15nF capacitors in parallel with each 2.2k resistor.
Make sure you do a lot testing afterwards, as you may experience small kicks when exiting coast at high speeds
Nevermind, it would be best to simply leave the hardware alone as the firmware is sensitive to minor changes in hardware.
People would be more apt to but your shit if you weren’t such an ass. Helping folks understand ASS a bit more than just dodging questions would be a good start.
I mean, I can help explain exactly whats going on, but there’s two components, the BEMF, which is easy to understand but is actually the smaller component, then something else entirely that requires a really stupid amount of trigonometry, advanced algebra, and circuit theory to understand.
better not to uncover the mystery of ASS 
Either it requires phase filtering for optimal performance…or it doesn’t. This should be known unambiguously by the developer. And I’ve asked questions about this before as pointed out in this thread.
I can however say that users have had success using ASS on my CFOC2 design of which has no phase voltage filtering. So it seems filtering helps and is ideal but maybe is not absolutely required.
If you want to track the rotor down to zero RPM filtering is needed. There is no way around it. Regular sensorless startup also profits from the A.S.S algorithm, but you need some RPM at least. The tricky bit is zero RPM startup with high torque. And that only works if the HW is designed accordingly.
Seems like properly addressing this would be a core tenant of free and open source don’t you think?
And opposing this in any way would be contrary to free and open source?
You can play around with the voltage sample timings to get somewhat acceptable results.
Or manually insert offset voltages in the algorithm to compensate for said inverter nonlinearities.
Vedder himself said A.S.S. has problems tracking at standstill (because it doesn’t actually track).
All it is doing is lowering the minimum signal-to-noise ratio (i.e. minimum speed) for BEMF tracking to work, but there’s actually a separate phenomenon occurring that he doesn’t seem to be aware of.
Inserting the HFI blip is a “perturb and observe” method of getting the motor to vibrate and spin a little bit, regardless whether it is in the right direction or not.
#justtrampathings
What if you battle harden your motors with red insulating varnish - you can’t not be a fan of that? - it’s littlerly the best stuff around. And doesn’t require a shitton of epoxy it’s the professional solution.
You can test that and see how it affects thermal properties. In my experience coating always impacts cooling.
Well, open source means that the source is open and that is obviously 100% the case here.
As far as I know Vedder was also pretty generous with sharing his knowledge on top of the sources.
Open sourcing things doesn’t mean that you need to be the free of charge R&D department for everyone. People are free to design/ create their own HW variants and test them and finally bring them to the market. And I would say the VESC-Project is a great resource if you want to design your own HW.
How far you can get to zero RPM tracking depends on how good your HW is and that is a combination of controller design and motor amongst other things. We can get close enough to zero to allow a high torque startup without sensors. A little HFI-bleep certainly helps to get all the way down to zero. That’s why the BEAT FW uses the term HFI-Start.
One key stipulation of Free and Open Source is the free replication and redistribution of the intellectual property in question, and that no information critical to the above activity is withheld or intentionally obfuscated.
Intentionally obfuscating the requirements of hardware in order to run supposedly open source software is entirely antagonistic to this. This is what you are doing by claiming other versions do not support A.S.S. and not being clear what is required to support A.S.S.
Not only that, the entire trademark fiasco that happened a few months back is a clear indication that VESC project or someone part of VESC project is not open to the free replication of the IP in question, which goes against the key stipulation of Free and Open Source.
How far you get with zero RPM is how lucky you get on the stats roll that is parameter detection.
Look Trampa, we understand you want to be the best product that exists for this field and you don’t want other people to profit for free from your own investments.
This is completely understandable and makes 100% business sense.
What doesn’t make sense is trying to do this while at the same time claiming that your IP is Free and Open Source.
There is no shame in being closed source. If you make a good product, people will respect that.
A.S.S. is not jet part of the stable release. It is only implemented in BETA FW that can be found in the VESC-Project Forum: Firmware 5.03 | VESC Project
The method to activate it is selecting HFI-Start in the sensor mode slection dropdown. Make sure to write the motor configuration after selecting the correct mode.
I’m not really sure if you know that Vedders runs a forum.
And from what I’ve heard from others around here is that he’s rather inaccessible there.
But, that is not relevant to the discussion. You claim to be a part of a project, so what you do with regards to the project is also subject to the stipulations of Free and Open Source.
The reason why people have an issue with these kinds of things is that you do these things and at the same time claim VESC is Free and Open Source.
If you spun your own closed source product and did these same things, none of us would have an issue with that.
Code is open, numerous HW schematis exist, BETA forum is there, Vedder answered all questions etc. If that is not enough I don’t know what to say.
What Vedder does is not pertinent to this discussion.
what is pertinent is what YOU are doing since you are a part of the project.
I use that exact stuff on my stators. Yes, they don’t bleed heat as fast, but they are a lot more reliable, if a slightly lower wattage.
I think this is the main incorrect assumption here.
My understanding (which may be incorrect) is that phase filtering may or may not be needed at different times during startup. This is why the VESC phase filters are switchable.
I believe at 0rpm a small hfi pulse is used to work out initial motor position with the filters off, then the filters are enabled and bemf is used to track motor position once it starts moving.
@Gamer43 I don’t think recommending people to make hardware mods to all VESC designs is a good idea without really understanding each one, which obviously is not possible without published schematics which you don’t have.
If you could point us to any schematic including ASS hardware then all the arguments here and what Hardware was required would Finnish.
You cant because you took a open Hardware design based on share alike licence, made changes and have not released the source. You can’t come here preaching how the current VESC and it’s features are open source whilst not releasing your own modified schematic for the hardware needed to support these features.
EDIT: i found that vedder has actually posted the MK5 schematics so it is possible to see the phase filters.
I’m glad you fixed the 'harare’s. I was confused as fuck
Edit: you didnt get all of them