Need to drive a mosfet to run a 50VDC 10amp Load ! of a Pixace 08m2

[Grant666]

Hi All

Im making a home pinball project, and shall be running some magnets off a 08M2.

Problem is that the magnets are 4.6 Ohm load at 50 Volts, so close to 10amps drawn at 500Watts. I have tried to use transistors to drive the magnet, but the wattage is killing them... So I think i can use a mosfet to do this. problem is, I have no idea how to drive the mosfet off the 08M2. So if anyone has done similar or can suggest a circuit, that would be wonderful. I could try , but i have a bad habit of letting the Smoke Genie escape from electronic parts.

I have attached my current circuit using the transistor, but it is no good. Fine on low load, but crashes at high current.

 

[techElder]

First thing I notice is that you are using a TIP36C to drive your load. That is a PNP transistor, but you have placed it on your schematic in a typical "pull to ground" configuration where I would expect to see an NPN transistor.

Perhaps you should have used the complementary TIP35C?

DATASHEET: http://datasheet.octopart.com/TIP36C-STMicroelectronics-datasheet-12516500.pdf

 

[erco]

Use a relay.

 

[nick12ab]

Problem is that the magnets are 4.6 Ohm load at 50 Volts, so close to 10amps drawn at 500Watts. I have tried to use transistors to drive the magnet, but the wattage is killing them... So I think i can use a mosfet to do this. problem is, I have no idea how to drive the mosfet off the 08M2. So if anyone has done similar or can suggest a circuit, that would be wonderful. I could try , but i have a bad habit of letting the Smoke Genie escape from electronic parts.

I take it this means the load won't need to switch on and off very quickly (kHz+), in which case you should be fine directly driving the gate of the MOSFET from a PICAXE output pin. Do check the gate capacitance though - if the gate voltage can't rise fast enough then the MOSFET will briefly have a higher internal resistance when turning on/off.

The most important MOSFET property to look for is Rds(on), lower is better. This is given in the datasheet for different gate voltages (VGS), you may need to specifically look for 'logic level' MOSFETs to get one which performs adequately at 5V VGS. Modern MOSFETs are available with RDS(on)s of tiny fractions of an ohm, though these are usually in surface mount packages. Once you have that, you can work out how much power will be converted into heat since you know how much current will flow through it, and you can find in the datasheet the absolute maximum which you mustn't exceed.

When the MOSFET is turned off, VDS will be 50V, so check that the absolute maximum is higher than this!

I have attached my current circuit using the transistor, but it is no good. Fine on low load, but crashes at high current.

The PNP transistor should be used as a high-side switch. As a low-side switch, it will perform poorly in a circuit like this. The opposite applies to NPN transistors.

 

[Circuit]

Hi All
I have no idea how to drive the mosfet off the 08M2. So if anyone has done similar or can suggest a circuit, that would be wonderful. I could try , but i have a bad habit of letting the Smoke Genie escape from electronic parts.

The RFP30N06LE Logic-level MOSFET by Fairchild can handle 30 amps at 60 volts max. - Might do the trick. The RDS(on) value is 0.047 ohms - a pretty low figure as mentioned by @Nic12ab. This also has ESD (Electrostatic discharge) protection making it easier to handle. IRL540NPbF http://docs-europe.electrocomponents.com/webdocs/0791/0900766b80791393.pdf rated for 100 volts and 36 amps with logic level gate drive but does not have ESD protection so make sure that you have anti-static gear on when handling (conductive wrist-band with 1Megohm to earth etc.).

As for "no idea how to drive the MOSFET off the 08M2", have a look at PICAXE Manual 3; page 8; "Standard Circuits 4 - The Power MOSFET Interfacing Circuit".

 

[fernando_g]

I only want add to the thread that you have not considered the requirements of the free-wheeling diode across the electromagnet itself.

That diode must be able to conduct the full current also (10 amps), and if the repetition rate is high enough, it will also require heat sinking.

 

[tmfkam]

As you have (wisely) used an opto isolator to interface the signal from the PicAxe into the solenoid drive circuitry, I'd suggest that you supply the 5V for the PicAxe from a totally separate supply, removing any shared ground connections from the solenoid and PicAxe sides of the circuit. Use a separate transformer completely where possible.

When that solenoid is switched off, there will be one heck of a lot of stored energy released into the 50V supply through the free-wheel (or flyback) diode. This will cause the 50V supply to jump, potentially (pun intended) along with the shared 'ground' and could well introduce problems into the PicAxe part of the circuit. By isolating the grounds for the PicAxe and solenoid sides, this will be dramatically improved.

This does of course mean that any inputs (switches) to the PicAxe will require that they too do not share any common connections with the 50V solenoid circuit. Again, opto isolators may be needed here.

I might also be tempted to place another diode with it's cathode connected as close as possible to the main switching transistor's collector, anode to the emitter (or equivalent on a mosfet device) to reduce the chance of failure within the switching device.

