Induction brass annealer redux

[hdmunger]

I'm not expecting to have the flame sensor operate a relay big enough to switch the ZVS board, just as a continuity switch to trigger either the 'stop' on the Sestos timer or the drop solenoid for the case to be removed from the coil...

 

[itchyTF]

You could run something like this directly off the 393 if the coil current is low enough. 5v and 500 ohms would probably be ok. Make sure you put a diode across the coil.

8L01-05-111 Coto Technology | Relays | DigiKey

Order Coto Technology 8L01-05-111 (306-1022-ND) at DigiKey. Check stock and pricing, view product specifications, and order online.

www.digikey.com

 

[radmcg]

Hey Guys. I know this is an old thread but I am quarantined in my house with covid for the second time and had some time to kill. Are there any schematics of plans for this annealer? I have read most of the 139 pages of post and see the photos and build list but no actual plans, ie wiring diagram. I'm sure I'm just missing it but see references to rev 2 in the original post but don't see actual schematics. As for the last 20 pages that is all way over my head I just need a wiring diagram to follow for rev 2. If Ive missed something I apologize.
Rad

 

[hdmunger]

Hey Guys. I know this is an old thread but I am quarantined in my house with covid for the second time and had some time to kill. Are there any schematics of plans for this annealer? I have read most of the 139 pages of post and see the photos and build list but no actual plans, ie wiring diagram. I'm sure I'm just missing it but see references to rev 2 in the original post but don't see actual schematics. As for the last 20 pages that is all way over my head I just need a wiring diagram to follow for rev 2. If Ive missed something I apologize.
Rad

Gina says there's one on page 110

 

[hdmunger]

Sorry this one is so long...

With the parts I have, this is the diagram of how I am thinking this would work (looking for a simple, cheap solution with no computational overhead that can easily be implemented in a variety of setups).



Flame Sensor - https://www.amazon.com/Infrared-Sensor-Detector-Smartsense-Arduino/dp/B07D1HNHN4

Specifications:
The working voltage: 3.3V-5V
The output format: DO digital switching outputs (0 and 1) and AO analog voltage output (not on this unit)
Comparator: wide voltage LM393
Length: 4.5cm / 1.77in
Weight: 4g

Product wiring description:
VCC: 3.3-5V power supply "+"
GND: power supply "-"
DO: TTL switch signal output.
AO: analog signal output (voltage signal) (again, not on this unit)

Instructions for use:
1. the flame sensor is most sensitive to flame, for ordinary light, it is also responsive
, generally used as a flame alarm purposes.
2. the output interface can be directly connected to the micro-controller IO port.
3. The sensor must be kept at a distance from the flame to avoid damage to the sensor.Lighter flame test distance of 80cm, for the greater flame, the farther the distance test
4. Analog output and the AD conversion process, can achieve higher accuracy
5. When the sensor detects a flame, sunlight, or infrared light, if achieve potentiometer setting threshold, the green indicator light will ON, the DO output low level (0-0.1 - V), the green light is OFF, the DO output about 3 V high voltage level


Relay - https://www.amazon.com/Valefod-1-Channel-Optocoupler-Isolation-Trigger/dp/B07WQH63FB

Specifications:
- Relay Type: 5V 1-Channel
- TTL gate circuit
- Quiescent Current: 5mA
- Maximum Current: 190mA
- Trigger Current: 3 - 5mA
- Trigger Voltage(LOW): 0 - 1.5V
- Trigger Voltage(HIGH): 2.5 - 5V
- NO Maximum Load: AC 250V, 10A or DC 30V, 10A
- Power Indicator: Green
- Relay Status Indicator: Red
- Size(L x W x H): 51*26*19mm
- Weight: 17g

How high/low level triggers selection terminal
- High level trigger acts when jumper wire and high pin are short connected
- Low level trigger acts when jumper wire and low pin are short connected

Input interface:

DC+ : positive pole of power supply (VCC, voltage=5V)
DC- : negative pole of power supply(GND)
IN : connection to control relay pickup by high/low level trigger

Output interface:
NO : connection of normally open state(the relay switch is dangling before pickup, the switch will connect with "COM" after pickup)
COM : common interface of relay switch
NC : connection of normally closed state(the relay switch is connected with "COM" before pickup, the switch will be dangling after pickup)

So, where am I going wrong? Where would a diode be placed in this setup? What other parts am I missing?

Also, can the IR sensor on the flame sensor unit be removed from the PCB and the wiring extended so that the sensor can be more easily placed to 'see' the case neck in the coil (without compromising its ability to reliably sense)?

 

[itchyTF]

That looks like it would work. The red board probably already has a diode on it if the output device is a mechanical relay.

You could probably move the ir led but don’t let the leads get too near the inductor coil. It might generate a signal to the flame detector.

 

[hdmunger]

That looks like it would work. The red board probably already has a diode on it if the output device is a mechanical relay.

You could probably move the ir led but don’t let the leads get too near the inductor coil. It might generate a signal to the flame detector.

Good point...I'll probably have the sensor at least 2 or 3 inches from the coil and pointing down at about a 45-degree angle to keep it out of the magnetic field when that time comes (unless being mounted almost directly above the coil would be better? due to the 'toroidal' shape of the field, if I recall correctly)

The parts 'should' be here Monday, so I'll play with it and let you know...

 

[itchyTF]

Check out post 2,543. His is a little more than an inch away.

 

[hdmunger]

Check out post 2,543. His is a little more than an inch away.

I see that...should make it easier to get a good reading...my intent is to put the sensor (remotely from the PCB as discussed) inside of a tube pointed at the case mouth area so that any extraneous sensor input can be reduced.

In the post before that, 2542, in the second paragraph, he talks about the analog vs. digital outputs, and his take is the opposite of what you and I are understanding. Do you think he just misspoke?

 

[itchyTF]

Since I don’t have the 4 pin type I can’t say for sure, but I *think* the analog output comes right off the ir sensor (LED?). I believe his reference to 10 bit resolution (0-1023) is from the A/D input capability of the Arduino (or whatever board he is using).
The LM393 is just a comparator. The pot just sets an analog reference which is compared to the output of the ir sensor. The digital output is 1 bit (high/low).

 

[hdmunger]

Thats my understanding also. The specifications for the flame sensor shown in my above post are apparently for the 4-pin model (even though listed on the page for the 3-pin) since they reference both the analog and digital outputs. This leads me to believe that there is no difference in the functions from model to model, only whether the analog is also included or not...

Since the digital-only models are less expensive, my hope is that it all functions as shown in the above diagram. That would mean you could fabricate 5 sensors using the packages I bought for $25 ($5 each) or less if you found a better deal on components.

 

[whatsupdoc]

For those using the Infrared Flame Sensor couple of questions.
How do you compensate for different brass heights?
Does ambient light affect the sensor.

 

[itchyTF]

I used Lok-Line (modularhose.com) to position the sensor. Easy to adjust for X, Y & Z.
I also designed a holder for the board (which a friend printed) that shields the led from the sides.

 

[ottsm]

I'm going to have to break down and get a 3D printer one of these days.

 

[hdmunger]

Well, the combination I posted above in #2765 does indeed appear to work.

I cobbled it together and tested it so that the relay switched the drop solenoid, and the case dropped out annealed. Some observations of note:

- the polarity of the digital output on the flame sensor is as depicted...relay set to low voltage activation, flame sensor outputs high signal until hot case sensed then the low signal causes the relay to close, dropping the case

- setting the pot on the flame sensor came down to turning it fully clockwise (most sensitive) then backing it off around 1/12 to 1/8 of a turn (only has about 3/4 of a turn from stop to stop). I arrived at this initial setting by powering on the circuit at full clockwise, the relay closed, backed it off until the relay opened, then a hair more

- enclosing the flame sensor in a black plastic tube made it more reliably function without random activations from ambient IR sources

- when testing, I was reminded of how important setting the coil height is...I initially was annealing about halfway down the body before I remembered to set it correctly, then it worked well with the flame sensor causing the case to drop out of the coil at about the same point it would have with timing controlling the drop

- I set the timer to power the ZVS for a couple seconds longer than I needed, so I got the following progression: annealing starts, flame sensor circuit dropped case at anneal completion, timer ran for the remaining 2 seconds or so then shut down the ZVS, the timer activated the drop solenoid again while coil was empty, and the timer cycle completed

This first test used a 4-pin flame sensor with both digital and analog outputs, and tomorrow I should receive a few 3-pin sensors for final testing...I don't expect there to be any difference since I am just ignoring the analog output and only bought it before it occurred to me that I might be able to accomplish this without the overhead of an Arduino, just these 2 individual circuits.

I need to design/print a mounting enclosure for these two PCB's and explore making the IR sensor on the flame sensor a remote fixture that doesn't get in the way of loading cases into the coil, but is robust enough to not be in danger of being damaged.

