kWeld - "Next level" DIY battery spot welder
- Thread starter tatus1969
- Start date
ElectricGod
10 MW
flippy said:
The SE series shut off if you over load them. Is that what you mean by "current limiting"? OR does the HRP series actually cap current and keep on running?
Regardless of that fact, the idea is to actually get the super cap bank charged gracefully hence the resistor bank. The resistor bank cost me nothing, but a little time to solder together.
It is 100% probable that I'll use the PSU for several things and 12v is much more useful than is 7.5v. For example, I have an iCharger 4010duo. It is generally fine on a smaller meanwell PSU, but there are times when I charge that the other Meanwell PSU can't deliver the current. This PSU will see double duty for the welder and the charger.
9.44v is low enough for the super caps and 16.8v (top voltage for the PSU is high enough for the iCharger. I essentially kill 2 bids with one stone.
Many here know this stuff much better than me, wrote this up to clarify my recent consolidation of the ideas, any feedback welcome.
Current Limiting is an overloaded term.
The best kind is "adjustable current"when the power source lets the user explicitly adjust the maximum available, easily on a per session basis, as with Lab style PSUs with an Amps dial on the front, or industrial units controlled by e.g. CAN messaging.
Related is the various methods of OCP, and Mean Wells can be all over the place sometimes even within the same series.
Latching means OC shuts the power output off, may need to be manually power cycled to resume.
Hiccup mode is power output halts for a time, then resumes once components cool down. Sometimes full shutdown after X attempts.
"Current limiting" style OCP, means the output gets greatly de-rated but continues, when demand lowers, returns to normal output. This may involve dropping output voltage, sometimes so much below spec, that charging in effect is halted.
The device must be **able** to drop voltage low enough in order to continue, if not then hiccup or latching should kick in as a backup OCP.
Finally the abstract EE design concept of a "current source" can pertain, but IRL the above are all within the CC/CV operation implemented by all power sources suitable to use as battery chargers, even if classified as rectifiers, power supplies, DC-DC converters etc.
______
A useful PDF from Texas Instruments
Know Your Limits
http://www.ti.com/lit/an/snva736/snva736.pdf
From Vicor
http://www.vicorpower.com/documents/application_notes/an_ConstantCurrent.pdf
Also a good video from Great Scott
https://youtu.be/8uoo5pAeWZI
_____
"OCP type" SE is likely hiccup or just plain latching.
HRP are "current limiting" type.
HLG are actually fully adjustable current output, thus inherently current limiting, the best kind for flexibility, as with a lab style PSU, or even better for charging batteries a Grin Satiator, which adds automated charge termination when the target battery is Full.
Current Limiting is an overloaded term.
The best kind is "adjustable current"when the power source lets the user explicitly adjust the maximum available, easily on a per session basis, as with Lab style PSUs with an Amps dial on the front, or industrial units controlled by e.g. CAN messaging.
Related is the various methods of OCP, and Mean Wells can be all over the place sometimes even within the same series.
Latching means OC shuts the power output off, may need to be manually power cycled to resume.
Hiccup mode is power output halts for a time, then resumes once components cool down. Sometimes full shutdown after X attempts.
"Current limiting" style OCP, means the output gets greatly de-rated but continues, when demand lowers, returns to normal output. This may involve dropping output voltage, sometimes so much below spec, that charging in effect is halted.
The device must be **able** to drop voltage low enough in order to continue, if not then hiccup or latching should kick in as a backup OCP.
Finally the abstract EE design concept of a "current source" can pertain, but IRL the above are all within the CC/CV operation implemented by all power sources suitable to use as battery chargers, even if classified as rectifiers, power supplies, DC-DC converters etc.
______
A useful PDF from Texas Instruments
Know Your Limits
http://www.ti.com/lit/an/snva736/snva736.pdf
From Vicor
http://www.vicorpower.com/documents/application_notes/an_ConstantCurrent.pdf
Also a good video from Great Scott
https://youtu.be/8uoo5pAeWZI
_____
So, check out the Mean Well datasheets.ElectricGod said:
"OCP type" SE is likely hiccup or just plain latching.
HRP are "current limiting" type.
