Notice however that Tahmids circuits the VSS and the COM are common'd up.
For the isolated inputs to be active, they would not require to be common.
Max.

 

Notice however that Tahmids circuits the VSS and the COM are common'd up.
For the isolated inputs to be active, they would not require to be common.
Max.

Thanks for pointing that out; realistically though, if I use this IR2110 driver, it wouldn't need isolation. Also, I just noticed that the Bootstrap Calculator only works above frequencies of 2500 Hz, but I think that whatever method of calculation Tahmid was using, the relation between frequency and C1 [Cb] is not linear as I'd mistakenly thought at first glance: i.e. he increases C tenfold for reduction in frequency by 1000 x (to 50Hz in his example).

So, if I'm reducing frequency by another 10 x to say 5Hz, then I only have to increase the capacitance of C1 by something like the cube root of 10, 2.154. Around 100 microF. Getting ahead of myself a bit probably.

 

I have never used the IR2110 without separate isolation if needed, I will have to study the specs and see if it is possible, but just at refresh, it may not be possible, i,e in and out not isolated.
Max.

 

I have never used the IR2110 without separate isolation if needed, I will have to study the specs and see if it is possible, but just at refresh, it may not be possible, i,e in and out not isolated.
Max.

Well, I started this thread wondering about a full bridge driver, and have since learned a fair bit about how to use a half bridge driver such as this IR2110. I'm sure there are by now some good substitutes incorporating isolation; although perhaps not for the IGBTs I'm trying to use (e.g. TI has half bridge drivers for MOSFETs in the parameters required, but not for IGBTs).

The fact that I'm using only 12V Vbus for IGBTs is also considered peculiar and I'm not sure why that is.

 

james31207

Yes & no at the same time.
I myself am trying to construct 200VDC Bridge Driver to run a DC motor.
After I have tried a IC type drivers I was extremely dissapointed that the slow rate of discharge combined with the ESD from the motor itself causes gate threshold breakdown & short circuits.
I have even tried building them from 2 N chanel & 2 P channel mosfets straight bridging the gates with the same short circuit results.
I have now resigned to the fact I will be using the same 2N & 2P mosfets but will be using an external microprocessor like an Arduino to drive them so I can give a small wait time to let everything discharge before switching polarities.

I found a better solution that may help you to.
It's simply a "solid state DC relay"
Hope this helps

 

The IR2110 is not an isolated gate driver, so I wonder whether it is suitable to drive these IGBTs.

I'm sure both International Rectifier(the ones who developed the driver) and now Infineon Technologies who bought the design, will be heartbroken to find out the IR2110 is not an isolated driver. Which it is by the way, the switched voltage, 500V shows you or anyone wondering that it is an isolated driver.

As to the suitability of working for IGBT's yes it is suitable. At one time the IGBT was plagued with something called tail currents, that didn't allow them to switch off fast. That problem has been solved but it still shows up in old references on the web about IGBT's. The stuff that once hits the web takes on a life of its own.

 

Same guy, same idea (which is not to produce large quantities of electrical power or motor torque)-- and yes, the relation between the rotor core (permanent magnets) and the stator coils is similar to a stepper motor; however the design is not intended as a motor, but as a generator in which the winding is placed between the two, carried on the rotor. Since the structure of the core is critical, there is no place for soft iron etc. in that rotor core; and it is also important to appreciate the geometric configuration of the field coils to see how these will drive the rotor, and act as the principal part of the inductive field.
I have NOT managed yet to make the rotor spin past 135 degrees for want of the switch I need to build or otherwise acquire -- which is why I'm asking questions here in as sensible and useful a fashion as I can. And if ever the contraption 'works', meaning that some discernible current, however small, is generated in the winding -- which I reiterate is exposed in its entirety to a changing magnetic field at all times (unlike conventional machines) --, you'll be the first to know. At no stage incidentally have I suggested that this is in any sense a 'greater than unity' device and wonder why people are forever suspicious of such intent in the face of unfamiliar notions. And it's an experimental apparatus, singular; wood is good enough for that single prototype.

