PDA

View Full Version : Poor man's induction heater power supply question.



DICKEYBIRD
03-20-2019, 03:18 PM
After seeing one of the techs in our shop use a "Bolt Buster" portable induction heater to heat some corroded & seized 21mm nuts to red hot in under a minute, I decided to look around & ended up getting one of these widgets:

https://www.amazon.com/s?k=fosa+1800W&i=tools&ref=nb_sb_noss

The shop threw out 2 battery chargers last week so I snatched 'em up. I haven't stripped them down yet to look at the transformers but the specs look like the available power should be in the ballpark without dropping a bundle on a commercial PS. They are identical chargers & both have dead digital control/display boards. The transformers should be identical and still good since there was no magic smoke or smell when the chargers died.

Here's the spec label:

http://i57.photobucket.com/albums/g227/DBAviation/Charger_zpsqirhrbrm.jpg (http://s57.photobucket.com/user/DBAviation/media/Charger_zpsqirhrbrm.jpg.html)

I should be able to make a PS using the 2 TX's in series, a suitable rectifier & capacitor, yes? The primaries would be wired in parallel & the secondaries in series, correct? Does the whole idea seem feasible?

MaxHeadRoom
03-20-2019, 04:29 PM
You may be able to do it with those, what VA does the transformers appear to be?
For a HF unit Google Royer HF heating unit .
Make sure you phase them correctly.
http://kaizerpowerelectronics.dk/general-electronics/royer-induction-heater/
Max.

darryl
03-21-2019, 02:22 AM
It is going to work- are you planning to gut the power supplies and use them in brute force mode?


This is interesting- if your chargers have a simple, two lead output winding from the transformers, then you can wire them in series and just use one bridge rectifier and one large output filter cap. But if they have tapped secondary windings, you could wire them in parallel, ignoring the taps, and then use the same bridge rectifier and cap.

I like the idea of only needing one filter cap- and make it a big one, like 50,000 mf or even more. Be aware that the no-load voltage could be up around 40 volts or so- you'll want to know to make sure your capacitor has a high enough voltage rating.

And of course you would phase the windings properly :)

markx
03-21-2019, 04:15 AM
Does that induction heater module have a power regulation option? I guess it does not as they are quite simple on most instances. In that case you would preferably need a DC source that has smoothly variable voltage. Also you need a panel volt and ampmeter to indicate wether you are within the safe zone. Flying blind with such a setup usually ends in magical smoke, tears and sobbing :D

About the dead chargers....are them switch mode devices or iron core? If they are switch mode and busted, then fixing them up can be quite a task. Also you can effectively forget about wiring them switch mode contraptions in a combined bundle to up the power. From iron core devices you can save the transformers+rectifiers+possibly caps and use these in a combined manner to create a new power supply of suitable parameters. This will also be quite a build, but when done properly shall yield a foolproof and hardy universal power supply that can be used for numerous projects.

For practical purposes I would not strive towards pushing these induction heater modules much in excess of 1000W regarding prolonged use and high duty. Previous experience has shown that the declared wattage is grossly overrated and one can strip a third off it as starters if any reasonable service life is expected out of these.
This means that a smps DC power supply of 24V 1200W would be very much adequate for driving the setup. There are standard units available on ebay/amazon and these run in the price range of about 50-100 bucks per piece, depending on how good a deal you can get. Also these have a pot that can be used to adjust the output voltage....not in a very wide range, but appreciably. If you value your own time in any sensible measure and also take into consideration that the power supply build will require extra parts and accessories (some of which shall be burnt in the process and will need to be replaced) then I do not think you can win much in terms of money with rebuilding the old chargers vs. acquiring new one.

Although in the range exceeding 1000W the cost of contemporary units for generating and converting the power goes up really fast, so a DIY approach with building your own "monster power supply" seems a tempting approach. Also you could try a lower wattage supply first....say in the range of 600W capacity. Just to get to know the system and see how it behaves under load. These are usually much cheaper and you could avoid spending a lot of time trying on getting the chargers operational again.


