Fiddys Tiny Tesla & TS Build/Feedback

Show off your finished kit here!

Re: Fiddys Tiny Tesla & TS Build/Feedback

Postby loneoceans » Sun Mar 22, 2015 2:00 pm

Fiddy wrote:
SystemsGuy wrote:Would y'all mind posting the part numbers? I'm happy to add to my next mouser order - cheap insurance is good! :)


Im going to use MUR1560 ultra fast diodes. Im not real keen on the 1.5KE510CA's as seen on the schematic.


Why are you adding the diodes when the FGA60N65SMD (I presume these are the IGBTs used) already come with comparable, in not better, internal co-pack diodes?
User avatar
loneoceans
Magnificent MOSFET
 
Posts: 103
Joined: Mon Feb 11, 2013 12:39 pm

Re: Fiddys Tiny Tesla & TS Build/Feedback

Postby Fiddy » Sun Mar 22, 2015 10:49 pm

loneoceans wrote:
Fiddy wrote:
SystemsGuy wrote:Would y'all mind posting the part numbers? I'm happy to add to my next mouser order - cheap insurance is good! :)


Im going to use MUR1560 ultra fast diodes. Im not real keen on the 1.5KE510CA's as seen on the schematic.


Why are you adding the diodes when the FGA60N65SMD (I presume these are the IGBTs used) already come with comparable, in not better, internal co-pack diodes?


They are typically faster, handle more current and keep the extra heat away from the IGBTs.
Pico SGTC
Nano SSTC
Audio Modulated SSTC
Audio Modulated Bipolar SSTC
TinyTesla SSTC
Onetesla TS DRSSTC
OneTesla DRSSTC
EVR MicroBrute DRSSTC
User avatar
Fiddy
Terrific Tube
 
Posts: 89
Joined: Tue Jan 15, 2013 6:31 pm
Location: Australia

Re: Fiddys Tiny Tesla & TS Build/Feedback

Postby loneoceans » Mon Mar 23, 2015 2:33 am

Fiddy wrote:
loneoceans wrote:
Fiddy wrote:Im going to use MUR1560 ultra fast diodes. Im not real keen on the 1.5KE510CA's as seen on the schematic.


Why are you adding the diodes when the FGA60N65SMD (I presume these are the IGBTs used) already come with comparable, in not better, internal co-pack diodes?


They are typically faster, handle more current and keep the extra heat away from the IGBTs.


But they're really not faster nor do they handle more current than the co-pack diode which was specifically designed for the 60N65, as can be seen from the specifications. You're simply adding extra unnecessary inductance in the circuit which can lead to higher Ldi/dt switching spikes.
User avatar
loneoceans
Magnificent MOSFET
 
Posts: 103
Joined: Mon Feb 11, 2013 12:39 pm

Re: Fiddys Tiny Tesla & TS Build/Feedback

Postby Fiddy » Mon Mar 23, 2015 3:15 am

But they're really not faster nor do they handle more current than the co-pack diode which was specifically designed for the 60N65, as can be seen from the specifications. You're simply adding extra unnecessary inductance in the circuit which can lead to higher Ldi/dt switching spikes.


FGA60N65SMD recovery time = 47 - 212ns
MUR1560 recovery time = 35 - 50ns

How is that not faster?

How much current and/or power do you think the the body diode can dissipate?

I dont think they will add much parasitic inductance anyway.
Pico SGTC
Nano SSTC
Audio Modulated SSTC
Audio Modulated Bipolar SSTC
TinyTesla SSTC
Onetesla TS DRSSTC
OneTesla DRSSTC
EVR MicroBrute DRSSTC
User avatar
Fiddy
Terrific Tube
 
Posts: 89
Joined: Tue Jan 15, 2013 6:31 pm
Location: Australia

Re: Fiddys Tiny Tesla & TS Build/Feedback

Postby loneoceans » Mon Mar 23, 2015 4:14 am

Fiddy wrote:
But they're really not faster nor do they handle more current than the co-pack diode which was specifically designed for the 60N65, as can be seen from the specifications. You're simply adding extra unnecessary inductance in the circuit which can lead to higher Ldi/dt switching spikes.


FGA60N65SMD recovery time = 47 - 212ns
MUR1560 recovery time = 35 - 50ns

How is that not faster?
How much current and/or power do you think the the body diode can dissipate?
I dont think they will add much parasitic inductance anyway.


Still incorrect. The FGA's copack diode is listed at 47ns vs the MUR's 60ns at 25C, but this isn't even really the point since they are not the same 'trr' times - the I_F and dI/dt test values are completely different, where the FGA is characterized at significantly higher I_F and dI/dt. The copack's diodes are in fact designed specifically to match the IGBT and to have soft recovery characteristics. You will find that the die area of the diode is almost, if not as large at the transistor itself since it needs to be rated similarly to the IGBT.

This will be a different story if you were using transistors with no reverse diode of course. Anyway my point was that there is no need to add an additional diode, and doing so will only add inductance to the reverse diode loop and may increase the voltage transients. If you're worried about heat, you may find that a TO220 diode mounted above the IGBTs without heat-sinking may even lead to that diode getting hotter than if the diode was co-pack in the first place (fortunately, it is!)
User avatar
loneoceans
Magnificent MOSFET
 
Posts: 103
Joined: Mon Feb 11, 2013 12:39 pm

Re: Fiddys Tiny Tesla & TS Build/Feedback

Postby Fiddy » Mon Mar 23, 2015 4:28 am

loneoceans wrote:
Fiddy wrote:
But they're really not faster nor do they handle more current than the co-pack diode which was specifically designed for the 60N65, as can be seen from the specifications. You're simply adding extra unnecessary inductance in the circuit which can lead to higher Ldi/dt switching spikes.


