PowerVelocity controller review

Get real world experience and user feedback on the electric bicycle products.
User avatar
ElectricGod
100 kW
100 kW
Posts: 1347
Joined: Nov 01, 2015 5:24 pm

Re: PowerVelocity controller review

Post by ElectricGod » Dec 03, 2017 12:11 am

fechter wrote:
Dec 02, 2017 10:54 pm
I'll be interested in the testing. I've used lots of 4110s with good results.
I loved your quote..."One test is worth a thousand opinions"...exactly my point.

I agree...so have I. IRF4110's have served me pretty well. Same for AOT290's. I haven't scrutinized IRF4110's hardly at all. I've only done any level of significant testing on AOT290's . Despite the inconsistent and less than stellar test results I've had, they have been a better mosfet than the IRF4110. I wouldn't dare run IRF4110's in a 12 fet controller at 4kw, but with AOT290's...not a problem.

I just put in an order for 25 of the IXYS IXFP180N10T2-ND mosfets with Mouser. I'll have them in a few days. Give it a little time and I'll do another order of them from DigiKey. They want a minimum order of 50 of them so that's going to set me back $200+. Hopefully ordering from 2 places will be similar to waiting 6 months or a year to make another order. How likely is it that Digikey and Mouser would send me mosfets from the same batch?

The AOT290's I currently have on hand are from 2 batches both purchased from the same place, but almost a year apart. The IRF4110's I have are from 3 or 4 orders purchased over 4 or 5 years time. My samples for both of these mosfets are fairly diverse.

Anyway, I've put my money where my mouth is...
I've purchased another component tester. I've purchased more mosfets. I'm going to test just like the spec sheet says to test and compare the results with the $30 tester. Proof is in the testing. That's lots more than anyone else is doing.

User avatar
madin88
10 MW
10 MW
Posts: 2328
Joined: May 27, 2013 2:02 am
Location: Austria

Re: PowerVelocity controller review

Post by madin88 » Dec 03, 2017 5:58 am

Powervelocity.com wrote:
Dec 02, 2017 4:57 pm
I never mentioned the risk of FETs failing due to inconsistent supply. It's rather consistent and comes form a single supplier.
I did mention that there may be just bad mosfets in batches and that's why I warrant controllers against defects.
My english is not perfect, but that there may be just bad mosfets in batches is exactly what i meant with "inconsistent supply".
I have repaired a few (not really many) controllers free of charge even though in some cases customer admitted it was their fault.
Nice to read, however i have asked you if you could send me a few AOT290 so i can repair the controller by myself, but no response from you.

Now the FETs are on the way from digikey together with some CSD19506KCS.
Will test the Rds(on) with my "ebay milliohm meter" with 10V on the gate and compare the results.

thats the meter i am using:

https://www.eevblog.com/forum/projects/ ... tach=66971
https://www.eevblog.com/forum/projects/ ... meter-kit/
http://bbs.38hot.net/thread-40484-1-1.html

it measures the shunt value written on my 12F controller with an accuracy of a few 0,001mOhm so i think it is accurate enough.

User avatar
ElectricGod
100 kW
100 kW
Posts: 1347
Joined: Nov 01, 2015 5:24 pm

Re: PowerVelocity controller review

Post by ElectricGod » Dec 03, 2017 7:38 am

I decided to buy a milli ohm meter. More confirmation is better than less.

This is the meter I bought. Hopefully it will work well enough. I have a bunch of precision .005 ohm shunts. If it measures them correctly, then it ought to do fine for Rds on mosfets. Yes...it's a 2 wire, not a 4 wire milli ohm meter. If it works, it works, if not, I can always send it back and get a 4 wire unit.

https://www.ebay.com/itm/MESR-100-ESR-L ... 2749.l2649

User avatar
Alan B
100 GW
100 GW
Posts: 7024
Joined: Sep 11, 2010 7:43 am
Location: San Francisco Bay Area, USA
Contact:

Re: PowerVelocity controller review

Post by Alan B » Dec 03, 2017 9:12 am

That's an ESR meter. Designed to measure series resistance of capacitors. Seems to have a resolution of 0.001. Not very adequate resolution for measuring a value of 0.0032 vs 0.0035.

You have no idea what controller companies are doing. You assume they are doing nothing, just slamming parts out of tubes into PC boards. The better manufacturers will do what it takes to have a reliable product, and if that means 100% testing of the FETs before they go into the PCB, they will have an automated test setup that will do exactly that. If the parts are as bad as you claim, they would definitely be doing that. They could choose to test the boards after assembly instead, but if the failure rate was very high the scrap or rework rate would be too much if the parts consistency was that poor.

They will also use good ESD practices, and if they find that the FETs are good enough right out of the tubes they will not need to test each one. If the failure rates are low enough of the final product, why test for something that isn't a problem?