E.g. here is a small universal lab power supply that I assembled from contemporary affordable modules. Smoothly regulated DC 1-50V output with about 5A max continuous current. I packed together a small 12V DC smps, a 130W SEPIC converter (up and downconverter) and some panel meters to monitor the values. If I remember correctly the plastic case for the power supply cost more than the internals and these were about 30 bucks altogether. The induction heater supply could be done in roughly the same manner, but with much beefier modules of course:

http://bbs.homeshopmachinist.net/attachment.php?attachmentid=4889&d=1553156657

http://bbs.homeshopmachinist.net/attachment.php?attachmentid=4890&d=1553156671

Noitoen
03-21-2019, 06:50 AM
For max power on that unit I think it requires 45V 40A

MadsBarnkob
03-21-2019, 08:53 AM
I spotted a little traffic from here to my website, www.kaizerpowerelectronics.dk and came here to give me view on these IH.

We have had a long discussion and sharing of experience with the 1000/1800/2500 Watt chinese induction heaters here: https://highvoltageforum.net/index.php?topic=530.0

I made 5 videos about the 1800 Watt model and two different cheap power supplies, one made from laptop chargers and another made from server power supplies.

Here are the 5 videos on different topics:

1) Unboxing a Chinese 1800 Watt Induction heater: https://www.youtube.com/watch?v=E0Wxj0_slSc

2) 2000 Watt laptop charger for induction heating: https://www.youtube.com/watch?v=5iMFYdDMfXo

3) Chinese 1800 Watt Induction heater, 5 minute stress test: https://www.youtube.com/watch?v=jmAZhjl8Ybk

4) 2200 Watt server power supply for induction heating: https://www.youtube.com/watch?v=xRAlXCzsY3Q

5) Chinese 1800 Watt Induction heater, 10 minute stress test at 1500W: https://www.youtube.com/watch?v=v-TsBLFp8yo

I hope this will help you build a cheap IH :)

DICKEYBIRD
03-21-2019, 12:04 PM
I hope this will help you build a cheap IH :)You are the man! Lotsa new stuff to learn.:) (...and you have to explain that forum name);)

JRouche
03-21-2019, 01:02 PM
I spotted a little traffic from here to my website,

Great info! I have also been interested in the IH for sometime but decided I just dont have the knowledge to make one. I may revisit that project. JR

DICKEYBIRD
03-21-2019, 02:01 PM
Thanks for all the input guys!

I stripped one of the chargers down last night and wasn't expecting the XFMR to have this many windings & connections on it. The overall size & weight of the thing appeared to my untrained eye to be manly enough by itself but after watching Mads' videos, it looks like I'll need MORE POWER! I have no clue if it's possible to connect the secondary windings in series and/or remove some primary turns to get around 36 to 40V & 40 amps out of 2 of these things in series. Opinions?

http://i57.photobucket.com/albums/g227/DBAviation/TX1_zpsdgjm3lvd.jpg (http://s57.photobucket.com/user/DBAviation/media/TX1_zpsdgjm3lvd.jpg.html)

http://i57.photobucket.com/albums/g227/DBAviation/TX2_zpst9falqzu.jpg (http://s57.photobucket.com/user/DBAviation/media/TX2_zpst9falqzu.jpg.html)

http://i57.photobucket.com/albums/g227/DBAviation/TX3_zpsgqlf2dvp.jpg (http://s57.photobucket.com/user/DBAviation/media/TX3_zpsgqlf2dvp.jpg.html)

http://i57.photobucket.com/albums/g227/DBAviation/TX4_zps9wb4whop.jpg (http://s57.photobucket.com/user/DBAviation/media/TX4_zps9wb4whop.jpg.html)

markx
03-21-2019, 04:24 PM
Allright, an iron core monster! :D Them things are "pliable" and you can connect them in a number of ways to get the desired end result. Usually the secondaries on these are tapped in steps to get the varied output voltages needed for the charging. Also they are quite bombproof, so if you mess up the phasing it will not instantly blow out the smoke like smps style contraptions do.
I do not know how long you intend to operate the induction heater in one setting, but if you foresee hours worth of cooking at >1kW then honestly the transformers are going to burn up if you do not immerse them in a massive oil bath. If we are talking mere minutes of operation at a time then you can get away with it. Most of the heating shall occur in the transformer core that is confined and since it is massive, will not cool effectively even under forced ventilation. The modern rectifiers are not really a problem in terms of heat dissipation.