 

[Grant666]

Hi All

Thanks for the advice. i do have a flyback diode on the magnet coil. I did use a relay at one stage , but I can only use this on the DC side ( I have a very nice 50Volt 10amp Switchmode supply )...I found the air gap in the relay was not sufficient enough to break the DC current, creating a lovely green plasma arc that melted the relay ! This why I'am looking for a solid state alternative. I shall keep the opto isolation, so as to save the picaxe if nasty things happen.
Oh here is a nice picture of the relay vaporising itself...argh

 

[rq3]

Hi All

Thanks for the advice. i do have a flyback diode on the magnet coil. I did use a relay at one stage , but I can only use this on the DC side ( I have a very nice 50Volt 10amp Switchmode supply )...I found the air gap in the relay was not sufficient enough to break the DC current, creating a lovely green plasma arc that melted the relay ! This why I'am looking for a solid state alternative. I shall keep the opto isolation, so as to save the picaxe if nasty things happen.
Oh here is a nice picture of the relay vaporising itself...argh

What a great photo! Also a learning moment. Snubber circuits are used in situations like this to suppress the arc. Without going into the math, the snubber is generally something like a 10 ohm resistor in series with a 0.1 uF capacitor across the relay contacts (very rough component values). Even (or especially) with a solid state "relay", the snubber should be considered a required part of the solution.

 

[erco]

Yup, great pic, and yup, you need a capacitor snubber across your relay, just as old cars with breaker points controlling the ignition coil always needed a capacitor to prevent arcing and metal transfer.
http://www.electro-tech-online.com/threads/automobile-ignition-question.114642/

From a few different automobile posts, the optimum cap value increases with voltage.

6V systems use 0.021-.028 uF
12V systems use 0.047uF

Likely the coil inductance is a determining factor too. But if you want to use a relay, some type of cap will help prevent that undesirable but lovely arc reactor of yours.

BTW I have a 1977 Gottlieb "Jungle Queen" pinball game. That's the last year of all-mechanical construction, no solid state components. The insides have that delightful aroma of heated plywood from warm wires, tons of relays, solenoids, switches, rotating contacts, and numerous #47 incandescent bulbs.

 

[tmfkam]

Hi All

Thanks for the advice. i do have a flyback diode on the magnet coil. I did use a relay at one stage , but I can only use this on the DC side ( I have a very nice 50Volt 10amp Switchmode supply )...I found the air gap in the relay was not sufficient enough to break the DC current, creating a lovely green plasma arc that melted the relay ! This why I'am looking for a solid state alternative. I shall keep the opto isolation, so as to save the picaxe if nasty things happen.
Oh here is a nice picture of the relay vaporising itself...argh

As I understand it, it isn't so much switching the current at 50V that causes the arcing within the relay, so much as the stored energy within the coil being released. This is large enough to break down the insulation of the air between the contacts striking the initial arc, the arc ionises the air making it conductive which passes current and maintains the arc. If you had a vacuum between the relay contacts, the arc might never strike, and you'd have no problems. If you switched a resistive load with your relay, there would be no back EMF and (again) you'd have no problems.

Of course using a transistor in place of the relay may release the same amount of energy across the transistor junction causing similar problems.

One way to reduce this is to increase the time taken for the coil to be switched off, gently decreasing the current rather than attempting to stop it dead.

This from learn about electronics: This back EMF has an amplitude which is proportional to the RATE OF CHANGE of current (the faster the rate of change, the greater the back EMF) and a polarity which opposes the change in current in the inductor that caused it initially.

A possible solution to this is to use a 50V AC supply for the coil. As the AC supply passes through zero each half cycle, the current flow in any arcing falls to zero and the arcing ceases. The coil itself (normally) won't care if it gets AC or DC, the magnetic 'pull' will be almost identical. [Removed reference to zero crossing on the advice of Texasclodhopper] Full wave rectified 50Hz (or 60Hz) would do if you feel you need DC, as long as the supply for the coil has no reservoir capacitance on it.

 

[techElder]

Just an experienced note to the above.

If you switch an inductive load at the "zero cross" of an ac waveform, you will create the most current in the inductor because there is no magnetic field to oppose the change in current.

I used to drive high current switches/relays (1000's of DC amps) by driving them both ways; thus speeding up the transition from closed to open and reducing the arcing at the contact. Didn't rely on just the relay springs to open the contacts.

The contacts were always silver coated, too.

 

[premelec]

@Tex - I'm not following that - "will create the most current in the inductor" as I thought most SS relays are SCR so must go below holding current to turn off - and SCR doesn't care where the current is coming from - inductor or power supply. I don't get near 100s of amps DC ever too scary - was that vacuum relays made by Kilovac for instance?

 

[tmfkam]

Just an experienced note to the above.

If you switch an inductive load at the "zero cross" of an ac waveform, you will create the most current in the inductor because there is no magnetic field to oppose the change in current.

Ahh... I've fallen into a trap there haven't I. Sorry. Though using AC should allow any arcing to self extinguish.

