Discussion in 'Reloading Forum (All Calibers)' started by Gina1, Aug 26, 2016.
The goal of my design was an air cooling system.
Read the Wikipedia on "Litz wire" and how it works. They also mention silver plated copper tubing to reduce resistance at high frequency. I was also thinking of using water cooling since I already made a standard water cooled system. David101 talks about the pipe not working on page 78 of this post, not sure what the issue was. I was thinking of just using the water cooling on the wires going to the coil and not the Litz coil itself. The idea being that the heat would transfer from the Litz wire back through to the pipe with the water cooling on the circuit boards +/- terminals, though the heat transfer may not be enough to be worth it. I may be overthinking the issue but if I jumper the +/- leads using a piece of insulated tubing that is long enough the resistance of water increases such that the effect on the circuit is minimal (the radiator itself is metal with insulated tubing connecting the +/- terminals also, and the demin water eventually becomes "more" conductive when exposed to air, the metal of the radiator, and the inhibitor additive). Fluxeon makes a ferrite that is water cooled but its a different version vs the "concentrator" version and my belief is that it is no better than the standard coil (though I could be wrong).
I tried the pipe around the ferrite to see if it would work. It didnt. I was some what surprised and being an electronics person I didnt consider the 160kHz a high enough frequency to really have any "skin" effect so I was skeptical of the Litz wire.
I also have some very fine multiple strand copper wire in quite a few different diameters that I used at work. for a different reason. I used it as flexible high current wire it has quite thick rubber or silicone insulation. One day I am going to get a piece take off the insulation and slip on some of the fiberglass insulation and try that as Litz wire just to see if that works the same. I have lots of it and in quite a few different sizes.
I would suggest having a look on the "bay" and "Ali" stores at ferrite square section. Certainly now after trying a bought one. The square section can be cut to the lengths you want to create your Horse shoe design. The parts can be glued together using something like JB weld or one of the other Cold Weld epoxies.
What I would also be experimenting with is the shape of the ferite that is presented to the neck of the case. For example it may be a 1 inch square section of ferite but when you get to the ends near the neck I would shape this to perhaps half the thickness. I actually like the square section better than the fluxon circular section of ferite at this point. The circular makes the winding the wire around the ferite easier but I dont think it is the best shape to be aimed at the case.
So I have completed the bulk of my build. Using a 4mm 8 turn coil at 45v I am showing 7.5 amps with no case and 10.4 amps with a Hornady 6.5 Creedmoore case. Using 750F Tempilaq it takes 5.2 seconds to anneal a 6.5 Hornady case and 7.6 seconds to anneal a 6.5 Lapua case. This is with the case mouth almost even with the top of the coil. I also have a 7 turn coil I will try to see if it makes a difference.
You should not be surprised. The coil impedance (Z = R + jX) determines the current in the coil. The Litz wire has many individual "skin effect" Z-s connected in parallel (overall very low resistance) compare to only one "skin effect" Z tubing coil. In addition, the inductance in the core is proportional to the number of the coil turns. May be, if the coil has higher number of turns, the result would be more encouraging.
I totally agree that square profile of the ferrit core is a better choice.
As far as cooling the entire induction assembly - a suitable non magnetic enclosure filled with circulating cooling oil (like power distribution transformers) would do the job.
Mouser carries a large variety of pre-formed and machineable ferrite cores.
Hello from Poland, I built my GinaErick annealer.
Gina1 and Hollywood thank you very much.
Anyone considered the use of a thermopile camera for further automation? I'm currently ingrossed in some Youtube edumacation to up my scripting game enough to be able to tie one of these into my planned build (if and when I get the time). I'd like to bypass the use of Tempilaq and sacrificing brass to the annealing gods if possible.
Right off the top, one of the problems would be the case is inside the induction coil. The IR camera would not have a clear picture of the case as it was heating up.
With reference to my build,
As Gina said, thermal imaging or using an infrared thermometer will not work due to the coil being in the way, so Tempilaq is really the most obvious way to go. Also, why are you sacrificing brass?
On my unit (using the Sestos timer) I use a momentary switch connected to the gate switch contacts to freeze the timer as soon as the Tempilaq changes color (hold it), then punch the aux drop button to drop the case out of the coil, and then the relay disable switch to release the relay powering the heater board so its no longer generating a field. Then I look at the timer (which is paused by the gate switch) and note the elapsed time to melt the Tempilaq, then release the gate switch and press stop. I usually drop a tenth or two off of the displayed time for my reaction time and retest another case with a normal cycle at that duration (I start the first test with an arbitrarily high time, like 15 seconds, so I know it will not time out before annealing temperature is reached).