On one project we made elaborate test setups to measure many thousands of FETs for extremely low noise characteristics, something that was really beyond their specs, to find the lower noise units required for physics experiments with tens of thousands of channels. You determine what is important to the project, and you do that. Not every part needs that scrutiny, just the critical ones.

That's why companies employ engineers and even scientists in many cases to determine what needs to be done to make a reliable product.

User avatar
madin88
10 MW
10 MW
Posts: 2328
Joined: May 27, 2013 2:02 am
Location: Austria

Re: PowerVelocity controller review

Post by madin88 » Dec 03, 2017 11:21 am

Alan B wrote:
Dec 03, 2017 9:12 am
That's an ESR meter. Designed to measure series resistance of capacitors. Seems to have a resolution of 0.001. Not very adequate resolution for measuring a value of 0.0032 vs 0.0035.
I told him already that this meter is the wrong device. It should have the 4-wire method and much more resolution. 1mOhm steps is bullshit.

User avatar
ElectricGod
100 kW
100 kW
Posts: 1347
Joined: Nov 01, 2015 5:24 pm

Re: PowerVelocity controller review

Post by ElectricGod » Dec 04, 2017 3:36 pm

Alan B wrote:
Dec 03, 2017 9:12 am
That's an ESR meter. Designed to measure series resistance of capacitors. Seems to have a resolution of 0.001. Not very adequate resolution for measuring a value of 0.0032 vs 0.0035.

You have no idea what controller companies are doing. You assume they are doing nothing, just slamming parts out of tubes into PC boards. The better manufacturers will do what it takes to have a reliable product, and if that means 100% testing of the FETs before they go into the PCB, they will have an automated test setup that will do exactly that. If the parts are as bad as you claim, they would definitely be doing that. They could choose to test the boards after assembly instead, but if the failure rate was very high the scrap or rework rate would be too much if the parts consistency was that poor.

They will also use good ESD practices, and if they find that the FETs are good enough right out of the tubes they will not need to test each one. If the failure rates are low enough of the final product, why test for something that isn't a problem?

On one project we made elaborate test setups to measure many thousands of FETs for extremely low noise characteristics, something that was really beyond their specs, to find the lower noise units required for physics experiments with tens of thousands of channels. You determine what is important to the project, and you do that. Not every part needs that scrutiny, just the critical ones.

That's why companies employ engineers and even scientists in many cases to determine what needs to be done to make a reliable product.
THat ESR meter does measure low resistances. I get that it was designed for measuring internal resistance of caps. I did read the description. So why does that matter specifically? If I put a shunt across it's 2 leads, will it measure the resistance of the shunt? My suspicion is that it will. So then, it ought to measure the junction resistance of a mosfet as well. Just because something was designed for a specific purpose doesn't mean using it for some other purpose is bad. I recognize it has limited resolution. I'm not expecting much out of a $40 tool, but if it's adequate for my purposes...so be it! Worst case...as I've already said...I send it back...no big deal. AND as I already said, I am fully aware that it isn't a 4 wire tester. Where's the problem here? I recognize what I'm doing and have readily admitted as much. There's really nothing more to say here regarding this meter. Besides, it's a $40 meter. If I toss it in the trash, that's my $40 to throw away...it could happen, but I doubt it.

Am I ignorant and stupid or am I making deliberate and knowing choices? I'm an experimenter and I enjoy modding things and I like learning...so I have to say that I'm making deliberate and knowing choices and not out of stupidity.

I'm reading everything that people write. I do take it to heart! I don't need to be right or wrong. If I'm wrong I'll just say I am. If I find out that the ESR meter I bought is crap and flat out won't work...I'll post that it won't work. Same for the Chinese component tester I bought. BUT!!! IF they do work, well I'm going to post that too!!! Your opinions and criticism about my choices have been noted, but I'm still willing to find out if other options will work.

Questions about static electricity and grounding has come up several times. I don't have a meter or ever bothered trying to detect static charge beyond seeing the tiny arc come off my finger. I do use a grounded Hakko soldering station and work at a grounded desk and generally use grounded test equipment where it is applicable. I discharge myself to my desk pretty often, but I don't wear a ground strap since I'm AT my desk and touching it's bare metal frequently. I do have a ground strap, but haven't used it in a very long time since grounding to my desk has been adequate as far as I can tell. There's not much opportunity for me to build up any static charge in this environment. Sitting at my desk and touching the desk pretty much won't allow any static build-up. If I get up from my desk and walk around my house and then return to my desk, it happens that there will be a spark, when I touch my grounded desk. Static DOES happen and I am aware that it is there. If I roll my chair a few feet, static build up can be significant enough that I will get a spark when I touch my desk. There's not a question of it's existence...hence the grounded desk.

