Author Topic: Mains switching research break out  (Read 8382 times)

0 Members and 1 Guest are viewing this topic.

Offline oschonrockTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: gb
Mains switching research break out
« on: August 02, 2020, 09:57:45 pm »
Input appreciated please.

There are lots of really dodgy "switch your XXX mains appliance from your YYY uC" circuits and schematics out there. Many of them dangerous. A lot of videos do this stuff on breadboard  :-[

I am trying to design a "breakout module" that does this "properly" and safely. It may/will ultimately need to come in a proper physical case, properly isolated, protected & appropriately fused etc so that people can do these things without killing themselves or the equipment.

The concept has 2 parts:
1. The HV switching part
2. a slave uC which you can talk to via I2C,  analog, or whatever

I am focused on part 1 here.

Attached is an early sketch of the basic switching architecture. Text annotations are the questions I would like input on.... Also see more verbose description of goals/intentions and circuit operation below.

Many thanks

EDIT:
--------

My original circuit description and outline of intention/goals were not adequate and caused some confusion below. So here is a fuller description:

I am not making a simple SSR, nor am I making an AC motor drive. Variable speed AC drives (VSDs) have a full rectification stage first then a DC bus with huge caps and then an H-Bridge type switch topology. The advantage of that is that they can make AC of any frequency, and this is required to properly control the speed of AC induction motors. My circuit can't do that, but it can operate in the following three "switching modes":

1. ON / OFF multi cycle period (typically 1-2 seconds, less and, depending on load size, you get into trouble with standards in some regions,( eg EN61000-3-3, AKA "the flicker standard").  The uP can (optionally) do the careful timing to ensure it switches the MOSFETs off during zero cross and hence causes less harmonics. This mode does the job of an SSR - if this is all I wanted I should use an SSR - and is quite trivial.

2. What is often referred to as chopping / dimming / phase angle switching etc. Means turning the devices on / off once per 50/60Hz half-cycle. This can be done on the leading edge, ie wait after zero cross before switching on, or on the trailing edge, ie turn the devices off early before the half cycle completes. The former (ie leading edge) is what triac based dimmers do, but they can't do trailing edge which has advantages in many circumstances.

3. PWM of the sine wave at frequencies of multiple kilohertz. This is different to the AC variable speed motor drive because it is not switching DC to make AC of any frequency. It is switching the AC sine wave. That means we can "smoothly vary the effective amplitude" of the sine wave, but we cannot change its frequency.  -- Note that while the attached circuit sketch can do this in theory, in practice, a proper isolated MOSFET gate driver is needed for this mode, as discussed below.

So proper induction motor control is beyond the scope of this "research break out board". It's purpose is to introduce hobbyists / students to the various kinds of AC switching you can do. It should be able to demonstrate the above 3 modes, and how different kinds of loads respond to them. The purpose is to educate and provide information on strengths weaknesses, from a "ready made box" which can do all 3 modes and be SAFE. Someone below mentioned current/voltage monitoring and that would be a great addition. Non-trivial due to the isolation requirements, but I might need to go there.

Some notes on intended circuit operation:

Unlike in a VSD H-bridge, the 2 MOSFETs will always "switch on at the same time".

Actually that is not quite correct. more precisely, in the "upper half cycle"  (when Vline > Vneutral the lower MOSFET on my schematic (Q2) will be in "body diode bypass", ie whatever you send to its gate is irrelevant.  In that upper half cycle the "triggering which is important is on Q1. In the lower half cycle, the roles are reversed. --

EDIT2: The above paragraph is wrong (I learned something), the body diode is not used like that. Both MOSFETS are really on together in that circuit, because MOSFETS can conduct current in REVERSE (ie from source to drain) when they are on. The Ron of a MOSFET will usually result in a lower forward voltage at a particular current than the body diode, so if it's switched on, almost all the current will flow through the main channel, not the body diode, and that's good => lower power dissipation.