Your suggestion of a stepper motor driver might be just what's required; and I wish you'd mentioned it earlier. As you realize, the rotor needs to spin in one direction continuously, and this requires that all the electromagnets [6 N, 6S] are switched on at all times, each set of 6 switching polarity every half-cycle of 180 degrees, but each set 90 degrees out of phase with the other. Bearing in mind then that this set-up would be akin to driving a bipolar permanent magnet stepper motor except that the electromagnets must be charged continuously -- not on/off etc. --, can you suggest a stepper motor driver that can be programmed to do that?

Can you clear up the question I asked you though, whether the driver output circuit and the H-Bridge are basically in parallel?

Last question first. A H-bridge is made up of two half bridges. And the half bridge can also be made into a 3phase or more driver by just adding more into a suitable circuit.


Now to what you said in the first paragraph. Why now as in your last thread can't/won't you answer the direct questions? Questions that if answered with any kind of reasonable answer would put the controversy of this to rest.

If you stop and think about what you propose it makes no sense, flies in the face of all proven physics. Your internal magnets yes will have some effect on making this into a very inefficient motor. I say inefficient because of the distance between the internal permanent magnets(PM's), and the solenoid electromagnets. You are making with that configuration something similar to a BLDC motor, but with much distance between the different magnets.

But a BLDC only makes a very small BEMF while running just enough that some types of ESC's use it to calculate where and when to turn on and off the coils. That is called ,Electronic Commutation, other BLDC motors use Hall sensors and PM's to do the same thing.

Now as to why I say you won't make any electrical output from this thing. It takes either a magnet field moving in a coil or a coil moving in a magnetic field to generate electricity. To the best of my knowledge and from anything I can find, other than the link you posted about this "project". PM's moving with a coil can't/won't generate a voltage. Now it may give again a very small, but measurable, voltage when the outside EM(electromagnet) is on between the coil and PM (and I'm not even sure that is true, but will give the benefit of doubt to it) the voltage made will be nowhere near by many magnitudes the voltage put into the EM's to cause the spin of the motor.

So after saying this I ask what is the purpose of this project? You would be better off joining forces with lowQcab and use his magic mystical motor to make a generator.

 

Only 300 years ago you would have been put to death for committing heresy if you
simply stated that the Earth was round, or that you could produce Lightning in a bottle.

And only around 3 years ago this type of stuff wouldn't even been tollerated on this forum. But for whatever reason it's now allowed.

 

Last question first. A H-bridge is made up of two half bridges. And the half bridge can also be made into a 3phase or more driver by just adding more into a suitable circuit.


Now to what you said in the first paragraph. Why now as in your last thread can't/won't you answer the direct questions? Questions that if answered with any kind of reasonable answer would put the controversy of this to rest.

If you stop and think about what you propose it makes no sense, flies in the face of all proven physics. Your internal magnets yes will have some effect on making this into a very inefficient motor. I say inefficient because of the distance between the internal permanent magnets(PM's), and the solenoid electromagnets. You are making with that configuration something similar to a BLDC motor, but with much distance between the different magnets.

But a BLDC only makes a very small BEMF while running just enough that some types of ESC's use it to calculate where and when to turn on and off the coils. That is called ,Electronic Commutation, other BLDC motors use Hall sensors and PM's to do the same thing.

Now as to why I say you won't make any electrical output from this thing. It takes either a magnet field moving in a coil or a coil moving in a magnetic field to generate electricity. To the best of my knowledge and from anything I can find, other than the link you posted about this "project". PM's moving with a coil can't/won't generate a voltage. Now it may give again a very small, but measurable, voltage when the outside EM(electromagnet) is on between the coil and PM (and I'm not even sure that is true, but will give the benefit of doubt to it) the voltage made will be nowhere near by many magnitudes the voltage put into the EM's to cause the spin of the motor.

So after saying this I ask what is the purpose of this project? You would be better off joining forces with lowQcab and use his magic mystical motor to make a generator.

The purpose of the project is unclear (and I should remind you too that at no stage did I 'present' this idea to this forum or attempt to solicit interest in it. I simply answered questions put to me about the reasons for my own questions concerning how to reverse polarities in electromagnets; which is fair enough as far as it goes).