HLG are actually fully adjustable current output, thus inherently current limiting, the best kind for flexibility, as with a lab style PSU, or even better for charging batteries a Grin Satiator, which adds automated charge termination when the target battery is Full.
flippy
1 MW
- Joined
- Aug 12, 2015
- Messages
- 2,351
ElectricGod said:
yes. just straight into the caps and it will just go into CC mode until you reach the set voltage. so no: it will not disable if it detects a dead short. wich empty caps bascially are.
so no reason to burn power away on a resistor bank
ElectricGod
10 MW
This mod tightened things up a lot. There's still enough room in there so I can get my fingers around the bullets to push them together or pull them apart. I'll do something like it for the second super cap board when I put it in parallel. Snug everything together as much as possible and yet make it all easy to take apart.
The positive connection is so short you can just pull it apart or push it together. The negative connection runs diagonal across the bottom of the super cap board so you have to pull or push it with your fingers.
Not super important, but since I had everything taken apart, I also coated all the small electronics in conformal to protect them. I tried to scrub the first super cap board with some electronics cleaner to see if the solder would clean up on all the small parts. Whatever that chemical stuff is inside the super caps, it oxidizes solder and turns it dark grey. I've seen 90 year old solder that is still shiny, silvery and clean.
My Meanwell SE-600-12 disables output power if I go over current. I'd call it hick-up mode. Just tonight, I attempted to charge direct to the caps by bypassing the precharge resistor bank. The PSU delivered current, then dropped off, then delivered current, then dropped off. It did this several times before it finally just disabled the output. When it was delivering current, it was well above the 50 amp specs for the PSU. I had to de-power the PSU to get it running again. This PSU works exactly like the 8 SE-600-24 PSU's I have been using for several years as EV chargers. I limit their current with a bank of resistors too. It works VERY well if you are using PSU's that don't do constant current.
If your PSU can do constant current, good for you, set the current and you are good to go. If it can't do CC, use a resistor bank to limit current like I did.
I can bypass the precharge resistors at 7.6v and just stay under the limits of the PSU.
The resistor bank is pretty close to dead on for maximum current of the PSU. I gracefully ran the super caps down to 100mv through the precharge resistors. At 100mv I was still seeing 2.5 amps discharge on my meter. I opened the resistor bypass switch and had my meter ready to go clamped on the positive wire going to the supercaps. I saw a peak of 52 amps for a couple of seconds just as I plugged in the PSU to a single super cap board. Before I could grab my camera to take a picture it had already dropped to 41.3 amps.
The lowest voltage the PSU can deliver is 9.44v. There was 1.2 amps current draw despite the LED volt meter reading 9.44v. As it turns out it reads a few millivolts high. Not too bad for a meter that cost me less than $2! Later current draw zeroed out. I've been just letting the whole thing sit there while writing this post and it is staying at 0.0 amps. While gracefully charging the super caps, they warmed up just slightly. Sitting there quiescent attached to the PSU, they are back down to room temperature again.
The positive connection is so short you can just pull it apart or push it together. The negative connection runs diagonal across the bottom of the super cap board so you have to pull or push it with your fingers.
Not super important, but since I had everything taken apart, I also coated all the small electronics in conformal to protect them. I tried to scrub the first super cap board with some electronics cleaner to see if the solder would clean up on all the small parts. Whatever that chemical stuff is inside the super caps, it oxidizes solder and turns it dark grey. I've seen 90 year old solder that is still shiny, silvery and clean.
My Meanwell SE-600-12 disables output power if I go over current. I'd call it hick-up mode. Just tonight, I attempted to charge direct to the caps by bypassing the precharge resistor bank. The PSU delivered current, then dropped off, then delivered current, then dropped off. It did this several times before it finally just disabled the output. When it was delivering current, it was well above the 50 amp specs for the PSU. I had to de-power the PSU to get it running again. This PSU works exactly like the 8 SE-600-24 PSU's I have been using for several years as EV chargers. I limit their current with a bank of resistors too. It works VERY well if you are using PSU's that don't do constant current.
If your PSU can do constant current, good for you, set the current and you are good to go. If it can't do CC, use a resistor bank to limit current like I did.
I can bypass the precharge resistors at 7.6v and just stay under the limits of the PSU.
The resistor bank is pretty close to dead on for maximum current of the PSU. I gracefully ran the super caps down to 100mv through the precharge resistors. At 100mv I was still seeing 2.5 amps discharge on my meter. I opened the resistor bypass switch and had my meter ready to go clamped on the positive wire going to the supercaps. I saw a peak of 52 amps for a couple of seconds just as I plugged in the PSU to a single super cap board. Before I could grab my camera to take a picture it had already dropped to 41.3 amps.