The optocoupler & +12V supply referenced to joined sources and that pull down resistor mean that the  controller/uP doesn't have to care which half cycle it is in. It would just look at the zero cross info from U1 and decide when to trigger U2 - upper or lower half cycle, don't care.

Discussion below introduces a proper isolated MOSFET driver which fundamentally works in a similar way but drives the devices more efficiently and therefore able to achieve Mode3 = PWM > 1kHz.

It is also possible that IGBTs would be a better choice of switching device, particularly for mode 3. Any input on this appreciated. -- EDIT2 discussed below: No. IGBTs are not good for this, because they can't conduct in reverse, and hence need a separate anti-parallel diode (there is such a thing as a RC-IGBT, reverse conducting, but this is not a mature technology).
« Last Edit: August 04, 2020, 02:22:11 pm by oschonrock »
 

Online coppercone2

  • Super Contributor
  • ***
  • Posts: 11345
  • Country: us
  • $
Re: Mains switching research break out
« Reply #1 on: August 02, 2020, 10:03:23 pm »
you want safe triggering put  industrial levels of protection on the opto input, they are a diode after all
 

Offline oschonrockTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: gb
Re: Mains switching research break out
« Reply #2 on: August 02, 2020, 10:05:30 pm »
you want safe triggering put  industrial levels of protection on the opto input, they are a diode after all

Thanks. What do you suggest? TVS etc?
« Last Edit: August 02, 2020, 10:09:16 pm by oschonrock »
 

Online coppercone2

  • Super Contributor
  • ***
  • Posts: 11345
  • Country: us
  • $
Re: Mains switching research break out
« Reply #3 on: August 02, 2020, 10:09:01 pm »
esd diode on the logic side (off board connection since its a module)

on the mains side I don't really know. Maybe a MOV, fuse, etc
 

Offline oschonrockTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: gb
Re: Mains switching research break out
« Reply #4 on: August 02, 2020, 10:14:12 pm »
esd diode on the logic side (off board connection since its a module)

on the mains side I don't really know. Maybe a MOV, fuse, etc

Makes sense. Thank you. Will add.
 

Offline OM222O

  • Frequent Contributor
  • **
  • Posts: 768
  • Country: gb
Re: Mains switching research break out
« Reply #5 on: August 02, 2020, 10:56:48 pm »
what on earth are you trying to achieve  :palm:

Don't use fets and simple opto coupler. You should look into using opto triac and triacs for the actual switch. This class AB amplifier you made has some significant drawbacks and most likely overheat the fets (not to mention zero crossing distortion etc) you also forgot protection features like movs, X and Y class rated caps depending on phases / netrual, common mode choke / ferrites for noise.

Look into construction of surge protectors or any reputable mains PSU, you'll find all of those features.
 

Offline oschonrockTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: gb
Re: Mains switching research break out
« Reply #6 on: August 03, 2020, 12:19:32 am »
what on earth are you trying to achieve  :palm:

Don't use fets and simple opto coupler. You should look into using opto triac and triacs for the actual switch. This class AB amplifier you made has some significant drawbacks and most likely overheat the fets (not to mention zero crossing distortion etc) you also forgot protection features like movs, X and Y class rated caps depending on phases / netrual, common mode choke / ferrites for noise.


Perhaps I should have been more explicit. The object is go beyond triacs, which don't work with many modern lamps types to start with, and create evil harmonics (leading edge on, zero cross off).

I was clearly concerned about the commutation (class AB as you called it), hence the question on that. The idea for that architecture is not from me ;-)

But if it's seriously flawed please be specific. What's the alternative? Separate and retrigger?

The other most common/sane architecture (other than full blown industrial AC drive style DC Bus and re-chop) is MOSFET wrapped in a in a bridge rectifier, which has the same challenges ( I believe) with lots more parts that don't add anything..

