Induction brass annealer redux

[David101]

Experimenting with different coils and the Flux concentrators has been very interesting. I found that these open Ferites only draw about 3 amps with no case in the path and 7 amps at 30 V with a case such as a 223 -308 size. The 223 was annealing in about 2.5 seconds and the 308 in 4.5 seconds. Only doing a few at a time there is no appreciable increase in temperature on either the coil/ferite or on the induction board. this is all air cooled as well. Draw back is that the case is very close to the ferite so more care is needed also the case must be held in place well as when the induction starts it pushes the case out of the field. In comparison I made the 8 turn 1 /8 using 1/8 copper tube this maxes out my current power supply (while waiting for the 48V unit.) Also this coil really heats up without any water. I have it all plumbed for water but again waiting on a more suitable reticulation water unit. The time it takes to anneal is also longer.

 

[SGK]

Ask Stanley?

http://www.stanleyzinn.com/induction-heating/coil-design-1.html

I don't think anyone here really has a clue, myself included.

 

[hambone1971]

Experimenting with different coils and the Flux concentrators has been very interesting. I found that these open Ferites only draw about 3 amps with no case in the path and 7 amps at 30 V with a case such as a 223 -308 size. The 223 was annealing in about 2.5 seconds and the 308 in 4.5 seconds. Only doing a few at a time there is no appreciable increase in temperature on either the coil/ferite or on the induction board. this is all air cooled as well. Draw back is that the case is very close to the ferite so more care is needed also the case must be held in place well as when the induction starts it pushes the case out of the field. In comparison I made the 8 turn 1 /8 using 1/8 copper tube this maxes out my current power supply (while waiting for the 48V unit.) Also this coil really heats up without any water. I have it all plumbed for water but again waiting on a more suitable reticulation water unit. The time it takes to anneal is also longer.

David do you have to rotate the brass in the flux concentrator to get a even anneal? Would it take different ones to anneal larger brass? Where do you get these to experiment with? Looks interesting. Thanks HB

 

[SGK]

BTW how many in this thread have enjoyed the wealth of information provided herein and yet haven't made a contribution to maintaining the forum via upgrading? I finally got off my butt and upgraded. I was perhaps 'cheap' and only went to silver but every little helps. Come on guys, step up. This thread alone is well worth the $20.

 

[hambone1971]

I donated before I knew how to properly do it. No $$ on my name for it. HB

 

[SGK]

Contact Admin - I believe they can upgrade you manually.

 

[David101]

They seem to be evenly annealed. Really the brass is such a good conductor of heat the if heat is applied on both sides the other two are just heat conductivity heated. I have to different size gap the 308 shell only just clears the smaller of the two and the 223 works well in this. The other is about 3 mm bigger and a 223 takes about a half second longer and I think the 308 probably works better certainly easier to use. You do need a good base or other device to hold them in the center of the ferite as when the power is turned on the case wants to be pushed out of the field.

 

[GrocMax]

I once, and only once, managed to have a case drop into the coil a fraction off line and this caused it to lean against the coil causing a short. This completely destroyed the power supply even though it is supposed to have overload protection. Costly mistake.

Thinwall vinyl or PTFE heat shrink, or even just spray on laquer/clearcoat should eliminate this.

 

[GrocMax]

I notice that there has been a few builds made using Heat sinks under the board. I know there are some available that have heat transfer type stickers to hold them in place. But I was looking at just using old fashioned paste (not sure how to attach to board as yet).
Then I was also reading that the boards use a large ground plane as part of the circuit to obtain frequency. Has anybody noticed a change when applying the heatsinks.
Another concern I have with putting heatsinks on the bottom is the construction of the board looks like it has two halves in a layer of the board and placing a heatsink on one side would cause not issue I think but on the other side it would sit across the gap in the two halves. This could cause an issue changing the capacitance value. Again anyone notice and changes?
Sorry for all the questions I have read this entire thread at least 3 times have learnt a lot from everyone posting their experiences on what does and doesn't work. Hopefully my questions will help the next guy. When completed I will also be able to give back but then from an experience.

In most cases the caps have 'overcoats' which impede any air cooling, besides in this application the heat is generated inside the caps and travels down the very thick leads into the board. The ground plane side of the caps will spread the heat into the board fast, and heat sinks on the bottom with good airflow are really effective at cooling the unit. If you can't get the right sized heat sinks with the thermally conductive, electrically insulating double sided tape on them already, buy the tape separate, or there are also thermal conductive/elec insulating epoxies. Just make sure they are exactly that- conducts heat, and NOT electricity. This means NO general purpose epoxies with metal fillers.