The lowest voltage the PSU can deliver is 9.44v. There was 1.2 amps current draw despite the LED volt meter reading 9.44v. As it turns out it reads a few millivolts high. Not too bad for a meter that cost me less than $2! Later current draw zeroed out. I've been just letting the whole thing sit there while writing this post and it is staying at 0.0 amps. While gracefully charging the super caps, they warmed up just slightly. Sitting there quiescent attached to the PSU, they are back down to room temperature again.
ElectricGod
10 MW
Since I have been making changes, it was time to re-calibrate.
I thought I'd see what happens on a lot higher power settings with thicker metal
This is .02mm nickel welded to .6mm steel.
I didn't realize until I was done that I was cooking my bench top too. You can't really see the underside of the welds through all the cooked formica on the metal. LOL!
Cleaned up a bit. Left to right is 100 joules. Each weld is a 50 joule increment up to 300 joules at the right most weld. I can see welding much thicker stuff than the flimsy .02mm nickel I'm currently using that was the maximum the Sunnko could handle. That right most weld glowed red for a second. At some point, I'm going to see about spot welding a couple of thicknesses of sheet steel.
I thought I'd see what happens on a lot higher power settings with thicker metal
This is .02mm nickel welded to .6mm steel.
I didn't realize until I was done that I was cooking my bench top too. You can't really see the underside of the welds through all the cooked formica on the metal. LOL!
Cleaned up a bit. Left to right is 100 joules. Each weld is a 50 joule increment up to 300 joules at the right most weld. I can see welding much thicker stuff than the flimsy .02mm nickel I'm currently using that was the maximum the Sunnko could handle. That right most weld glowed red for a second. At some point, I'm going to see about spot welding a couple of thicknesses of sheet steel.
That approach would be suitable for stuff that is not potentially dangerous like ultracaps or Lipos, but in this case I have to raise theElectricGod said:
flag again and strongly discourage from doing this. Please don't treat this as an offense towards you personally, but as this is the 'official' kWeld support thread and a lot of other users/customers read the information posted here, I need to address this to keep others from attempting to experiment with this system like this. Trying out things in order to see if it blows up or not should not be encouraged in any way. All others: please follow the manuals for these components closely. The kCap modules may not be put in series without additional balancing circuitry. Do not charge them to more than 8.1V +/- a few percent. The maximum ratings may not be exceeded at any time.There are a few false assuptions here that I need to address:ElectricGod said:
- yes, this forms a capacitive voltage divider, but you may get significant unbalance that would lead to overvoltage of one of the two boards.
- it is not resistance (ESR) variation that causes unbalance, but capacitance variance along with activity profiles of the integrated balancers, as well as variation in leakage current of the capacitors.
- Looking at the datasheet of the used BCAP0310-P270-T10 capacitors shows a capacitance tolerance of +0 / -20% and a leakage current of 0 to 0.45mA. They don't specify a maximum initial capacitance which makes things even worse. 20% less capacitance basically translates to 20% more voltage across the two modules. This is a theoretical situation as it would only happen when one of the boards is old and worn out, and the other is new, but just to show the relationships. Voltages in a capacitive divider distribute reciprocally to their capacitances (rule of thumb), meaning that the capacitor with 20% less capacitance will get 20% more voltage, or 9.72V. The other, even worse issue, is leakage current that discharges them slowly. If that differs just a little bit, then over time one of the modules will gain more and more voltage (the one with less leakage), whereas the other's voltage will drop. Both effects lead to overvoltage of one of the modules, which must be avoided.
- 8.1V is the maximum that must not be exceeded. You keep saying that lots of people run these Maxwell capacitors at 11.1V. Can you post some links that shows evidence of this? This is unheard for me, and it would be very dangerous to do so.
Yes you did that, but on the other hand you also keep saying that lots of people would use these capacitors at higher voltages without problem. And you post pictures of your setup. I'm afraid that some may miss the point and copy that, and eventually expose themselves to danger. That's why I need to keep raising warnings against this practice. It's fine to experiment with this by yourself as you are the only one who is responsible for yourself in the end, but I kindly ask you to stop posting the results of this in this kWeld support thread.ElectricGod said:
That's 9.44V and way above the rated kCap voltage. This Meanwell PSU is not appropriate for kCap, and must not be used.ElectricGod said:
Hey ElectricGod,
Just one additional reminder or warning.