What you say now is generally true (except for your characterisation of the idea as 'nonsense', 'flying in the face of physics' and so on; if you knew me at all, you wouldn't say things like that), or anyway, the issues you raise are valid and pertinent, but what you are neglecting is that most of the winding -- which you're calling a coil -- is being exposed not only to a changing magnetic field [dB/dt in the Maxwell-Faraday equation etc.], but one capable of generating an EMF in that winding. The EM inductive effect is not restricted just to those points in that winding immediately adjacent to the permanent magnets of the core (although it is important that every vertical component of the winding does abut the core PM at some point; in fact, each successive component is adjacent to a PM component incrementally shifted in latitude and longitude, so to speak. Do you see that?).

At the same time, although the rotor is not envisaged as spinning much faster than say 480 rpm [for switching at 8Hz], it is supposed that the dB/dt will be somewhat greater than expected due to the changing obliquity of the lines of force arising between the PMs and the stator coils through which much of the winding effectively passes. Imagination is necessary here; you have to imagine the magnetic propeller as a kind of lever within the effect, like any propeller (like a sail or a screw), moreover behaving continuously in the same direction of rotation (unlike conventional loop AC generators knocking the copper back and forth like a row-boat). There is a dramatic change in the direction of components of EMF in the winding in the vicinity of the core PM, and a complete flip at certain points.

Of course all this is somewhat hypothetical, but the whole thing is just an experiment upon that notion: that because the winding is singular and progressive, and is exposed continuously to a potentially EM inductive effect, all these components of inductance will add up, or integrate, to create a voltage wave which is quite distinct from AC generated conventionally. I could probably just let it go and say what the heck, but something seems to prevent that, to maintain the obsession; maybe something to do with the way the whole thing dropped into my head about 30 years ago last Thursday).

Again, as far as the aim is concerned, it is not to generate quantities of electrical power, not at any scale; but if even a small current of the kind I am IMAGINING is found, then its expected properties will be rigorously investigated and if determined to be useful, scaled up.

Now, are you gonna answer my simple electronics toddler questions or not? So far you've been quite helpful in that respect, and I genuinely appreciate it. Just leave the physics, as outlandish as it might seem on the face of it -- it isn't really much of a stretch -- and all the rest of the apparatus to me will you. What I need to do is put together my H-Bridge with the right gate driver.

I'm relieved that the IR2110 is what it's supposed to be after all; it was just something in the blog by Tahmid to which MaxHR directed me where he seems explicitly to state, in a blog actually about the IR2110, that it is a 'non-isolated' driver; so unless I misread it, his credibility is now in about the same class as Pinocchio. [And indeed, the IR2110 is according to the specs, non-isolated].

The other thing which is bothering me is the practical difference between a high/low side driver like the IR2110S and a half-bridge driver, which seems to come down to the difference between the SD and L[in] pins on the input side. A good short answer could save me days of misery here dude.

Did you look up 'Faraday's Paradox'? Not that it necessarily has much bearing on all this, but it does show you how peculiar the relation between ferromagnetic fields and copper can be.

Nice talking to you Shortbus: Greetings from Sydney. G'day.

 

I have never used the IR2110 without separate isolation if needed, I will have to study the specs and see if it is possible, but just at refresh, it may not be possible, i,e in and out not isolated.
Max.

Apparently the IR2110 is in fact an isolated gate driver. I was misled by Tahmid who seems to state the contrary in his blog, in the text between figures 4 and 5.

 

The other thing which is bothering me is the practical difference between a high/low side driver like the IR2110S and a half-bridge driver, which seems to come down to the difference between the SD and L[in] pins on the input side.

In designs so far I have assumed that total galvanic isolation does Not exist between input and output and therefore used isolation
Just to get clarification I have submitted a technical request to Infineon Tech queries for a definitive answer.
Max.

 

You might want to look at "microstepping" https://www.linearmotiontips.com/microstepping-basics/ You would have some complications to deal with, but the concept might be valid.

