ADVrider

Go Back   ADVrider > Gear > Vendors
User Name
Password
Register Inmates Photos Site Rules Search Today's Posts Mark Forums Read

Reply
 
Thread Tools Search this Thread Display Modes
Old 09-13-2013, 02:13 PM   #496
Gany
Gnarly Adventurer
 
Joined: Apr 2007
Location: Melbourne, Au
Oddometer: 299
I have limited knowledge so could be making things up, but would rigging up a battery powered stick welder give you the 100amps you are trying to draw or just not controlled enough?
Gany is online now   Reply With Quote
Old 09-13-2013, 02:53 PM   #497
Trl Rdr
Big Red Bird
 
Trl Rdr's Avatar
 
Joined: Aug 2007
Location: Lincoln, NE - The Tax me state
Oddometer: 2,027
Quote:
Originally Posted by Gany View Post
I have limited knowledge so could be making things up, but would rigging up a battery powered stick welder give you the 100amps you are trying to draw or just not controlled enough?
Several 12 volt batteries in parallel would get enough amps if you can find enough load to attach to it. Voltage will drop as time passes.

Sent from my DROID RAZR HD using Tapatalk 2
__________________
2 XR650r(s) 2001 & 2002 Plated
GL18HPM9 Titanium (Gold Wing)
CRF230F for the tight shit!
Baja Mexico Trail Ride Report
Big Bend Ranch Drive and Ride Report
Trl Rdr is offline   Reply With Quote
Old 09-13-2013, 07:33 PM   #498
craftycoder OP
Motobrain PDU
 
craftycoder's Avatar
 
Joined: Oct 2009
Location: Atlanta
Oddometer: 2,173
Quote:
Originally Posted by Gany View Post
I have limited knowledge so could be making things up, but would rigging up a battery powered stick welder give you the 100amps you are trying to draw or just not controlled enough?
I need independent loads of no more 15A each. I can do it and I know how. The issue is that it isn't easy because most ways of burning off 1400W is dangerous. Either hot or spinning fast etc. doing it in a safe and controlled way is important.

Batteries would work as a source, but they will drain quickly and i Want to do long term burn in testing for obvious reasons. I will get that Agilent power supply and build a new 100A test rig when the time comes. It will be done indoors and safely controlled. I've got plans. I suppose my point was just to point out that there are good reasons why I have not yet done all the testing I will do before I am done.
craftycoder is offline   Reply With Quote
Old 09-14-2013, 12:47 AM   #499
RoTTe
n00b
 
Joined: Sep 2013
Oddometer: 4
Quote:
Originally Posted by craftycoder View Post
I need independent loads of no more 15A each. I can do it and I know how. The issue is that it isn't easy because most ways of burning off 1400W is dangerous. Either hot or spinning fast etc. doing it in a safe and controlled way is important.

Batteries would work as a source, but they will drain quickly and i Want to do long term burn in testing for obvious reasons. I will get that Agilent power supply and build a new 100A test rig when the time comes. It will be done indoors and safely controlled. I've got plans. I suppose my point was just to point out that there are good reasons why I have not yet done all the testing I will do before I am done.
Hi there!

Send me a sample! I have all juice that this needs ( lab PSUs, 15kW each, 90A max for channel at 168V ).

http://www.elektroautomatik.de/en/pr...er/ps8-3u.html






Btw, I think that the MOSFET -PSMN1R0-30YLC from NXP- maybe it's too much. In my little experience designing electronic I learned that when you oversize something too much, it doesn't apply better safe than sorry. Usually gives more headaches than help. This mosfet had an absurd low impedance for gate ( don't mistake with static resistance ), and an ultra-high input/output capacitance. Is this bad perse ? nop. All power mosfet had this characteristics. Do You need to oversize each channel 660% ? I don't know.

There is better parts on market:

PSMN0R9-25YLC: Same package, better price, less Vbr -bad thing-. But the reduced Rds (8% less) minimize the total power loss ( each uOhm count ).

Also, -call me paranoid- but I think it would be good to consider automotive parts -neither better nor cheaper but with a lot of manufacturer stress test-

BUK961R4-30E: worst... all ? price, Rds, different package. But Q101 compliant.
BUK9Y3R0-40E: Same package, improved dynamic characteristics, higher Rds, but again Q101 compliant.


Just an opinion! and good work!!
RoTTe is offline   Reply With Quote
Old 09-14-2013, 01:58 AM   #500
craftycoder OP
Motobrain PDU
 
craftycoder's Avatar
 
Joined: Oct 2009
Location: Atlanta
Oddometer: 2,173
I have a bag of BUK9Y3R0-40 on my desk. They are going to be on the next board with a smaller shunt. They have worse transient current performance, if memory serves they blow in the 400-500A range, but if want to test them against my analog current clamp. If it works I am inclined to go withy the automotive part for the reasons you suggest.

