting to the set point) how hot it will get with 100% heater duty cycle. The situation gets even worse if a small nozzle is installed. I believe this is what may have happened.
I wonder if the motor drive IC eventually failed because they didn't add the RC filters?
Maybe. Could have been the (incorrectly specced) supply voltage, or maybe some form of SCR latchup (ironically) that could have killed the chip, or something else. The diode and RC filters would be there for a reason.
Dave, why didn't you just connect the whole station as it was to the heat gun, just with the heater replaced with a light bulb? Then you could check if it controls the fan and reacts on thermocouple temperature.
Because that is a repair mans mind - Dave, brilliant as he is, just hasn't got one of those as shown on many of his 'repairs'.
i'd replace it anyway, even if it was just a hand unit failure. most of the triacs in these aren't too expensive, so it just makes sense to make sure its going to be okay. with a new handheld bit, it should be right as rain.
Isn't the temperature sensor a thermocouple, not a thermistor ?
Did I say thermistor? Oops.
@@EEVblog Yes, even a few times in a row...
It could have been the control loop. I have played with the firmware and replaced the micro on on my atten 858d+, and I remember even with full airspeed, full trottle on the heater ramps up the temperature really quick. If something goes wrong with the regulation it will go red hot even with the fan on, i think. I bought some thermal fuses, but found no good way to mount them in the heater module.
I have a version of this. The iron and rework station actually work quite well.
But I had to fix things before use (like they fused the ground instead of mains) and I installed a quality mains switch from Mouser.
Still, I never leave it plugged in, because I don't trust it 😆
3:12 Hi Dave, can you give us the download link for the nice and cute "DaveCAD" Software :-D
Did you actually try the fan with a power supply? You mention is as if you just tested continuity, but that's not representative...
Try watching the video before commenting
@@ernieschatz3783 Watch the _other video_ before commenting, he said he "measured that fan and it showed open circuit", suggesting he used continuity mode on a multimeter. But that wouldn't be representative, since the fan is driven by an IC the multimeter's low voltage wouldn't suffice to make it start doing anything. So when testing that way it could very well be showing "open circuit" even if it worked fine.
Best video for ages, ta Dave
Make sure to only download original DaveCad™ software from the vendor website. You can check the authenticity of your product by examining the logo, it should have a letter D with a small face drawn inside of it.
Should have tested the hall sensor to see if that was the fail point.
I might have to change the fan in mine to avoid a similar failure even though it is a later model it won't hurt to check.
For Pete's sake Dave look up the meaning of 'simplistic'. It doesn't mean what you think it does.
Why not to make some dummy heat gun to test if there main unit has something wrong? I mean with high resistant you should be able to demitre if triac works. I also think that you have such nice adjustable voltage source that you could simulate thermal voltage of thermal element.
Probaby the red fan cable was broken and the holder did not activate the reed switch that stops the heater.
Your wide-range precision resistance substituter has surpassed its re-calibration due date 😉
The magical smoke didn't escape, she'll be right.
It is likely that without the resistors and capacitors installed as specified in the data sheet, transient voltage spikes hit the driver chip and popped it.
Am I the only one who learns something new about electronics, immediately feels like I'm a genius cause its just crazy amazing, then immediately feels like a moron cause this is the BASIC, elementary level stuff AND this is stuff that's been around for about 250years already?
Dave, I downloaded and installed DaveCAD, but Im having trouble getting my Papermate graphics and line tracing hardware to work. Any ideas? Are there additional plug-ins I need? Also I updated to Inkjoy, if that helps.
well that advice is good,but bleeding! obvious. Trap for young players is forgetting tongue direction and depth!
i recently decided to buy a "my first" rework station and glad i decided against this one. i noticed about 5 different "brands" with this same form factor and thought twice about it.
It could still be a high voltage break down, I have seen high voltage transistors fail only under high voltage and tested OK at low voltage, drop 240V onto it and see how it goes, this is exactly why I built my high voltage leakage tester project: th-cam.com/video/_N-gh16ADH0/w-d-xo.html
Thats a crappy fan for that application!! No overheat melt fuse and the con trol system failed and the fan failed...
Did you check out the reed switch to see if it was damaged?
If the fan failed, the heat could have risen enough to thermally cause the reed to open, maybe? If so, then that switch opening would signal the unit to turn on the heating element
My money is on just a fan failure not providing enough heat over the sensor to control the heater , if you hook up the heater wrapped inside something to simulate the original wrapping and check the sensor I would bet you`ll find the sensor does NOT operate like it would with air flowing over the unit
Sometimes you over complicate the simplest things
All I have, had and will have on that.
