Fluxer Heaters, induction heaters for Dynavap

mr_cfromcali

Accessory Maker
Accessory Maker
I prefer a working reliable heater and i would rather wait for this than have one that i would feel unsure about.:tup:

Are the impatient ones canceling their orders :rolleyes:

Not many are cancelling, but it is a long list, it grows daily, and people become frustrated that it isn’t moving more quickly.

I am also frustrated at times that this process isn’t quicker. It’s hard to set appropriate expectations when you don’t know how long it will take to truly resolve an issue.

At the same time, no one wants a product that doesn’t work, and I have no interest in selling junk. My goal is to make a fucking amazing and reliable little Dynavap-specific IH. I’m close on the former, and working hard on the latter.

I’ll get there. Eventually.
 

Jill NYC

Portable Hoarder
@mr_cfromcali - thanks for the update.
I guess my patience doesn’t count for much since my #12 has worked flawlessly thus far but...

I just think about how long it has taken Dynavap to bring their own IH to the market... and I think it’s safe to say they are bigger than a 1-man operation and have enough funds to do beta tests and go back to the drawing board when issues are found. Not to mention this isn’t just a side gig for them.

Your honesty, perseverance, transparency and problem solving skills are admirable! I am rooting for you every step of the way.
 

MothChewMoth

Gamer Extraordinaire
I prefer a working reliable heater and i would rather wait for this than have one that i would feel unsure about.:tup:

Are the impatient ones canceling their orders :rolleyes:

Means less people ahead of me!

In all seriousness though as antsy as I am to order, I'd never consider cancelling because the creator was focused on making something work better. Hopefully others feel the same.
 

mr_cfromcali

Accessory Maker
Accessory Maker
Hi all,

It might be time to make some more popcorn, lol.

:popcorn:

The new TVS diodes arrived, and they are looking pretty good!

Test bed is a 14mm coiled FD, w/fresh batteries.

These traces show the voltage on both heater MOSFET Gate pins, before the addition of the new TVS diodes. Each pin has a max of 20V per the spec sheet. These are showing >27V, which is bad.

Before installing TVS diodes on the MOSFET Gate pins:

SDS00145a.png



After installing TVS diodes on the MOSFET Gate pins:


SDS00146a.png


They work! :D

So they work in the absolute sense of, "can they perform the task," which is critical, of course. The next step is to see if they can do this without breaking something else, and my fingers are crossed that this will prove to be true. But in terms of expected results, I'd call these are "better than I hoped for," and I am very encouraged.

More to come... :science:
 
Last edited:

mr_cfromcali

Accessory Maker
Accessory Maker
Maybe you've addressed this already but are you installing the new battery pack electrical insulator in the Flux Deluxes you're making now?

Yes. The battery protectors have been hand made up to this point, but I recently realized I could think of them as an unpopulated PCBs, and draw them up on the computer instead. So I did that, and the first batch of those arrived a few weeks ago. I have been using them on repairs but have been too head down in the fix-it stuff to parade the new stuff...but there are a few new items to show off.

For now, I'll start with just the new battery protector:

IMG-7982-private.jpg

The new part.

IMG-7983-private.jpg

The new part installed.

I'll have some more exciting features (IMHO) to share in the future, but from my vantage point, NOTHING will be sexier than basic reliability ;)

More to come...
 
Last edited:

mr_cfromcali

Accessory Maker
Accessory Maker
I have emailed a few times and never get a reply

Guilty. Sorry. I only have so much bandwidth, and have been giving priority to existing customers and to troubleshooting/engineering a solution for the recent spate of MOSFET and diode issues; see the last 10 pages of this thread for more info on that effort.

If you are an existing customer and I missed your issue, I am very, very sorry! PLEASE email me again, and make it plain in the subject that this is in regards to a warranty issue! I will get back to you within half a day, usually sooner.

If you asked to be added to the mailing list, I have already added you, but silently due to the volume of requests I have been receiving. Thank you for your interest in this project! I will contact you again when I am closer to building your heater.

If you had written with another question, I apologize for the frustration of not responding in a timely manner. My volume of email is high at the moment and there is only one of me, and my Fluxer stuff is a hobby, NOT my day job. I give this stuff as much time as I can, but I have a family, a job, and a life outside of this project, and each of those also takes some of my time. Please try again to reach me.

It was extremely important to the project that I work out a solution for the MOSFET issues that have been affecting the heaters. No matter how cool these things may look, they are worthless if they can't heat a Dynavap reliably. I am sorry this meant that new emails got short shrift, but I hope to get back to a normal email response time soon.

Thanks!
 

mr_cfromcali

Accessory Maker
Accessory Maker
Hi folks, back with another update. It's what I do these days. ;)

I'm feeling more confident that this current TVS fix is the right one - it's been very stable so far. Of course, I've said that several times before and been proven wrong every time (to this point), so we will have to see what more time and testing reveal.

FWIW, I feel confident enough of this TVS configuration that I revised the PCB to accommodate the new TVS diodes, and ordered a new batch of PCBs. The boards will be here in a week or so. I'm hopeful this solution will still be looking this good when they get here. :tup:

I also worked out how I can retrofit these new TVS diodes into/onto existing heaters. It's a little trickier than the first TVS diode install, but it's similar. Each repair will take 30-45 minutes.

This is a screenshot of my test unit with the TVS diodes retrofitted as I intend to install them to fix existing heaters. To my eyes this graph with diodes permanently installed looks even better than the previous graphs, when the solution was simply prototyped:


SDS00155.png



Purple line above is the MOSFET A's Gate pin, yellow line is MOSFET B's Source pin. The Gate pin has the lower, 20V voltage limit. A max of <20V on that pin is fantastic! It's suppressing the inductive voltage surge in a single pulse lasting <2us (us = microseconds), and in doing so it is keeping it under the 20V max for rating the MOSFET Gate pin; 15mm coil unit used in above test.

To my eyes, this graph looks really, really good. If this fix is stable, it's all but perfect. I'm both thrilled and a tiny bit scared, lol.
:tinfoil:

I will be pounding on this some more before I release it as a fix and install it as a repair, but so far it has been working perfectly. I am going to offer a cautious, "so far, so good." Yay!

Stay tuned.

:popcorn::popcorn::popcorn:

Cheers,
:leaf:
 
Last edited:

Moses Baca

Colorado State Reformatory #8755
Fingers crossed!

I appreciate your continued efforts and how open you are with the entire process. Very commendable. I was reminded of the story of WD40--it took 40 attempts to get the water displacement formula correct, thus the name. Keep at it, friend!
 

mr_cfromcali

Accessory Maker
Accessory Maker
Thanks, @Moses Baca ! I would truly like to find the right solution and shut the book on this issue, so I can get back to making heaters instead of studying their faults. The current fix does exactly what I want it to do from a voltage suppression standpoint, so I'm now on the hunt for negative side effects or unintended consequences. I haven't found any yet, which is another good sign. We'll see if any find me, lol. ;)

I have been plugging away at this issue in different forms since mid-February, when it appeared as a consequence of the hotter 14mm and 15mm coils. My finish line for this device is on the other side of resolving this issue. It's been a struggle, but I want to get there, and I want it to be clear when I do. This EE stuff is a black box to most of us, so by posting the oscilloscope graphs and such I am trying to pull back the curtain and show my work. The funny (?) thing is, each of my previous solutions also "worked", they just didn't work in the real world, and/or for very long.

The current solution is still looking good, though. :tup:

Cheers,
:leaf:
 

mr_cfromcali

Accessory Maker
Accessory Maker
MrC, keep at it. I very much appreicate all your efforts to get this device up, running and fully stable. Esepecially, as this is just a side thing you are doing and at this moment is more beneficial to the community then it is to you.

Thanks, @LongDraw! WRT your last point, I may be funding this like it's a vanity project, but I would love to sell a bunch of these once I can stabilize the design and improve reliability! A lot of people seem to want one, so we'll see what happens.

:popcorn:

Cheers!
 

testymctest

Well-Known Member
@mr_cfromcali - thanks for the update.
I guess my patience doesn’t count for much since my #12 has worked flawlessly thus far but...

I just think about how long it has taken Dynavap to bring their own IH to the market... and I think it’s safe to say they are bigger than a 1-man operation and have enough funds to do beta tests and go back to the drawing board when issues are found. Not to mention this isn’t just a side gig for them.

Your honesty, perseverance, transparency and problem solving skills are admirable! I am rooting for you every step of the way.

Mr C,
Just curious if you know why some of the heaters like mine #22 14mm (to hot for me) converted to 15mm (perfect) work flawlessly and others are failing (knock on wood that asking this doesn't jinx my unit :huh:)
 

mr_cfromcali

Accessory Maker
Accessory Maker
Mr C,
Just curious if you know why some of the heaters like mine #22 14mm (to hot for me) converted to 15mm (perfect) work flawlessly and others are failing (knock on wood that asking this doesn't jinx my unit :huh:)

Thanks for the question, @testymctest . It is a head scratcher. :hmm: I have some ideas, but unfortunately these things don't write log files that explain why they failed, etc. So my post-mortem analysis involves a mix of forensics and guesswork.

In general, the heater MOSFETs are the most vulnerable components in the circuit, followed by the power relay MOSFETs. Sometimes these parts will stand up to tremendous abuse, like a big voltage spike, as if they were designed for it. Other times they will fold like the proverbial cheap suit. Variable parts tolerances, weather (!) - especially static electricity, batteries, and variable user technique probably all play roles in whether a given heater is a reliable device or a MOSFET serial killer.

If I knew what to tell you to do or not do, I would absolutely communicate that. :nod:

Since I don't know why a specific part may have failed, I have been looking at the circuit in context (i.e., on an oscilloscope during use) to see how it performs, and if it is working as expected. I did not rely on the scope as much as I should have early in this project, but I am making up for that now. The scope is very informative, and makes it plain that releasing the tactile switch leaves a lot of electrical chaos in its wake, with some 14mm coil heaters producing a voltage surge as high as 100V+ (for fractions of a second). Suppressing that voltage spike is the battle I have been fighting.

Some heaters stand up to a ton of this sort of abuse, but they do so for reasons I can't explain. In the face of a 70V power surge they should most likely fail, but like a boxer they can somehow take a big hit and stay on their feet. I don't know why. :shrug:

Not the clearest answer, but that's what I can tell you. Hope it was helpful?

I will add that the current TVS solution continues to work flawlessly on my bench, and looks very, very good on the scope. Unless something changes, I expect to begin installing this repair on customers' heaters after the weekend.

Cheers,
:leaf:
 

LongDraw

Well-Known Member
Thanks, @LongDraw! WRT your last point, I may be funding this like it's a vanity project, but I would love to sell a bunch of these once I can stabilize the design and improve reliability! A lot of people seem to want one, so we'll see what happens.

:popcorn:

Cheers!


I have no doubt you will sell a bunch, you already have a long list of ppl interested. my guess is that is only gonna grow once stablized. plus you are being smart, not just putting out a bunch of heaters with fingers crossed they don't have issues and/or if they do worrying about it at a later time.
 

ClearBlueLou

unbearably light in the being....
Keep on doin’ like you’re doin’, everything will be alright...after previous IH heater experience, this thread displays clarity, intelligence, dedication, and an organized mind...and increases my satisfaction with how you conduct yourself and your business.

I have no doubt you’ll earn back your development expenses in fairly short order: your better mousetrap assures it.
 

mr_cfromcali

Accessory Maker
Accessory Maker
Hi all,

Time for another update.

:popcorn:

MOSFET issues: Like a good mystery story, it seems that there's been more than one MOSFET killer lurking in the Flux Deluxe's circuit board design. :ninja:

I've spent the last 2.5 months focused on taming the electrical aspects of this heater, especially its susceptibility to voltage spikes. It took multiple tries, but as of last week I truly believe I turned the corner on the voltage spike issue. My current TVS diode solution continues to look good on the scope and work as expected. As far as I can tell, that issue is genuinely fixed. Huzzah!

So you can imagine the churn in my stomach yesterday when my test unit (with the lastest TVS diode fix) died mid-heat cycle, absolutely NOT due to any sort of voltage spike.

WTF??!? :bang::goon:

It turns out there's more than one way to kill a MOSFET. If you don't have any excess voltage at hand, excess heat will also do the trick. Thanks, physics. :rolleyes:

All ZVS induction heaters are susceptible to this, and unfortunately I now see this is an area of design to which I should have paid more attention earlier in the project. Poor MOSFET heat management wasn't a show-stopper with the 16mm coil FDs, but the 14mm and 15mm coils pull more power through their MOSFETs than the original 16mm coil. I didn't realize how detrimental this could be until yesterday, when I finally saw a MOSFET fail on my bench. Which sucks, but yay! for seeing it happen, and being able to eliminate excess voltage as the cause. Knowledge doesn't come cheaply, sometimes, but that's how you learn stuff. :sherlock:

Unfortunately, this situation was hidden behind the voltage spike issue, so it took me by surprise. :worms:

The good news - and once again I may be squinting to see the upside - is that there are other, smarter people that manage hot electronic circuits for a living, and they have proven strategies for mitigating this sort of thing. With a quick search I found the formulas for adding the right number of vent holes under my MOSFETs, as well as how to mate them to add'l surface area on both sides of the PCB, how to add more copper to the PCB surfaces for better heat exchange, etc.

I had done NONE of these things to this point, and as a consequence the Flux Deluxe's MOSFETs can only shed heat through their small footprint on the PCB. :haw:

I now know better. :nod:

Last night, I revised the FD's PCB to incorporate more surface area to vent heat, and more copper to facilitate better heat transfer. On both sides of the PCB. Previously, each MOSFET only had its 1cm^2 footprint for cooling. Now each MOSFET has 20cm^2 (!) of surface area for cooling, 10cm^2 on each side of the board.

These changes should provide for better heat regulation and allow the MOSFETs to stay below critical temperatures.

The new boards should be here in a week or so. I think they will be HUGE improvement over the current PCBs in this regard.

What does this mean for existing FD owners?

If your heater is working, that's great! Long may it run! :rockon:

If your heater has failed, please contact me to arrange the repair. Repairs are on hold until the new PCBs get here and test out OK, but a stable circuit will be worth the wait. I will know more once the new PCBs get here. It's always possible there's yet another big issue hiding behind these first two. I don't think there is, but this project has been sneaky that way, so we'll need to wait and see.

Finally, production remains ON HOLD. I am not making ANY new heaters until I get this sorted out.

I assure you, once I get these issues sorted, I will be happy to crank out as many of these as I can, as they will be good to go.


I will be putting together a post soon to show off the new features I'll be offering, which are incorporated into the new PCBs. They are more evolutionary than revolutionary, but they make the device more complete.

Sorry to post a cliff-hanger instead of a season finale, but that's how it goes sometimes.

:popcorn::popcorn::popcorn:

Cheers,
:leaf:
 
Last edited:

started@52

Well-Known Member
Hi all,

Time for another update.