You can connect these transformers in a number of ways and feed them to a common rectifier and cap bank e.g. :

http://bbs.homeshopmachinist.net/attachment.php?attachmentid=4908&d=1553198697


Note: several smaller electrolytic capacitors in parallel are performing better than one huge specimen (the cap internal resistances become paralleled and are thus in sum reduced)

From the cap bank you can try to directly feed the induction heater, but if you are not able to get desired voltage from the combination of windings then you can include a stepup or stepdown converter after the cap bank to deliver adjustable voltage. But that converter has to be a beefy one to withstand the currents.

Something of the following sort, but preferably more powerful:

https://www.ebay.com/itm/1500W-50A-DC-Boost-Converter-Step-up-Power-Supply-Module-IN-10-5-60V-OUT-15-70V/252553726589?hash=item3acd600a7d:g:xYgAAOSwCGVX5UU j&frcectupt=true

PStechPaul
03-21-2019, 04:29 PM
The size of that transformer looks about right for the continuous specification of 12V at 20 amps, or 240 watts. So you would need eight of them to get the 1800 watts for the induction heater. Also, the welded laminations will make the transformer saturate and essentially limit the output current (which is fine for a battery charger), but not for a fairly "stiff" voltage source.

You might do better to use a couple 12V batteries with the chargers working as such and connected in series. You should be able to draw 75 amps for 5 minutes or so from 12 or 17 A-h SLA batteries, or typical 50-100 A-h automotive car starter or deep cycle batteries. But make sure the output to the induction heater is properly fused in case of a magic smoke event. And make sure the device is properly contained in some sort of "blast shield" such as a NEMA electrical box.

DICKEYBIRD
03-25-2019, 03:02 PM
Recognizing that the necessary batt. charger transformer mods are way above my knowledge level, I ordered a half dozen 750W server power supplies on ebay for a little over $9 each. I watched Mads' great video where he describes in detail how to isolate the cases & boards of the same model HP power supply that I bought & he connects 3 of them in series for a 36V 60A supply. Looks like a great (ahem) bang for the buck!:eek:

He mentions a 20000uF capacitor bank in series with the output for "load sharing." I found a deal on ebay for a 30000uF/65v cap. Would the larger value be a problem?

I kept one of the batt charger bodies and it looks like it should be big enough to mount the PS, the IH unit, the cooling water reservoir/radiator/fans & plus an ammeter & voltmeter. It already has a power cord & wheels to roll it around in the shop!:)

PStechPaul
03-25-2019, 04:30 PM
I think that means a capacitor should be connected across the output of each power supply, but I don't see how that will help load sharing. It could also be a single capacitor across the series connected power supplies, but, again, not sure how that will help load sharing. Which of Mads' videos describes this? The individual 12V supplies will always see the same current. A larger capacitor might cause issues with a large surge current at turn-on, so perhaps add a 1 ohm power resistor in series with the output to charge the capacitor at 65 amps, and then a switch or relay across the resistor to bypass it once the capacitor is charged. You might be able to make a simple surge limiter using a 12V relay coil and a 24 volt zener in series that will pull in and bypass the resistor once output on the capacitor reaches about 32 VDC. The following circuit also has an automatic discharge function, which is also a good idea.

http://enginuitysystems.com/pix/electronics/Precharge_Surge_Discharge_Simple.png

DICKEYBIRD
03-25-2019, 05:24 PM
Thanks Paul, I always appreciate your electronics advice. :) The 1st mention of a cap is in this video: https://www.youtube.com/watch?v=xRAlXCzsY3Q and this one mentions the value: https://www.youtube.com/watch?v=v-TsBLFp8yo

So it could be OK to run these in series without a capacitor? It seems to my definitely un-trained mind that whatever magic a cap provides should already be in there seeing as how these things are made & used by the ba-zillions. I certainly don't know and was trying to emulate Mads who knows his stuff.

darryl
03-25-2019, 07:32 PM
Using the server power supplies, they already have enough filtering to keep the ripple voltage on their outputs within some limit. They will automatically load-share by virtue of being voltage regulated, and because the current is the same through all of them because their outputs are wired in series. This is unlike the battery charger brute force 60 hz 'old style' models which actually need a high value of capacitance after the rectifiers to keep ripple voltage down to design levels.