I'm going to be UBER careful with the IXYS mosfets, touch the grounded desk a lot, discharge the ESD bag or tube to the desk before opening it and removing the mosfets, etc. I won't even pull them from the tube or ESD bag until I have the new component tester in my hands and have tried it out on several other mosfets. Just to be sure it really works I'll also compare it's results to my other 2 component testers. If everything matches up for the same mosfet on all 3 component testers...great! I'll also test those same mosfets via this test method...just like I've already said I would do! The results will be the results! THEN, I will pull out a few IXYS mosfets and do the same tests. The test results will be the test results! I will also try out that ESR meter and see if it can do what I hope it can do. If not, oh well. Again...the results will be the results...whatever they are.

Image

I think I'll abuse some mosfets and see what happens when they are exposed to static. Yes I know static is bad for electronics and specifically more so for CMOS components. That's the point! I will take some mosfets (I have lots of them that don't meet manufacturer specs) and test 10 of them. I have AOT290's, IRF4110's and a few others that are all new. All I care about is finding out what they actually test to. I can record that information before the test begins. Each mosfet can be uniquely labeled to eliminate confusion and to keep them straight for later testing. Then I'll take 5 of those mosfets and put them on a piece of wire. I can walk around my house with them in my pocket, wash the dishes, do laundry with them strapped to my wrist and so on. Just expose them to the natural household environment. I'll take another 5 and tie them together on a wire and toss them in with the cloths so that they get dried and thoroughly abused by static in there! Finally, I'll test all 10 mosfets again. Did they suffer harm? Do some of them no longer work at all? Do they now test differently than before? My expectation is that they will suffer some damage, but the proof is in the test results.

I just got an email from Mouser. My IXFP180N10T2 mosfets have been shipped. I also bought $150 of XHP70.2 and XP-L2 LED's from digiKey. The XP-L2's will go in a flashlight that currently has 18 Chinese XML's in it. They are both the same sized die so this will work well. I'll probably need to upgrade the driver board too. The current 18 XML's in it are dimmer than a single legit XHP70. Everything is running at manufacturer spec'd voltage and current. 18 XP-L2's should make it far brighter than the XHP70. The XHP70.2 LED's will go in some Chinese aluminum bike lights for my EV's. A friend suggested using an RC BEC as an LED driver. I'll have to give that a try.


You guys are going to have to try harder! If I bought the cheapest 4 wire tester, that would get criticized. If I bought a mid-range 4 wire tester, that would get criticized. If I bought the best unit on the planet and then tested exactly like "industry" supposedly does, still something would get criticized. YAWN! Do real testing! How about posting real test results because you spent the time and money to do it? I'd really love it if someone else would do actual testing. Feel free to use any method you want. I'd LOVE to see refuting or corroborating results based on actual testing. Have I said essentially "do real testing" enough yet?

Madin...you have a 4 wire resistance tester. PLEASE post your results for the AOT290's you just bought. Test other mosfets please. Put your AOT's in your component tester and post the results you get. Let's compare what you get with what I get.

Alan...you've tested mosfets before to find the least noisy of the batch. Test some more and look to see if they meet the manufacturer specs. Post your results! I don't know what you have on hand for parts, get some AOT290's and IRF4110's and XFP180N10T2 mosfets. Back your opinions about manufacturer practices with your own test results.

If Addy is still reading this thread...test mosfets and post your results.

Do actual testing guys!!! Quit looking for fault in what I'm doing. Actually prove that you get different or similar results by TESTING!!! Lets work together rather than against each other. The results will be the results. I'm going to post my results regardless of what anyone says. Put your money and time where your mouth is. That's all I'm doing! This whole thing started with me stating that I was looking for another mosfet option because my tests on AOT290's were showing that they are inconsistent and diverged from manufacturer specs by a lot.

No one except me to date has even expressed that they are taking up this challenge...not ever....not even to the point of buying a $30 component tester! However, since this whole discussion started, I've not only bought more mosfets, but I've bought test tools too!

What have you done besides criticize and tried to find fault?

User avatar
Alan B
100 GW
100 GW
Posts: 7024
Joined: Sep 11, 2010 7:43 am
Location: San Francisco Bay Area, USA
Contact:

Re: PowerVelocity controller review

Post by Alan B » Dec 04, 2017 9:37 pm

Damaging the gates of these mosfets only takes a small, low voltage static charge. They are only rated to withstand 20 volts and with gigohms of resistance it takes very few electrons to pop their thin little oxide layer. That is a very, very small static charge. They are quite fragile. You might consider taking full ESD precautions as specified - antistatic mats, wrist straps, grounding, etc for these test results to have any validity. Any worker not following those procedures on the assembly line would lose their job.

I will do some testing when I need to. At the moment I don't have that need or test setup. I do have some test equipment for things I need, like a true DC RMS meter (most meters marked RMS are AC RMS only and fail to get the right answer on PWM), (and that meter will measure down to the microvolt region which could be used for Rds) and I have a voltage calibration standard so I can actually check my meters for accuracy which most people never bother to do. I also have antistatic mats and wristbands and surface mount rework equipment and stereo microscope and use them when needed. When I get ready to build the power stage for my DIY controller I'll set something up. Right now I've got a number of 3D printer projects to get done, and product to get out the door, unrelated to ES, so the que is pretty full. Plus it is time to go to tennis practice and get some fun and exercise.