If I recall correctly annealing is accomplished in the range of 700 to 900 F, so if you are off a little one way or the other from 750 F, its not a big deal, as long as you are consistent with that timer setting for that brand/lot/caliber of brass from then on.
Non-contact temp measurement can be precisely aimed. Its just a question of getting the right probe $y$tem.
Do you have an example in mind? and at what price point? The one he linked to seemed awfully pricey at $277, and looks to require some sort of software solution to interpreting the image it provides.
If there is a reasonable way to detect temp of the case in the coil and have it trigger a drop and stop heating, you could do away with the timer altogether and just control cycle timing with manual/automatic feed/start.
Timed and/or non-contact IR-switched induction annealer control.
1. Timed, no case, no IR switch.
2. IR switched, multiple times. Case starts cold then warms up.
3. IR switched.
4. IR switched.
5. Obstructed IR switch cooks case. Obstruction removed, IR sensor interrupts timer.
6. Obstructed IR switch cooks case. Obstruction removed, IR sensor interrupts timer.
The yellow green light is the work coil on. The blue light is the timer powering the tank circuit.
Only the IR sensor's perceived temperature of the brass matters, not the acutal temperature. It only matters that it switches consistently. IMO the timer is less accurate especially with high-wattage annealers on small cases with different neck/shoulder-thicknesses. Calibrate the sensor with Tempilaq then leave it alone for a specific cartridge.
IR control is better than a timer because:
• Can do mixed headstamps.
• dirty brass with foreign residue (water, dirt, lube) reach the same temperature.
• Ambient/brass temperatures do not affect degree of annealing.
• With long annealing strings annealer hardware components heat up and can change effective output.
^ I like this...seems like once you get the IR sensor set up correctly, other than coil height for different sized cases, you would never need to change anything regardless of brand/cart/lot.
What parts did you use?
I'd probably use a Fluxeon induction coil assembly and have an open field of view to the case side. But the thermopile camera looking down on the coil from a slight angle should present enough case neck and shoulder for a few pixels and have the control set to only requiring a few pixels at annealing temp to drop the case out. Still spitballing at this stage though
Thanks to McFred's work I went ahead and tried a simple flame sensor just to see if that would work. These are dirt cheap and consist of a 4 pin connector (+5VDC, COM, Digital Out, Analog Out), they have no focusing mechanism and are a simple IR sensitive phototransistor. I connected up the digital output and adjusted the pot so it was just on the edge of triggering. The AD converter on the Arduino can be a little slow so I figured the digital signal would be better. It was difficult to get in the correct position for it to see the brass, once done I could get it to trigger just as the case started to glow. The unit also needed to be positioned far enough away that the induction coil didn't fry it. I agree, the Fluxeon unit would be a better way of focusing the phototransistor. I ordered the Fluxeon unit a few weeks ago and it finally arrived. I need to get some additional time to take apart my water cooled unit to try this out. McFred, what IR unit did you use?
I have a question and I don't mean to be lazy about finding an answer, but I've read this thread a couple of times and after receiving the first shipment of parts the question came to me.
What gauge of wire is needed to connect the 110v the Power supply?
That being asked, I purchased a 48v 20A 1000W power and will be using the 1800W induction board. So, anybody who has built something similar, your feedback would be awesome. Anybody who can give me a solid answer, your feedback is also greatly appreciated.
For the 110 volt power, from the 110 volt AC power to the 48 volt power supply (going into your build) use a 14 gauge stranded line cod. Stranded because you want it to be flexible where you plug it in to the 110 volt outlet.
Since you have a 1000 watt power supply AND the 1800 watt induction PCB I would go with 12 or 10 gauge wire. That much current, any small loss (in the wire) will throw your current readings off. Can be stranded or solid wire, as your dealing with DC at this point.
Hope this helps
Gina1 thank you and Thank Hollywood for making this open source machine available. You two are the tops! Also, thank you for the feedback, I found the DC chart so I settled on 10AWG but that cord was what was getting me. I wanted to make sure I wasn't building an impending electrical fire. I asked for the Heaviest cord the local electronics shop had and they gave me a multi-strand 14 AWG so it looks like I lucked out and don't need to make another trip. Thank You again for your feedback. I can't wait to scrap my little DIY propane annealer, the cheap motors are already running out of true and one of them just cant keep time anymore.
Separate names with a comma.