 

A more comprehensive description of the necessity for a torque/current constant.
https://www.controleng.com/articles/stepper-motor-torque-basics/

 

A more comprehensive description of the necessity for a torque/current constant.
https://www.controleng.com/articles/stepper-motor-torque-basics/

Returning to the previous subject of Tahmid's blog, do you agree that the large resistors across the gate-source pins are necessary in this case [of IGBTs], in which the Vce [Vbus, rail voltage, whatever it's called, the Vds] is only 12V?

 

I have dealt mainly with Mosfets, and experience with IGBT's limited.
I have these that may, or not help.
Max.

 

What I need to do is put together my H-Bridge with the right gate driver.

The other thing which is bothering me is the practical difference between a high/low side driver like the IR2110S and a half-bridge driver, which seems to come down to the difference between the SD and L[in] pins on the input side. A good short answer could save me days of misery here dude.

A high/low side gate driver is a half bridge driver. Putting two in a circuit with a signal on the Hin of one and the Lin of the other makes them into a H-bridge driver. Turning both Hin and Lin on only one driver lets out smoke.

If you read the data sheets of those drivers or any of the type , you will see some of them have other added things and some are simple drivers. The SD stand for shut down, letting you stop the driver without turning off the power to it. There is something else some have that shuts down the circuit when the voltage level goes too low.

 

So far you've been quite helpful in that respect, and I genuinely appreciate it. Just leave the physics, as outlandish as it might seem on the face of it -- it isn't really much of a stretch -- and all the rest of the apparatus to me will you. What I need to do is put together my H-Bridge with the right gate driver.

I helped before you went into the fringe realms. Doing an H-bridge is easy but yet you struggle with it. But you seem to think you've found a way to get the "holy grail", without even having an understanding of simple electonics ?

 

@james31207 This is the answer from Infineon:


"The input and output of IR2110 is not isolated.
If your application requires input/output isolation, then you may add isolation circuit before input."
Max.

 

@james31207 This is the answer from Infineon:


"The input and output of IR2110 is not isolated.
If your application requires input/output isolation, then you may add isolation circuit before input."
Max.

While that may be the answer to the Galvanic isolation question on these drivers it isn't what they state in the App notes for them being isolated in the common sense of the word, and common use of them, non galvanic isolation that allows them to work with 500 to 1200 in some cases of volts.

Quote, "2.3 High-Side Channel
This channel has been built into an “isolation tub” (Figure 3) capable of floating from 500 V or
1200 V to -5 V with respect to power ground (COM). The tub “floats” at the potential of VS.
Typically this pin is connected to the source of the high-side device, as shown in Figure 2 and
swings with it between the two rails." From App note AP-978 https://www.infineon.com/dgdl/Infin...N.pdf?fileId=5546d4626c1f3dc3016c47de609d140a

This is what I was reffering to as them being isolated.

 

I helped before you went into the fringe realms. Doing an H-bridge is easy but yet you struggle with it. But you seem to think you've found a way to get the "holy grail", without even having an understanding of simple electronics ?

Fair enough, although I don't think of this as any 'holy grail'. As for electronics, I've not been required to have any knowledge of electronics whatsoever until last year, nor have I ever had any interest in it, and certainly no practical interest. (I'm a doctor, an MD as you say in the States). Any understanding I have of electronics is broadly theoretical and comes from reading Feynman's lectures (especially on QM), as well as bits and pieces gleaned from forums and the internet -- sometimes when I've got an hour or two to spare. Nor do I have a teacher to smooth the way or to answer typical questions quickly.

Obviously therefore, I have zero practical experience in electronics and although I've got a pretty good idea now about what the circuitry of this whole H-Bridge switch will look like, the prospect of soldering everything together correctly terrifies me. I only learned to solder 3 months ago from my father who learnt from his father when he was about 12. He's 90 now with hands too shaky to do the job himself (although he doesn't know much electronics). We even used the same tin of flux from my grandfather, you know, from 1920 or so.

I will get it done though, one way or another, and if you were in town sir, I could very easily be persuaded to pay you handsomely to build the thing for me. Evidently there's a bit of confusion however about the isolation characteristics of the IR2110, although probably this is inconsequential here, and seems suitable for the job. I'll study the link. Thanks again.