I bought two of the 20A/1000W Sainsmart units, and have not had any issues with the first one, and run it HARD.

I'm not familiar at all with the other newer boards people are having problems with.

 

[GrocMax]

Depending on where you live, Mouser Electronics (my fave) carries a wide variety of ferrite cores. The ferrite core concentrates the flux field. If you are in AUS/NZ or Europe I'm sure there are comparable electronics retailers, but for me Mouser in Mansfield TX rocks, reasonable shipping, get it next day, and they don't screw up.



I was under the impression that if it was done correctly the case would self-center when the magnetic field comes on.

If you wind a coil with a CCW first 4 loops, and CW last 4 (or is that vice-versa?) the workpiece will self-center and levitate.

Beauty of the current limiting supplies or some way of current limiting is you can experiment with all kinds of crazy stuff and not worry about 'releasing the schmoo'.

 

[GrocMax]

If you are worried about flyback spikes going into the PS just put some good sized flyback (recirculation) diodes in, this keeps that big voltage spike from happening. But I believe the better PS units will already have them in place. Can't say I've ever measured mine for flyback, once I found the right relay and the high pressure pump its been 104% reliable.

I use it to silver braze small stainless tubes and fittings on occasion, inductive works awesome for making a really good silver braze joint.

 

[GrocMax]

The PS needs to be at 20v plus and the power needs to hit the board quick or the tank circuit will not start 'ringing' and the board will blow up (mosfets mostly). Lots of videos of people blowing up these boards trying to test them with a 12v or 18v battery, or using a test PS and trying to 'ramp up' slowly. The 'quick' part can be accomplished by either an SSR or a std coil and contact relay.

I tried using the switching function built into the RSP-750 PS and this was a complete fail as the PS tries to ramp up the output voltage and current slowly, fortunately for me the limiting features saved the board, the PS outputs shut down immediately.


I don't know if it makes a difference electrically but I chose to use the relay on the negative leg of the board only, I'm not sure there is any benefit at all to using a relay pole on both the + and negative sides, other than if one sticks or fails to release the other will. Maybe a real EE could chime in here.

Hi Guys,

I started planning a build for an induction annealer a while back and stumbled onto this thread discovering that all my research was unnecessary as you guys essentially proved my idea would work... So thanks for such a great resource!

I have a technical question I'm hoping one of you might help with.

My build is slightly different in that in between my cheap Chinese PSU and the induction board, I'm using a DPS5020 (https://www.banggood.com/RUIDENG-DP...ply-p-1181200.html?ID=514816&cur_warehouse=CN) this is a buck converter essentially so I can do current and voltage limit/adjust (as well as it reads the draw) - very cool bit of kit for the price and certainly reduces some complexity in the design.

At the moment I'm also using an SSR to switch the power to the inductor, but the DPS5020 has a built-in on off for its output.... I remember reading somewhere it's important to have the supply ramp up fast for inductor resonance (hence my test used SSR to make sure components work) otherwise something will let out magic smoke?

I'd like to see if the DPS5020 is able to power on fast enough that I can skip the SSR (the DPS has a usb/serial interface that you can use to turn it on/off so i can simplify design if SSR not needed) - but I'm not sure how to test this and don't want to fry anything. I don't have access to an oscilloscope that I know of (which I have a feeling is the easy way to do this).. so alternatives on a postcard please

if the oscilloscope is the only way could someone detail this test also (and I'll ask around the club see if anyone has one I can make use of).

Thanks again and kind regards

Tim

 

[GrocMax]

Don't know about the rest of you, but for me, mostly thinner wall 223 based brass, trying to get full anneal in the neck is a big fail, as they get too soft and have issues with collapsed necks (even with light bullet 'grip') during reloading- TEST FIRST before you do all your valuable competition or rare or expensive brass!

My advice would be not to go full 750F tempilaq, but sneak up on it. Maybe it will need it, but you CAN'T undue it easily if its too soft! Figure out how much anneal time it takes to get 700+ on a test piece, then back off. Once its ruined, its pretty much ruined. You can always make it softer, the ONLY way to harden it is work hardening!

 

[GrocMax]

These are generally identical parts rebranded, you may be able to find proper diagrams from a different 'brand'. Or take a close look at the board, quite often the pin assignments are silk screened onto the board next to the connector or elsewhere on the board.

Need some help here for a builder. Any one seen this v/a type of meter? Screen name of builder is Sprin Go


Hi Regina,

Thank you for your previous answer, I got some thick wire and doing the wiring now.
According to somebody's wiring diagram from the forum, it says that the volt/amp meter has black, yellow and red wires.
Well, mine has blue, green and yellow.
As expected, I got no user's manual with it.
Can you please look at the photo below and try to figure out which wire is which?
Thanks a million in advance!
Dejan

Sprin Go, Today at 3:44 A

 

[GrocMax]

This seems to be the center for electronics enthusiasts here, I'd like to share something I've had for several years that I found has a much higher value than its price-

Digilent Electronics Explorer unit.