Thanks for your thoughts and the ideas on protection protection features particularly. I had not yet added any, as trying to validate architecture first ("early sketch..."), but those are all certainly valid suggestions.
« Last Edit: August 03, 2020, 12:31:44 am by oschonrock »
 

Online BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: Mains switching research break out
« Reply #7 on: August 03, 2020, 12:34:05 am »
If you want to completely get rid of switching harmonics like you claim, your 2 choices are mechanical relays or 2 photovoltaic optocouplers, each one powering the gate of each mosfet completely eliminating their loss of power during the AC transition and both mosfets' gates would always be on under both phases.  You are basically engineering an AC/DC opto-mosfet relay switch.

Your other choice is to use 2 gate transformers with 2 separate outputs each driven with a good current square wave high frequency AC signal with full wave bridge output rectification plus pull-down resistor to energize the gate when you want it on, then cut the waveform when you want it off.

There also exist isolated gate drivers with isolated gate power supply for this purpose, but they may be costly and not normally stocking type ICs.

Last note, Triac AC switches can be made to not interfere with LED lamps, you just need a snubberless tolerant triac and remove the snubber circuitry.  There should no longer be enough current flowing through your AC switch to make an LED lamp shine dim or flicker when the switch is turned off.

If the 2 mosfet design is a must, I would use the photovoltaic optocouplers and get rid of that full wave bridge rectifier and loading circuit which wont always power both gates completely when on.  Just make sure the photovoltaic output voltage is high enough to completely turn of the gate of the selected mosfets.  Maybe use logic level mosfets to ensure this happens.  Perform proper measurements at a complete temperature range check, otherwise, under certain circumstances, you may not be completely turning on the mosfets.

Example:
https://www.arrow.com/en/products/tlp3906-tple/toshiba
This one's output is a good solid 5v @ 12ma @125 degrees Celsius.  Panasonic also has one with similar specs.
These are too slow for PWM and slicing strategies.  You need to use a gate transformer or a second optocoupler or re-thing your posted schematic to achieve that properly.  Even if your current design would be considered a safe functional circuit, that 4.7k pulldown on the gate would be too slow for PWM strategies while slicing would work, turning off the mosfets would be so gradual and slow, expect them to get really hot.  They need to turn off fast, like 100 ohm pulldown at least somewhat equivalent to the turn off speed of a proper mosfet gate driver.

If you want a fast turn off instead of a 1-10 megaohm load on your gate to source to turn off the mosfet, you will need a second transistor/mosfet output optocoupler to discharge your gate quickly when it is time to turn off the mosfet.

Otherwise, just use a high voltage/high current photo-fet AC/DC relay IC/component which has this all inside already.
« Last Edit: August 03, 2020, 01:12:37 am by BrianHG »
 

Offline oschonrockTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: gb
Re: Mains switching research break out
« Reply #8 on: August 03, 2020, 01:08:41 am »
If you want to completely get rid of switching harmonics like you claim, your 2 choices are mechanical relays or 2 photovoltaic optocouplers, each one powering the gate of each mosfet completely eliminating their loss of power during the AC transition and both mosfets' gates would always be on under both phases.  You are basically engineering an AC/DC opto-mosfet relay switch.

Your other choice is to use 2 gate transformers with 2 separate outputs each driven with a good current square wave high frequency AC signal with full wave bridge output rectification plus pull-down resistor to energize the gate when you want it on, then cut the waveform when you want it off.

There also exist isolated gate drivers with isolated gate power supply for this purpose, but they may be costly and not normally stocking type ICs.

Last note, Triac AC switches can be made to not interfere with LED lamps, you just need a snubberless tolerant triac and remove the snubber circuitry.  There should no longer be enough current flowing through your AC switch to make an LED lamp shine dim or flicker when the switch is turned off.

If the 2 mosfet design is a must, I would use the photovoltaic optocouplers and get rid of that full wave bridge rectifier and loading circuit which wont always power both gates completely when on.  Just make sure the photovoltaic output voltage is high enough to completely turn of the gate of the selected mosfets.  Maybe use logic level mosfets to ensure this happens.  Perform proper measurements at a complete temperature range check, otherwise, under certain circumstances, you may not be completely turning on the mosfets.