What would be good is to have some capacitance across the full output voltage, and probably tie that in right at the input to the induction heater module. There should be some kind of input filtering on the module already so it can be stable- I would think so anyway. Perhaps some attention should be paid to this area, just to make sure the module and power supply don't freak out with each other. A huge capacitance added across the power supply would probably freak out the power supply- it will be faced with operating into what will appear to be a dead short. Some will not like this, others will give reduced current until the output voltage is able to rise. In any event the ripple current either from the power supply or reflected back to it by the induction heater load is at a high frequency, and you would need to use a cap meant for switching power supplies in order to make much difference anyway.

A thought about the power supplies in series- if any one of them was slow to start relative to the others, it would see an abnormal voltage across it for a time if the load was on when they're powered up. I would make sure that you use a power-up sequence so load is always off before power up or power down. If you use the extra cap at the heating module, the cap stays wired to the module and the load switch is between the input wire and the power supply. The cap could be 50 or 100 uf, or 1000 uf- it's just going to serve to keep noise down on the wiring.

DICKEYBIRD
03-25-2019, 08:40 PM
Thanks Darryl, the info I've seen warns that the IH unit should never see a low voltage. If using switching PS's, you have to give them as much time as needed for the voltage to be fully up to max & stabilized before switching on the IH unit with a switch or relay. In the video, Mads powers up his 3 PS's individually & verifies the voltage is stable before switching on a breaker between the PS & the unit.

markx
03-26-2019, 03:43 AM
I see from the amazon pictures of the IH unit that there is a small (supposedly) voltage converter hanging out from the side of the unit. This is most likely a stepdown converter to power the internal driver circuitry with a stable voltage of around 12V dc. As the unit is meant to be powered in the input voltage range of 12-48V, I suspect the small converter is tuned to have a hair bit below 12V output. Since usually a "garden variety" stepdown converter does not operate if the input is below the supposed output voltage of the converter and we shall have a problem when the main power supply sinks below the "hair bit less than 12V dc condition". In that case the small voltage converter would stop converting and leave the driver circuitry and switching elements floating in a strange manner. Usually that condition let's out the magic smoke very nicely. Just a speculation regarding the mentioned problems with low input voltage condition....

To escape such a predicament, I usually implement a voltage converter with SEPIC configuration in my own contraptions (it can up and downconvert througout the whole voltage range with little regard to input voltage values). In that case the main power supply voltage can float in either directions, but the driver circuitry still gets the correct voltage if there is any input from the main supply. Very useful in battery operated contraptions, where the main cell voltage in a fully charged condition can be higher than driver circuitry needs (so it has to be downconverted). And also the main cell voltage in a depleted condition can sink below the critical needs of driver circuitry ( hence needs to be upconverted). The up/downconverter takes care of both situations and avoids any trouble caused by the possible fluctuations.

MattiJ
03-26-2019, 04:04 AM
I see from the amazon pictures of the IH unit that there is a small (supposedly) voltage converter hanging out from the side of the unit. This is most likely a stepdown converter to power the internal driver circuitry with a stable voltage of around 12V dc. As the unit is meant to be powered in the input voltage range of 12-48V, I suspect the small converter is tuned to have a hair bit below 12V output. Since usually a "garden variety" stepdown converter does not operate if the input is below the supposed output voltage of the converter and we shall have a problem when the main power supply sinks below the "hair bit less than 12V dc condition". In that case the small voltage converter would stop converting and leave the driver circuitry and switching elements floating in a strange manner. Usually that condition let's out the magic smoke very nicely. Just a speculation regarding the mentioned problems with low input voltage condition....

Could be even just for the fans.

PStechPaul
03-26-2019, 04:46 AM
I don't think the Royer oscillator needs a separate gate drive supply for the NMOS transistors. Here is a simulation:

http://enginuitysystems.com/pix/electronics/Royer_NMOS_Simple.png

Slightly different circuit:
http://enginuitysystems.com/pix/electronics/Royer.png

J Tiers
03-26-2019, 11:37 AM
The classic Royer does not. The power dissipation depends on the saturating core to "snap" quickly, but they generally are not "quick" on the sort of time scale really needed.