Only Dave would substitute two decade resistance boxes for a mere potentiometer... 🤓 As Homer Simpson would say, “Doh!”
hey dave, yesterday i watched ur tutorial playlist, and i found this video unlisted for some reason in the playlist tutorial, just want to let you know
I've had a TRIAC fault where there was a partial short from gate to anode which was enough to keep it switched on regardless of the gate signal. There was no short from Anode to Anode.
YT had a video replay from a security PVR in someones house showing one of these cheap tools catching the bench on fire. I tried searching for it but couldn't find it... could someone post the title, if they've seen it? The owner just made it to put-out the fire.
Can ya test the coils? I watched till the end to see if the coils were actually bad, or a transient took out the hall effect chip. Wonder if the filtering circuitry that was left out was responsible for filtering out any reverse transients from the coils somehow.
You know you have just too much equipment laying around when it's faster to wire up 2 decade resistors instead of a simple pot.... :-)
Fan fail - I wonder if the fan wires are cheesy aluminium and one has gone o/c maybe - you could stick 12v directly on the fan board and just check it still doesn't spin up Dave (I noted the comments above about fan chip voltage range also)......But it seems odd that we have two failures happening at the same moment here, so I wonder if something in the power supply has gone intermittent and is knackering multiple things, like fan and triac drive/control loop at the same time ?
I have a cheapest&noname 858D station and there're not 24V on the fan. It regulates within about 4-17V through linear regulator. Btw, the fan is marked as 24V. I believe the IC inside is the same as your one, with 3-16V power supply. So, probably you killed it instantly☻ I wonder why they mark it as 24V? Probably because the entire thing is made of cheapest crap, so they're using some standart rotor, that's just cheaper I suppose and no one cares about its marking.
It should have still done something (even briefly) at 24V, but it didn't, it was dead.
@@EEVblog Anyway, there're no fan fault detection.The MCU can only turn it off and on. The fan does not have hall output for rpm detection, only power rails and nothing else. And there're no thermal fuse near heater. So I guess in normal condition closest side of the heater is cooling by airflow and only because of that the plastic around is not melting. If the fan fails, the whole heater became hot and plastic starts to melt. That's your case. And you're lucky! In case of TRIAC fail, the heater will became red in seconds and the plastic around will start burning☻ (say hello to Lukey solder stations!☻)
Please look at this video and you will understand what I'm talking about: v=m5kzyUFb2Sg
its not uncommon in 3D printers to have the thermister fail open and the MCU attempt to heat the heater back up and cause thermal runaway
thats why under temp protection is so important
On a 3d printer with the proper firmware, the printer will refuse to heat if it can't sense the temperature, if the temperature is too high or too low.
unfortunately this feature tends to be disabled to prevent false triggers
another even lesser used feature is the hotend not heating up when it should be to indicate a thermistor or heater cartridge that has fallen off
all fairly simple to implement in anything like this
Anyone know why the first thing checked was not the thermocouple and its wiring back to the processor? If the thermocouple actually failed first and it thought it was cold, would that not have shut the fan off? I know the fan is dead now but i would imagine a vast proportion of the heat generated after the initial fault happened would have exhausted out of the fan side as it is the pass of least resistance as it has bigger holes than the nozzle.
Would have liked to have seen an insulation test done on the small motor
Why? Ffs its an open circuit.
Test the thermocouple. I had a weller that wound up turning the power on to the iron full time and it turned out the thermocouple had failed open circuit.
The visible damage is a result of the failure. It seems that you have not yet found the "why" pointing to the root cause. Several more "why" cycles are probably needed...
He mentions it right at the beginning but is obviously too smart to realise that without airflow the gap between set and is temperature id high enough so the set temperature can never be reached, he just say the thermocouple is close enough and dismiss this possibility, because also too much engineer to know the heating element is heating from it's center, right there where the metal pipe begins to glow.
Waiting for Triac-testing part 3 since there was some issues with the testing procedure. Also want to see measuring the coils of the fan.
There are so many things that you exclude in the videos that at least i want to see when diagnosing. Still i have no evidence what went wrong. Just some guessing.
You just need to watch carefully.
Open circuit on fan.....
And the issues raised with the test are pointless.... the triac works...