Reading further, it seems as though inductive loads are best switched at peak supply voltage, for the reasons Texasclodhopper stated. This from TE:"A zero-crossover solid-state relay may be the worst possible method of switching on a transformer or a highly inductive load. Evidence has come to light that zero-crossover turn-on of such loads can cause a surge current of perhaps 10 to 40 times the steady state current, whereas turn-on at peak voltage results in little or no surge." The article can be found at http://www.te.com/commerce/DocumentDelivery/DDEController?Action=srchrtrv&DocNm=13C3206_AppNote&DocType=CS&DocLang=EN I am assuming that as Texasclodhopper suggests, this applies equally at switch off.

I used to drive high current switches/relays (1000's of DC amps) by driving them both ways; thus speeding up the transition from closed to open and reducing the arcing at the contact. Didn't rely on just the relay springs to open the contacts.

The contacts were always silver coated, too.

Driving them both ways? Are you talking about AC and DC?
Reading your reply again, I wonder if you are suggesting the relay coil be switched on with (for example) a positive supply, then off by being driven with a negative supply?

 

[Grant666]

Yup, great pic, and yup, you need a capacitor snubber across your relay, just as old cars with breaker points controlling the ignition coil always needed a capacitor to prevent arcing and metal transfer.
http://www.electro-tech-online.com/threads/automobile-ignition-question.114642/

From a few different automobile posts, the optimum cap value increases with voltage.

6V systems use 0.021-.028 uF
12V systems use 0.047uF

Likely the coil inductance is a determining factor too. But if you want to use a relay, some type of cap will help prevent that undesirable but lovely arc reactor of yours.

BTW I have a 1977 Gottlieb "Jungle Queen" pinball game. That's the last year of all-mechanical construction, no solid state components. The insides have that delightful aroma of heated plywood from warm wires, tons of relays, solenoids, switches, rotating contacts, and numerous #47 incandescent bulbs.

Thanks for the advice. Also I have 12 pinballs, 1 is an EM "Aladdin's Castle" Yes it is a distinctive smell. I also have a Bally Bingo Pinny, now there's a mechanical work of art, no picaxe could emulate.... Also next week we are going to look at 6 EM machines that have been stored in a Shipping container with a bloody hole in it ! So moisture and humidity damage, hope we can save them ! I might try a relay with a snub circuit t see what happens ! Oh the massive green arc didn't smell bad, but hissed like a Cobra then melted the relay ! oopsy

 

[techElder]

Ahh... I've fallen into a trap there haven't I. Sorry. Though using AC should allow any arcing to self extinguish.

Driving them both ways? Are you talking about AC and DC?
Reading your reply again, I wonder if you are suggesting the relay coil be switched on with (for example) a positive supply, then off by being driven with a negative supply?

I'm not suggesting anything. What I used to do involved a relay with dual coils; one "above" and one "below" the armature. I could drive it both ways ... hard and fast.

A relay contact with that much current should be switching AC not DC for all the reasons above.

EDIT: Hope my comment didn't sound curt.

A little historical aside:

I developed a "secret" driver for a large magnetic coil that needed to saturate a heavy iron pipe. My competitors were building bigger and bigger high voltage DC power supplies to do the same.

My "secret" driver was a zero-crossing solid state relay that skipped a half cycle of the AC supply. As tmfkam learned above, the first half cycle surge current was so strong that the magnetic field was off the chart for this application. No one knew what I was doing for years.

 

[Grant666]

I'm not suggesting anything. What I used to do involved a relay with dual coils; one "above" and one "below" the armature. I could drive it both ways ... hard and fast.

A relay contact with that much current should be switching AC not DC for all the reasons above.

EDIT: Hope my comment didn't sound curt.

A little historical aside:

I developed a "secret" driver for a large magnetic coil that needed to saturate a heavy iron pipe. My competitors were building bigger and bigger high voltage DC power supplies to do the same.

My "secret" driver was a zero-crossing solid state relay that skipped a half cycle of the AC supply. As tmfkam learned above, the first half cycle surge current was so strong that the magnetic field was off the chart for this application. No one knew what I was doing for years.

No, all good, no curtness detected !

Yes the trick is to switch the AC through the relay , then rectify it to run the DC Load. We do similar with Elevator Door cams, that are basically a big ass DC coil. But the pinny actually only runs a few switch mode supplies, and no AC transformers to be seen ! Thats why i'm just switching the DC. So I think the mosfets will do the trick, I have sourced the ones that drive directly off a TTL 5V signal, they are 80watt so would look at paralleling them up, probably 8 to allow for a bit more wattage leeway. But if you want to sketch me up a nice circuit, then please do. I have a basic level of electronics, but give me a circuit and i can build it.... keep the feedback coming.thx

 

[tmfkam]

I'm not suggesting anything. What I used to do involved a relay with dual coils; one "above" and one "below" the armature. I could drive it both ways ... hard and fast.

That is similar to how I had started to think you were explaining. I hadn't realised you could get such dual coil relays and it took a while for my fuddled brain to clear the mists of confusion and catch up with your (much faster) train of thought.