I'll look at the other FEt you suggest as well. I feel like I've read every powerfet data sheet but that part number doesn't ring a bell for me at the moment. It is the 4 o'clock hour though. I'm surprised I can remember anything.

Nice looking power supply! Many thanks on the tips.

Quote:
Originally Posted by RoTTe View Post
Hi there!

Send me a sample! I have all juice that this needs ( lab PSUs, 15kW each, 90A max for channel at 168V ).

http://www.elektroautomatik.de/en/pr...er/ps8-3u.html






Btw, I think that the MOSFET -PSMN1R0-30YLC from NXP- maybe it's too much. In my little experience designing electronic I learned that when you oversize something too much, it doesn't apply better safe than sorry. Usually gives more headaches than help. This mosfet had an absurd low impedance for gate ( don't mistake with static resistance ), and an ultra-high input/output capacitance. Is this bad perse ? nop. All power mosfet had this characteristics. Do You need to oversize each channel 660% ? I don't know.

There is better parts on market:

PSMN0R9-25YLC: Same package, better price, less Vbr -bad thing-. But the reduced Rds (8% less) minimize the total power loss ( each uOhm count ).

Also, -call me paranoid- but I think it would be good to consider automotive parts -neither better nor cheaper but with a lot of manufacturer stress test-

BUK961R4-30E: worst... all ? price, Rds, different package. But Q101 compliant.
BUK9Y3R0-40E: Same package, improved dynamic characteristics, higher Rds, but again Q101 compliant.


Just an opinion! and good work!!
craftycoder is offline   Reply With Quote
Old 09-14-2013, 04:36 AM   #501
RoTTe
n00b
 
Joined: Sep 2013
Oddometer: 4
Quote:
Originally Posted by craftycoder View Post
I have a bag of BUK9Y3R0-40 on my desk. They are going to be on the next board with a smaller shunt. They have worse transient current performance, if memory serves they blow in the 400-500A range, but if want to test them against my analog current clamp. If it works I am inclined to go withy the automotive part for the reasons you suggest.

I'll look at the other FEt you suggest as well. I feel like I've read every powerfet data sheet but that part number doesn't ring a bell for me at the moment. It is the 4 o'clock hour though. I'm surprised I can remember anything.

Nice looking power supply! Many thanks on the tips.
There are -in my opinion- other reasons for automotive parts -eg: maintenance issues-. For a pure resistive load -fairies world with unicorns running around - the worst case scenario: happy mosfet. This is neither real nor enduring. Nightmares occur -using a PWM with FAN load its a good example-, and if you filled the block with epoxy -who said replace the damaged part ONLY ?- murphy said if anything can go wrong it will go wrong.

The repetitive changes for inductive loads (real world: aka all loads more or less) it's the worst case scenario for a power mosfet. The ruggedness comes from a part that assure you that you can put the mosfet on its limits without damage ( working only with "Absolute Maximum Rating" values usually comes to parts dying or degraded performance over the time ).

NXP had some APPs notes about this, pulse and overshoot models:

http://www.nxp.com/documents/applica...te/AN10273.pdf

If I'm wrong, please correct me (I love when I'm wrong and learnzt things). I think the whole problem comes from the electromigration and damage to the insulation layer(s) which occurs on the die when induced the worst conditions which usually ending badly if the mosfet die with a short drain-source (you can turn off the thing and so on...). it's uncommon to see an inmmediate change -like a cliff-, it's more like a down road... to hell -and fire-.

The first rule about the fuses: you have a PCB to protect them. How many roasted PCBs with intact fuses ?. The AEC compliant parts, and specially the above mosfet with automotive grade comes with greater insulation layer -i hope - and some internally tricks that makes this things more reliable. When all manufacturers have a specific automotive grade costing twice, not worth the money when replace whole -unfixeable- part that worth hundred times more.

RoTTe screwed with this post 09-14-2013 at 04:43 AM
RoTTe is offline   Reply With Quote
Old 09-14-2013, 05:44 AM   #502
craftycoder OP
Motobrain PDU
 
craftycoder's Avatar
 
Joined: Oct 2009
Location: Atlanta
Oddometer: 2,173
I agree with you RoTTe. That is why I have the bag of new auto grade FETs on my desk (I had the wrong part number this morning actual photo below) and the intention to make use of them. I'm using a Transorb on each output to protect the power FETs from transients (which have worked well for voltage spikes up to 450V) and I also have dual pfets to protect the error amplifiers (current sense amps) from negative transients. I also encourage users to put diodes across inductive loads for safety but will they really do it? Somehow I doubt it...



