ZapPat
10 kW
Mr. Mik said:
GREED.
Same as for the 40$ software bug fix I guess...
Standardizing the measurement of internal resistance
- Thread starter oofnik
- Start date
GREED.
Same as for the 40$ software bug fix I guess...
Doctorbass
100 GW
Guys, i would really love to build my own CBA amplifier! 8) .. but some here have more knowledge than me in electronic for using analog transistors and gain and all that calculation stuff... these electronic classes are far from me...
BUt i have 2 great heavy transistor rack!!! for a total of 5000W of dissipation!! it's two larhe heatsing with 10x MJ15024 and the other with 10x MJ15025 with all their shunt resistors!!
But i only gave a general idea on how to build a currant amplifier.. first.. it seems very simple to me.. but i'm sure there is always something i forget about the desing..
Can some of you help me with that desing!
i'll post some great pics of them
BUt i have 2 great heavy transistor rack!!! for a total of 5000W of dissipation!! it's two larhe heatsing with 10x MJ15024 and the other with 10x MJ15025 with all their shunt resistors!!
But i only gave a general idea on how to build a currant amplifier.. first.. it seems very simple to me.. but i'm sure there is always something i forget about the desing..
Can some of you help me with that desing!
i'll post some great pics of them
recumbent
100 kW
Can we use an electric baseboard heater rated at 1,500 watts, as a load? I realize it's AC but still don't know the answer.
My 48 volt battery records about 1500 watts (on cycle analyst) under full load when riding, How can i duplicate this load at home without my bike.
Pardon my supreme ignorance on this matter
, but i just don't know.
My 48 volt battery records about 1500 watts (on cycle analyst) under full load when riding, How can i duplicate this load at home without my bike.
Pardon my supreme ignorance on this matter
, but i just don't know.
Heaters, blow driers, toasters etc. all use nichrome wire to heat up. Resistance is proportional to the length, and power is equal to the current squared times resistance, or voltage squared divided by resistance. So if you have a heater rated for 1500W that runs on 120 volts, that means the resistance is 9.6 ohms. If you were to hook this up directly to say a 3.3v lithium cell, you would only be pulling about 1.1 W! So what you have to do is take the unit apart and use maybe 1/10th the length of the wire to get the right amount of current flow and thus power.
All math was derived from Ohm's law, V = current * resistance, and power P = current * voltage.
All math was derived from Ohm's law, V = current * resistance, and power P = current * voltage.
ZapPat
10 kW
Doctorbass said:
That sounds like some very serious heat sink, Doc!
I'll think about your current amplifier dilema, but I must admit I am already quite distracted by my BLDC controler project. Plus I don't have a CBA myself. I actually would be more inclined to make a CBA from scratch, but only after I'm done with my controller!
Anyhow, I understand your amplifier needs to:
- For every 1A that the CBA draws itself, the amplifier will draw 10A (9X more for the amp itself than the CBA).
- We need to make sure that the voltage presented to the CBA itself isn't affected.
Anything else to take note of?
AndyH
10 kW
Doctorbass said:
Here you are Doc. First, a quick overview. Follow the links for the 4.5MB versions. I take requests if you want a better look at something. :wink:
http://www.absoluteefficiency.com/images/amp1.jpg
http://www.absoluteefficiency.com/images/amp2.jpg
http://www.absoluteefficiency.com/images/amp3.jpg
View attachment amp3_s.jpg
View attachment amp1_s.jpg
Tiberius
10 kW
Trimpots
Aargh! They are the work of the devil.
Nick
Aargh! They are the work of the devil.
Nick
Doctorbass
100 GW
Thanks you very much AndyH !!
I now can figure how it is made and i can see that it's like a basic CBA multiplyed 5X for the transistor stage..
I wonder if i could do something similar with all these 10 x MJ150025 rated 250W each for 18A.. their SOA curve seems interesting
I now can figure how it is made and i can see that it's like a basic CBA multiplyed 5X for the transistor stage..
I wonder if i could do something similar with all these 10 x MJ150025 rated 250W each for 18A.. their SOA curve seems interesting
Attachments
BatteryMooch
10 kW
Thanks for taking those pictures! I've been dying to know how it was laid out for a long time.AndyH said:
Could you confirm that the op-amps. e.g, U1, are TLC272C's and what the MOSFET's are? I couldn't quite make it out in the photos.
The CBA uses the TLC272C so that was expected but it appears that the MOSFET's are from ST Micro and not IR's IRL2910 (as used in the CBA).
