My DIY Annealer Goals & Plan Summary
I reread this thread again now to update my annealer plans. I'll review here what I'm planning to do for my setup.
Many 7mmTCU cases were splitting necks. These are formed from 223 brass and the necks are very thin and work harden in the process. I had done propane torch annealing of cases by hand in the distant past, but consistency was not great and it was not convenient. This triggered the research and review of annealing setups including this massive thread on the GinaEricks, so I started planning my own and collecting parts. My first post here was #2025 on p103 in 2020. The project was interrupted and we're just getting back to it.
Variant on a DIY Inductive Annealer
One of the nice things about this project is we can each adjust the design to suit our own abilities, interests and needs.
Ideas for this version have been gleaned from many places including but not limited to this thread and forum, plus a few of my own twists based on my experience as an engineer. A big thanks to everyone who has shared their plans and experiences with their projects online and provided an opportunity to learn by studying their results.
Project Goals
Extend life of brass.
Improve consistency.
Fit from 221 Fireball to 300 Win Mag.
Automate - but don't go too fast.
5-9 seconds anneal time is desired for best consistency.
Control the power to adjust time.
Sense the case glowing if possible.
Use simple hardware, put any complexity into the software and 3D printed components that are easy to adjust and improve.
Build in Equipment Protection and make the machine reliable.
Component Choices
Most parts from Amazon. Many were collected in 2020, some recently and a few still to get.
1/8" or 3mm CU tubing for initial coil, 7.5 turns single layer, 1" ID, wound on a 3D printed removable mandrel, possibly prefilling tubing with salt to maximize coolant flow channel and control kinking. (3D print a filling funnel for salt). Requires 775mm length of tubing including 50mm ends, solder into 1/4" tubing for ZVS and coolant interface. A spreadsheet was used to analyze many different coil configurations and choose this initial one.
Use the common 50V 20A 1000W ZVS pcb, lots of experience and availability, some quality issues - plan to carefully QC, resolder, pretest, and modify the pcb for direct gate PWM control.
Get Spare ZVS FETs and diodes (IRFP260N, 1N4742A and FR307) for repairs of this PCB if needed. Perhaps keep a full spare ZVS pcb on hand, this seems to be a primary failure point.
Use two power supplies:
PS1 12V/5V MW RD-65A 65W for logic and fan/pump power includes PFC (power factor correction) and automatic input voltage switching.
PS2 48V MW RSP-750-48 750W 15.7A for ZVS Power, has remote on-off control, PFC, automatic input voltage switching and current limiting control options.
Pi Foundation Pico dual core 133 mHz microcontroller, w/ MicroPython for all timing and programmed controls. Note this is not a Raspberry Pi, but a relatively new Microcontroller produced by the same organization, and it has a lot of excellent features, I/O, libraries and documentation at a very reasonable price point. It can be programmed in several ways including Arduino C/C++, however I prefer to use the MicroPyton environment which avoids the C/C++ complexity and compile cycles. It is easily programmed via USB serial connection from Thonny (free) or other tools can be used.
Robotics pcb for the Pico w servo & motor controls is a convenient and cost effective off the shelf pcb that has a lot of specifically useful features for this project and accepts the Pico microcontroller.
2.4" I2C OLED and rotary encoder with push switch for the user interface. The Pico is available with Wifi and Bluetooth capability for those wanting to interface to phones etc but I don't plan to do that. The display and knob is planned instead.
Digital Sensors DS18B20 for ZVS pcb, Coil and coolant temperatures. Monitoring temperatures is important to insure the machine is operating correctly and avoid some common failure modes.
Current Sense pcb 30A, Voltage Sense pcb 20:1 - Monitoring the ZVS power current and voltage is critical to insuring proper operating and shutting down quickly when problems are developing. This is also useful for displaying diagnostics and status and avoids the need for a separate current shunt and meter while reducing overall project wiring.
Perisaltic Pump 12V - These pumps have low flow but develop pressure suitable for pushing coolant through the small ID of the coil tubing.
Computer Liquid Cooling Radiator w 12V fan.
12V fan to cool ZVS pcb - between the capacitors heating and the I squared R losses in the ZVS pcb there is a lot of heat in the pcb which needs to be monitored and dissipated for reliable operation.
2 metal gear RC servos for 3D Printed Case Feeder and Dropper - Readily available, inexpensive and simple to control while metal gears insure long life. Unlike motors and solenoids that require special drivers and are often discontinued without suitable replacements RC servos are standardized and available from many sources with compatible mounting and wiring. No spring is required, they have capability to push and pull, and their endpoints and motion speed can be adjusted in software.
2 small opto isolated 5v coil relays, one to control the 48V PS remote input, the other to turn on bias power to ZVS PCB gates - these commonly available relay pcbs with onboard drivers are used for Arduino and robotics projects and are suitable for driving low power controls from the Pico microcontroller.