The 2 Mosfet design is not a must and I did not "make any claim to completely get rid of harmonics". Zero harmonics are not possible if you chop up a sine wave.

Triacs draw currents with high harmonic content. This is a very old problem. The LED issues have just been added recently.
How mechanical relays are a solution I don't get. We want "half on" as well as off and on...and all points in between.

The idea of the sketch I posted is to allow the mosfets to commutate (with small cross over interruption at very low currents, unless v/i phase angle is large then this could be a real problem) while the optocoupler is high. Just Ron and a body diode in series, plus the commutation losses.  But if < 100% power is desired then the uC would turn the optocoupler off at x% of T/2 after the zero cross trigger. Basically this is what "trailing edge dimmers" do, so it's not new.
 
Unfortunately I don't understand your comment on gate transformers, it sounded like it could be interesting. Any sketch/link/part number?
« Last Edit: August 03, 2020, 01:20:56 am by oschonrock »
 

Offline Monkeh

  • Super Contributor
  • ***
  • Posts: 8135
  • Country: gb
Re: Mains switching research break out
« Reply #9 on: August 03, 2020, 01:13:59 am »
How mechanical relays are a solution I don't get. We want "half on" as well as off and on...and all points in between.

Up to this point you've never suggested you wanted a dimmer instead of a switch.
 
The following users thanked this post: BrianHG

Offline temperance

  • Frequent Contributor
  • **
  • Posts: 687
  • Country: 00
Re: Mains switching research break out
« Reply #10 on: August 03, 2020, 01:29:05 am »

I am trying to design a "breakout module" that does this "properly" and safely. It may/will ultimately need to come in a proper physical case, properly isolated, protected & appropriately fused etc so that people can do these things without killing themselves or the equipment.


-Perhaps the best place to start:
-EN/IEC 60950-1
-EN/IEC 61000-4-5
-EN/IEC 61000-4-4

For something you want to sell you don't want to settle for anything less.

Solutions for what you try achieve do exist and are called solid state relays. You will find some with MOSFET's made by Panasonic, IXYS, Omron,... But those are only used in well defined environments.

To give you an idea about what you might encounter in a not so well defines environment: a simple refrigerator can create an over voltage with enough energy under the right conditions to kill any MOSFET if it's not properly protected.
 

Offline oschonrockTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: gb
Re: Mains switching research break out
« Reply #11 on: August 03, 2020, 01:37:16 am »

I am trying to design a "breakout module" that does this "properly" and safely. It may/will ultimately need to come in a proper physical case, properly isolated, protected & appropriately fused etc so that people can do these things without killing themselves or the equipment.


-Perhaps the best place to start:
-EN/IEC 60950-1
-EN/IEC 61000-4-5
-EN/IEC 61000-4-4

For something you want to sell you don't want to settle for anything less.

Solutions for what you try achieve do exist and are called solid state relays. You will find some with MOSFET's made by Panasonic, IXYS, Omron,... But those are only used in well defined environments.

To give you an idea about what you might encounter in a not so well defines environment: a simple refrigerator can create an over voltage with enough energy under the right conditions to kill any MOSFET if it's not properly protected.

Thanks. I am well familiar with SSRs. I used to work for Omron a long time ago ;-)

They can't turn off half half-way through the cycle to my understanding, so they cannot "duty cycle/ dim etc" , so this is a totally different application.
 

Offline oschonrockTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: gb
Re: Mains switching research break out
« Reply #12 on: August 03, 2020, 01:38:40 am »
How mechanical relays are a solution I don't get. We want "half on" as well as off and on...and all points in between.

Up to this point you've never suggested you wanted a dimmer instead of a switch.

I am sorry. Not explicit enough. I thought the "zero cross trigger" would be a give away. If I just wanted to turn stuff on/off I would use an SSR.
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 13218
Re: Mains switching research break out
« Reply #13 on: August 03, 2020, 01:43:57 am »
Errrr Guys . . . . . . .

