I’m aware of that. You can also put small and ultra-cheap capacitors in parallel with the large ones to alter the spectral characteristic as well, even tailoring it to your specific needs.
1 Like
That business approach did work of course, still there are people who prefer to spend with vendors who maintain standards from a business ethics and tease out quality in their product line rather then depending on marginal profits from pumping product.
This is why I decided to pick up a Unity(& some couch foxboxes) from Enertion instead. I’m pretty sure @onloop realizes he found himself a unicorn 
with @Deodand
4 Likes
Yeah for sure, I’m 100% in on the unity as well. I run nothing but focboxes and unities except one odd dual escape. It’s a good product!
5 Likes
The latest new trend in the capacitor world shows that indeed Polymer based capacitors perform a lot better then the regular, I am working on a new controller as well (solar car application) and am seeing the similar things as @secupol on their performance and size. Please keep in mind that most of the time in motor controller world you don’t pick the bulk capacitor because of its capacitance, but mainly because of its current rating in combination with the impedance across a frequency range. Polymer capacitors will outperform every other capacitor style (accept for ceramic) in many cases by at least a factor of 3. As secupol said, please look into polymer caps specs and maybe order one to do some measurements, you’ll be surprised.
Could we also please lower the tone of criticism? Or translate it a bit into a constructive one preferably based on facts and comparisons?
Shouting a lot and shouting hard will not prove your opinion, it will just crowd the topic and drown the opinions of others who don’t spend their full day on the forum shouting back 
.
(Yes there is a drawback as well, so far the noticeable drawback of polymer caps is the higher leakage current, but that is mainly an issue for long life battery powered application, which esk8 is not xD )
28 Likes
Any idea when you will have an initial batch for sale?
1 Like
The problem you have here is that the leads are WAY too long on the larger capacitor. The leads look to be at least 10mm long which hugely increases your lead inductance. Remember that these caps are supposed to be mounted through a PCB, meaning the lead length is only ~1mm.
Also, how are you measuring the voltage rail? What is the loop area of your scope probe ground? You really need to have a custom scope lead soldered to a test point on your board to make this measurement properly. If you don’t , the huge switching current from the motor leads will be inductively coupled to the scope probe.
2 Likes
What is the value from 1mm and 10mm? Did you actually experiment with the difference between 10mm and 1mm?There was little difference in my experience.I would be grateful if you could convince yourself that there are other differences.
It’s easy to measure if you have an inductance meter. The value will be ~100nH difference by adding the 10mm of wire. The most significant effect of this will be ringing in the main voltage when switching.
Also think about the resistance, how much current are you expecting these caps to take? Ideally the caps will take the fundamental frequency of the FETs PWM, so at 50A continuous and 50% duty, you are getting 100A in the cap. I would be interested to see how warm your thin cap wires get. Of course you haven’t loaded your system yet which is another huge problem.
1 Like
About the capacitors, yes it is about farads, but also the ESR and the peak currents, thats why the “modern” capacitors perform better, but this is application dependant. I guess in a simple power supply application it does not matter so much, in a motor drive application it does.
e.g. the picture Secupol posted about the ripple at the caps. the electrolithic vs polymer, the polymer is much more capable of delivering peak energy
2 Likes
Some of us in this thread are doubting that measurement though. Sure, modern polymer capacitors are good but 220uF vs 2200uF is a big difference. My problems with the measurement
- The lead length on the 2200uF cap
- How was the scope probe connected? Can’t use a ground clip for this.
- Load on the circuit. I don’t believe the circuit has been loaded sufficiently. Need to run at 50% duty, large current.
- 1uS per division is showing us the gate edge. Actually I don’t believe it’s showing us the gate edge, its showing the inductively coupled edge->probe loop. The purpose of the bulk capacitors is to take the actual FET current, so you really want to look at a whole switching cycle.
1 Like
It all smells like Marketing Jive to me personally and less like actual science.
I’m a little more charitable, I think OP is just making the incorrect measurement and also not measuring properly. I have run into a similar problem in quadcopter racing. Measuring voltage rails with a scope is incredibly difficult when you have 50+ amps being switched so fast.
1 Like
I don’t necessarily think it’s OP that’s doing it…
I think it all starts with the seller or stakeholder of the company that makes the new polymer capacitors. It’s 2019, everything is marketing and you have to sort the truth out yourself.
Generally, if it’s less farads I am extremely skeptical. Last person I heard the same spiel from was trying to sell dirt-cheap ESCs.
If it really worked better at storing charges then it would be more farads. You can always add a dirt-cheap picofarad and nanofarad ceramic in parallel to handle the higher frequencies.
1 Like
Why are we not using 1fahrad caps if capacity is the only thing that matters? From my own testing on big quadcopters I had much better result with many small low esr caps then with a similar big one. It’s more about the low esr part than the capacity itself. These caps are not big enough to take care of a voltage drop it’s to remove harmful spikes and for that it needs to be “quick”.
A bigger capacitor in parallel with a smaller capacitor is “quick”.
ahh sorry, I misinterpreted you. Yes absolutely the specs of components are out of wack with reality (see mosfet ratings). I think we are identifying two separate problems, one with capacitor specs, one with the tests/conclusions made in this thread.
Nobody said capacity is the only things that matters. Everything matters.
- cost
- size
- availability
- temperature rating
- package
- capacitance
- power handling
- ESR
- etc…
Your suggestion of 1F cap fails on cost, size, availability, package at least.
1 Like
And no 1F cap is available at 60V unless it’s super-huge and super-heavy
Also, even placing a bunch of 350F 2.7V super caps in series/parallel you have the huge problem of the inductance of the leads to those capacitors. The problem of gate edges and 30khz 100A switching is solved by a combination of low inductance ceramic caps very close to the fets combined with large (~2000uF) electrolytics (polymer if you like) as close as possible to the fets. Vedder did a good youtube video showing how to simulate this kind of thing.
2 Likes
@secupol, I think your product looks great, as I read this thread, your results are being questioned, probably because you have no track record with proven information. Why not invite a few well respected people from this community for beta testing?
1 Like