Emergency crash off switch interrupts mains power - These machine safety switches are often required on equipment and are a convenient way to quickly remove power in case of operational problems. They can handle 10 amps AC at line voltage which is plenty of mains power for the annealer. They can be used as the primary power switch, or combined with a master power switch and used only for the crash off function.
Power Inlet with 10A Fuse. Available with or without master power switch. Uses standard removable power cords.
IR Flame Sensor - these appear to be useful for sensing or even controlling the annealing temperatures, and further testing is warranted.
Skipped Components
Some components have been avoided in this design.
No timer - the Pico can handle precise timing and avoids the procurement and wiring associated with a separate timer.
No DC SSR, Contactor or Relay feeding the ZVS is used in this system - instead the ZVS FET gates are clamped off, 48V power is turned on by a low level signal using a power supply remote control feature, after it stabilizes the ZVS gates are PWM controlled for variable power control and pulse timing. A number of folks have done this, it does require some minor surgical modifications to the ZVS pcb.
No solenoids - solenoids are not standardized, require significant drive current, and are mechanically problematic as well as being difficult to second source. They require springs that can be problematic to source as well. Metal gear servos are used here instead.
No separate shunt and DC volt/amp meter is used. The Micro really should monitor these values, and if this is done the separate metering is redundant and requires a lot of extra wiring which can be avoided.
Electrical Safety
No exposed voltages over 50V - keep the AC power wiring covered for personnel safety. Fused mains power. Emergency Shutoff switch.
Monitoring
Voltage, current and temperature monitoring are all handled with sensors connected to the microcontroller and displayed on the OLED display. This minimizes wiring and allows the software to handle problems quickly and hopefully avoids some of the damage that results from heat buildup or slow response to problem conditions like power supply overload or shutdown.
Coolant
The plan is to keep the liquid cooled items low in the frame over an integrated drip pan and put electronics above the liquid cooling to keep any leaks or drips contained and out of the electronics. Open frame construction similar to 3D printers is planned but details will be worked out as the project goes forward.
Power Supply Current Limiting
There has been some confusion about current limiting on power supplies. The only thing a DC power supply can do to limit current is to reduce its voltage output. It has no other control mechanism. In this case reducing voltage below 12V can destroy the ZVS pcb so while it is useful initially to protect the FETs from overcurrent it is ultimately not the sole mechanism to protect the hardware. The supply chosen here has programmable current limiting but we're not planning to use it other than at full output for overload protection, and the micro will be monitoring current and voltage to shut down the ZVS gates and the power supply itself if limits are approached.
PWM Control
PWM controlling the ZVS pcb FET gates is a way of effectively turning the voltage down to control power levels. Normally the voltage adjustability of the power supply is limited and not practical to change on a per cartridge type basis. The primary power adjustments were making different work coils or changing the position of the brass in the coil which is necessary but not really sufficient. By using direct gate PWM and reducing the duty cycle the equivalent of a wide range voltage adjustment is enabled, and it can be a parameter optimized for each cartridge as anneal times are fine tuned in the desired range. This allows extending too-short anneal timings as well as moderating heat buildup in the ZVS pcb. It also facilitates testing at lower power and safely creeping upward on power and heat levels. If the IR flame detector is used it allows controlling the actual temperature of the brass in the desired temperature range for real time control of the annealing cycle temperature profile.
Coil Physics
There are a lot of misunderstandings about induction coils and magnetics. One example that is commonly mentioned is that the field falls off with distance squared or cubed. This is not the case inside the coil where the field is fairly constant. The field outside the coil does fall off rapidly. Another fallacy is that making the coil smaller increases the field, but the equations for field strength do not involve coil diameter, only current and turns per unit length. Making the coil diameter smaller reduces resistance, inductance and increases frequency which increases FET and capacitor losses. The reduction in resistance increases current when voltage is constant so this can increase the field but not if the coil is lengthened to maintain inductance and frequency. Many folks put the entire case into the coil, but we don't actually want to heat the lower part of the case. This puts excess load on the ZVS board and heats things we don't want to heat. Only the neck/shoulder should be in or near the coil. Overlapping windings and reducing coil length or increasing current in the coil increases field strength. Higher voltages will drive higher currents into the coil providing the supply and ZVS pcb can handle it. Coil current and turns per unit length are the keys to field strength. Overlapping turns does increase turns per unit length but these longer outside turns add a lot of resistance which reduces current. Look for study material on solenoid coils and read about magnetic fields for more information. In any case many interesting experiments have been done here and the results are useful even if not completely understood...
3D Printing
The 3D printer is the best tool I have ever purchased. If you like to make things and want a little robot to do most of the work for you - look into 3D printing. It can do more than you realize, is less expensive than many other workshop tools and has a wider range of applications.
So that's my current plan, slowly evolving and hopefully taking shape soon. The pile of components is steadily growing. Thanks to everyone contributing their experiences, comments and knowledge to this and related internet discussions. Any comments or feedback on this plan are encouraged. Thanks in advance.
