Future of the inverter kit

[johu]

I am stocking parts for 14 more "traditional" revision 2.1 kits. The plan is to stop selling this version and re-combine the kit comprised of the following components:

3x SCALE-2 gate drivers (https://www.digikey.de/product-detail/e ... ND/7347485)
2x Melexis current sense modules
1x voltage isolator
1x Either V3 main board or CAN only board based on Blue pill

I do consider not running the gate drivers via me as they are an expensive stock item. Instead people can just get them on digikey/mouser etc. The only missing link is the 15V supply voltage for the drivers.

As much as I like the V3 main board as much it is expensive to manufacture because of the low quantity. So I do consider a very simplistic design based on the Blue Pill board that has over current protection, CAN, 6 PWM outputs, resolver exciter and internal analog inputs. So a CAN-only version comprised of through-hole components similar to Damiens Prius board.

So this minimalist kit (sense boards, Blue Pill main board, wifi board) would certainly be sub 100€ with forum-only support.

Please let me know your thoughts on this.

[arber333]

I think it makes sense. I agree wholly on the driver part. However I dont think $25 for dual driver board is that much. I paid like 150€ for my drivers, but i dont mind it since the experience thaught me broken IGBT and car stuck in the middle of the road can be more expensive. As you see in the photo...

I would recommend though you draw an interface board for three of those drivers with 15V supply installed and make it plug and play with the same connector you have on main board. That way people would fuss about drivers the least. They would just buy them and carry on with their project.
The intended assembly:

Yes those interleaved tracks were waterjet cut and painted. They were a lot of work, but in the end no EMI to bother the brain and i havent had IGBT dead since!
Embodied....

[doobedoobedo]

Anything which makes it more affordable I think is a good idea. CAN only may be an issue for some people I guess, but for many it won't be a problem. More options is always good.
The less cash you need to have tied up in inventory can only be a good thing for you too.

[JaniK]

Damiens prius board type openinverter kit for sub 100€

Whats not to like.
Would that too drive stock inverters trough can bus?
Or need own diy power stage?

I would say that a noob like me wants to buy as many parts that fit together in one kit. You have allready done the work to figure out what works. So I would say if 100pack of gate driver go for 16,46/pc - 1.646,72€/100pack

I think you selling them bundled with kits for sub 100€/set
of 3pc would be great value for kit buyer and some stocking fee for you too.

Good to see you thinking of many ways to improve the package. I am so impressed about the openinverter consept, just found it few weeks ago.

[dima]

It might be a good opportunity to include a CAN USB adapter. I recommend (non-endorsed, just my own experience) uCCB https://ucandevices.github.io/uccb.html

$30 CAD (€20 EUR)
https://www.tindie.com/products/lll7/ca ... erter-uccb

[arber333]

It might be a good opportunity to include a CAN USB adapter. I recommend (non-endorsed, just my own experience) uCCB https://ucandevices.github.io/uccb.html

P1010218.JPG

$30 CAD (€20 EUR)
https://www.tindie.com/products/lll7/ca ... erter-uccb

Nice find Dima!
I also agree with through hole components. It is much often for majority of people to feel humbled by SMD. DIP feels them (myself as well) more secure at work...
I think Blue Pill could well be next milestone in inverter development. Simple, easy to implement and cheap spare parts. If we use the new SCALE drivers we can move inverter safeties from brain to drivers. Then whatever brain does, drivers can check or ignore it. Maybe we would only need a single Fault pin that would have functions of MPROT, ERR, Desat Overtemp and UVLO.

I will still wait for someone to try BP with PMSM motor and FOC code before i pass my judgement in full though.

[johu]

Anything which makes it more affordable I think is a good idea. CAN only may be an issue for some people I guess, but for many it won't be a problem. More options is always good.
The less cash you need to have tied up in inventory can only be a good thing for you too.

Yes I'm already doubting that decision. I'll probably do like Damien, bring out precharge and DC switch as outputs and start, fwd, rev, brake as inputs. Also throttle and throttle2. Thats still low component effort but makes it easier to get started.

Would that too drive stock inverters trough can bus?
Or need own diy power stage?