Meanwell IRM-02-12S is a 12V power supply, not a bridge rectifier. 

Now we've cleared up that confusion,   there are three things I don't like about it.

1. No suppression of back-EMF spikes from inductive loads.  Unless you use avalanche rated MOSFETs you *NEED* a beefy bidirectional TVS diode across the pair of MOSFETs.  It may also benefit from a RC snubber to control dV/dt below the TVS breakover voltage.

2. Heavy load on the opto + limited CTRR at elevated temperature makes it too easy to get insufficient gate drive if your drive to the opto LED is weak.   Unless you can guarantee sharp 10mA drive pulses, it would be a good idea to run the opto with a higher load resistor and add a Schmidt trigger input gate driver between it and the MOSFETs.

3. That simple ZC opto circuit is going to give a rather mushy ZC pulse. You may wish to consider using one of the many snap-action active ZC detector circuits, possibly separate ones for rising and falling zero crossing  Their optoisolator phototransistors can always be paralleled if you don't want to discriminate between the crossing directions.
 
Adding load current monitoring could also be rather useful.   It also opens up the possibility of protecting the MOSFETs by detecting over-current and rapidly driving the gates off, then locking out the drive till the next ZC pulse.   Some additional inductance in series with the load may be required to give the over-current trip time to react.
 

Offline oschonrockTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: gb
Re: Mains switching research break out
« Reply #14 on: August 03, 2020, 01:44:37 am »
photovoltaic optocouplers, each one

Actually these devices are new to me. You mean this yes?

https://www.vishay.com/docs/83469/vom1271.pdf

Not cheap, and not sooo available (yet?), but they look fab! I can ditch the whole annoying isolated 12V SMPS! Not sure of the limitations yet, ie I suspect that they top out at a certain size MOSFET? Too much energy required to switch efficiently - gate capacitance?

Thank you. That might really move it forward.
 

Online BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: Mains switching research break out
« Reply #15 on: August 03, 2020, 01:45:31 am »
Unfortunately I don't understand your comment on gate transformers, it sounded like it could be interesting. Any sketch/link/part number?
You know, any low voltage, IE 15 vin 15v out transformer.  Put AC 15v in and get AC 15v out, galvanically isolated.

Some people find or have custom transformers made.  Some use telecom transformers or network transformers.  It might make you like easier just to use an isolated DC-DC converter, 5v in, 15v out.

You can have one designed with 1 primary and 2 secondary, to be used in conjunction to power a true high speed optocoupled mosfet gate driver which drives the gate high and low, like a logic gate output.  Such a device is designed for high speed switching, but you would also need 2 of them for each gate.

This is what a 2 channel HV isolation driver looks like.  Though, you would be wiring the output a little differently and the input side would need inverted inputs.

https://www.mouser.com/ProductDetail/Texas-Instruments/UCC21220ADR?qs=qSfuJ%252Bfl%2Fd5KgCu6IE%2FAWg%3D%3D

With 40ns delay and 5ns switching, you may now officially PWM cleanly into the 100KHz region, maybe 1MHz.

But, the HV side needs a continuous clean +/-15v isolated DC-DC power supply to create the +/- gate voltages centered from the mosfet's central 'source'.  Something like this:
https://www.mouser.com/ProductDetail/Cincon/EC1SA05N?qs=%2Fha2pyFaduhjhDCN%2Fb9F8eSzvtBCYVIMWKpaSLbxmGc%3D

Otherwise you would need 2 of your power supplies, one for each gate.

Now, I need not say that you need to know what you are doing at every junction and how things function at high AC mains voltages, otherwise, such a project could end in electrical shock or fire.

« Last Edit: August 03, 2020, 01:48:00 am by BrianHG »
 

Offline oschonrockTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: gb
Re: Mains switching research break out
« Reply #16 on: August 03, 2020, 01:50:14 am »
Errrr Guys . . . . . . .

Meanwell IRM-02-12S is a 12V power supply, not a bridge rectifier. 

Now we've cleared up that confusion,   there are three things I don't like about it.