It will be interesting to see what results you get. Unfortunately if they indicate problems with the devices there is going to be a strong tendency to suspect the test gear or handling procedures. But I think it is great that you are taking the initiative to work on this.

User avatar
madin88
10 MW
10 MW
Posts: 2328
Joined: May 27, 2013 2:02 am
Location: Austria

Re: PowerVelocity controller review

Post by madin88 » Dec 05, 2017 12:12 pm

ElectricGod wrote:
Dec 04, 2017 3:36 pm
THat ESR meter does measure low resistances. I get that it was designed for measuring internal resistance of caps. I did read the description. So why does that matter specifically? If I put a shunt across it's 2 leads, will it measure the resistance of the shunt? My suspicion is that it will. So then, it ought to measure the junction resistance of a mosfet as well. Just because something was designed for a specific purpose doesn't mean using it for some other purpose is bad. I recognize it has limited resolution. I'm not expecting much out of a $40 tool, but if it's adequate for my purposes...so be it! Worst case...as I've already said...I send it back...no big deal. AND as I already said, I am fully aware that it isn't a 4 wire tester. Where's the problem here? I recognize what I'm doing and have readily admitted as much. There's really nothing more to say here regarding this meter. Besides, it's a $40 meter. If I toss it in the trash, that's my $40 to throw away...it could happen, but I doubt it.
Seriously, or a try to whitewash that buy?
This meter is not able to measure mOhm accurately so it will not help you anything.
The specs and lack of 4wire measurement does tell me this. Thats how it is, but try it out :P

User avatar
ElectricGod
100 kW
100 kW
Posts: 1347
Joined: Nov 01, 2015 5:24 pm

Re: PowerVelocity controller review

Post by ElectricGod » Dec 05, 2017 1:39 pm

Put peoples feet to the fire and all of a sudden they disappear or have reasons why they can't be involved.

"What?! You can't expect me to do more than be critical of what you are doing? I can't be bothered with actually doing anything constructive!"

Truly weak and pathetic!

User avatar
Alan B
100 GW
100 GW
Posts: 7024
Joined: Sep 11, 2010 7:43 am
Location: San Francisco Bay Area, USA
Contact:

Re: PowerVelocity controller review

Post by Alan B » Dec 05, 2017 1:50 pm

This is your project. Why do you suddenly assume it is everyone's?

People are trying to help you, but it doesn't mean they're going to do your project.

User avatar
ElectricGod
100 kW
100 kW
Posts: 1347
Joined: Nov 01, 2015 5:24 pm

Re: PowerVelocity controller review

Post by ElectricGod » Dec 05, 2017 1:55 pm

madin88 wrote:
Dec 05, 2017 12:12 pm
ElectricGod wrote:
Dec 04, 2017 3:36 pm
THat ESR meter does measure low resistances. I get that it was designed for measuring internal resistance of caps. I did read the description. So why does that matter specifically? If I put a shunt across it's 2 leads, will it measure the resistance of the shunt? My suspicion is that it will. So then, it ought to measure the junction resistance of a mosfet as well. Just because something was designed for a specific purpose doesn't mean using it for some other purpose is bad. I recognize it has limited resolution. I'm not expecting much out of a $40 tool, but if it's adequate for my purposes...so be it! Worst case...as I've already said...I send it back...no big deal. AND as I already said, I am fully aware that it isn't a 4 wire tester. Where's the problem here? I recognize what I'm doing and have readily admitted as much. There's really nothing more to say here regarding this meter. Besides, it's a $40 meter. If I toss it in the trash, that's my $40 to throw away...it could happen, but I doubt it.
Seriously, or a try to whitewash that buy?
This meter is not able to measure mOhm accurately so it will not help you anything.
The specs and lack of 4wire measurement does tell me this. Thats how it is, but try it out :P
Go back to my first post on this meter. Reread it. I've never changed my tune on what I thought of it or my expectations.

So...are you going to actually test mosfets and post your results or is that too much to ask? Being critical and finding fault is easy. Actually doing work requires effort and commitment!

User avatar
ElectricGod
100 kW
100 kW
Posts: 1347
Joined: Nov 01, 2015 5:24 pm

Re: PowerVelocity controller review

Post by ElectricGod » Dec 05, 2017 2:08 pm

Alan B wrote:
Dec 05, 2017 1:50 pm
This is your project. Why do you suddenly assume it is everyone's?

People are trying to help you, but it doesn't mean they're going to do your project.
To be honest, I don't expect or assume anything other than bitching about what I'm doing. But if putting peoples feet to the fire shuts up all the critical BS, well fine! SO be it! Frankly I am tired of all the criticism and fault finding. No one wants to do the expensive thing and actually be constructive by testing and showing real test results. What you call "help" isn't helping. Help by actually doing tests on mosfets! PROVE that your opinion is accurate! I've at least tested mosfets and that is why I said what I said...that not all mosfets from the same batch are created equal and often times some are way out of spec. THOSE ARE FACTS FROM MY TESTING.