:popcorn:

MOSFET issues: Like a good mystery story, it seems that there's been more than one MOSFET killer lurking in the Flux Deluxe's circuit board design. :ninja:

I've spent the last 2.5 months focused on taming the electrical aspects of this heater, especially its susceptibility to voltage spikes. It took multiple tries, but as of last week I truly believe I turned the corner on the voltage spike issue. My current TVS diode solution continues to look good on the scope and work as expected. As far as I can tell, that issue is genuinely fixed. Huzzah!

So you can imagine the churn in my stomach yesterday when my test unit (with the lastest TVS diode fix) died mid-heat cycle, absolutely NOT due to any sort of voltage spike.

WTF??!? :bang::goon:

It turns out there's more than one way to kill a MOSFET. If you don't have any excess voltage at hand, excess heat will also do the trick. Thanks, physics. :rolleyes:

All ZVS induction heaters are susceptible to this, and unfortunately I now see this is an area of design to which I should have paid more attention earlier in the project. Poor MOSFET heat management wasn't a show-stopper with the 16mm coil FDs, but the 14mm and 15mm coils pull more power through their MOSFETs than the original 16mm coil. I didn't realize how detrimental this could be until yesterday, when I finally saw a MOSFET fail on my bench. Which sucks, but yay! for seeing it happen, and being able to eliminate excess voltage as the cause. Knowledge doesn't come cheaply, sometimes, but that's how you learn stuff. :sherlock:

Unfortunately, this situation was hidden behind the voltage spike issue, so it took me by surprise. :worms:

The good news - and once again I may be squinting to see the upside - is that there are other, smarter people that manage hot electronic circuits for a living, and they have proven strategies for mitigating this sort of thing. With a quick search I found the formulas for adding the right number of vent holes under my MOSFETs, as well as how to mate them to add'l surface area on both sides of the PCB, how to add more copper to the PCB surfaces for better heat exchange, etc.

I had done NONE of these things to this point, and as a consequence the Flux Deluxe's MOSFETs can only shed heat through their small footprint on the PCB. :haw:

I now know better. :nod:

Last night, I revised the FD's PCB to incorporate more surface area to vent heat, and more copper to facilitate better heat transfer. On both sides of the PCB. Previously, each MOSFET only had its 1cm^2 footprint for cooling. Now each MOSFET has 20cm^2 (!) of surface area for cooling, 10cm^2 on each side of the board.

These changes should provide for better temperature regulation and allow the MOSFETs to stay below critical temperatures.

The new boards should be here in a week or so. I think they will be HUGE improvement over the current PCBs in this regard.

What does this mean for existing FD owners?

If your heater is working, that's great! Long may it run! :rockon:

If your heater has failed, please contact me to arrange the repair. Repairs are on hold until the new PCBs get here and test out OK, but a stable circuit will be worth the wait. I will know more once the new PCBs get here. It's always possible there's yet another big issue hiding behind these first two. I don't think there is, but this project has been sneaky that way, so we'll need to wait and see.

Finally, production remains ON HOLD. I am not making ANY new heaters until I get this sorted out.

I assure you, once I get these issues sorted, I will be happy to crank out as many of these as I can, as they will be good to go.


I will be putting together a post soon to show off the new features I'll be offering, which are incorporated into the new PCBs. They are more evolutionary than revolutionary, but they make the device more complete.

Sorry to post a cliff-hanger instead of a season finale, but that's how it goes sometimes.

:popcorn::popcorn::popcorn:

Cheers,
:leaf:
I give you props for sticking with this! Also give you props for the clarity, honesty, transparency of all your hard work, and the class you present it all with. Keep on keeping on!
 

Chandler

Well-Known Member
16mm coil owners, How have your experiences been to this point with your FD? I've been thinking the whole time I need the 14mm turbo charged Ferrari, but as I medicated last night it occurred to me that an even faster heatup time than my psm isn't for me. My psm heats hot to the click. I don't go past it or combust. I enjoy it thoroughly, but I could do with a more gentle heat source.
@mr_cfromcali might be a dumb question but is the battery life of the 16mm coil better than the other FD's?
 
Top Bottom