Oh, is that where the rectifier came from. Thanks for clearing that up, and sorry to cause confusing. Note to self: Give more verbose description of circuit.


1. No suppression of back-EMF spikes from inductive loads.  Unless you use avalanche rated MOSFETs you *NEED* a beefy bidirectional TVS diode across the pair of MOSFETs.  It may also benefit from a RC snubber to control dV/dt below the TVS breakover voltage.

2. Heavy load on the opto + limited CTRR at elevated temperature makes it too easy to get insufficient gate drive if your drive to the opto LED is weak.   Unless you can guarantee sharp 10mA drive pulses, it would be a good idea to run the opto with a higher load resistor and add a Schmidt trigger input gate driver between it and the MOSFETs.

3. That simple ZC opto circuit is going to give a rather mushy ZC pulse. You may wish to consider using one of the many snap-action active ZC detector circuits, possibly separate ones for rising and falling zero crossing  Their optoisolator phototransistors can always be paralleled if you don't want to discriminate between the crossing directions.


Adding load current monitoring could also be rather useful.   It also opens up the possibility of protecting the MOSFETs by detecting over-current and rapidly driving the gates off, then locking out the drive till the next ZC pulse.   Some additional inductance in series with the load may be required to give the over-current trip time to react.

These are all super useful. Thanks and I agree. I don't quite understand point number 3. Will look into that. Or do you have a useful link?

Do you have experience with those photovoltaics? like that VOM1271 in above post?  Looks almost too good to be true? How can that have enough energy to switch properly? Even with the 12V SMPS in there you make some valid point about snappy efficient switching...so these photowhatsits...??
 

Online BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: Mains switching research break out
« Reply #17 on: August 03, 2020, 01:52:02 am »
photovoltaic optocouplers, each one

Actually these devices are new to me. You mean this yes?

https://www.vishay.com/docs/83469/vom1271.pdf

Not cheap, and not sooo available (yet?), but they look fab! I can ditch the whole annoying isolated 12V SMPS! Not sure of the limitations yet, ie I suspect that they top out at a certain size MOSFET? Too much energy required to switch efficiently - gate capacitance?

Thank you. That might really move it forward.
Those are great for ON-OFF switching and chopping, but useless for multi-KHz high speed PWM.  For high speed PWM throughout the sine wave, you need the item I listed in 1 post above.

They are only $1.67 each (for 10) here:  (~1$ or less in significant quantity)
https://www.ttiinc.com/content/ttiinc/en/apps/part-detail.html?mfrShortname=VOP&partsNumber=VOM1271T&autoRedirect=true&minQty=2000
At that price, plus you don't need that 12v power supply, you save money by using 2 of these, 1 for each gate.  Remember, they guarantee a minimum of 7.8v for your gate voltage, so, choose your mosfet wisely, or run 2 in series.

Remember, with these optos, you want to drive the LED with 30-50ma each, that's 60-100ma total for both.  This means coming out of your UC, I would add a transistor or strong small mosfet (one of those cheap 1.2amp sot23 logic level mosfets) with 2 separate series resistors on the drains to drive each opto.
« Last Edit: August 03, 2020, 02:27:42 am by BrianHG »
 

Offline oschonrockTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: gb
Re: Mains switching research break out
« Reply #18 on: August 03, 2020, 02:04:40 am »
Yeah, the gate transformers what I thought, and had seen that before, don't fancy it.

But this:

This is what a 2 channel HV isolation driver looks like.  Though, you would be wiring the output a little differently and the input side would need inverted inputs.

https://www.mouser.com/ProductDetail/Texas-Instruments/UCC21220ADR?qs=qSfuJ%252Bfl%2Fd5KgCu6IE%2FAWg%3D%3D

With 40ns delay and 5ns switching, you may now officially PWM cleanly into the 100KHz region, maybe 1MHz.