AND, I'm going to do the expensive thing and actually test MORE. Show me one person anywhere in this thread that can say that besides me. You can't! BUT you can look back through the many posts and see all the criticism and fault finding that so many people have been more than willing to post.

My project? Really? More back peddling and no commitment! LAME! The ONLY reason why I'm doing this is because of all the crap I got for stating the facts of my previous tests!

User avatar
Alan B
100 GW
100 GW
Posts: 7024
Joined: Sep 11, 2010 7:43 am
Location: San Francisco Bay Area, USA
Contact:

Re: PowerVelocity controller review

Post by Alan B » Dec 05, 2017 3:35 pm

Help can be for you, or it can be for others who read the thread.

You choose how to take suggestions.

Test results are affected by test procedures and equipment. A flawed test shows what?

A test result is a test result. Not necessarily a fact.

Addy
1 W
1 W
Posts: 64
Joined: Aug 11, 2017 11:15 am

Re: PowerVelocity controller review

Post by Addy » Dec 05, 2017 4:22 pm

ElectricGod wrote:
Dec 04, 2017 3:36 pm
If Addy is still reading this thread...test mosfets and post your results.

I'm not convinced that I need to, not that it matters because I don't have a decent quantity of mosfets on hand to test. When it comes time to put together the controller I'm laying out right now, I may do some testing. I just found this document though:

Nexperia (NXP) AN11599 - Using power MOSFETs in parallel

I haven't finished reading it but there are a couple interesting points already:

"There is a range of values for the parameters of the MOSFET dies on a wafer. There is a wider range for MOSFET dies on different wafers in the same production batch. The range is even wider for all batches even though the MOSFETs are the same type. MOSFETs with parameter values outside the data sheet limits are rejected."

also

"It is technically and commercially undesirable to have to select MOSFETs and it should be
unnecessary.
"

User avatar
ElectricGod
100 kW
100 kW
Posts: 1347
Joined: Nov 01, 2015 5:24 pm

Re: PowerVelocity controller review

Post by ElectricGod » Dec 05, 2017 6:51 pm

Addy wrote:
Dec 05, 2017 4:22 pm
ElectricGod wrote:
Dec 04, 2017 3:36 pm
If Addy is still reading this thread...test mosfets and post your results.

I'm not convinced that I need to, not that it matters because I don't have a decent quantity of mosfets on hand to test. When it comes time to put together the controller I'm laying out right now, I may do some testing. I just found this document though:

Nexperia (NXP) AN11599 - Using power MOSFETs in parallel

I haven't finished reading it but there are a couple interesting points already:

"There is a range of values for the parameters of the MOSFET dies on a wafer. There is a wider range for MOSFET dies on different wafers in the same production batch. The range is even wider for all batches even though the MOSFETs are the same type. MOSFETs with parameter values outside the data sheet limits are rejected."

also

"It is technically and commercially undesirable to have to select MOSFETs and it should be
unnecessary.
"
Thanks Addy for posting that article. Unfortunately we all know that no component manufacture does 100% testing. They test more like 1 in 10,000 or 1/100,000 or whatever is the minimum they can get away with. Everyone wants to control costs and testing costs money!

This statement is flat out erroneous for any component that skates through untested. "MOSFETs with parameter values outside the data sheet limits are rejected." IF they test 1:10,000, then 9,999 components are untested and therefore unknown quantities. I'm sure, if a mosfet is in fact tested, that if it is outside of the specified parameters, that the one mosfet that failed is discarded. It might also be true that if a reject is found, that spot checking gets increased to something like 1/1000 for the rest of the mosfets on that one wafer. This still leaves lots of room for mosfets that don't meet spec to slip through. The manufacturer is playing the odds...and that's all they are doing!

The second statement is wishful thinking and complete nonsense under specific circumstances. Can you "get away with" just putting components on boards and then testing the final product? Absolutely, and it happens all the time. Every board manufacturer does it and most of the time it usually works out OK. Can I buy 50 mosfets and just put them on a board and will they probably work? Yup...why not...manufacturers do exactly that all the time. Will they be matched and close to typical specs? Who knows since they were never tested in that manner. BUT!!! If you are looking for matched components...such as 12 AOT290's that are close to typical values, then this statement is fiction. If it happens by chance, well you got lucky!

Anyone care to actually refute me with facts by actually testing components? All I hear so far is excuses. I don't have... "Fill in your excuse here". Everyone want's to scrutinize my testing...whatever that may be and poke holes in it, but no one wants to put their name on the line. I believe I hear clucking sounds! LOL! How often do you guys need to scrub your backs to keep that yellow streak looking more skin colored?