I don't claim it is lab quality or anything (not that I could ever tell with the relatively slow stuff I work on), but if you do projects you could use several power supply outputs (-10v to +10v, 5v, 3.3v, logic level, some with current limiting), 4 channel scope, 2 channel waveform generator, sniffing out digital communications, all on a std breadboard format, it might be for you. They are being discontinued, only ~500 left. $400.00 USD.

https://store.digilentinc.com/elect...lloscope-multimeter-workstation-limited-time/

 

[SGK]

They seem to be evenly annealed. Really the brass is such a good conductor of heat the if heat is applied on both sides the other two are just heat conductivity heated.

I found that only a few millimetres off-centre with a cylindrical work coil led to heating further down one side of the case than the other. This is why I went to the trouble to add the inserts to my design, even one for 308 cases which reduced the drop hole diameter only about 1.5mm.

 

[SGK]

Don't know about the rest of you, but for me, mostly thinner wall 223 based brass, trying to get full anneal in the neck is a big fail, as they get too soft and have issues with collapsed necks (even with light bullet 'grip') during reloading- TEST FIRST before you do all your valuable competition or rare or expensive brass!

Where are you positioning the shoulder-body junction in the work coil? Have you tried shifting the case higher?

(It is almost cheaper for Mouser to deliver to Europe than to Florida - and faster. And they ship to Europe from TX.)

 

[David101]

Like you I have two of the 20A/1000W Sainsmart units one is still stock standard the other is getting some mods. The bigger heatsinks on the mosfets is not necessary but looks good . I used the two heatsinks taken off the mosfets and cut them to size to put on the bottom of the PCB. I used normal heatsink paste (as that is what I have) in the middle of the heatsink and some araldite at each edge to glue them to the board. I am going to add two more capacitors as well as I think a lower frequency Not by much just more around the 100kHz is where I want to be. Using the coil yes the work piece wants to self center using the ferite it pushes out. Using the ferites the current is so much less no fans are needed no water cooling the board and caps stay cool. There is a slight unevenness of the temperature around the case as I just went and checked specifically. But I still have reservations although I am sure the AMP unit is based on them the Annie and Fluxeon have gone away from them. I can not recommend Fluxeon even though that is where I got the ferites they do not answer any email questions or pre sales questions.

Very interesting your comments about the 223. I have just been doing some looking at the tempilaq change also in a dark room for a brass colour change. I found the temilaq changes well before I could see a colour change. But when the neck did change it went from nothing to a full bright red for a instant with an extra 0.1 sec there was probably a 0.5 sec spread from tempilaq change or more getting to that point however.

 

[GrocMax]

Its totally a result of over-annealing in my case, not the coil. We really don't want a full dead soft neck anyways, it wasn't mfg that way in the first place. Different cases react differently. I'm just saying sneak up on it, its not easy to reharden via neckup/neckdown operations, easier to re-anneal with a little more time added. Its probably more accurate to say I'm normalizing/stress relieving more than annealing. Softens it a little, not to dead soft condition.

My coil has a 'bend the field sharply' loop at the top and the bottom, and is shaped to hit the shoulder and neck more evenly, also I can do very short cases, and it concentrates the field in a much narrower vertical distance. As long as there are 8 full coil loops, and the tubing is the same, and the avg diameter is the same, it will behave with the same inductance.

 

[David101]

Its totally a result of over-annealing in my case, not the coil. We really don't want a full dead soft neck anyways, it wasn't mfg that way in the first place. Different cases react differently. I'm just saying sneak up on it, its not easy to reharden via neckup/neckdown operations, easier to re-anneal with a little more time added. Its probably more accurate to say I'm normalizing/stress relieving more than annealing. Softens it a little, not to dead soft condition.

My coil has a 'bend the field sharply' loop at the top and the bottom, and is shaped to hit the shoulder and neck more evenly, also I can do very short cases, and it concentrates the field in a much narrower vertical distance. As long as there are 8 full coil loops, and the tubing is the same, and the avg diameter is the same, it will behave with the same inductance.

But its also the way the case heats using the ferites as I have not used the coil much yet due to no water coiling the case heats very much from the top down. Perhaps more experimentation with case height? not sure. Does the shoulder really need to be that soft to bump it back if needed? perhaps we are concentrating too much on trying to get that factory look which over does the neck?