Induction brass annealer redux

[Cedric]

Hi guys,
i watched a bit through this threat already. Have build it similar to the original version from GinaErik and it worked very fine. Building now a 2nd one for a friend who wants to use it all the way from 6.5CR till 50BMG.
The coil is now a little trouble to find the right size but will work that out.
Have noted that some of you exchange the contactor by Fotek SSR's. Why did you do that? Simply because of the higher amps and the SSR's can take that better than a contactor?

I had to buy all components for 220V as i am from germany, but all worked fine. Inlcuding the ABB Contactor (coil on 220v and it connects the 48V @ 20A perfectly without arc). So i am just curious why you changed it to an SSR.

Cheers

 

[David101]

I went with the SSR because I thought there would be less arc and thinking if we are trying to switch in such small increments of time it would be more accurate. Turns out I have had more trouble with the SSR most likely because they are Chinese and cheap as opposed to the more expensive contactors I purchased.

 

[Cedric]

All right. Then I stick with the contractor in connection with a varistor.

i got one more question. Just made another inductor coil with 4mm tube, 7 turns (length 35mm, 1.37“) and inner diameter 35mm. Tube is insulated with heat shrink tube.
It goes with 50BMG up to 22amps. Just little over my 20amps which the power supply and inductor board can handle. And this at 45Volts. (But actually Dame current as on 48V).

i have also a coil with a 3mm tube, inner dia of 28mm / 1.10“ with 8 turns.
Here the current goes over 30amps with 50BMG.
But also without any case it’s already at 17amps.
I got the exact same coil on my second annealer, not insulated with 8 turns and that’s without any case at some 8-9 amps.

how is that possible? The insulation on or off doesn’t make a difference. Checked it already.
I wonder about the high currents for the new coils at idle mode (without case). And i can’t find out what I’m doing wrong

 

[whatsupdoc]

So I have started collecting the various components required to construct this annealer,
The ZVS unit i purchased is this.
https://www.banggood.com/Low-ZVS-12...=105229698&sc_uid=YZigA1aAEN&cur_warehouse=CN
https://www.banggood.com/Low-ZVS-12...=105229698&sc_uid=YZigA1aAEN&cur_warehouse=CN

The copper coil consists of 5mm tube where the inside diameter is 40mm the height the coil is 50mm.

Has anyone used this size coil?

After going through the miles of posts on this thread there seems to be various coil sizes used, so for 308/6.5 cases what seems to be the correct coil size?

 

[David101]

The coil that comes with the ZVS is for heating / melting other things. The coil in the first page is a sure winner for the 6.5/308 case. However I went to the metric size up from that 1/8 to 4mm which allows a bit more water flow. I also only went with 7 turns to concentrate the flux a little more. I stuck with the ID of about 27 mm

 

[Cedric]

@David101: what’s your Amps reading then?

cheers

 

[mktacop]

Very nice !!
Interesting that you put your display off to the right side. Lot of innovation there.

Thank you Gina. That decision was made so that it would be easier to add a case feeder later.

 

[Cedric]

found my fault.... so simple and hard to see... simply the wire size in the 48V circuit. I used only 1.5mm² and on my working unit I used 2.5mm². So i change the cable and tada... just 6amps in idle and with 50BMG (30mm coil with 8 turns now) it goes up to 17amps.
So for the future guys, use a cable as big as possible for the 48V circuit and all your dreams come true ;9
Working perfectly now

Job done and my friend is happy with his new unit

 

[Kilohertz]

wow! Well I have read thru the first 10 pages then I skipped to the end to see how the story ends, but it doesn't, still going, cool.

Now I am going to build one and have a question about the power supply. It's possible the answer is in the 93 pages but I know someone can answer quickly.

Having been in the radio/telecom industry most of my working life, I have collected a TON of stuff. I have 3 power supplies which may work but need some guidance which would be better.

I have a Nortel 48VDC 25A monster, rack mount, linear, weighs a ton, but is here and free. My other option is a Glenayre 48VDC @ 7.5A supply, or a Glenayre 28VDC @ 14A, both the Glenayres are 2 rack space units, the Nortel is 6 RU's. I would prefer to use one of the Glenayre units, they are switch mode and have adjustable voltages, could probably fudge the current up a bit as I think I have the manuals and schematics around somewhere.

So is 28V at 14A enough, or would I be better with 48V and 7.5A?

Brass will be .243 up to 30-06 and 7mm Rem Mag, nothing bigger.

Looking forward to getting this project rolling.

Cheers

Here is the Nortel supply.

https://www.dropbox.com/s/05ukec4w46ps6ih/nortel ps.jpg?dl=0

PS well shoot if I had read to page 13 of the 93 pages I would have seen that around 700W is really needed as minimum so I guess the big Nortel or I could series a pair of the 28V 14A Glenayres. Boy there is a lot of info in this thread. Thanks to Gina and Hollywood and everyone else who contributed to this. Will be fun to build.

 

[SGK]

You will need more power than the Glenayre will deliver. Think c48V and c15A as the zone. A 308 case in the recommended coil will draw about 14.5-15 amps.

 

[David101]

It might depend on how the power supplies work in an overcurrent event. Most people myself included are turning the 48V down as much as possible to only about 45 V. 36-40V is a good area to be working in. The natural current that the unit will want to draw at that voltage would probably be about 20A at max. It ramps up. If the power supply current limited at 15 and kept delivering at that 28V 15 A I think you would have a good working unit.