Let's replace the fan and see if the unit works, if it doesn't work then my first target would be the heating element power drivers, after that the thermistor and finally the micrcontroller .
Many many thanks 😊 🙏
Power supply failure? Regulation failed and too high voltage fried the fan and logic? Maybe.......
Triacs don't work for dimming line voltage led lights generally. I had to get a new dimmer for mine.
Maybe it comes with a countdown algorithm. If the gate ic failed would that be the same as 100% power on?
Is it just me, or did anyone else see a few turns on the motor coil looked loose. Maybe that is where it failed!!!
What is the processor locked up, and it kept outputting a PWM signal without reading inputs anymore
what I am missing is a simple resistance test on the 2 coils of the stator. Did one fail ? we want to get to the bottom of this surely ;)
Both measure ok (100 ohms)
Yes....also my first reaction....measure the Fan coils. 😏. By the way..i have the same only Atten heatgun station for years....still working fine....i would like to buy an expensive Weller....but then i have to put and build an extra fuse on it.... 😏
Can't you get spare guns? You could have fixed it. They should only cut the heating element and it should not be via reed switch it should be an arm on the side of device with sprung microswitch. Or at least put the reed switches in several positions or put them in the station and magnet on the gun. if fan stops working there should be a detection and standby with warning lamp for fan fault. Just bought a most likely fake quick ali express special €135 857DW
So, does this mean the driver failed, but the motor is fine? I know you hooked it up the power supply, but if you hooked up the motor bypassing the driver would it work? Curious, not sure how these work.
You need the driver, it alternates the drive to the coils so it can spin.
@@EEVblog Gotcha. Interesting. Thx for getting back :)
The coils measure ok BTW.
Here's an off-topic comment: Wow, TH-cam gets really annoying! With uBlock Origin set as a "general cross site request blocker" and then allowing all the different youtube and googlevideo domains for the site to actually work, it really pushes in unwanted video content a lot! I don't want to be innundated with crap (every advertisement is crap, that's a general rule which never gets broken), and yet I have to reload the page 3 to 4 times and jump back to the time code where the video that I actually wanted to watch was interrupted by unwanted bullshit. I wish it were possible any other way (but then, paying isn't possible either because you'd lose your anonymity towards Google, which is an even bigger no-go). Some aspects of this modern world really are to be hated whole-heartedly.
I do wonder if there may be something wrong with the position of the thermocouple. Because the way it is mounted it may just look at the temperature of the air around it, and not the temperature of the heating element. So if tried to keep the temperature of the air but if there is no airflow (if the fan breaks or ...), then it would overheat the element and still think that the temperature of the element is ok.
Yes, that's it, the difference between set and is temperatures is without airflow big enough so the set temperature will never be reached, the heating element heats from it's center when there is no airflow, clearly visible by the glowing on the metal pipe, so the difference is huge.
Possibly thermal sensor or temperature controller. I don't speak English or translate google
A stupid diode missing can burn down your house, lab , whatever! Crazy!
Nice explanation Dave
don"t assume the fan was fed with 24v !, also are the coils in fan have continuity to "ic" driver?. If coils ok measured at driver ok then sus ic driver failure (knock on effect)
What driver?
Could a short in the fan disable the control loop by overloading the low voltage supply?
Would need to reverse engineer the circuit to see.
Dave measured the motor resistance with a multimeter it was open circuit.
It is now open circuit, but that doesn't mean, it failed short and then due high current something burned and turned open circuit.
@@morantaylor Well that pretty much settles that.
Yes, it adds the extra thermal resisrtance between heater ane the heat sensor.
The heater assembly should have a one-shot or self resetting mechanical thermal cut-off placed near the middle of the heating coil body where it would do the most good. Pretty much any 'approved' heater would have this design. They can be as small as the reed switch they used. Even bargain basement hair driers have one. As it was built to a super low price, I get leaving it out, but would have saved a lot of unnecessary excitement. But then we don't get a laugh and a Dave video.
The failure mode with these is quite simple. The fan is individually controlled and there is no feedback. The thermal regulation is done by a microcontroller using the couple and triac. When airflow stops, the airflow from the heating element to the thermocouple also stops, making the thermocouple read low less, causing the heater to switch fully on. Since the heat will never quite reach the thermocouple to get it to the set temperature, this thing will go full meltdown.
All it would have needed is fan rpm sensing and maybe even a thermal fuse in the heater block.
I would guess that device is reliant on the fan to reach thermal equilibrium and then when the fan fails or is "stuck off" it has no working feedback to know the fan has failed. I noticed that the fan only has 2 wires so likely no tachometer feedback. I also suspect that the safety feedback devices do not actually work or it is "pot luck" if they are actually connected to anything on the circuit board.
Are these devices not designed to antisipate situations like a fan stopping. Maybe there should be a tiny tachometer attached to the fan to give feedback to the microcontroler.
Inside the motor driver there’s a diode in series with the VCC. You checked it with an ohmmeter. The diode might not conduct with the ohmmeter. Did you test it with power?
Please make a video specifically for triacs, I don't wanna hear what other people has to say about traics other than you.
I don't understand why they didn't use a bimetal thermal switch to cut off the power when heater get overheated like they do in hair dryers!?
At which temperature? In normal use they get pretty hot to get the airflow at 500 degrees even with an environment of less then 10 degrees C. Glowing-hot I guess on the element, actually you see it glowing quite clear when they are in normal use.
A hairdryer is just getting 'warm' compared to this.
If you had one, where to install? Maybe at 2000 degrees C, close to the core but in which direction? Before it reacts when it is on the bottom of the heater, much plastic will have been melted already, probably shorting out the contacts or making a connection with the mains to the metal (death trap for electrocution) or you have to get a real expensive one.
And that will be the main reason... heating handle inclusive transport over the whole world is about 5 -10 Euro's -not much room for expensive safety-things...
just measure the coil resistance of the fan to determine whether the coil is internally broken or blown.
Reading the comments, i say that this requires more diving.
mine stop working something shorted, i determined it was my triac but it still won’t work after replacing it, what could be the problem? moc3041??
Hello my hot air gun failed a few years ago similar with what David story was and it turn out to be the optocoupler (it failed as a short)
TRIACs stop conducting on the "zero crossing" of the cycle
Hello, where should the oscilloscope probes be connected? to avoid short circuit
9:25 THere is fat chance that this is older than me XD
Intrested in the PID szoftwere insaide inthe chip what does when the fan fails?
what's about the triac in the "white" opto-triac ???
To To You mean phototransistor?
@@johnfrancisdoe1563 Yes
A fan without RPM monitoring has this side effect.
Sometimes triacs/SCRs and cmos logic can have junction latchup that will clear itself once it’s fully powered down. It’s rare, but they can fail to fully shut off after the zero crossing occurs... I’m not too sure how to test for that fault though... seems to be a scary horrible design if it burns your house down if the cheap ass fan isn’t flowing enough air... it would be interesting to test just where the thermocouple actually shut the thing down... he said it was glowing red to white hot so surly the thermocouple was pretty damn hot...
That part on fan may be not capacitor actually is fuse or resistor
Probably the fan, but could the heat coil get shorted to the metal shield and find the way to a closed circuit, maybe also fry the fan?
Are there any sensors to detect fanfailure ??
For this ic distance to the rotor magnet is critical
Hi Dave! I did some work on TRIACs. One very common reason why they stop turning off is if they are on square wave power supply like old style inverters or UPS. They need time to turn off which they get on sign wave as it goes through zero crossing. There is almost 1ms time during 50Hz power line. While 50Hz square wave do not give this time for TRIAC to turn off! It do not damage the TRIAC but give 100% duty cycle to the load!
Can you verify the state of power supply at the time it went kaboom!
I'll add my vote for a video on going down the rabbit hole on the workings of a TRIAC. Explaining the quadrant thing etc. Cheers
QUESTION ! ! ! The bearing in the center of the motor at the end, was that "wallered" out by wear or did it get warp by the teardown? Worn bearing in motor cause it to hang up on something prevent it from turning? My guess actual cause was motor but it failed mechanically not electrically. Set it down but instead of sense it off it actually locked the motor. Sorry I know bad wording.
:) I have on my table right now a fan control unit that uses triac to control fan. And its weirdest possible malfunction. It works fine up to certain fan speeds. After that it goes crazy. :) Atmel mcu controls triac. I have no idea what the ef.... is happening.
An the next video: let's rebuild
I have some Stagelighting-Dimmerpacks that utilize the BTA16 and I must say everytime a Dimmerchannel blew up on me it resulted in the Triac staying on, dead short leaving the lights at full intensity. It’s an easy fix though by simply changing the damn thing but still every single one of the countless times it has happened failed in its on-state