These Maxwell Supercaps contain as you know some nasty chemicals. What you might not know is, if you overheat the Cap to arount 150°C there is cyanid created in the Supercap. I don't think that you want to have cyanid in your house so I would look for a very well ventilated spot to overvolt my Supercaps or just stick with the correct voltage.
Just one additional reminder or warning.
These Maxwell Supercaps contain as you know some nasty chemicals. What you might not know is, if you overheat the Cap to arount 150°C there is cyanid created in the Supercap. I don't think that you want to have cyanid in your house so I would look for a very well ventilated spot to overvolt my Supercaps or just stick with the correct voltage.
ElectricGod
10 MW
I'm not sure where I saw people running from a car battery or 3S lipo through the super cap module.
Going back through all the pages and skimming through this thread, I didn't see what I remembered. It could have been in youtube videos.
What I saw was to run them at higher voltage...AND even then...keep it down as low as possible that you needed a fan.
I didn't see anybody try what I tried. However I bet there is at least one other person out there that has done what I did.
Regardless, I'm not doing 11-12v and they were getting pretty warm even with a decent fan at that voltage.
I don't recommend running the super cap board at 11-12v! I did it for a couple of minutes. Longer is probably bad!
Yes I know this is OVER specs.
So far no problems running at 9.44v. The caps get slightly warm...like 80F and only when they are charging from 0v at 50 amps. At a regular weld cadence of a weld every second or 2, they never warmed up at all. Even when I was doing those heavy welds I presented a few posts pack, they stayed at room temperature. Next time I pull out the welder, I'll also pull out my laser temp meter and see what the actual temps are for all the caps as they charge up from 0v. To the touch they feel like 80F. Whatever I actually measure, it's still 100F UNDER what the caps were at when I deliberately tested to failure. They were VERY HOT before they burst.
As a side measure, something that I have assumed is that all the caps are charging to the same voltage and actually balancing. I don't really know if that's true or not. Not all caps are created equal...not even really good ones like Maxwells. I'll check the voltages across the caps. I think I'll implement this with a 3S balance cable and a battery tester capable of dealing with lower voltages. This is an easy add-on so I can know actual super cap status quickly. I'll post pics after a while. There's no rocket science here, just a simple JST balance cable.
Should people run their super cap modules over the recommended voltage? NO!!!! Especially not if you are risk averse.
I'm testing and seeing how this pans out. So far it's fine. I'm obviously not risk adverse.
I apologize if I'm making people uncomfortable, that is not my intention. I want to see what can really be done and NOT blow stuff up like my initial test eventually achieved. Maybe I'm posting this content in the wrong thread? I could have backed off the voltage when I noticed the caps were getting warm. BUT I'd never be able to show you what happens when they blow if I didn't. Now everybody knows...and again...do NOT do what I did!!!
My tests and experiments do not in anyway diminish the quality of the KWeld. It IS a very nicely engineered spot welder. Just run them as specified and you are fine.
Manufacturer recommended voltage for these super caps is 2.7v. It's printed right on the side of the super caps!
3 in series = 8.1v.
I'm running at 9.44 volts which is 3.15v per cap. This is OVER the spec'd voltage by .45v per cap. It will probably be fine, but that doesn't mean do like I do. I push everything...you probably shouldn't.
Going back through all the pages and skimming through this thread, I didn't see what I remembered. It could have been in youtube videos.
What I saw was to run them at higher voltage...AND even then...keep it down as low as possible that you needed a fan.
I didn't see anybody try what I tried. However I bet there is at least one other person out there that has done what I did.
Regardless, I'm not doing 11-12v and they were getting pretty warm even with a decent fan at that voltage.
I don't recommend running the super cap board at 11-12v! I did it for a couple of minutes. Longer is probably bad!
Yes I know this is OVER specs.
So far no problems running at 9.44v. The caps get slightly warm...like 80F and only when they are charging from 0v at 50 amps. At a regular weld cadence of a weld every second or 2, they never warmed up at all. Even when I was doing those heavy welds I presented a few posts pack, they stayed at room temperature. Next time I pull out the welder, I'll also pull out my laser temp meter and see what the actual temps are for all the caps as they charge up from 0v. To the touch they feel like 80F. Whatever I actually measure, it's still 100F UNDER what the caps were at when I deliberately tested to failure. They were VERY HOT before they burst.