Buy parts, set up your test gear...do actual testing and stop looking for excuses or how I'm doing something wrong to weasel out of it! You're just deflecting folks...I see right through it. Do the expensive thing...the thing that actually matters...the thing that is actual evidence...TEST components! Yeah...your excuses are lame and I'm holding your feet to the fire!

Addy
1 W
1 W
Posts: 64
Joined: Aug 11, 2017 11:15 am

Re: PowerVelocity controller review

Post by Addy » Dec 05, 2017 7:04 pm

Thanks for the laugh ElectricGod. You think you know better than a major semiconductor manufacturer? :lol:

That application note is more factual than your tests. But hey, at least you get to brag about actually doing tests!

User avatar
ElectricGod
100 kW
100 kW
Posts: 1347
Joined: Nov 01, 2015 5:24 pm

Re: PowerVelocity controller review

Post by ElectricGod » Dec 05, 2017 8:03 pm

Addy wrote:
Dec 05, 2017 7:04 pm
Thanks for the laugh ElectricGod. You think you know better than a major semiconductor manufacturer? :lol:

That application note is more factual than your tests. But hey, at least you get to brag about actually doing tests!
You are welcome! However, you know I'm NOT wrong. Manufacturers test 1:10,000 or less! This is common practice. Ever work for a semiconductor company? I have...Microchips...in Arizona. I saw the entire manufacturing line from ingots to IC's. I know what their testing policy was in the 90's. I haven't worked for a a semiconductor company since, but the practice hasn't changed. So yeah...the vast majority of components walk out the door flat out untested and unknown, but probably good because the manufacturing process is consistent enough to minimize defective parts.

Still looking for ways to weazle out of testing I see and still deflecting by disparaging me...YAWN! How boring!

So then...care to put your money where your mouth is and actually test? No? Well that comes as no surprise...opinions are cheap...actual work isn't.

Addy
1 W
1 W
Posts: 64
Joined: Aug 11, 2017 11:15 am

Re: PowerVelocity controller review

Post by Addy » Dec 05, 2017 9:35 pm

ElectricGod wrote:
Dec 05, 2017 8:03 pm
However, you know I'm NOT wrong. Manufacturers test 1:10,000 or less! This is common practice. Ever work for a semiconductor company? I have...Microchips...in Arizona.
If you really did work for a semiconductor manufacturer then you should know why they test less than 1:10,000 parts. Here's a reliability report from Alpha & Omega themselves: http://www.aosmd.com/media/general-reli ... report.pdf

From table 2, under "Summary of AOS Device / Process Qualification Monitor Results", for N channel mosfets, out of a sample size of 198,352 the number of failures was 0. They give a FIT (Failures In Time) rate of 0.06, which means that they expect 1 failure for every 16.67 billion device-hours of operation.

Maybe you're right though. You should phone them and tell them that they should buy chinese ebay test equipment and step up their testing.

User avatar
ElectricGod
100 kW
100 kW
Posts: 1347
Joined: Nov 01, 2015 5:24 pm

Re: PowerVelocity controller review

Post by ElectricGod » Dec 05, 2017 10:06 pm

Addy wrote:
Dec 05, 2017 9:35 pm
ElectricGod wrote:
Dec 05, 2017 8:03 pm
However, you know I'm NOT wrong. Manufacturers test 1:10,000 or less! This is common practice. Ever work for a semiconductor company? I have...Microchips...in Arizona.
If you really did work for a semiconductor manufacturer then you should know why they test less than 1:10,000 parts. Here's a reliability report from Alpha & Omega themselves: http://www.aosmd.com/media/general-reli ... report.pdf

From table 2, under "Summary of AOS Device / Process Qualification Monitor Results", for N channel mosfets, out of a sample size of 198,352 the number of failures was 0. They give a FIT (Failures In Time) rate of 0.06, which means that they expect 1 failure for every 16.67 billion device-hours of operation.

Maybe you're right though. You should phone them and tell them that they should buy chinese ebay test equipment and step up their testing.
I tell you all this one thing...
I don't need your acceptance nor am I seeking it. Say whatever you want. All I see is smug and arrogant people that are cock sure they are right. I have no such delusions. I expect to be wrong in some way. I look forward to failing in some way and needing to do further work. I frankly WANT to be wrong...somehow. A humble person learns from his mistakes and is willing to expose them as mistakes and as a learning experience. I have presented this theme over and over and it continually falls on deaf ears. I have no ego to bruise or need for prideful rightness. I'm perfectly comfortable in my own skin knowing I did due diligence and that I followed through with what I said. My results will be what they are. I will do the testing even though all of you are smug and arrogant and will refuse to acknowledge my results regardless of what they are or how I might test. This is what I expect...pettiness, critical and fault finding responses, closed minded and total inability to accept anything but your own cock sure, "can't possibly be wrong" attitudes. I choose to question things. I choose to be willing to be wrong and to be OK with that and to learn from it. I choose to invest myself in NOT being arrogant or smug and instead to try different things and to NOT be closed minded or cock sure that I am right or that I know a thing. I used to be like you guys...cock sure, arrogant, dead sure of my rightness and smug. I find it disgusting and revolting now. I don't EVER want to be that person again. Looking back at myself, I'm ashamed of who I used to be.