BatteryMooch
10 kW
I'm a little late to the thread but I just joined the forums.Doctorbass said:
I wanted to say great work on those mods! I had my fuse/fuseholders drop out about 1Ah in to a 29A, 100W test a few weeks ago and ended up doing a few mods myself. I'll post them soon. The CBA is such a great device, but often sooooooo frustrating.
AndyH
10 kW
CamLight said:Thanks for taking those pictures! I've been dying to know how it was laid out for a long time.AndyH said:
Could you confirm that the op-amps. e.g, U1, are TLC272C's and what the MOSFET's are? I couldn't quite make it out in the photos.
The CBA uses the TLC272C so that was expected but it appears that the MOSFET's are from ST Micro and not IR's IRL2910 (as used in the CBA).
The MOSFETs are P120NF, and at least one of the op-amps is a TLC272CP. I'll have to get the rest of the info when the amp returns -- it's on its way 'home' to have Q3 replaced and to be recalibrated. I stepped on a powerstrip during a battery test and 'cycled power' on the amp. POP POP and out comes the magic smoke.
Doctorbass
100 GW
For info i confirm that my mosfet are IRL2910 on my CBA II
Doc
Doc
BatteryMooch
10 kW
AndyH said:The MOSFETs are P120NF, and at least one of the op-amps is a TLC272CP. I'll have to get the rest of the info when the amp returns -- it's on its way 'home' to have Q3 replaced and to be recalibrated. I stepped on a powerstrip during a battery test and 'cycled power' on the amp. POP POP and out comes the magic smoke.
Ouch! I hope you get it back soon Andy.
Thanks for the part numbers Doc and Andy! I wanted to compare the specs for the FETs in the Amp with the IRL2910's in the CBA II, especially to see if the Amp's FETs were rated for continuous use biased in their linear region (i.e., used as a load). The IRL2910's aren't.
steveo
100 kW
I have a cba 2 Which blow a fet & fuse. I will be attempting to fix tonight!!!..
Awsome Doc, You already have done modification!!!! nice
-steveo
Awsome Doc, You already have done modification!!!! nice
-steveo
steveo
100 kW
FYI for everyone...
A 4110 mosfet will work as a good replacement mosfet for CBA 2!!!
-steve
A 4110 mosfet will work as a good replacement mosfet for CBA 2!!!
-steve
BatteryMooch
10 kW
Be careful though...steveo said:
The IRFB4110 has great thermal resistance specs (a Rjs of about 0.9 degrees-C/W vs. the IRL2910's 1.25 degrees-C/W) but the Maximum Safe Operating (SOA) graph from the data sheet is worrisome (http://www.irf.com/product-info/datasheets/data/irfb4110pbf.pdf, FIgure 8 on page 4).
Check out the DC plot line. It shows that when using the IRFB4110 as a load (i.e., operating in its "linear region") as is done in the CBA, you must derate it down to about 6.5A at a pack voltage of 10V (65W). At 15V, you're down to 1.5A (22.5W). And at about 17V, you can't exceed 1A (17W).
The IRFB4110 is a true switching MOSFET, that's why it's so great for speed controllers. But it hates operating as a load and must be severely derated if it does.
Exceeding these derated specs risk the chance of hotspotting inside the MOSFET die and thermal runaway, which will fry the MOSFET. It might fry in a minute, or it might take a year, but it will happen.
Steve, as a test could you take your CBA up to 100W with a decently high voltage pack? At least 24V, 48V would be great. The thermal resistance specs for the IRFB4110 will allow it to run cooler than the IRL2910 but the higher voltage will test the MOSFET's susceptibility to hotspotting.
Since the IRL2910 in a stock CBA doesn't even have a DC plot line in its SOA graph and we don't know how far it might need to be derated when used as a load, IMHO, this is one of the possible causes for CBA MOSFETs burning out when used at power levels significantly under the rated continuous max of 100W.
ZapPat
10 kW
Thats very interesting indeed, Camlight!CamLight said:
I've hardly ever used FETs in their linear region, and didn't really think about this much. Indeed it would explain the many FETs destroyed by west mountain CBA users. Do you know what mecanism in the FET limits it's use in it's linear region to such a great extent? It sounds kinda odd to me, who always uses FETs in switching circuits.
If this is true, this companie's CBA has a major design flaw!
I thought they were using the FET in a buck converter setup at first, but then didn't see any inductors nor anything that would be used as a load inside. No doubt Docbass's good old transistor bank would be much better than those FETs for this application! He could use them as a stand alone simple current-regulated load with a few extra parts I think. I seem to recall some such circuits in my electronics encyclopedias...
