So you're saying at 27 mV my starting current is 36 amps?

It certainly looks that way, although I'll admit it's a higher number than I would've guessed.

 

Thanks. Now I am thinking my 30 amp MOSFET won't handle this air injector pump.

 

I have in my possession some IRG4PSH71K IGBT's. The data sheet says:
Continuous Collector Current 78A @ 25˚C
Continuous Collector Current 42A @ 100˚C
Pulsed Collector Current 156A
Clamped Inductive Load Current 156A

Now, the hard part - - - I don't know how to wire an IGBT. Should it switch the positive? Or negative? And given that I'm controlling this from PWM sinked line. Should I "Pull UP"?

 

I have in my possession some IRG4PSH71K IGBT's. The data sheet says:
Continuous Collector Current 78A @ 25˚C
Continuous Collector Current 42A @ 100˚C
Pulsed Collector Current 156A
Clamped Inductive Load Current 156A

Now, the hard part - - - I don't know how to wire an IGBT. Should it switch the positive? Or negative? And given that I'm controlling this from PWM sinked line. Should I "Pull UP"?

Let me start by saying that I've never used an IGBT. Nevertheless, I think I understand the concept. Basically you'll wire them the same way you'd wire their MOSFET equivalent.

So, in the case of your N channel device, you'll want to connect your load to the supply voltage on one side, and the collector of your IGBT on the other. Then you'll connect the IGBT emitter to ground. Finally, the IGBT gate will be activated by positive voltage, up to an absolute max of 20V above emitter.

If your PWM controller only sinks, then you'll need a pull up resistor to your 12V supply.

 

@ebeowulf17 Thanks. That's what I was thinking. Always good to make sure I'm not doing something entirely wrong.

 

Always good to make sure I'm not doing something entirely wrong.

What do you have against sparks and smoke?

 

Hey Tony. Your 100 Amp 75 mV shunt will output 0.000075 Volt per Amp or 0.075 mV per Amp. A 100 Amp 100 mV shunt would be 1.0 mV per amp. You have 75 mV / 100 Amps = 0.075 Mv per Amp.

The best you can resolve as to reading current will depend on the uncertainty and resolution of the meter you use to measure the shunt plus any uncertainty of the shunt.

So if your meter reads 7.5 mV then yes, you have 10 amps. The problem comes when you get low current like 1 Amp on a 100 Amp shunt because you have such a low output and then there is some noise which generally makes for reading low currents like a few amps on a 100 amp shunt a bit of a pain. My best resolution on my hand held Fluke 87 DMM is 0.1 mV so with your shunt and my DMM I would have problems seeing 1.0 Amp. This is where I go to a better bench type DMM for better resolution and uncertainty (accuracy).

<EDIT> Well while I fed the dogs and slowly typed I see everything is answered. </EDIT>

Ron

 

Hey Tony. Your 100 Amp 75 mV shunt will output 0.000075 Volt per Amp or 0.075 mV per Amp. A 100 Amp 100 mV shunt would be 1.0 mV per amp. You have 75 mV / 100 Amps = 0.075 Mv per Amp.

The best you can resolve as to reading current will depend on the uncertainty and resolution of the meter you use to measure the shunt plus any uncertainty of the shunt.

So if your meter reads 7.5 mV then yes, you have 10 amps. The problem comes when you get low current like 1 Amp on a 100 Amp shunt because you have such a low output and then there is some noise which generally makes for reading low currents like a few amps on a 100 amp shunt a bit of a pain. My best resolution on my hand held Fluke 87 DMM is 0.1 mV so with your shunt and my DMM I would have problems seeing 1.0 Amp. This is where I go to a better bench type DMM for better resolution and uncertainty (accuracy).

<EDIT> Well while I fed the dogs and slowly typed I see everything is answered. </EDIT>

Ron

I believe you slipped a tens place on a few of those: it's 0.75mV for each 1A, not 0.075mV.

 

Voltage across shunt when:

Starting = 27 mV
Running = 12.7 mV

Don't give me the answer, give me the formula I'm so badly missing. It's probably stupidly obvious, which is why I can't see it.

EDIT: I'm coming up with 20.25 Starting Amps and 9.5 running amps. Am I even close? It sounds reasonable. Just want to be sure.

So the starting current is 27mv/75mv x 100 Amps and the running current is 12.7mV/75mV x 100 Amps. And since the units all work the number is reasonable.
In addition, 36 amps is a fairly believable starting current for a fair sized DC motor .
But when the motor is built into an automotive electrical system you will discover that the starting current is less because the wiring harness resistance drops the voltage a bit, and so the actual current is less. I fought with that for a couple of weeks when we built a production tester for automotive heater/AC assemblies. We had a fast responding very well regulated 39 amp supply and they were telling us that was what they wanted, but that our current readings were all far off on the high side. So I asked them to monitor the motor voltage at the motor terminals in an actual vehicle, the same as they were monitoring the current, It was quite an education for them as there voltage dropped by almost 3 volts for quite a few milliseconds when the motor first started. Suddenly they decided to change the test specification a bit.

 

Hey Tony. Your 100 Amp 75 mV shunt will output 0.000075 Volt per Amp or 0.075 mV per Amp. A 100 Amp 100 mV shunt would be 1.0 mV per amp. You have 75 mV / 100 Amps = 0.075 Mv per Amp.

The best you can resolve as to reading current will depend on the uncertainty and resolution of the meter you use to measure the shunt plus any uncertainty of the shunt.

So if your meter reads 7.5 mV then yes, you have 10 amps. The problem comes when you get low current like 1 Amp on a 100 Amp shunt because you have such a low output and then there is some noise which generally makes for reading low currents like a few amps on a 100 amp shunt a bit of a pain. My best resolution on my hand held Fluke 87 DMM is 0.1 mV so with your shunt and my DMM I would have problems seeing 1.0 Amp. This is where I go to a better bench type DMM for better resolution and uncertainty (accuracy).

<EDIT> Well while I fed the dogs and slowly typed I see everything is answered. </EDIT>

Ron

For low amps I can use my meter. It has a 10 amp setting and a mA setting. With the motor - I wanted to know something about it so I could build a circuit that could handle the current. From what I learned this morning, the MOSFET isn't going to work. Instead, I have several IGBT's I can use. As for a flyback diode, I have some 50 amp diodes out of a welding machine. A bit bulky, but it will do the job.

I believe you slipped a tens place on a few of those: it's 0.75mV for each 1A, not 0.075mV.

Yeah, thanks. Decimals are my downfall sometimes.

So the starting current is 27mv/75mv x 100 Amps and the running current is 12.7mV/75mV x 100 Amps. And since the units all work the number is reasonable.
In addition, 36 amps is a fairly believable starting current for a fair sized DC motor .
But when the motor is built into an automotive electrical system you will discover that the starting current is less because the wiring harness resistance drops the voltage a bit, and so the actual current is less. I fought with that for a couple of weeks when we built a production tester for automotive heater/AC assemblies. We had a fast responding very well regulated 39 amp supply and they were telling us that was what they wanted, but that our current readings were all far off on the high side. So I asked them to monitor the motor voltage at the motor terminals in an actual vehicle, the same as they were monitoring the current, It was quite an education for them as there voltage dropped by almost 3 volts for quite a few milliseconds when the motor first started. Suddenly they decided to change the test specification a bit.

I love it when some college educated engineer starts talking facts when he knows only book knowledge. It's the old timers who KNOW what's going on. But would they listen? No. Not until they're slapped in the face with reality.

Hey all; thanks for the help. I can move forward with this project in the near future. I still have to decide on a housing and plumbing hookups. But this one will get done this year for sure (hopefully).

 

For low amps I can use my meter. It has a 10 amp setting and a mA setting. With the motor - I wanted to know something about it so I could build a circuit that could handle the current. From what I learned this morning, the MOSFET isn't going to work. Instead, I have several IGBT's I can use. As for a flyback diode, I have some 50 amp diodes out of a welding machine. A bit bulky, but it will do the job.

Yeah, thanks. Decimals are my downfall sometimes.

I love it when some college educated engineer starts talking facts when he knows only book knowledge. It's the old timers who KNOW what's going on. But would they listen? No. Not until they're slapped in the face with reality.

Hey all; thanks for the help. I can move forward with this project in the near future. I still have to decide on a housing and plumbing hookups. But this one will get done this year for sure (hopefully).

OK, Tony, except that I don't know how to take the remark. I am BOTH a college educated electronic engineer and an experienced old timer. Most of my jobs were designing production line test equipment sold to the big USA auto companies. Those folks are quite picky about accuracy, repeatability, and reliability. And also they are quite willing to ask for what can not be achieved reasonably, at least some of them are.
That equation I presented translates to "(Actual reading / full scale reading)x (full scale value)", and when the units come out right the answer is usually correct.
And along the way I also learned to design the hydraulic and pneumatic parts of the systems, and the framework and rotating stuff as well.

 

I believe you slipped a tens place on a few of those: it's 0.75mV for each 1A, not 0.075mV.

Yeah, I believe I did do that. Not like my calculator made a mistake either.

Thanks for pointing that out....
Ron

 

Tony, if this is just going to be a one time test I have a Fluke Y8100 DC/AC current probe you are welcome to borrow. It's a clamp on for use with a DMM and has two ranges of 0 to 20 amps and o to 200 amps with a 0 to 2.0 volt output. This way using it on the 0 to 200 amp range you have 0 to 2.0 volts out to work with or 10 mV per amp. This works out well if for example you are running the output into a data logger or scope so you can see and plot a motor start up current or inrush currents.

Anyway, if you feel this would work out well you are welcome to borrow it and return it when you are done using it. The above link is to the data sheet so you have an overview and operating destructions. Another nice feature is you can loop the current carrying line through the jaws twice and read double the actual current.

Ron

 

The Fluke Y8100 is a good probe. A bit bulky and heavy with the batteries, but it does the job.

 

OK, Tony, except that I don't know how to take the remark. I am BOTH a college educated electronic engineer and an experienced old timer.

Sorry MisterBill, the comment wasn't intended toward you or anyone else here. It's just that I've dealt with these educated people before where their solutions defy logic and good sense. And having book knowledge is one thing but having real world practical experience is another. As I have with many here, I've come to appreciate and respect your knowledge on many levels. My sincerest thanks to you and all others who have some practical "real world" guidance for a guy like me who never made it through college and never went beyond DC theory.

Tony, if this is just going to be a one time test I have a Fluke Y8100 DC/AC current probe you are welcome to borrow.

Thanks for the offer Ron. As for my needs, I've ascertained enough usable information from the shunt. The #1 thing I wanted (needed) to know was whether the circuit I was about to build would survive in live application. Everything works on paintbrush, but when the rubber hits the road - I hate wasted effort.

I've re-designed my circuit using an IGBT I happen to have about 8 of. The only thing I need to come up with is a suitable diode to handle the current from BEMF.

The days are warm and I'm not planning on building any fires for a while. But it will be nice to have something ready to go when the nights turn colder. Or to use to extract heat from the fireplace. Exact use of my project has yet to be determined; but for now I have a workable circuit that can handle (think I read) 70 plus amps.

 

Some IGBT's have diodes made in them for emf.
here some good reading if you want.
https://www.onsemi.com/pub/Collateral/AND9088-D.PDF

 

@be80be: I'm not highly adept in reading data sheets. I can manage a few facts, but the data sheet linked in post #43 does not suggest this device has a freewheeling diode. Besides, wouldn't the diode sit across the motor leads? The IGBT is just a switch. Perhaps if it does have a diode in it similar to the MOSFET I was originally thinking of using, that will protect the device, but not handle any BEMF. Am I wrong in my assumptions?

And thanks for the link. Not sure where, but someone directed me to that at least once.

[edit] here's a link to another of my posts regarding this issue:
https://forum.allaboutcircuits.com/threads/flyback-diode-needed-or-recommended.160195/

 

I don't no if you have posted a link to the IGBT's your using.
You may want use a MOV across your motor

The MOV (VDR) has a rated voltage. Below this voltage it has a very high resistance. So by selecting a MOV with a rated voltage slightly higher than the normal supply voltage it may be safely connected across the coil of the magnet and just like the diode has no affect while the supply is connected.

Metal Oxide Varistor (MOV)
When the supply is disconnected, the Back EMF will rise to the rated voltage of the MOV. At this point the MOV will start to conduct and clamp the voltage to just above this value.

 

Know next to nothing about MOV's. I'll have to look to see if I have one. I probably do - with all the scrap stuff I have hanging around. The trick will be understanding the numbering on them.

Thanks.

 

Know next to nothing about MOV's. I'll have to look to see if I have one. I probably do - with all the scrap stuff I have hanging around. The trick will be understanding the numbering on them.

Thanks.

The Marvelous MOV. Metal Oxide Varistor. Their most common use in household common devices is likely surge protectors used for protecting electronic equipment. The only problem I see is I "think" the more times a MOV reaches breakdown the less reliable they become. You may wish to think about Snubber Considerations for IGBT Applications. I just see using a snubber type design verse a MOV as more viable. Not to say it is but likely the direction I would look.

Ron