On a slightly different note...
I'm also sure that there are quite a few people that know more than me...lots more than me. You guys might be some of those people. Who am I to say that you are or aren't.

Cheers!

User avatar
madin88
10 MW
10 MW
Posts: 2328
Joined: May 27, 2013 2:02 am
Location: Austria

Re: PowerVelocity controller review

Post by madin88 » Dec 06, 2017 12:18 pm

ElectricGod wrote:
Dec 05, 2017 1:55 pm
Go back to my first post on this meter. Reread it. I've never changed my tune on what I thought of it or my expectations.
there stands "more confirmation" and "hopefully it will work" :roll:
Don't get me wrong, i was going to prevent you from buying the wrong device for measuring Rds On of Mosfets. Nothing else.

Do you know now whats the problem if you try to measure very low resistance with 2 wires?

User avatar
ElectricGod
100 kW
100 kW
Posts: 1347
Joined: Nov 01, 2015 5:24 pm

Re: PowerVelocity controller review

Post by ElectricGod » Dec 06, 2017 3:47 pm

madin88 wrote:
Dec 06, 2017 12:18 pm
ElectricGod wrote:
Dec 05, 2017 1:55 pm
Go back to my first post on this meter. Reread it. I've never changed my tune on what I thought of it or my expectations.
there stands "more confirmation" and "hopefully it will work" :roll:
Don't get me wrong, i was going to prevent you from buying the wrong device for measuring Rds On of Mosfets. Nothing else.

Do you know now whats the problem if you try to measure very low resistance with 2 wires?
My fourth post about the meter: (And hopefully the last repeating what I already said)

I knew what I was buying...I've said that already. My thought was simple...let's see if it will work. I said that already too.

My first post about this 2 wire meter:

"This is the meter I bought. Hopefully it will work well enough. I have a bunch of precision .005 ohm shunts. If it measures them correctly, then it ought to do fine for Rds on mosfets. Yes...it's a 2 wire, not a 4 wire milli ohm meter. If it works, it works, if not, I can always send it back and get a 4 wire unit"

My second post about the meter:

"THat ESR meter does measure low resistances. I get that it was designed for measuring internal resistance of caps. I did read the description. So why does that matter specifically? If I put a shunt across it's 2 leads, will it measure the resistance of the shunt? My suspicion is that it will. So then, it ought to measure the junction resistance of a mosfet as well. Just because something was designed for a specific purpose doesn't mean using it for some other purpose is bad. I recognize it has limited resolution. I'm not expecting much out of a $40 tool, but if it's adequate for my purposes...so be it! Worst case...as I've already said...I send it back...no big deal. AND as I already said, I am fully aware that it isn't a 4 wire tester. Where's the problem here? I recognize what I'm doing and have readily admitted as much. There's really nothing more to say here regarding this meter. Besides, it's a $40 meter. If I toss it in the trash, that's my $40 to throw away...it could happen, but I doubt it."

My third post about the stupid meter:

"Go back to my first post on this meter. Reread it. I've never changed my tune on what I thought of it or my expectations. "


Why is any of this difficult to understand?

User avatar
ElectricGod
100 kW
100 kW
Posts: 1347
Joined: Nov 01, 2015 5:24 pm

Re: PowerVelocity controller review

Post by ElectricGod » Dec 08, 2017 4:16 pm

The new component tester arrived last night. I wasn't expecting much from it beyond what the other component testers do. And honestly, I bought it just because it was described as showing Rds. I can confirm that is does display Rds values for mosfets. Whether they are accurate measurements or not will take further testing via other means...such as voltage drop across the mosfet or use of a milliohm meter. I pulled out several mosfets and tested them on all 3 of my testers and the measured results were all very close. For example, the oldest tester would say that a mosfet turned on at 3.7 volts, and the previous tester would say 3.8 volts and the new one would would measure 3.75 volts. THat's pretty close to the same. I tested a bunch of BJT's as well and the results that each tester displayed were at most 2-3% different. These testers are from 2015, 2016 and 2017 and completely different designs. I'd say that this level of consistency on 3 less than $30 component testers is pretty good! Last night, I was a testing fool. LOL! I have transistors that I have been scrounging for many years. I've never bothered to test all of them, just the one that will do what I need at the moment. I am going to guess that I tested 500 transistors last night. The new component tester has a rechargeable LIPO cell in it. I completely ran it down twice. While it was charging, I'd switch over to my next oldest component tester.

