| | |
I'm building L297/L298 driver combo to run small Nema 17 stepper motors. I don't have fast diodes but I have standard n4007 diodes. Is it OK to use them? Standard diodes have less recovery time than fast diodes. What is the highest RPM I can achieve with standard diodes before they start having problem to recover? Any formulas? Thank you in advance.
Short answer is not really... the recovery time has nothing to do with the RPM - its to do with the efficiency of the reverse current dumping capability and the power dissipated in the diode during the transition. A normal diode will fry quickly due to it being in the transition state for several microseconds instead of a few tens of nS.
So what you are saying it is ok to use standard diode until it fries. Even though diode is not fast but it is still switching fast and it should perform just as good as fast diode until it goes bad. Is it correct?
Dear Almaz,
Use Schottky diodes. They cost a bit more, but it may be money well spent. The L297/L298 suppliers will give you a part number.
Best wishes,
Martin
it'll fry virtually immediately... there's a lot of power flying around... so NOT ok cos once the diode fries the output devices in the L298 will be subject to reverse voltages and then they'll fry... not an area to skimp on, and fast recovery diodes are cheap and easy to source...Originally Posted by Almaz
I've been testing everything on breadboard using n4007 diodes and so far they are working. Using small nema 17 steppers I can get about 40-60 rpms per minute. I'll try to check the frequency later on today using sound card oscilloscope.
small motors with low inductance - you might get away with it for a while on low driver volts and no load... wait till your motors are generating back-emf on deceleration under load....I've been testing everything on breadboard using n4007 diodes and so far they are working
Quick tutorial on diode speed:
Diodes are like one-way valves. The permit current to flow in one direction but not the other. "Fast" and "slow" refers to how quickly the diode wakes up and realizes current is trying to flow in the wrong direction. This is called "reverse recovery time".
A 1N4004 takes 4 microseconds to realize "Hey, current is flowing in the wrong direction! Quick, shut it off!". During that time (the 4uS), the diode looks like a short-circuit. A "fast" diode is more alert. It typically realizes current is going the wrong way and reacts in 40 nanoseconds (0.04 microseconds or 100 times more quickly).
You say "4 microseconds or 0.04 microseconds. Who cares? It's quick either way." Not true.
You are building a chopper drive. The diodes turn on and off 20 thousand times a second. That's 50 microseconds of time between diode short-circuits or 8% (4/50) of the time! Your L297 gets hot, your diodes get hot. Bring the current up to 2A and things mysteriously burn up.
Replace them with "fast" diodes and they are short-circuits only 0.08% of the time. 100 times less and completely negligible for bad effects.
OK, if some is good, more is better right? Let's use Schottky diodes; they have zero reverse recovery time! Not good either. Schottky diodes are expensive, have low reverse voltages (30 to 40VDC) and are "leaky" (they conduct some current in the wrong direction all the time). You have already taken care of 99% of the problem by using fast diodes; you are left chasing the remaining 1%. That remaining 1% incurs drawbacks that result in a less than optimal solution.
"OK, I'm sold on fast diodes. Question: What idiot would want to use slow (standard diodes) then?" Standard, slow diodes are used for 60Hz rectification of AC into DC. The diodes conduct 120 times a second or every 8,333 microseconds. 4 microseconds reverse recovery results in 0.05% wastage in the rectifiers. That is too small of a number to worry about at 60Hz, kind of like the 0.08% number is too small to worry about in your L297 bridge. Good enough is good enough.
Hope that helps.
Mariss
Dear Mariss,
Thank-you. I am not worthy...LOL
Best wishes,
Martin
Martin,
That's cool. I love what I do so I tend to humanize stupid silicon parts because it helps me to understand what the dumb little critters do. I strive to see the whole circus alive in my head when I look at a schematic.
Mariss
Mariss...
I think that's what I said, but in an easier to understand and more comprehensive formwere you a teacher in another life?
Irving...
Posting Permissions
| | |
LinkBack URL
About LinkBacks