BatteryMooch
10 kW
Check out Fig. 3 on page 5 of the datasheet. You can see that when the MOSFET is biased with a gate voltage below approx. 5.2V that any increase in temperature will result in the MOSFET trying to pull more current. And since MOSFETs are rarely biased at a voltage that high when used as a load, the MOSFET will always try to conduct more current as it heats up.ZapPat said:
This isn't much of a problem in a switching application since the MOSFET spends so little time in its linear region, but it's why you need to switch a FET on/off as fast as possible because it spends time in this linear region, quickly heating up as it does. But when used as a load, always biased in its linear region, inevitably one part of the MOSFET's die will heat up more than another. This leads to a lowering of the threshold voltage for that part of the die and results in it hogging a bit more current, which heats it up more, etc. It's called "hotspotting" and can lead to thermal runaway and destruction of the MOSFET. The construction of linear MOSFETs (vs. a switching MOSFET's "trench" construction) helps to prevent this from happening but linear MOSFETs are typically much more expensive and often have very high voltage, but very low current ratings compared to switching MOSFETs.
Fig, 3 also shows you why you can't reliably parallel MOSFETs when they're biased in their linear region. They are never driven hard enough, over 5.2V for the IRFB4110 for example, to allow them to "automatically" compensate for differences between them and to share current equally. One will always try to hog it all and burn out. I'm not saying it can't be done, it can, it's just problematic and you have to be very careful to select the MOSFETs from the same batch, wire them exactly the same way, etc.
The problems you can encounter when using switching MOSFETs as a load are not commonly known, even for engineers in the field. Many SOA graphs don't even have a DC plot line. This leads to the, mistaken, conclusion that derating isn't needed for use as a load or that they can use the "last" line in the graph, typically the 10mS on-time plot line.
The CBA uses a wonderfully simple setup for their MOSFET. Use a PWM output from the microprocessor, filer it to smooth it to DC, pass it through an amp to lower its impedance, and then directly drive the MOSFET with the amp's output. No digital-to-analog converter needed.
Having said all this, switching MOSFETs can be successfully used as a load. I've been designing loads for years now, up to several kilowatts, and none of them have blown when used within their ratings. Ya' just gotta' pick the right MOSFET and have a very firm grasp of the thermal environment that MOSFET is operating in.
I did some looking for a MOSFET that would work as a replacement for the CBA's IRL2910 and the only one that looked good was the IPP048N06L. It's rated for use as a load and even with the derating, it can run as a 100W load at 40V. I've made the swap, along with several other people I know, and so far it's worked wonderfully with up to 125W loads. Time will tell though as many CBAs don't fail for years.
steveo
100 kW
CamLight said:
Hey
thanks for your feedback!, I used a 4110 as per it is the best of mosfets i've ever seen; and i have a bunch of sprares kicking around; my second choice would of been a PNF75NF mosfet... but i wanted better! ..
I will try 12v tonight at say 8amp Which is 96W; That is pushing it IMO;
I will series 2 12v packs for total of 24v @ 4 amps which is 96W
I have 1 or 2 30 & 35 amp fuses left at my house lol .. i will need to buy some more .. I think i will stop by canadian tire tonight to pick up a bunch!!!
What is the wattage limit of the cba 2 for single cell? 100w to 150W is was there paper says?..i prefer to stay on the 100w end .. as doc showed in an earlier post .. the pcb need to have thicker wire added for more current capability, lower resistance & temp!
-steveo
Doctorbass
100 GW
steveo said:
Steveo.. When you say: "Will work" does it mean you didn't tried yet? if not, is it succesfull for you? any test done?
Doc
BatteryMooch
10 kW
That 24V test sounds good. The wattage will still be within the CBA's ratings but it will test the 4110's ability to be used as a load.steveo said:
150W is the rated max for the CBA (for short periods), 100W continuous. Based on my tests though, you should never run it above 86W or so with the stock IRL2910 MOSFET. With the lower thermal resistance specs for the 4100, you can exceed that 85W recommendation. But, based on the 4110's SOA graph, you can only do that for battery voltages below 5V or so. That's why I'm so interested in testing a 4110 at power levels still well within the CBA's rated max but way beyond what the datasheet says the 4110 can handle (at higher pack voltages).
steveo
100 kW
Hey
doc/camlight
ok i was able to do one torcher test; 12 @ 9 amps = 100w approx.. about a miniute or so in .. mosfet blew taking out the fuse also..
i personally think at 86w it would work perfectly fine .. I think i was pushing the limits .. but your right it should take 100w continuous !!!
i may try the P75nf75 mosfet as a replacement instead of wasting expensive 4110's on this bad boy..