But, the HV side needs a continuous clean +/-15v isolated DC-DC power supply to create the +/- gate voltages centered from the mosfet's central 'source'.  Something like this:
https://www.mouser.com/ProductDetail/Cincon/EC1SA05N?qs=%2Fha2pyFaduhjhDCN%2Fb9F8eSzvtBCYVIMWKpaSLbxmGc%3D


That looks like a "proper" solution.  These are the sorts of drivers that will give this much more flexibility in terms of switching strategies.

That little power supply is quite fine. It would be annoying to gave to use it in addition to the 230VAC SMPS I have there already, but if I "require" 5V on the "safe right-hand side" of the schematics and use that - by relying on the 1000V insolation which that little block provides...then that's a gain. ..?
 

Online BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: Mains switching research break out
« Reply #19 on: August 03, 2020, 02:10:03 am »
Yeah, the gate transformers what I thought, and had seen that before, don't fancy it.

But this:

This is what a 2 channel HV isolation driver looks like.  Though, you would be wiring the output a little differently and the input side would need inverted inputs.

https://www.mouser.com/ProductDetail/Texas-Instruments/UCC21220ADR?qs=qSfuJ%252Bfl%2Fd5KgCu6IE%2FAWg%3D%3D

With 40ns delay and 5ns switching, you may now officially PWM cleanly into the 100KHz region, maybe 1MHz.

But, the HV side needs a continuous clean +/-15v isolated DC-DC power supply to create the +/- gate voltages centered from the mosfet's central 'source'.  Something like this:
https://www.mouser.com/ProductDetail/Cincon/EC1SA05N?qs=%2Fha2pyFaduhjhDCN%2Fb9F8eSzvtBCYVIMWKpaSLbxmGc%3D


That looks like a "proper" solution.  These are the sorts of drivers that will give this much more flexibility in terms of switching strategies.

That little power supply is quite fine. It would be annoying to gave to use it in addition to the 230VAC SMPS I have there already, but if I "require" 5V on the "safe right-hand side" of the schematics and use that - by relying on the 1000V insolation which that little block provides...then that's a gain. ..?
You can find 240v side supplies which have a center-tap +/-15v output.
That TI chip drives the gate both high and low with a good 4amp surge making a true high frequency switching solution possible and 15v means the cheap bottom end mosfets will be completely turned on and off.  Since you arent driving the output 2-n channels shorting a DC output high and low, you only need a series resistor of something like 20ohm between each TI chip output and each gate of each mosfet.

The TI chip is only rated 3KVrms for 1 minute only between logic side and the isolated sides.
That little supply is rated 1kv minimum continuous.  I think that the TI chip is also only 1kv continuous.
« Last Edit: August 03, 2020, 02:17:46 am by BrianHG »
 

Offline oschonrockTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: gb
Re: Mains switching research break out
« Reply #20 on: August 03, 2020, 02:14:20 am »

Those are great for ON-OFF switching and chopping, but useless for multi-KHz high speed PWM.  For high speed PWM throughout the sine wave, you need the item I listed in 1 post above.

They are only $1.67 each (for 10) here:  (~1$ or less in significant quantity)
https://www.ttiinc.com/content/ttiinc/en/apps/part-detail.html?mfrShortname=VOP&partsNumber=VOM1271T.&customerPartNumber=&minQty=10&customerId=
At that price, plus you don't need that 12v power supply, you save money by using 2 of these, 1 for each gate.  Remember, they guarantee a minimum of 7.8v for your gate voltage, so, choose your mosfet wisely, or run 2 in series.

Remember, with these optos, you want to drive the LED with 30-50ma each, that's 60-100ma total for both.  This means coming out of your UC, I would add a transistor or strong small mosfet (one of those cheap 1.2amp sot23 logic level mosfets) with 2 separate series resistors on the drains to drive each opto.

Yeah that makes total sense. 2 good options there.

1. photovoltaics  for chopping (lose the SMPS), or
2. +/-15V isolated supply powered from low side with a HF MOSFET driver (not much more expensive either) to add PWM capability as well

Both are better than what I have. Thank you.