I would say that a noob like me wants to buy as many parts that fit together in one kit. You have allready done the work to figure out what works. So I would say if 100pack of gate driver go for 16,46/pc - 1.646,72€/100pack

I think you selling them bundled with kits for sub 100€/set
of 3pc would be great value for kit buyer and some stocking fee for you too.

Yes it still could be used with stock inverters or home brew power stages, no change in concept here. I'd rather link to digikey for the drivers out of said reasons.
And I won't possibly sell 80€ worth of components for sub 100€! This is a one off "business" so you need higher profit factors for it to have any viability at all. Remember that you can always resort to making PCBs and sourcing parts yourself so you should be ready to pay extra for the service of a kit.

It might be a good opportunity to include a CAN USB adapter.

Nice find indeed! Not sure if I want to include it by default because it also adds support burden and is not strictly needed for operation. I would also link to it however.

[Jack Bauer]

There are a lot of inverters out there these days whose power stages can be re used. I'd support the idea of a simple blue pill based kit. I'm looking at making an attempt at something like this for my Tesla boards. The small drive unit board in particular has tons of space for through hole parts.

[ChazFisher]

Anything which makes it more affordable I think is a good idea. CAN only may be an issue for some people I guess, but for many it won't be a problem. More options is always good.
The less cash you need to have tied up in inventory can only be a good thing for you too.

Yes I'm already doubting that decision. I'll probably do like Damien, bring out precharge and DC switch as outputs and start, fwd, rev, brake as inputs. Also throttle and throttle2. Thats still low component effort but makes it easier to get started.

I really like this option. Being able to get a design going with a simple pot for the throttle input is important, I think.

[johu]

Here is the first suggestion. Not sure if the Molex connectors are practical?
So 7 pins for optional traditional IO, 8 pins for essentials: supply, CAN, motor temp, precharge and DC switch. 6 pins for resolver or encoder. Right now I haven't wired it up for ABZ encoders because I don't think they are any good. (Because the controller starts out blind)

I've kept gate drive and analog inputs on IDC headers but not quite finished there either. I might eliminate 5V from the gate drive header because the on-board regulator is only good for 5W. Then I could combine analog inputs and gate drive on a single 20-pin locking IDC.

I've also read up on the Scale-2 gate drivers. They aren't standalone, you need to add external components off-board for desat, gate resistance and Miller clamping. So another PCB after all, not what I expected.

[Jack Bauer]

Looks cool:) For the connectors I could recommend the Molex Mini Fit type that Tom uses on the simpbms board. Cheap , common and can be crimped with a long nose pliers.

[JaniK]

Looking good and I understand the point you made of keeping stock- a sale now and then -profit.. I have a diy computer case part that I sell few kits a month and I consider selling them as a hobby only because the low margins. Touche..

[johu]

Well, the kits are like merchandise: buying one from me supports the project financially (and motivates me). If I were to sell them with hardly any margin it would drain the life from the project because I'd end up sourcing parts, writing support documents, bundling parts to a kit, buying post stamps, doing book keeping etc. with nothing in return.

* angry rant redacted *

[johu]

Looks cool:) For the connectors I could recommend the Molex Mini Fit type that Tom uses on the simpbms board. Cheap , common and can be crimped with a long nose pliers.

So I checked them out, they are quite bulky. But then, who said the board must be 80x80mm? Might be worth extending it for better accessibility

[johu]

Similar product but with 2mm pin pitch: https://www.digikey.de/product-detail/e ... ND/1634757
Would that be good to handle or too small?

[Jack Bauer]

Would it fit on a 100mm x 100mm pcb? Those are still quite cheap.

[johu]

It would even fit on 100x80

[johu]

Reconsidering standard 2.54mm one row again. That way you can populate simple pin headers, Molex one row connectors or even small screw terminals.

[joromy]

2.54mm one row is king!! Give many options, if you give a little room behind them. Then there is possible to use cage clamp (wago or other)

Especially when doing testing and prototyping, it's a pain to crimp, pull out the pin, crimp again......

[Isaac96]

I have a DMOC power stage and am planning on buying either the v2 or v3 inverter kit.
Which is better?
Or can I purchase a blue pill board now?
I'm planning to buy within the next couple of weeks
EDIT: My motor (Siemens 1pv5135) has a quadrature encoder if that makes any difference.
Thanks
-Isaac