As a side measure, something that I have assumed is that all the caps are charging to the same voltage and actually balancing. I don't really know if that's true or not. Not all caps are created equal...not even really good ones like Maxwells. I'll check the voltages across the caps. I think I'll implement this with a 3S balance cable and a battery tester capable of dealing with lower voltages. This is an easy add-on so I can know actual super cap status quickly. I'll post pics after a while. There's no rocket science here, just a simple JST balance cable.
Should people run their super cap modules over the recommended voltage? NO!!!! Especially not if you are risk averse.
I'm testing and seeing how this pans out. So far it's fine. I'm obviously not risk adverse.
I apologize if I'm making people uncomfortable, that is not my intention. I want to see what can really be done and NOT blow stuff up like my initial test eventually achieved. Maybe I'm posting this content in the wrong thread? I could have backed off the voltage when I noticed the caps were getting warm. BUT I'd never be able to show you what happens when they blow if I didn't. Now everybody knows...and again...do NOT do what I did!!!
My tests and experiments do not in anyway diminish the quality of the KWeld. It IS a very nicely engineered spot welder. Just run them as specified and you are fine.
Manufacturer recommended voltage for these super caps is 2.7v. It's printed right on the side of the super caps!
3 in series = 8.1v.
I'm running at 9.44 volts which is 3.15v per cap. This is OVER the spec'd voltage by .45v per cap. It will probably be fine, but that doesn't mean do like I do. I push everything...you probably shouldn't.
ElectricGod
10 MW
HMT_01 said:
Please do NOT do what I did!!!
Are you saying cyanide?
Googling for cyanid didn't produce results.
Anyway, if you are talking about cyanide...then yeah...this is not particularly good stuff.
From wikipedia:
Acute exposure
If cyanide is inhaled it can cause a coma with seizures, apnea, and cardiac arrest, with death following in a matter of seconds. At lower doses, loss of consciousness may be preceded by general weakness, giddiness, headaches, vertigo, confusion, and perceived difficulty in breathing. At the first stages of unconsciousness, breathing is often sufficient or even rapid, although the state of the person progresses towards a deep coma, sometimes accompanied by pulmonary edema, and finally cardiac arrest. A cherry red skin color that changes to dark may be present as the result of increased venous hemoglobin oxygen saturation. Despite the similar name, cyanide does not directly cause cyanosis. A fatal dose for humans can be as low as 1.5 mg/kg body weight.[9]
Chronic exposure
Exposure to lower levels of cyanide over a long period (e.g., after use of improperly processed cassava roots as a primary food source in tropical Africa) results in increased blood cyanide levels, which can result in weakness and a variety of symptoms, including permanent paralysis, nervous lesions,[10][11][12] hypothyroidism,[11] and miscarriages.[13][14] Other effects include mild liver and kidney damage
ElectricGod
10 MW
tatus1969 said:That approach would be suitable for stuff that is not potentially dangerous like ultracaps or Lipos, but in this case I have to raise theElectricGod said:
There are a few false assuptions here that I need to address:ElectricGod said:
- yes, this forms a capacitive voltage divider, but you may get significant unbalance that would lead to overvoltage of one of the two boards.
- it is not resistance (ESR) variation that causes unbalance, but capacitance variance along with activity profiles of the integrated balancers, as well as variation in leakage current of the capacitors.
- Looking at the datasheet of the used BCAP0310-P270-T10 capacitors shows a capacitance tolerance of +0 / -20% and a leakage current of 0 to 0.45mA. They don't specify a maximum initial capacitance which makes things even worse. 20% less capacitance basically translates to 20% more voltage across the two modules. This is a theoretical situation as it would only happen when one of the boards is old and worn out, and the other is new, but just to show the relationships. Voltages in a capacitive divider distribute reciprocally to their capacitances (rule of thumb), meaning that the capacitor with 20% less capacitance will get 20% more voltage, or 9.72V. The other, even worse issue, is leakage current that discharges them slowly. If that differs just a little bit, then over time one of the modules will gain more and more voltage (the one with less leakage), whereas the other's voltage will drop. Both effects lead to overvoltage of one of the modules, which must be avoided.
- 8.1V is the maximum that must not be exceeded. You keep saying that lots of people run these Maxwell capacitors at 11.1V. Can you post some links that shows evidence of this? This is unheard for me, and it would be very dangerous to do so.