I wanted to weed out the bad parts and since they were all supposed to be BJT's, find anything that wasn't. I found 20 or so mosfets in my BJT's, several triacs, several diodes, voltage regulators and other things mixed in with them. I also found a few dead transistors. Some components these meters can't read correctly. For example, linear voltage regulators are too complex internally so the tester displays "unknown part". They also commonly display this for a component that is totally blown. As a result, if I got that message, I would find out what it was and then possibly toss it in the trash. I ran into a couple mosfets that detected as a typical mosfet with a resistor source to drain. Looking them up, they were not built that way so in the trash they went. Clearly a partly blown part! I guess the back diodes had partly failed, but not enough to create a dead short.

Inexpensive Chinese component testers are super convenient and easy to use. There's no set up at all. Plug in the part to be tested and hit the test button. In seconds, most 2 or 3 legged components are identified and their specs are displayed. Anyone that isn't super anal about following industry methods for testing parts will find these tools very helpful. For everyone else that despises this sort of tool...feel free to test components in the hardest way possible that takes lots of expensive and complex equipment. For the average person that does not want to get an EE degree, just get one of these component testers...they will do what you want.

This is the ebay auction for the new component tester I just got. The color display is cool, but who cares. It has a 150mah LIPO battery which is nice, but it runs down in about an hour of constant use. It really ought be more like 300-400mah. There's plenty of room inside the shell for a larger battery. I bought it since it does Rds, but how useful that is will take some testing the hard way to confirm. Same for the gate on voltage...which typically these testers show at somewhere around the actual spec sheet values. The new tester displays values very similar to my older testers.

https://www.ebay.com/itm/Transistor-Tes ... 2749.l2649
Last edited by ElectricGod on Dec 08, 2017 7:00 pm, edited 2 times in total.

User avatar
Alan B
100 GW
100 GW
Posts: 7024
Joined: Sep 11, 2010 7:43 am
Location: San Francisco Bay Area, USA
Contact:

Re: PowerVelocity controller review

Post by Alan B » Dec 08, 2017 4:58 pm

ElectricGod wrote:
Dec 08, 2017 4:16 pm
... I pulled out several mosfets and tested them on all 3 of my testers and the measured results were all very close. For example, the oldest tester would say that a mosfet turned on at 3.7 volts, and the previous tester would say 3.8 volts and the new one would would measure 7.75 volts. THat's pretty close to the same. ...
I'm trying to understand how 3.8 and 7.74 are the same? Hopefully a typo. What are the Rds readings you are seeing with it?

User avatar
ElectricGod
100 kW
100 kW
Posts: 1347
Joined: Nov 01, 2015 5:24 pm

Re: PowerVelocity controller review

Post by ElectricGod » Dec 08, 2017 6:56 pm

Alan B wrote:
Dec 08, 2017 4:58 pm
ElectricGod wrote:
Dec 08, 2017 4:16 pm
... I pulled out several mosfets and tested them on all 3 of my testers and the measured results were all very close. For example, the oldest tester would say that a mosfet turned on at 3.7 volts, and the previous tester would say 3.8 volts and the new one would would measure 7.75 volts. THat's pretty close to the same. ...
I'm trying to understand how 3.8 and 7.74 are the same? Hopefully a typo. What are the Rds readings you are seeing with it?
Ha! Good catch! I reread that paragraph 4 times before I hit submit and missed my typo...fixed now.
Since I was rereading, I fixed a couple more typos and added the ebay auction link which I had forgotten.

User avatar
ElectricGod
100 kW
100 kW
Posts: 1347
Joined: Nov 01, 2015 5:24 pm

Re: PowerVelocity controller review

Post by ElectricGod » Dec 11, 2017 12:25 pm

My testing efforts from the other night...
These are some of the BJT's and diodes scrounged from various boards that thanks to these 2 component testers are now known to be good. I have another parts drawer like this brimming full of smaller 3 legged parts of unknown origin or functionality. They will get tested later. Regarding the 2 component testers in the picture. The one in the clear shell has seen use for most of this year and been very helpful and relaible. The one in the white shell is the tester I just bought. It has a color LCD and measures Rds and IR signals, but otherwise does exactly the same thing as the older one. The clear shelled unit has an small function generator in it. I've never used it so I have no idea how good or bad it is, but it's supposed to be able to make a few common wave forms. I forget the frequency range, but something like up to 5mhz.

Image

The IXYS mosfets arrived. They will stay exactly like this until I'm ready to test them. God forbid they get exposed to the air and get fried with errant electrons! I checked my grounding over the weekend and I'm all good and I added additional grounding to the front edge of my desk so that there was even less chance of static discharge. It's just 1/8" thick aluminum angle that's 5 feet long and covers the front edge of the desk, but it's grounded and that's the important thing. It's pretty hard to NOT stay discharged with it right there.

Image
Image

Still waiting on the cheepie resistance meter. That will be another week...not that it's important for testing at the moment. I have everything else already to do so called "real" testing of mosfets.

Post Reply