A more efficient version would use a better gate driver.

But a decently designed circuit should pull everything to an "off" condition if the control voltage is low/absent. The same condition of missing control voltage is often present when power is applied, so it is pretty important to start up with a defined "off" condition for any of the switching devices, at least.

The Royer typically has a bias for one transistor that ensures it is always the one to conduct first. That then activates the positive feedback via the saturating core and begins the switching action. Many of the induction heater circuits are not a Royer, but actually are more a variety of power astable flip-flop, such as Paul's circuit, with a tuned output setting the frequency. they typically have no trouble with low or missing voltage, because there is no separate control power supply.

If the main power supply is high enough, then a separate supply may be necessary, or some system for limiting gate voltage is needed, such as the resistor divider Paul shows, or a zener to limit voltage. Those can cause even slower gate voltage changes.

The ideal gate voltage circuit would be very fast, for least power dissipation. It can be necessary, if you need to meed EMI limits, to trade off some extra dissipation for lower EMI. Really fast switching produces a heck of a lot of EMI.

DICKEYBIRD
03-26-2019, 12:17 PM
Could be even just for the fans.Yup, exactly that.

PStechPaul
03-26-2019, 02:50 PM
Some simulations of a Royer oscillator using LTC1693-1 dual gate driver:

http://enginuitysystems.com/pix/electronics/Royer_NMOS.png

http://enginuitysystems.com/pix/electronics/Royer_NMOS_3.png

MadsBarnkob
03-27-2019, 05:17 AM
Thanks Paul, I always appreciate your electronics advice. :) The 1st mention of a cap is in this video: https://www.youtube.com/watch?v=xRAlXCzsY3Q and this one mentions the value: https://www.youtube.com/watch?v=v-TsBLFp8yo

So it could be OK to run these in series without a capacitor? It seems to my definitely un-trained mind that whatever magic a cap provides should already be in there seeing as how these things are made & used by the ba-zillions. I certainly don't know and was trying to emulate Mads who knows his stuff.

Darryl is right about the capacitors not being needed for current sharing when using the server supplies in series only, as there is only one current path through them all. The whole setup was from the 12 laptop charger power supply setup, so I just let the capacitors sit, but I still believe they are useful!

If you have a quick load change, the capacitors will deliver the peak current and that will give the power supplies more time to regulate onto the new changed load, a improvement to this would properly be a smaller capacitance bank from polypropylene capacitors instead of electrolytics. With a power film capacitor bank instead, you can properly do with 1/10 of the capacitance as they can deliver peak current faster.

In my experience, if you want to omit the capacitors, the power supplies can get into a battle of who is taking the load, if they are in parallel that is, and the result is over-current protection kicks in, the voltage drops, then the other power supply takes most of the load as its shared according to output voltage, then that drops and the result is unstable 2nd order sine waveform. You can also use diodes to "hide" the power supplies from each-other, if they have enough output capacitance to handle the current peaks.

A stable power supply is the most important thing to not destroy a Royer oscillator, so always use a switch after your DC supply, as you also mentioned that I did in the video. The IH would properly run just fine on 12V/60A with just one of them powered on, but if you were using a iron core transformer and it was loaded too hard so that voltage drops below 10VDC, you risk destroying your MOSFETs from inadequate gate drive and possible cross-conduction.

DICKEYBIRD
03-27-2019, 08:24 AM
Yay! Another electronics whiz on the forum to contribute to the argu...err, discussions!:p Seriously though, thanks very much for your input here.:) As all of our non-electronically challenged members will say, I know just enough to get myself in trouble!

Funny you should comment here today; I just showed my daughter & granddaughter the Take 5 video. They both were amazed!

As usual I have another question for you guys: I have a few 40 amp 220v circuit breakers in my junk box. The different characteristics of AC & DC affect switch & relay ratings, yes? Would one of those breakers' power handling ability be close enough to use between the PS's & IH unit as a switch?