Yes you did that, but on the other hand you also keep saying that lots of people would use these capacitors at higher voltages without problem. And you post pictures of your setup. I'm afraid that some may miss the point and copy that, and eventually expose themselves to danger. That's why I need to keep raising warnings against this practice. It's fine to experiment with this by yourself as you are the only one who is responsible for yourself in the end, but I kindly ask you to stop posting the results of this in this kWeld support thread.ElectricGod said:
That's 9.44V and way above the rated kCap voltage. This Meanwell PSU is not appropriate for kCap, and must not be used.ElectricGod said:
I acknowledge what you wrote and agree.
People...do NOT do what I do!
Keep your super cap board AT the recommended voltage.
ElectricGod
10 MW
Perspective is everything...
Not risk adverse:
Colin Furze...
https://www.youtube.com/watch?v=s4QBhNkaAk8
https://www.youtube.com/watch?v=LDM9aXVAoOw
Risk adverse:
Sloooooooow drivers...
https://www.youtube.com/watch?v=hnml1rxiyIw
Guess which one applies to me?
Not risk adverse:
Colin Furze...
https://www.youtube.com/watch?v=s4QBhNkaAk8
https://www.youtube.com/watch?v=LDM9aXVAoOw
Risk adverse:
Sloooooooow drivers...
https://www.youtube.com/watch?v=hnml1rxiyIw
Guess which one applies to me?
I'm definitely of the first category. I deliberately never purchased a motorcycle because I know that I will drive myself to dead very quickly.ElectricGod said:
I have been driving the Ariel Atom on public driving events from 6th Gear in UK a few times, and most of the times the race instructor got mad at me because of ridiculous driving. I think that I am scaring them off too much when pushing down the gas pedal all the way in third gear to get into a little drift (that car is *well* capable of doing that). Conclusion of this: buy my own road legal Atom and get rid of the race instructor.
If you have noticed my Youtube channel (WeAreTheWatt) then you know that I'm not of the second category 8) The kart that we have been building uses a 700W servo motor cranked up all the way to 4kW (4x rated current, and rpms way above the rated mechanical max). The planetary gear also didn't like that amount of speed and torque on its input shaft much either :wink:
My warnings here are definitely not for guys like you and me, but as this thread is also there to inform other existing and potential users of this stuff, I see my responsibility to keep the supplied information clean. Otherwise a less experienced user may put himself at risk by not reading closely enough. Just your one picture showing 9.44V on kCap can be enough to mislead someone.
Thanks for your last posts!
edit - p.s. Colin Furze - love his channel
edit2 - pp.s. the Youtube channel name was given by my son after a 150W power resistor literally exploded in our living room after the kart's motor controller pushing 1kW of power into it. It was an extremely loud bang, the high voltage must have caused arcing inside the resistor after the metal had melted, causing high pressure to build up. These resistors are very rugged and hermetically sealed, and the aluminium chassis was bent wide open after that. After us having calmed down from the shock, he just meant to me: Resistance is futile - We Are The Watt. Couldn't stop laughing for quite a while 8) And guess what, it happened again (that time in our garden), until I finally found the problem, a mistake in a formula that had a - where a + was needed.
Just wanted to share a piece of work with you that I am doing for a customer. These teflon handles will provide additional heat insulation and should allow significantly faster welding speed. The assembly also uses 6AWG leads, which should help keeping them cooler as well.
ElectricGod
10 MW
Last night I was redoing welds on 3 20S3P packs that I had done with my Sunnko welder. The KWeld does a far better job! I have yet to parallel the second super cap board. For now since my welds are into .02mm nickel, I don't really need the power a second board will offer. Later...well that's another story. I intend to be pushing some serious amperage.
My cadence last night was something like a weld every second or a bit faster. I noticed that the Meanwell PSU slowly fell behind, but not enough that it kept me from welding continuously. I'd do 30-40 welds in a row and the voltage on the super caps would slooooowly drop. It may have been that I forgot to bypass my precharge resistors. By the time I got to 40 welds in quick succession, the weld pens were pretty hot so I'd need to pause anyway. I was working at 50 joules. Any more power than that and I have the possibility of blowing through the can on the Panasonic 18650 cells. My weld pens have 5 layers of heat shrink on them. They get pretty hot before I stop welding! The super cap fan turns on briefly and at low fan RPMs by the time the weld pens are too hot to keep going. I had an issue with the weld pen tips getting dirty. since the 3 packs had already been welded and I had soldered balance wires onto the nickel already, any flux or solder the tips came in contact with would scorch on the weld tips and I'd need to use some 22 grit sand paper to clean them off periodically. In 2 hours I did something like 2 welds per cell end each pack is 60 cells so that's 120 welds per pack and I redid 2 of them or 240 welds.