FGA60N65SMD recovery time = 47 - 212ns
MUR1560 recovery time = 35 - 50ns

How is that not faster?
How much current and/or power do you think the the body diode can dissipate?
I dont think they will add much parasitic inductance anyway.


Still incorrect. The FGA's copack diode is listed at 47ns vs the MUR's 60ns at 25C, but this isn't even really the point since they are not the same 'trr' times - the I_F and dI/dt test values are completely different, where the FGA is characterized at significantly higher I_F and dI/dt. The copack's diodes are in fact designed specifically to match the IGBT and to have soft recovery characteristics. You will find that the die area of the diode is almost, if not as large at the transistor itself since it needs to be rated similarly to the IGBT.

This will be a different story if you were using transistors with no reverse diode of course. Anyway my point was that there is no need to add an additional diode, and doing so will only add inductance to the reverse diode loop and may increase the voltage transients. If you're worried about heat, you may find that a TO220 diode mounted above the IGBTs without heat-sinking may even lead to that diode getting hotter than if the diode was co-pack in the first place (fortunately, it is!)


Still incorrect, data sheet says typical 35ns MAX 50ns, the max for the FGA60N65SMD is a whopping 212ns at 175degC.

the MUR1560 can handle up top 150A <10us; rep. rating, much higher than the 30A. then there's the whole power dissipation factor where its not being dissipated inside the TO-247 package with the IGBT.

Im leaving them in there either way.
Pico SGTC
Nano SSTC
Audio Modulated SSTC
Audio Modulated Bipolar SSTC
TinyTesla SSTC
Onetesla TS DRSSTC
OneTesla DRSSTC
EVR MicroBrute DRSSTC
User avatar
Fiddy
Terrific Tube
 
Posts: 89
Joined: Tue Jan 15, 2013 6:31 pm
Location: Australia

Re: Fiddys Tiny Tesla & TS Build/Feedback

Postby loneoceans » Mon Mar 23, 2015 12:44 pm

Fiddy wrote:Still incorrect, data sheet says typical 35ns MAX 50ns, the max for the FGA60N65SMD is a whopping 212ns at 175degC.

the MUR1560 can handle up top 150A <10us; rep. rating, much higher than the 30A. then there's the whole power dissipation factor where its not being dissipated inside the TO-247 package with the IGBT.

Im leaving them in there either way.


My last comments for this topic and as a caution to everyone else reading - As I said and from my experience in power engineering, the t_rr is only given at for I_F = 1A, di/dt = 50A/us and T_j = 25C. This changes significantly with temperature and di/dt and so on, usually getting significantly longer. There are varying datasheets on the MUR1560 (see attached below). The FGA's diode is characterized at 200A/us and I_F = 30A. Typical t_rr for the MUR160 at 28A 200A/us (100C) then becomes 140ns. The FGA's trr would then be somewhere between 47(25C) and 212ns(175C) typical, looking closer to less than 140ns if you really want to compare t_rr.

The same idea goes for the recovery charge. So my point is, the comparison is not straightforward as comparing numbers, and you will also be more interested in the t_a and t_b or diode reverse recovery periods, which are not in the FGA datasheet and will affect power dissipation and recovery characteristics.

Will adding the diode be harmful to your circuit? It depends on when the IGBTs do the switching among others, but a really quick calculation shows the lead inductance (vs on-die) of the off-board diode is about an order of magnitude greater (though still small), so more of the current is going to flow through the co-pack diode at the high frequencies during switching anyway. Then depending on how much current is flowing through the outside diode during switching, the added inductance can only increase the voltage spikes! - making your overall fix less robust than if you had left out external diodes in the first place. This is a big enough problem for careful inductance management to be a big part of industrial design, as well as highlighted in many application notes where even the leads of TO247 packages cause a big enough inductance in high current switching leading to significantly higher voltage spikes as well as switching losses.

You can do the same exercise with power dissipation w.r.t conduction losses (about the same V_F at same I_F) as well as switching losses due to reverse conduction and reverse charge. The thermal resistance of your TO220 to air will also be about an order of magnitude more than the co-pack's diode thermal resistance to the heatsink. I will refrain from going through more specifics but the internet has lots of good reverse recovery diode application and technical notes.

I know you have good intentions to make the coil more robust, but I didn't want to encourage everyone else to add their own unnecessary reverse diodes, or do something else which might inadvertently make the coil less robust. You'd be better off buying a better IGBT instead.

Diode: http://www.micropik.com/PDF/mur1560.pdf
IGBT: https://www.fairchildsemi.com/datasheet ... N65SMD.pdf
Last edited by loneoceans on Thu Apr 09, 2015 12:20 pm, edited 5 times in total.
User avatar
loneoceans
Magnificent MOSFET
 
Posts: 103
Joined: Mon Feb 11, 2013 12:39 pm

Re: Fiddys Tiny Tesla & TS Build/Feedback

Postby Fiddy » Mon Mar 23, 2015 12:49 pm

loneoceans wrote:I know you have good intentions to make the coil more robust, but I didn't want to encourage everyone else to add their own unnecessary reverse diodes, or do something else which might inadvertently make the coil less robust. You'd be better off buying a better IGBT instead.


Yes more robust, more robust than without them, its no mistake.
Pico SGTC
Nano SSTC
Audio Modulated SSTC
Audio Modulated Bipolar SSTC
TinyTesla SSTC
Onetesla TS DRSSTC
OneTesla DRSSTC
EVR MicroBrute DRSSTC
User avatar
Fiddy
Terrific Tube
 
Posts: 89
Joined: Tue Jan 15, 2013 6:31 pm
Location: Australia

Previous

Return to Completed oneTesla Kits

Who is online

Users browsing this forum: No registered users and 24 guests


cron