Quote:
Originally Posted by RoTTe View Post
There are -in my opinion- other reasons for automotive parts -eg: maintenance issues-. For a pure resistive load -fairies world with unicorns running around - the worst case scenario: happy mosfet. This is neither real nor enduring. Nightmares occur -using a PWM with FAN load its a good example-, and if you filled the block with epoxy -who said replace the damaged part ONLY ?- murphy said if anything can go wrong it will go wrong.

The repetitive changes for inductive loads (real world: aka all loads more or less) it's the worst case scenario for a power mosfet. The ruggedness comes from a part that assure you that you can put the mosfet on its limits without damage ( working only with "Absolute Maximum Rating" values usually comes to parts dying or degraded performance over the time ).

NXP had some APPs notes about this, pulse and overshoot models:

http://www.nxp.com/documents/applica...te/AN10273.pdf

If I'm wrong, please correct me (I love when I'm wrong and learnzt things). I think the whole problem comes from the electromigration and damage to the insulation layer(s) which occurs on the die when induced the worst conditions which usually ending badly if the mosfet die with a short drain-source (you can turn off the thing and so on...). it's uncommon to see an inmmediate change -like a cliff-, it's more like a down road... to hell -and fire-.

The first rule about the fuses: you have a PCB to protect them. How many roasted PCBs with intact fuses ?. The AEC compliant parts, and specially the above mosfet with automotive grade comes with greater insulation layer -i hope - and some internally tricks that makes this things more reliable. When all manufacturers have a specific automotive grade costing twice, not worth the money when replace whole -unfixeable- part that worth hundred times more.
craftycoder is offline   Reply With Quote
Old 09-14-2013, 05:59 AM   #503
craftycoder OP
Motobrain PDU
 
craftycoder's Avatar
 
Joined: Oct 2009
Location: Atlanta
Oddometer: 2,173
I am using that BUK7 rather than the BUK9 because the 9 has a VGS of +-10V and I am driving it at about +9V which seems like I am skirting a little too close to edge for comfort. It is 3.3mΩ RDSon which is twice that of the current FET but I may drop the shunt from 4mΩ to 1-2mΩ and end up with better performance, reliability, and temperature characteristics.
craftycoder is offline   Reply With Quote
Old 09-14-2013, 08:12 AM   #504
RoTTe
n00b
 
Joined: Sep 2013
Oddometer: 4
Quote:
Originally Posted by craftycoder View Post
I am using that BUK7 rather than the BUK9 because the 9 has a VGS of +-10V and I am driving it at about +9V which seems like I am skirting a little too close to edge for comfort. It is 3.3mΩ RDSon which is twice that of the current FET but I may drop the shunt from 4mΩ to 1-2mΩ and end up with better performance, reliability, and temperature characteristics.
I think it's a wise choice -BUK9 is referenced as logic level MOSFET rather than standard level of BUK7, wide safe zone-. As always with these things: test, test, test!. I hate this feeling when you face a design and you have to take the good with the bad -all choices had its trade-off and none is perfect!-

But the magic smoke is priceless Aside the money/time cost!
RoTTe is offline   Reply With Quote
Old 09-14-2013, 08:42 AM   #505
craftycoder OP
Motobrain PDU
 
craftycoder's Avatar
 
Joined: Oct 2009
Location: Atlanta
Oddometer: 2,173
Quote:
Originally Posted by RoTTe View Post
But the magic smoke is priceless Aside the money/time cost!
I've seen my share of magic smoke. My reflow oven blew up last weekend (right before I was going to build a new PDU prototype) and I had to do a complete rebuild of all the oven internals... I had 120AC coupled to the my 5V rail (I was unaware of this) and was getting a strange error from the temp sensor (probably related to the AC but I am uncertain). I decided to touch the business end of the thermocouple and it gave me a heck of shock (I'm literally burning the wick are both ends for this project). In the body spasm the followed the thermocouple touched the oven case (earth grounded) and all the low voltage electronics exploded. Oops... The problem turned out to be in the OEM oven DC power supply (which has a dead short between 120VAC and the output of the 7805 5V DC regulator) and so it was not directly my fault. Still, I should have done a better job understanding the device I was repurposing. I rebuilt the whole thing with junk drawer parts and I once again have a reflow oven without any high voltage AC coupled in. I did mention that my tools are almost all homemade didn't I

Now that I have an oven again, I expect to be building up the new PDU MCU board tomorrow. A fresh design with some more safety equipment.