I'm pretty happy with this welder!
If there is something to complain about it is the heat in the pens, but that's the nature of high current stuff so not really a complaint. I want to use 6 awg wire so there is more wire to sink off heat. Teflon ought to work, but I was thinking about finding silicon grips for them like soldering irons commonly have.
My cadence last night was something like a weld every second or a bit faster. I noticed that the Meanwell PSU slowly fell behind, but not enough that it kept me from welding continuously. I'd do 30-40 welds in a row and the voltage on the super caps would slooooowly drop. It may have been that I forgot to bypass my precharge resistors. By the time I got to 40 welds in quick succession, the weld pens were pretty hot so I'd need to pause anyway. I was working at 50 joules. Any more power than that and I have the possibility of blowing through the can on the Panasonic 18650 cells. My weld pens have 5 layers of heat shrink on them. They get pretty hot before I stop welding! The super cap fan turns on briefly and at low fan RPMs by the time the weld pens are too hot to keep going. I had an issue with the weld pen tips getting dirty. since the 3 packs had already been welded and I had soldered balance wires onto the nickel already, any flux or solder the tips came in contact with would scorch on the weld tips and I'd need to use some 22 grit sand paper to clean them off periodically. In 2 hours I did something like 2 welds per cell end each pack is 60 cells so that's 120 welds per pack and I redid 2 of them or 240 welds.
I'm pretty happy with this welder!
If there is something to complain about it is the heat in the pens, but that's the nature of high current stuff so not really a complaint. I want to use 6 awg wire so there is more wire to sink off heat. Teflon ought to work, but I was thinking about finding silicon grips for them like soldering irons commonly have.
ElectricGod
10 MW
flippy said:
I'll post my results after I try out some things, but I may have a much simpler idea that anybody can do.
I have multiple layers of heat shrink on mine and that does help, but the heat shrink still gets heat soaked. I think those teflon handles will essentially act like my many layers of heat shrink. All it does is prolong the time before you can't weld anymore.
I'm working on how to not have to stop at all.
The module does balancing, but it does it in a way that it equalizes the voltages across the capacitors (within a few ten millivolts). It does *not* limit to a given maximum additionally (I didn't see a point in adding this because a limiter wouldn't be able to fight an overvolting 50A+ charger anyway). So the balancer wouldn't complain about all cells being overvolted by the same amount, as long as the voltages are distributed evenly.ElectricGod said:
Sure, let me see first how they perform. ElectricGod's remark that they may just extend the use time might be true, maybe a different material would be better. These tubes are made from 90mm long teflon rods (20mm dia) with 10mm center drill, and another 10mm drill into the side for the set screws. Then I sharpened one end with a grinding machine.flippy said:
BVH
1 kW
- Joined
- Mar 26, 2009
- Messages
- 411
I once worked with some Fluorosint material. If I remember correctly, it is Mica infused PTFE Teflon with very high heat handling capabilities. You might look into this very easy to machine material.
https://www.mcam.com/na-en/products/engineering-plastics/advanced-325-425-f/fluorosintr-enhanced-ptfe/
https://www.mcam.com/na-en/products/engineering-plastics/advanced-325-425-f/fluorosintr-enhanced-ptfe/
ElectricGod
10 MW
tatus1969 said:
Aerogel is incredible stuff. I just saw that you can get it infused into cloth. It's not quite as good as the pure aerogel, but it still works as an incredibly good insulator. What I care about is not needing to ever stop welding. Aerogel infused cloth may be the answer.
https://www.ebay.com/itm/Aspen-Aerogel-SPACELOFT-Insulation-Hydrophobic-Mat-6mm-PER-LINEAR-FOOT-4-75-SQFT-/380906349501?_trksid=p2385738.m4383.l4275.c10
I bought some aerogel cloth. I know this stuff sheds bits of aerogel, but it will be encased in heat shrink so I think it will be OK.
I wonder if I can find a resin or epoxy with aerogel in it? That ought to be pretty good too.
Balancing...that is good to know. It's relative to the input voltage and divided by 3 instead of trying to maintain 2.7v per cap. The SMD resistors that bleed off voltage won't have a chance against a 12 amp PSU never mind 50 amps. I still want to see the voltages for each super cap.