The caveat like for all switching supplies is not to switch the supply use a contactor to avoid soft start, kills the oscillator board.

 

[Kilohertz]

Thx David. I pulled the cover off last night and there is a voltage and current trim pot inside which I will play with tonight when I get back to this. These are commercial supplies meant to live in a rack in a rugged site, unattended. Fan cooled with large external heatsinks, all thru hole components, and I have a number of them so if one should happen to fail, I can try something else. They are rated for 850W input so will probably handle short term overloads just fine. I'm sure the Nortel is adjustable as well but it is a beast.

Thanks again for the help.
Cheers

 

[David101]

Looking at my machine, other peoples and the commercial units 600W seems to be sufficient to perform the task. I have a 1500W power supply in one a huge overkill the little board will cook itself long before and a 1000W both Chinese units hence why I left some overhead.

I have found the Inductor coils to be the component that gets the hottest followed by the capacitors (Heat sinks on the bottom of the board like others have suggested very good idea) and probably in last place the mosfets. Keep them all cool and continuous operation is achievable.

If anyone has some ideas on cooling the Inductors passively please pass on the information. I am considering using larger gauge wire but that has issues with connecting to the PCB which is likely to break in the process.

 

[Kilohertz]

I spent an hour today playing with both the 48V and 28V Glenayre units and was able to dial back the 48 to 42V and dial up the 28 to 30V. There are current adjustments but I need a big dummy load to test them and adjust. They both start out on the AC side switching 2 MOSFETS for a high freq at 55VDC sounds like about 10-12KHz. That section feeds a xfmr which is wired parallel for the 28V unit and series for the 48V unit. Output from the xfmr is 17.5AC and 30AC for the 28 and 48 units. It looks like there is a voltage doubler in the 28V supply.

I was also contemplating just building a proper 110Khz switching supply using a 3524 or similar IC driving a bridge with the appropriate FETS or IGBTs, would eliminate the need for a separate supply, as it would BE the supply for the induction coil. Just thinking outside the box. Has anyone else tried this? I'm only up to page 17 of the thread.

Cheers

 

[mktacop]

Work in progress post.

Using the case feeder designed by @SGK, coupled with some new parts I designed, I am almost there on a case feeder for my annealer. @KeeWay helped immensely by updating his Arduino code. Once a new, slower motor arrives, this upgrade should be complete.

Edit: I should also say that the feeder mount is articulated so that it can be moved out of the way if desired/needed.

 

[SGK]

Nice. Recommend some "bumpers" on the feeder disc. https://www.amazon.com/dp/B07882C9KC/?tag=accuratescom-20

 

[mktacop]

Nice. Recommend some "bumpers" on the feeder disc. https://www.amazon.com/dp/B07882C9KC/?tag=accuratescom-20

Thanks!

 

[hdmunger]

Hello all...recently finished my version of the induction annealer and have a question regarding my design.

I elected to use a heavy gauge wire to connect my ZVS board to my induction coil to make it easier to construct the mechanism that allows for height adjustments of the coil (see pic 3 - don't remember the exact guage, but it is almost 1/4" in diameter). When doing initial testing, I noticed that that large guage wire would get slightly warm after only a few cases annealed.

My question is...would it be better if the wire was shorter (less resistance) or is longer better (more of a heat sink), or of a different/smaller guage altogether?

----------------------------------------------------------------------------
My basic design is this - 110VAC powering the timer and 48VDC 20A 1000W power supply, 48V power supply powering 48V to 12V buck converter, 12V powers fans, pump, relay, and 48V also powers ZVS board thru the relay. I also used 1/4" copper tubing in a dual layer 3+3 configuration for better coolant flow (seems to work OK...5.5 sec for .308 brass).

pic 1 - https://drive.google.com/file/d/1oLxBFMN-Ii4Zl__fuFg2fdR_0i7yeShW/view
pic 2 - https://drive.google.com/file/d/1r3dmNhzdVcBdeOlvzHxwTD2mJY-EwhFv/view
pic 3 - https://drive.google.com/file/d/1zcP4qw3_cz3kxK3hcfO7fqkFR2IhzZQL/view
pic 4 - https://drive.google.com/file/d/14dQPAQbQcteZ-bFoZZfh03z7QxdewjKs/view?usp=sharing

Thanks for your advice in advance.

 

[SGK]

I would op for shorter. Heat is power being dissipated (I x R). And I would twist any/all pairs of wires firmly together (everywhere) to cancel any parasitic inductances. Quite some time ago I showed a formula to compute/estimate the current flowing in the IC tank. Currents are really very, very high.

 

[hdmunger]

They are currently about 12" long and I could probably only reduce that to 8-9" and still maintain my adjustability. As far as twisting is concerned, I separated the total of about 13 or 14 strands into 3 separate twists at the ZVS end so that I could use all 3 mounting/connecting holes where the metal standoffs were mounted for the induction coil connections. At the other end I used some connectors I found at Home Depot that I think are for connecting a house ground to a 1/4" copper grounding rod...they are also made of copper and slip over the coil ends and clamp on with a decent 'bearing surface' (for lack of a better term).

Thanks for the response SGK.

PS - I think that that conversation regarding high current was what led me to choose such a large guage wire for this connection...