The source of the magic smoke! That is what it looks like when you pull 1500W though a pin designed for 150mW. The thermocouple amp looked a whole lot worse. I tried to take a photo of it too, but I just couldn't get my camera to focus on it through the microscope...
craftycoder is offline   Reply With Quote
Old 09-14-2013, 10:30 PM   #506
RoTTe
n00b
 
Joined: Sep 2013
Oddometer: 4
Quote:
Originally Posted by craftycoder View Post
[...]
I had 120AC coupled to the my 5V rail (I was unaware of this) and was getting a strange error from the temp sensor (probably related to the AC but I am uncertain). I decided to touch the business end of the thermocouple and it gave me a heck of shock (I'm literally burning the wick are both ends for this project). In the body spasm the followed the thermocouple touched the oven case (earth grounded) and all the low voltage electronics exploded.
[...]
Lucky you with 120VAC, try with 220VAC of Europe. But I've tasted worse. (Camera) Flash capacitors with ~300VDC -several times, I'm stubborn-. The worst by far was a neon inverter... the wire -rigid-, taped to table -cheap chinese tape- and the rest is up to your imagination.

Good luck!
RoTTe is offline   Reply With Quote
Old 09-15-2013, 06:16 AM   #507
craftycoder OP
Motobrain PDU
 
craftycoder's Avatar
 
Joined: Oct 2009
Location: Atlanta
Oddometer: 2,173
Progress report

We've reached 30% of our Kickstarter goal. Not bad for 5 days, but we still have a long way to go to reach our goal and see this project become a reality. My messaging has not been ideal. I think the video focuses too much on the smartphone angle and not not enough on the programming of the device which is where the real power lies. I was of the opinion that the programming would be hard to explain in 30 seconds so I might as well tease people with the fun stuff (the real time telemetry and interactivity) in hopes they invested the time to learn more from all the other available resources. It is not easy to confirm my hypothesis but based on the significant traffic and the comparatively few people who have decided to back the project I think missed the mark a bit. Monday morning quarterbacking...

Anyway, thank you very much for your support and advice. We continue to need your help spreading the word beyond the hallowed halls of ADVrider. Your help will be vital in seeing this project become a product.
craftycoder is offline   Reply With Quote
Old 09-17-2013, 02:59 AM   #508
EKinOR
Gnarly Adventurer
 
Joined: Mar 2012
Location: Central Oregon
Oddometer: 214
I'm spreading the word everywhere I can. This project needs to get funded!
__________________
- Eric
2002 BMW R1150GS
EKinOR is offline   Reply With Quote
Old 09-17-2013, 04:42 AM   #509
craftycoder OP
Motobrain PDU
 
craftycoder's Avatar
 
Joined: Oct 2009
Location: Atlanta
Oddometer: 2,173
Quote:
Originally Posted by EKinOR View Post
I'm spreading the word everywhere I can. This project needs to get funded!
Thanks man. We can certainly use the help.
craftycoder is offline   Reply With Quote
Old 09-17-2013, 07:29 AM   #510
beechum1
Beastly Adventurer
 
beechum1's Avatar
 
Joined: Oct 2007
Location: Herlong, CA. Look it up, I'll wait.
Oddometer: 7,172
Quote:
Originally Posted by craftycoder View Post
We've reached 30% of our Kickstarter goal. Not bad for 5 days, but we still have a long way to go to reach our goal and see this project become a reality. My messaging has not been ideal. I think the video focuses too much on the smartphone angle and not not enough on the programming of the device which is where the real power lies. I was of the opinion that the programming would be hard to explain in 30 seconds so I might as well tease people with the fun stuff (the real time telemetry and interactivity) in hopes they invested the time to learn more from all the other available resources. It is not easy to confirm my hypothesis but based on the significant traffic and the comparatively few people who have decided to back the project I think missed the mark a bit. Monday morning quarterbacking...

Anyway, thank you very much for your support and advice. We continue to need your help spreading the word beyond the hallowed halls of ADVrider. Your help will be vital in seeing this project become a product.
Some of us have other priorities for the $ just now. I'm getting ready for Scotty Breauxman's Baja Rally. All my extra money is going to needs for the 3 weeks. I do need two if these, one for each bike, but ........$$$
__________________
Rally Moto Kit /// Vendor Thread
Quote:
Originally Posted by Dread Pendragon View Post
Mention to HogWild which way the wind is blowing where you're at, wait 20 minutes, and he'll post a picture of the intersection your at and a Google Earth route of how to get there.
beechum1 is offline   Reply With Quote
Reply

Share

Thread Tools Search this Thread
Search this Thread:

.
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is On

Forum Jump


Times are GMT -7.   It's 05:59 AM.


Powered by vBulletin® Version 3.8.5
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Copyright ADVrider 2011-2014