 

[dskogman]

We’ll my latest version of the annealer ran for 3 years but I lost the power supply last week. It was a 36v 1000w chinesium so I decided to upgrade to a 48v 1000w Meanwell. Unfortunately it doesn’t seem to be working very well. The voltage drops to around 25v and takes longer to heat up. I’m guessing the power supply is cutting back the voltage to keep the current down so maybe the induction board is damaged which blew the cheap power supply ?

 

[oliverpsmile]

Well, the combination I posted above in #2765 does indeed appear to work.

I cobbled it together and tested it so that the relay switched the drop solenoid, and the case dropped out annealed. Some observations of note:

- the polarity of the digital output on the flame sensor is as depicted...relay set to low voltage activation, flame sensor outputs high signal until hot case sensed then the low signal causes the relay to close, dropping the case

- setting the pot on the flame sensor came down to turning it fully clockwise (most sensitive) then backing it off around 1/12 to 1/8 of a turn (only has about 3/4 of a turn from stop to stop). I arrived at this initial setting by powering on the circuit at full clockwise, the relay closed, backed it off until the relay opened, then a hair more

- enclosing the flame sensor in a black plastic tube made it more reliably function without random activations from ambient IR sources

- when testing, I was reminded of how important setting the coil height is...I initially was annealing about halfway down the body before I remembered to set it correctly, then it worked well with the flame sensor causing the case to drop out of the coil at about the same point it would have with timing controlling the drop

- I set the timer to power the ZVS for a couple seconds longer than I needed, so I got the following progression: annealing starts, flame sensor circuit dropped case at anneal completion, timer ran for the remaining 2 seconds or so then shut down the ZVS, the timer activated the drop solenoid again while coil was empty, and the timer cycle completed

This first test used a 4-pin flame sensor with both digital and analog outputs, and tomorrow I should receive a few 3-pin sensors for final testing...I don't expect there to be any difference since I am just ignoring the analog output and only bought it before it occurred to me that I might be able to accomplish this without the overhead of an Arduino, just these 2 individual circuits.

I need to design/print a mounting enclosure for these two PCB's and explore making the IR sensor on the flame sensor a remote fixture that doesn't get in the way of loading cases into the coil, but is robust enough to not be in danger of being damaged.

Good job! I did separate the sensor from the PC board. I you do so - keep the same polarity. I hope, that the new pot will be more precise then the original.

How do you evaluate your cases?

The only think left for my Arduino controlled built is to calibrate the flame sensor. Also, I had to replace the current sensor.

 

[hdmunger]

Good job! I did separate the sensor from the PC board. I you do so - keep the same polarity. I hope, that the new pot will be more precise then the original.

How do you evaluate your cases?

The only think left for my Arduino controlled built is to calibrate the flame sensor. Also, I had to replace the current sensor.

I noticed that the flame sensor PCB has a '+' and a '-' on the solder points for the IR sensor, so I will honor that when relocating the sensor.

Yes, the pot is not as precisely controllable as I would like, but the response time on my setup is probably quicker than if it had to be processed by an Arduino using the analog output.

At this point, I have only compared the approximate time the flame sensor dropped the case out of the coil to the time I used to have them drop at. Recent reading has me led me to believe that the old 750 Tempilaq target was probably lower than necessary for the actual time that the case is being heated and how quickly it cools when removed from the magnetic field. I have read that for the quick heat cycles we are using, 900 to 1000F is probably more appropriate. For me, the useful temperature window is wide enough that I am only looking for the top of the case to start to glow slightly, but as long as the result is not so soft that crushed necks result from sizing, I am happy.

 

[VenatusDominus]

I see that...should make it easier to get a good reading...my intent is to put the sensor (remotely from the PCB as discussed) inside of a tube pointed at the case mouth area so that any extraneous sensor input can be reduced.

In the post before that, 2542, in the second paragraph, he talks about the analog vs. digital outputs, and his take is the opposite of what you and I are understanding. Do you think he just misspoke?

I used the digital output so that I could adjust my threshold via the Arduino and not have to adjust the POT on the chip ( coarse adjustment compared with the 1024 available digital adjustments). But for you implamentation, the analog output and adjusting via the POT would be the way to go, especially if you changed the POT out for one with a finer adjustment.

 

[VenatusDominus]

For those using the Infrared Flame Sensor couple of questions.
How do you compensate for different brass heights?
Does ambient light affect the sensor.

My base where the case sits adjusts up and down to allow the case neck to sit in the same spot in the coil, the the sensor is always looking at the same area of the case regardless of case length. Yes ambient light affects the sensor and I adjusted mine for the ambient light that I anneal in.