 

Offline oschonrockTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: gb
Re: Mains switching research break out
« Reply #21 on: August 03, 2020, 02:28:55 am »

You can find 240v side supplies which have a center-tap +/-15v output.

Phew...Too many parts.... Link?   I agree that might be better. More robust and fewer bridges across the isolation.

That TI chip drives the gate both high and low with a good 4amp surge making a true high frequency switching solution possible and 15v means the cheap bottom end mosfets will be completely turned on and off. 

Yeah this is what we want. Also means I can easily produce multiple versions of this thing with different MOSFET ratings. 100W -> 2.5kW @ 240VAC. That driver will laugh at all those.

Since you aren't driving the output 2-n channels shorting a DC output high and low, you only need a series resistor of something like 20ohm between each TI chip output and each gate of each mosfet.

You mean I don't need shoot though prevention...? Or you mean high-side bootstrap?

Can ditch the optocoupler too...obviously

The uC now needs to provide 2 coordinated signals, but has total control for all switching / chopping / PWM  strategies?

Thank you. This has been very helpful.
« Last Edit: August 03, 2020, 02:34:27 am by oschonrock »
 

Offline oschonrockTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: gb
Re: Mains switching research break out
« Reply #22 on: August 03, 2020, 02:43:20 am »

You can find 240v side supplies which have a center-tap +/-15v output.

Phew...Too many parts.... Link?   I agree that might be better. More robust and fewer bridges across the isolation.


This works. Bit overkill at 340mA and expensive @ £8-9 each.

https://uk.farnell.com/recom-power/rac10-15dk-277/power-supply-ac-dc-2-o-p-10w/dp/2822844

If anyone has better options let me know.


 

Online BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: Mains switching research break out
« Reply #23 on: August 03, 2020, 03:03:28 am »
According to your circuit, your making an AC PWM on/off switch, not a variable output DC PWM supply.

A variable DC PWM output supply has a topology where when you turn on the low side and high side mosfet simultaneously, you short the V+ rail to your GND.  You are not making this, so in the TI datasheet, the example circuit on page 1 with the added diodes and resistors to tame the turn on and turn off speed do not apply to your design as both mosfets are either on, or off at the same time.  All you need is a series resistor from each output to each gate.

I made a little mistake, you only need 1 cheaper single DC-DC 15v isolated converter and a single mosfet driver as your mosfet sources are connected together.  You only need to go from 0v to 15v from the 'source' on both.

In this case, here you go, faster & cheaper & 3.75kv isolation: (remember, best to have a series resistor on the output pin driving the gates as you shouldn't drive them with an absolute 4 amp 8ns fall time pulse.)
https://www.digikey.com/product-detail/en/silicon-labs/SI8261AAC-C-ISR/336-4490-1-ND/8540659

Plus here is the cheapest isolated 3kv DC-DC converter: (Digikey Datasheet typo, this guy is only 1.5kv rated.)
https://www.digikey.com/product-detail/en/cui-inc/PDSE1-S5-S15-S/102-6294-ND/10229829

I have better faith in this one rated at 3kv isolation: (This one is authentic 3kv isolation for only a few cents more)
https://www.digikey.com/product-detail/en/cui-inc/PEME1-S5-S15-S/102-6340-ND/10229875
(You may need a parallel 15.2-16v zener diode on the output of these DC-DC converter as they have a minimum load of 7ma, otherwise the voltage may creep above the spec 15v.)
« Last Edit: August 03, 2020, 03:23:01 am by BrianHG »
 

Online BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: Mains switching research break out
« Reply #24 on: August 03, 2020, 03:34:41 am »
4 amp output version of the silicon labs gate driver IC:
https://www.digikey.com/product-detail/en/silicon-labs/SI8261BBC-C-ISR/336-5208-1-ND/9175803
Only ~13 cents more.

BTW, my above change means you only need 1 photovoltaic optoisolator as well, or 2 in series to get 15v output gate drive.
« Last Edit: August 03, 2020, 03:37:14 am by BrianHG »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf