PDA

View Full Version : How To Test FETs ???


shuter
2007.01.24, 11:58 PM
I have some 3010 fets that I installed on a board, discovered that the board did not work, removed the fets and reinstalled another pair. I am wondering whether the fets are still good or if I shorted them out. Is there a way to test them with a multi meter to see if they are good or not? Would hate to just throw them away but would hate even more installing them on another board only to find they are no good.

CrashTestJohnny
2007.01.29, 01:48 AM
I have some 3010 fets that I installed on a board, discovered that the board did not work, removed the fets and reinstalled another pair. I am wondering whether the fets are still good or if I shorted them out. Is there a way to test them with a multi meter to see if they are good or not? Would hate to just throw them away but would hate even more installing them on another board only to find they are no good.

Here's how the fets work:

8765
||||
HHH
||||
1234

1 N-channel Source
2 N-channel Gate
3 P-channel Source
4 P-channel Gate
5 P-channel Drain
6 P-channel Drain
7 N-channel Drain
8 N-channel Drain

Usually Pin 1 has a dot near it to identify it.

So there are actually two fets within each chip you want to test, one is an
N-Channel fet (the left half of the chip), which is used as an on/off switch for ground (0 Volts), and one is a P-Channel fet (the left half of the chip) used as an on-off switch for the positive voltage (5 Volts)..

On the chip, pins 1,2,7, and 8 are connected to the N channel fet, and pins
3,4,5, and 6 are connected to the P channel fet.

Basically the way the fet works is that, to make your car go forward, the
N-channel is turned on one chip, and the P-channel on the other chip is turned on, so.. the motor has 5V in one lead and ground in the other.


So what you should do is to test each fet seperately to make sure both are working. Make sure none of the pins on the chip are soldered together.
Get a 5-6 v power source (a couple leads from a set of 4 AAA or 4AA in series is fine). Also get your stock motor, (or a really mild motor).. You'll also need a few alligator clips, to connect things together..

Also before you touch your fets, touch a large metal object to ground yourself, and get rid of any static electricity. (Especially if you are in a carpeted room)

First you'll test the N-channel fet..
-Connect pin 1 to the Ground (negative of the power supply)
-Connect one lead of the motor to pins 7 and 8
-Connect the other lead of the motor to +5V (positive of the power supply)
Pin 2 can be now be used to turn the N-Channel fet on and off.
-Connect +5V to pin 2 to turn the fet on, (the motor should be spinning now!)
-Connect Ground to pin 2 to turn the fet off, (the motor should stop spinning as if disconnected from the battery)

As long as you are able to turn the motor on and off using the Pin 2, your N-channel fet is OK

Now you'll test the P-channel fet
-Disconnect everything you connected to the chip before.
-Connect Pin 3 to +5V
-Connect one lead of the motor to pins 5 and 6
-Connect the other lead of the motor to Ground
Pin 4 can now be used to turn the P-Channel fet on and off.
-Connect Ground to pin 4 to turn the fet on, (the motor should be spinning)
-Connect 5V to pin 4 to turn the fet off, (the motor should stop spinning as if disconnected from the battery)

As long as you are able to turn the motor on and off using Pin 4 now, your P-channel fet is OK

So if both the N-channel and P-channel fets work, you are probably all set.

You can also use the same technique to test out a fet stack you solder together before placing it on the board.


Some other suggestions I have is to use a current limiting lab power supply,
instead of a battery pack, and limit the current to 500mA or so if usign a motor.

Also, if you want to be really safe, you can use an 5 Volt LED (light emitting diode) in place of the motor first, and limit the current on the current limiting power supply to 50-100mA (enough so the LED is not full bright).

Actually, after fetting my car I always use a current limiting power supply in place of batteries to test it out first. Also instead of a motor for testing I connect two leds to the motor terminals on the board (one connected forward, and one connected in reverse.) One LED should light up for forward, and one should light up for reverse. When using the LEDs the current can be limited to 100mA or so, which shouldn't hurt the electronics if something is hooked up wrong since 5V at 100ma is only 0.5 watts, and the fets should be able to dissipate at least 1W of heat continously.












Here's some other info too:

http://www.repairfaq.org/sam/semitest.htm#stmos

shuter
2007.01.29, 02:12 PM
CrashTestJohnny - Thanks for the very detailed directions for fet testing. Not an easy to set up test for an electronically challenged person like myself but your instructions seem clear enough. It is an interesting project, I'll give it a try.

henal
2007.02.04, 05:19 AM
hi i have got 2 n chanel mosfet and 2 p chanel mosfet im not sure if there are blown when i used this methode of tesing them out i got like the source pin was shorted with the drain i test it out with a mulitmeter and i got a hige resistanse on one of them i got about 29 k omhs and the other one i got is like 300 k ohms and it dont metter if i use the gat pin it does do notting



now for the n chanel

theres seem to work but ever time i connet positive to the gate pin it turns on and never stops for a long time

and again there are total brand new bouth form maplin electronics in uk

the p chenal mosfet is bsp250
the n chenal mosfet is irfz24n

need all the help i can im not a noob in electronics i know alot about it i thinks there are blown but there are brand new also i know about the ESD so i did touch an copper pipe there time i was going to solder them and also i know how to soler as i was an ps2 chiper so i dint leave the soldering iron for too long

henal
2007.02.13, 08:02 AM
any one some one must know are there blown

Z_408
2007.05.09, 09:32 PM
Here's how the fets work:

8765
||||
HHH
||||
1234

1 N-channel Source
2 N-channel Gate
3 P-channel Source
4 P-channel Gate
5 P-channel Drain
6 P-channel Drain
7 N-channel Drain
8 N-channel Drain

Usually Pin 1 has a dot near it to identify it.

So there are actually two fets within each chip you want to test, one is an
N-Channel fet (the left half of the chip), which is used as an on/off switch for ground (0 Volts), and one is a P-Channel fet (the left half of the chip) used as an on-off switch for the positive voltage (5 Volts)..

On the chip, pins 1,2,7, and 8 are connected to the N channel fet, and pins
3,4,5, and 6 are connected to the P channel fet.

Basically the way the fet works is that, to make your car go forward, the
N-channel is turned on one chip, and the P-channel on the other chip is turned on, so.. the motor has 5V in one lead and ground in the other.


So what you should do is to test each fet seperately to make sure both are working. Make sure none of the pins on the chip are soldered together.
Get a 5-6 v power source (a couple leads from a set of 4 AAA or 4AA in series is fine). Also get your stock motor, (or a really mild motor).. You'll also need a few alligator clips, to connect things together..

Also before you touch your fets, touch a large metal object to ground yourself, and get rid of any static electricity. (Especially if you are in a carpeted room)

First you'll test the N-channel fet..
-Connect pin 1 to the Ground (negative of the power supply)
-Connect one lead of the motor to pins 7 and 8
-Connect the other lead of the motor to +5V (positive of the power supply)
Pin 2 can be now be used to turn the N-Channel fet on and off.
-Connect +5V to pin 2 to turn the fet on, (the motor should be spinning now!)
-Connect Ground to pin 2 to turn the fet off, (the motor should stop spinning as if disconnected from the battery)

As long as you are able to turn the motor on and off using the Pin 2, your N-channel fet is OK

Now you'll test the P-channel fet
-Disconnect everything you connected to the chip before.
-Connect Pin 3 to +5V
-Connect one lead of the motor to pins 5 and 6
-Connect the other lead of the motor to Ground
Pin 4 can now be used to turn the P-Channel fet on and off.
-Connect Ground to pin 4 to turn the fet on, (the motor should be spinning)
-Connect 5V to pin 4 to turn the fet off, (the motor should stop spinning as if disconnected from the battery)

As long as you are able to turn the motor on and off using Pin 4 now, your P-channel fet is OK

So if both the N-channel and P-channel fets work, you are probably all set.

You can also use the same technique to test out a fet stack you solder together before placing it on the board.


Some other suggestions I have is to use a current limiting lab power supply,
instead of a battery pack, and limit the current to 500mA or so if usign a motor.

Also, if you want to be really safe, you can use an 5 Volt LED (light emitting diode) in place of the motor first, and limit the current on the current limiting power supply to 50-100mA (enough so the LED is not full bright).

Actually, after fetting my car I always use a current limiting power supply in place of batteries to test it out first. Also instead of a motor for testing I connect two leds to the motor terminals on the board (one connected forward, and one connected in reverse.) One LED should light up for forward, and one should light up for reverse. When using the LEDs the current can be limited to 100mA or so, which shouldn't hurt the electronics if something is hooked up wrong since 5V at 100ma is only 0.5 watts, and the fets should be able to dissipate at least 1W of heat continously.












Here's some other info too:

http://www.repairfaq.org/sam/semitest.htm#stmos

CrashTest,

When are you going to drop by the store to race?

Z_408

HammerZ
2007.06.08, 01:24 PM
I lost forward on one of my cars. I pulled these off and put on a pair of 4562 fets that I had on hand. Usually you would see some type of damage to the fets in a burned place. These look a little too clean.

BryceVR
2007.12.18, 02:14 AM
Typically the gates of an IC are burned with 40-60v. A short or spike can drop one of the gates without manifesting external symptoms. (They don't all blow the top off) It's only the typical failures (too much current) which will "POP" the FET, other damage can be done with little indication.
shuter. Why go to all the headfork of putting in what are essentially stock FETs. Surely just go with some hot FETs and swap out a stock late model board if you want to go stock racing.
CrashTestJohnny Great synopsis. Probably one of the most succinct I've seen.
Can I use this for a tech article (pm me with credit info).

shuter
2007.12.18, 11:42 PM
BryceVR - Thanks for the information about how fets fail and clairification of their visual condition. I have had a fet stop working and could see no visual irregularities. I have not yet tried Crash Test Johnnys test procedure but may do soon soon. I bought the mini alagator clips needed to make the test connectors but am not very motivated to test as we lost our track location. One of the guys here has set up my track and lap counter in his basement but nobody is interested in racing much. Very small town. His basement is not quite big enough to run HFAY track layouts and there is little interest to continue racing here. Headfork?

I was interested in replacing everyone's 3004 fets with 3010 fets because 3010's had become the new HFAY Stock standard and are a clear a performance increase. When coupled with an MA-010 motor there is a substantial performance improvement over a MR-02 stock motor and 3004 fets. I have cars with higher capacity fets and more powerful motors but do not have the driving skills to make use of them.

New boards are roughly $100.00. Replacing fets is inexpensive.

You may not have noticed but this is an old thread. You have, however, advanced the ball and I appreciate the information.

marc
2008.09.07, 02:35 PM
It would be cool if somebody could invent snap-on FETS for those none-soldering kind of people.

Jshwaa
2015.02.25, 10:15 AM
I have some 3010 fets that I installed on a board, discovered that the board did not work, removed the fets and reinstalled another pair. I am wondering whether the fets are still good or if I shorted them out. Is there a way to test them with a multi meter to see if they are good or not? Would hate to just throw them away but would hate even more installing them on another board only to find they are no good.

*Disclaimer - This method is for experienced users of electrical equipment such as a power supply and knowledge of measurement techniques such as current measurements with a multimeter, do not do this is you have little or no experience in either, or of course, do no have the proper equipment. You could possibly make your FET situation worse or destroy perfectly good FETs.*

A proper test for the installation of FET's or integrity of existing FET's, is to set up a little circuit such as the one below. Of course you can just throw cells in your car and see if the motor runs in forward or reverse, but if not properly tested first, a bad FET installation could unnecessarily destroy the FETs and/or your mini-z elecs, so if done right this procedure can save you that outcome.

http://i83.photobucket.com/albums/j314/Jshwaa/Mini-Z/FET%20Test%20Circuit.jpg (http://s83.photobucket.com/user/Jshwaa/media/Mini-Z/FET%20Test%20Circuit.jpg.html)

1) Set the power supply to 4.8V and 0.1A max current.

2) Power up, and while monitoring current, throttle 'half-way' in forward and reverse. Do not full throttle, as you will draw max current and the voltage will drop and the elecs will shut off. Only throttle as long as it takes you to verify a current reading. If you get a current reading that is near the power supply current setting, then so far so good.

3) Feel FETs for heat. If you feel heat, wait for it to dissipate before continuing. If you feel abnormal heat or the FET's start to blister, then obviously stop as you've reached your answer...the FET's are in failure mode.

3) Increase current by 0.1A.

4) Repeat 2 through 4 until you reach 2A.

If your elecs responded by drawing current through each step, then they should be good. For those of you that would like to push them to the max, then keep repeating the current increments as far as you feel your FETs should handle, while maintaining the 4.8V. Most power supplies can't do what good cells can do in this regard, so don't feel bad if your 'bench' power supply doesn't fit the bill for this procedure, AND DON'T DO THIS WITH CELLS AS THEY ARE UNFORGIVING IN SHORT CIRCUIT SITUATIONS.

imxlr8ed
2015.02.25, 11:17 AM
For less than a dollar a pop on some electronics sites... better just to buy new fets than to bother testing anything or re-using possible faulty ones.

JMO :rolleyes:

Just ordered 25 for around 82 cents each.

LED
2015.02.25, 11:35 AM
*Disclaimer - This method is for experienced users of electrical equipment such as a power supply and knowledge of measurement techniques such as current measurements with a multimeter, do not do this is you have little or no experience in either, or of course, do no have the proper equipment. You could possibly make your FET situation worse or destroy perfectly good FETs.*

A proper test for the installation of FET's or integrity of existing FET's, is to set up a little circuit such as the one below. Of course you can just throw cells in your car and see if the motor runs in forward or reverse, but if not properly tested first, a bad FET installation could unnecessarily destroy the FETs and/or your mini-z elecs, so if done right this procedure can save you that outcome.

http://i83.photobucket.com/albums/j314/Jshwaa/Mini-Z/FET%20Test%20Circuit.jpg (http://s83.photobucket.com/user/Jshwaa/media/Mini-Z/FET%20Test%20Circuit.jpg.html)

1) Set the power supply to 4.8V and 0.1A max current.

2) Power up, and while monitoring current, throttle 'half-way' in forward and reverse. Do not full throttle, as you will draw max current and the voltage will drop and the elecs will shut off. Only throttle as long as it takes you to verify a current reading. If you get a current reading that is near the power supply current setting, then so far so good.

3) Feel FETs for heat. If you feel heat, wait for it to dissipate before continuing. If you feel abnormal heat or the FET's start to blister, then obviously stop as you've reached your answer...the FET's are in failure mode.

3) Increase current by 0.1A.

4) Repeat 2 through 4 until you reach 2A.

If your elecs responded by drawing current through each step, then they should be good. For those of you that would like to push them to the max, then keep repeating the current increments as far as your power supply will take you, while maintaining the 4.8V. Most power supplies can't do what good cells can do in this regard, so don't feel bad if your 'bench' power supply doesn't fit the bill for this procedure, AND DON'T DO THIS WITH CELLS AS THEY ARE UNFORGIVING IN SHORT CIRCUIT SITUATIONS.

I'm sorry but this must be the DUMBEST commentary I have ever read.
First off, nobody except the real electronics geeks have an adjustable power supply.
Secondly, suggesting to make a deliberate shortage is like saying get a lighter to see if you still have gas in your car...

Jshwaa
2015.02.25, 11:44 AM
I'm sorry but this must be the DUMBEST commentary I have ever read.
First off, nobody except the real electronics geeks have an adjustable power supply.
Secondly, suggesting to make a deliberate shortage is like saying get a lighter to see if you still have gas in your car...

Whoa, take it easy. Adjustable power supplies aren't rare by any means, and that is the reason for the adjustability, so you set a max current. No matter what, your elecs will only draw up to that setting. Nothing to worry about. The dumb part would be to not heed the disclaimer, and especially to attempt a FET surgery and test by way of roulette.

So, there are no electronics geeks that are into RC cars? I beg to differ, but I'm not here to argue or trash other people's input either, so...

Put a lighter to your car all you want, but doing so would suggest a lack of sanity on your own part or taking what was posted completely wrong on purpose. If you have a better way of testing FETs in the mini-z application I'm all ears, sir.

imxlr8ed
2015.02.25, 11:58 AM
I like your posts Jshwaa... very in depth and you're pushing things in your own way. But yes, for most of us it is a bit too much to set up a whole test rig for testing fets.

Keep pushing it though! It was nice to see an ocilliscope plot the output signal from a fet... that was a first in my book!

LED
2015.02.25, 01:17 PM
Whoa, take it easy. Adjustable power supplies aren't rare by any means, and that is the reason for the adjustability, so you set a max current. No matter what, your elecs will only draw up to that setting. Nothing to worry about. The dumb part would be to not heed the disclaimer, and especially to attempt a FET surgery and test by way of roulette.

So, there are no electronics geeks that are into RC cars? I beg to differ, but I'm not here to argue or trash other people's input either, so...

Put a lighter to your car all you want, but doing so would suggest a lack of sanity on your own part or taking what was posted completely wrong on purpose. If you have a better way of testing FETs in the mini-z application I'm all ears, sir.

Yes I do, put them in and see if the motor runs :-)
And I also agree with imxlr8ed. You provided some very useful info over the last months, and I also follow each one of your threads. That is why I am really really surprised to read such a comment from you. Because obviously you are one of those electronics geeks (no offense intended, I wish I was one to). But come, you must admit that using a shortage to test something is really far fetched.

Jshwaa
2015.02.25, 04:45 PM
Yes I do, put them in and see if the motor runs :-)
And I also agree with imxlr8ed. You provided some very useful info over the last months, and I also follow each one of your threads. That is why I am really really surprised to read such a comment from you. Because obviously you are one of those electronics geeks (no offense intended, I wish I was one to). But come, you must admit that using a shortage to test something is really far fetched.

I understand your hang-up with this, really. Shorting your motor leads is the last thing you'd want to do with cells, but under a controlled current source, it is just setting up a condition that your FETs are in all the time.

Let's say that you just did a FET job and you accidentally shorted a couple pins together. Throwing your cells in afterwards could be causing a worse condition than shorting your motor wires under a controlled current source.

If it would make you feel better, put a motor at the output, but you should still control the current so as to limit it to a level that won't damage anything in the event of a soldering error. And FYI, a motor only poses an ohm or so resistance to the FETs anyway, so your FETs see a 'near' shorted condition most of the time. 1 or 2 ohms is not much resistance.

I apologize if this is too much information, but this internet thing is pretty big. I'll try not to fill it up. ;) But there's nothing wrong with being skeptical either. I get it.

TheSteve
2015.02.25, 06:08 PM
Instead of a dead short on the output I'd suggest a low value resistor - such as 10 ohms. It could also be used to ensure the current limit on the supply is set correctly before starting the test and is just "good practice" in general. The resistor should be capable of at least 5 watts.

btw, since we're talking about testing FET's I'd not recommend the original procedure mentioned earlier in this thread. The gate of a FET should never be left floating - IE nothing connected to it. When you remove the positive voltage and apply a negative voltage or vice versa you are leaving the gate pin floating. Just touching the gate pin with your finger can cause it to partially turn on which can very easily lead to an instant catastrophic failure. Yes, I am not going into the details of why this occurs but can if needed.

Jshwaa
2015.02.25, 07:03 PM
Instead of a dead short on the output I'd suggest a low value resistor - such as 10 ohms. It could also be used to ensure the current limit on the supply is set correctly before starting the test and is just "good practice" in general. The resistor should be capable of at least 5 watts.

btw, since we're talking about testing FET's I'd not recommend the original procedure mentioned earlier in this thread. The gate of a FET should never be left floating - IE nothing connected to it. When you remove the positive voltage and apply a negative voltage or vice versa you are leaving the gate pin floating. Just touching the gate pin with your finger can cause it to partially turn on which can very easily lead to an instant catastrophic failure. Yes, I am not going into the details of why this occurs but can if needed.

Sure, you could put a 10 ohm resistor at the motor output, but then during test when you throttle, you'll be drawing 4.8/10 = 0.480A or 480mA at the motor end at most. This would be a good cursory test to begin with, but your motor may want a few amps from time to time, so it would be prudent to test in increments to a max value of what you believe your FETs SHOULD handle, which is a data sheet spec you should reference before testing. I'm thinking that a pulse in the forward and reverse direction could be handled by the FETs up to at least 2A. Time is crucial though. You can't keep high amperage on them constantly. Your motor doesn't either, because as your motor accelerates to top speed it draws less and less current, until the only current it draws is that which is required to overcome drag.

You can't up the voltage either, because the voltage is across the elecs, and there is a limit to what the components on the board can handle, so you must lower resistance to allow the elecs to draw up to, but not exceeding, the 'controlled' current setting, see how the elecs respond, and go from there. It is how the industry does it, however they go straight to max current spec and allow the defect to fail, and then recycle it through a repair process. My procedure is trying to avoid the shotgun approach and walk it through in increments so as to safely determine whether or not an error has occurred before going to full run-mode.

I remember back to when I first came to the boards and posted that I was changing my FETs out of the box. There was a reaction as if I was going to destroy my mini-z before I had a chance to drive it. I don't recall anyone deferring me to a safe procedure for changing FETs, and no, throwing your cells in after a FET job isn't safe, and as TheSteve indicated, neither are any of the previous methods introduced.

Here's a link to a good power supply which you can use universally to power whatever for the rest of your life. There's probably at least a metric ton of information to post on how you could use one of these in your mini-z hobby, and maybe, not to mention, a greater understanding of all things electrical.

POWER SUPPLY (http://www.jameco.com/webapp/wcs/stores/servlet/ProductDisplay?app.products.maxperpage=15&dp=category_root%24%2460%24%246040&rfTypRegular+Output+Current+1+%28AMPS%29=NUMBER&closedColumns=&ssr=1&jameco_page=52&catalogId=10001&history=jsl43ou0%7CfreeText~power%2Bsupplies%5Esea rch_type~jamecoall%5EprodPage~15%5Epage~SEARCH%252 BNAV%40anjq6y0g%7Cposition~1%5Eaflag~Y%5Erefine~1% 5EprodPage~15%5Epage~SEARCH%252BNAV%40y3nrwlim%7Cd p~category_root%2524%252460%2524%25246040%5Edc~60% 2524%25246040%2524%2524604010%5Edd~Test%252C%2BToo ls%2B%2526%2BSupplies%2B%252F%2BTest%2B%2526%2BMea surement%2B%252F%2BBenchtop%2BPower%2BSupplies%40k ok22r9n%7Cdp~category_root%2524%252460%2524%252460 40%5Edc~60%2524%25246040%2524%2524604010%5Esort_fi eld~Price%2B%2528Descending%2529%40mn4t7v64%7Cdp~c ategory_root%2524%252460%2524%25246040%5Edc~60%252 4%25246040%2524%2524604010%5Edd~%5Esort_field~Pric e%2B%2528Ascending%2529&storeId=10001&rfNamRegular+Output+Current+1+%28AMPS%29=Regular+O utput+Current+1+%28AMPS%29&isShowImages=0&position=1&langId=-1&productId=207635&search_type=all&rf=1&rfValRegular+Output+Current+1+%28AMPS%29=10+%3C%3D ++%3C+50&MinOrderQtyValue=Select&ShipAvailValue=Select&dc=60%24%246040%24%24604010&ddkey=http:ParametricSearchResultsView)

Drop in the bucket compared to what some of you have probably spent on mini-z's, and can enhance your hobby going further.

I'm sure some of you have messed around with a multimeter....

MULTIMETER (http://www.jameco.com/webapp/wcs/stores/servlet/Product_10001_10001_645618_-1)

Another universally useful device that has a multitude of purposes in aiding you with anything you put batteries in. Buy one. Learn how to use it and what the values are telling you. You may find yourself thinking of edges you can apply to your race scheme. :p You can bet some of the professional RC racers do.

imxlr8ed
2015.02.25, 10:25 PM
[QUOTE=Jshwaa;456259]
I remember back to when I first came to the boards and posted that I was changing my FETs out of the box. There was a reaction as if I was going to destroy my mini-z before I had a chance to drive it. I don't recall anyone deferring me to a safe procedure for changing FETs, and no, throwing your cells in after a FET job isn't safe, and as TheSteve indicated, neither are any of the previous methods introduced. [QUOTE]

Well now... ain't gotta be all attitude about it. There are a bunch of us on here who followed newbies over the years only to watch them get fet crazy and then disappear because they fried their car messing with it. Maybe if you would've put it in your sig that you are an electronics nut, I wouldn't have warned you about all that fryin stuff. ;)

Smart only gets you so far in life.

Jshwaa
2015.03.01, 06:02 PM
Ok, I suppose a 'better' test circuit would be to do this...

http://i83.photobucket.com/albums/j314/Jshwaa/Mini-Z/FET%20Test%20Circuitb.jpg (http://s83.photobucket.com/user/Jshwaa/media/Mini-Z/FET%20Test%20Circuitb.jpg.html)

However, circuit-wise, it is no different than the previously mentioned circuit, as the 'real' difference is where the current meter is placed, and the current meter is a short as far as the circuit is concerned.

mleemor60 was kind enough to donate a set of elecs with apparent range issues. I hooked them up to my supply and meter and found that the stock elecs are pretty stout in terms of current flow per heat dissipated at the FET's. A constant 2A didn't phase them...

So here is the setup...

The current meter is where a motor would normally be, essentially shorting the motor output through the meter....

http://i83.photobucket.com/albums/j314/Jshwaa/Mini-Z/DSCN2311.jpg (http://s83.photobucket.com/user/Jshwaa/media/Mini-Z/DSCN2311.jpg.html)

And then the power supply is connected to where the batteries normally connect.

http://i83.photobucket.com/albums/j314/Jshwaa/Mini-Z/DSCN2304.jpg (http://s83.photobucket.com/user/Jshwaa/media/Mini-Z/DSCN2304.jpg.html)

http://i83.photobucket.com/albums/j314/Jshwaa/Mini-Z/DSCN2306.jpg (http://s83.photobucket.com/user/Jshwaa/media/Mini-Z/DSCN2306.jpg.html)

At first I started with the lowest current setting my supply can be set to, which is about 0.25A. I do this by turning the 'Course' and 'Fine' adjustments of my supply all the way down. I can then press the 'CC Limit' button (middle of supply), and a dummy short is internally placed across the output, and the current output is read on the supply. I then know that a short circuit condition will only put out that much current.

http://i83.photobucket.com/albums/j314/Jshwaa/Mini-Z/DSCN2308.jpg (http://s83.photobucket.com/user/Jshwaa/media/Mini-Z/DSCN2308.jpg.html)

I can then throttle and monitor current, and see if the elecs get hot, or if they are OK. Then increment the current further.

http://i83.photobucket.com/albums/j314/Jshwaa/Mini-Z/DSCN2314.jpg (http://s83.photobucket.com/user/Jshwaa/media/Mini-Z/DSCN2314.jpg.html)

I did this up to 2A, and the elecs were fine, as they should be. I didn't do anything to them yet.

http://i83.photobucket.com/albums/j314/Jshwaa/Mini-Z/DSCN2320.jpg (http://s83.photobucket.com/user/Jshwaa/media/Mini-Z/DSCN2320.jpg.html)

I would do this before putting cells in, because in a short circuit condition the cells will dump current as fast as the chemistry will allow, which as you guys well know is a fair amount of current. Better to be safe than sorry.

arch2b
2015.03.02, 07:49 AM
I must admit, I really like seeing the 'science' behind mini-z. :) if you every want to do a series on this, just let me know as we would love to add this type of information to sticky zones.

Jshwaa
2015.03.04, 11:01 PM
So, I failed in trying to fix the range issue on this set of elecs that mleemor60 gave me. I thought that would be an opportunity to push the stock FET's to failure, just to see...

Using the method previously mentioned, I walked the current magnitude up in increments of 0.5A to start. I tested the stock FET's at a constant 2A, then 2.5A, then 3A, etc. and monitoring temperature vs. time...

http://i83.photobucket.com/albums/j314/Jshwaa/Mini-Z/DSCN2340.jpg (http://s83.photobucket.com/user/Jshwaa/media/Mini-Z/DSCN2340.jpg.html)

The stock elecs impressed me in how much constant current, and high temperature that they can withstand. At temperatures well over 300 degrees F, and currents in the 6A range, the FET's just took it. This tells me that a mini-z user would be challenged to destroy them by simply using their mini-z as prescribed, with any 4-cell power source, and virtually any motor on the market.

My Fluke has a temperature probe and datalogging capability, so I charted up the temperature profile of the FET's as I apply constant current.
Here is the stock elec's temp profile for a constant current of 6A...

http://i83.photobucket.com/albums/j314/Jshwaa/Mini-Z/stock6A.jpg (http://s83.photobucket.com/user/Jshwaa/media/Mini-Z/stock6A.jpg.html)



Here's the profile of the failure at 7A. Notice the peak. That is where they failed.....

http://i83.photobucket.com/albums/j314/Jshwaa/Mini-Z/stock7A.jpg (http://s83.photobucket.com/user/Jshwaa/media/Mini-Z/stock7A.jpg.html)

http://i83.photobucket.com/albums/j314/Jshwaa/Mini-Z/DSCN2378.jpg (http://s83.photobucket.com/user/Jshwaa/media/Mini-Z/DSCN2378.jpg.html)

The traces on the board got so hot from the current flow, that the solder joint holding the red battery wire to the board gave just after taking the above picture. Here's a close-up of the carnage after cleaning off the thermal compound...

http://i83.photobucket.com/albums/j314/Jshwaa/Mini-Z/DSCN2379.jpg (http://s83.photobucket.com/user/Jshwaa/media/Mini-Z/DSCN2379.jpg.html)



So my summary on the stock elecs is that nobody would need to upgrade them for 'more' current, more than to simply lower that resistance, and to get a lower temperature profile for your operating condition.
Their power handling has my thumbs up. If you can effectively desolder your power wires as the FET's fail, that's pretty good...

**Testing does not take into account inductive motor spikes. More testing needed.....

Felix2010
2015.03.05, 01:16 AM
Great thread with lots of cool info. I agree with Arch2b, the science of R/C cars and Mini-Z is a fun topic. Thanks for all the pics and info Jshwaa. :)

The donated Mini-Z board, what stock FET's are used on it? The ASF boards I have are 3010's, double stacked, one layer/two FET's per side. Is this board a Mini-Z SPORTS PCB? I know the SPORTS bords have one side/two FET's only (Not 4 like the ASF boards have). But it looks like these aren't 3010's?

arch2b
2015.03.05, 06:02 AM
Great baseline information to have on record!

Doesn't change user perception likely, especially given popping FETs is not a rare occurrence. In my experience alone, 1/2 if not more of all the cars i've ever owned have popped FETs with stock and never more than 70t motor. Maybe this simply suggests there are larger quality control issues with factory FET installations. i simply replace the FETs right out of the box on every car i own except for those that require stock FETs for class requirements as a result of the frequency in which this has happened to me.

i understand the data shows the electronics are robust enough that this should not happen, experience tells you this does happen often enough that its not an anomaly.

Jshwaa
2015.03.05, 08:37 AM
Great baseline information to have on record!

Doesn't change user perception likely, especially given popping FETs is not a rare occurrence. In my experience alone, 1/2 if not more of all the cars i've ever owned have popped FETs with stock and never more than 70t motor. Maybe this simply suggests there are larger quality control issues with factory FET installations. i simply replace the FETs right out of the box on every car i own except for those that require stock FETs for class requirements as a result of the frequency in which this has happened to me.

i understand the data shows the electronics are robust enough that this should not happen, experience tells you this does happen often enough that its not an anomaly.

Good feedback, arch2b. I would only add that a motor is a little more of a noisy load to the FET's, and I did not take into account the massive inrush of current that braking causes. Your motor essentially is a generator at that point, and could be producing spikes of current/voltage that have damaging effects, which my testing would not have brought to light. I had a soft gentle, quiet, 0.5ohm resistor as a load through most of my testing.

KWT
2015.03.05, 11:40 AM
I was going to bring up the spikes from the motor but you already mentioned it. What usually blows fets is the huge spikes on startup or a stalled motor. You need to test this with a modified motor attached. Also, with no load on the motor, it will not look the same.

TheSteve
2015.03.05, 01:34 PM
It is also possible the N and P channel fets get turned on at the same time due to poor hardware design, firmware bugs, extreme noise from the motor or static. This will generally kill fets instantly. I don't believe it happens that often in Mini-Z's but am sure it has claimed boards over the years.

arch2b
2015.03.05, 01:43 PM
contrary to that theory, most of my FET failures were mid stream driving and not at a start off or pinned at a rail. i've had no less than 2 F1's pop FETs on long straights. it looked awesome but not fun to DNF.

Jshwaa
2015.03.05, 03:21 PM
contrary to that theory, most of my FET failures were mid stream driving and not at a start off or pinned at a rail. i've had no less than 2 F1's pop FETs on long straights. it looked awesome but not fun to DNF.

Do you recall the cell/motor composition on the setups that you blew FET's on, and whether or not the motors had capacitors on them? I believe the caps are what's supposed to help suppress the huge spikes that occur; as the commutation switches poles and you have large current flowing through the motor windings.

Also, built into most FET packages is a diode...

http://i83.photobucket.com/albums/j314/Jshwaa/Mini-Z/NchanPchan.jpg (http://s83.photobucket.com/user/Jshwaa/media/Mini-Z/NchanPchan.jpg.html)

This diode is there to protect the FET from large spikes, and to shunt the large potential to a source, whether it be ground, or the Vcc rail depending on which FET you are referring to (N-channel or P-channel), in the event of an inductive spike seen at its load.

This is one huge advantage to stacking, as you are putting 'more' of these diodes in parallel as well, giving 'more' of a channel to shunt large spikes, and each individual diode is helping protect the sum.

In the case of a FET-failure when running an actual motor, it starts with the breakdown of these diodes. So it would be worth noting that the previous testing did not essentially stress these diodes at all. I was merely running clean current through the FET's drain-to-source channel.

I would give this another whirl with stock FET's AND a loaded motor, but you see what happened to my test specimen. When I 'finally' get my inertial load wheel from my machinist, I will try this again, with say the Si4562's or the 8858's. I got a 10-strip of each. Anyone prefer one over the other?

JesseT
2015.03.06, 01:24 AM
I always have a 1 Ohm 25W resistor i hand together with my mini-z stuff. Together with a multimeter, they are the best tool to investigate and performance test the FETS, battery connectors, wiring, etc.
Put the resistor in place of the motor, and just floor it. Then you can measure the voltage drop between any two places easily. Between battery and its contact, from end to end of a wire, from fet drain to source, and so on. Then you will start to understand where you are losing all the power, and how to build a proper Stock class car.