Doc..
I had tested a a123 cell at about 75w yesterday for about 5 mins and it ran with not problems!!! but this last test no sucessful
-steveo
doc/camlight
ok i was able to do one torcher test; 12 @ 9 amps = 100w approx.. about a miniute or so in .. mosfet blew taking out the fuse also..
i personally think at 86w it would work perfectly fine .. I think i was pushing the limits .. but your right it should take 100w continuous !!!
i may try the P75nf75 mosfet as a replacement instead of wasting expensive 4110's on this bad boy..
Doc..
I had tested a a123 cell at about 75w yesterday for about 5 mins and it ran with not problems!!! but this last test no sucessful
-steveo
BatteryMooch
10 kW
steveo
Thanks for sacrificing a MOSFET for the cause!
According to my calculations that IRFB4110 probably wasn't running above its absolute maximum operating temperature. IMHO, it blew because it hates being operated as a load.
An 86W test with a 12V pack? IMO, it would still blow, and fast too.
The IRFB4110 is only rated for about 48W at 12V. An 86W discharge is still way beyond its rated maximum.
The STP75NF75 also isn't rated for use as a load but its SOA graph looks a bit better than the IRFB4110...it might work.
If you want, PM me your address and I'll send you an IPP048N06L MOSFET to try. It's rated for use as a load at up to 100W at 40V. Several people I know are already using this MOSFET successfully at 125W....including me.
Thanks for sacrificing a MOSFET for the cause!
According to my calculations that IRFB4110 probably wasn't running above its absolute maximum operating temperature. IMHO, it blew because it hates being operated as a load.
An 86W test with a 12V pack? IMO, it would still blow, and fast too.
The IRFB4110 is only rated for about 48W at 12V. An 86W discharge is still way beyond its rated maximum.
The STP75NF75 also isn't rated for use as a load but its SOA graph looks a bit better than the IRFB4110...it might work.
If you want, PM me your address and I'll send you an IPP048N06L MOSFET to try. It's rated for use as a load at up to 100W at 40V. Several people I know are already using this MOSFET successfully at 125W....including me.
steveo
100 kW
CamLight said:steveo
Thanks for sacrificing a MOSFET for the cause!
According to my calculations that IRFB4110 probably wasn't running above its absolute maximum operating temperature. IMHO, it blew because it hates being operated as a load.
An 86W test with a 12V pack? IMO, it would still blow, and fast too.
The IRFB4110 is only rated for about 48W at 12V. An 86W discharge is still way beyond its rated maximum.
The STP75NF75 also isn't rated for use as a load but its SOA graph looks a bit better than the IRFB4110...it might work.
If you want, PM me your address and I'll send you an IPP048N06L MOSFET to try. It's rated for use as a load at up to 100W at 40V. Several people I know are already using this MOSFET successfully at 125W....including me.
Hey
I have sent you a pm
I have P75NF75 on discharge at 7.5amps @ 12v pulling approx 86w for over 20mins; temp farley warm but ok!..
-steveo
Doctorbass
100 GW
Nice experiment on different mosfet setup!! guys
By my side i can confirm that i did more than 20 complete discharge at 35A and max wattage to around 110W at starting disch and then 100W to around Ah/2 on the graph...
BUT i installed large cooper heatsing sheet on the OEM mosfet legs close to the to-220 case... and also on the fuse leg
and finally a 2.5" fan blowig on the the pcb side that cool down the cooper heatsings...
I have to say that i had problem before to do this mod... the overtemp message appeared several times during the 30A+ discharge and i never found the ARGGG*&&*%?&*%!!!!! thermal sensor (thermistor i guess) on the pcb so my last option was to cool it down...
Doc
By my side i can confirm that i did more than 20 complete discharge at 35A and max wattage to around 110W at starting disch and then 100W to around Ah/2 on the graph...
BUT i installed large cooper heatsing sheet on the OEM mosfet legs close to the to-220 case... and also on the fuse leg
and finally a 2.5" fan blowig on the the pcb side that cool down the cooper heatsings...
I have to say that i had problem before to do this mod... the overtemp message appeared several times during the 30A+ discharge and i never found the ARGGG*&&*%?&*%!!!!! thermal sensor (thermistor i guess) on the pcb so my last option was to cool it down...
Doc
Similar threads
