Awesome(!) Raspberry Pi Hat on Kickstarter

[Jelby-John]

I made it because I wanted one and it didn't exist. It's not new nor clever but it does what I wanted it to. I'm also asking for quite a lot because I need them made as I don't have the space for a reflowing setup yet. Perhaps if there isn't enough interest, I'll relaunch in the Autumn for less and make them literally 'in-house'. Anyway... {EDIT: It is an extension board for the Pi which adds 16 channels of PWM and 8 channels of ADC over I2C}

The Hatalogico for the Raspberry Pi on Kickstarter

I made a couple of mistakes on the prototype which I will fix if the campaign is successful:
* I used different sized vias. I didn't mean to, it just happened.
* I should have used 5V to power the chips so it could be easily used without a pi.
* Not sure if the 5V power-in has beefy enough tracks. They haven't melted yet...

I intend to make a few other tweaks too if I get chance, but even if it fails, I still have my (rather expensive) prototypes. 

There is also the Ardhat on KS at the moment which also looks good (I'm backing that also) but that is essentially an Arduino on a Hat. My first prototype was exactly that (without all the bells and whistles) but decided I didn't want two microprocessors. Anyway, check that campaign out too. Ardhat

Since the Onion just got a fairly raw deal from you guys and I thought/think it is a great campaign and product, I'm expecting the worst... I think I can handle it. 

Be gentle?
John Lumley

[Tac Eht Xilef]

Maybe you could tell us what it is?

All you've told us in your post above is that it's (a) not new, (b) not clever, and (c) not an arduino anymore. Hardly encourages anyone to go looking at the kickstarter page, does it?

(Sorry, it's a small dig, but a remarkably common mistake.)

[Jelby-John]

Fair points...
@cleaningOut: I can see how it might be confusing, will see if I can change it. Thanks for the input.
@TacEhtXilef: I can't believe I didn't tell you what it is. I was too busy telling you what it is NOT. 
@mojo-chan: It just adds dedicated PWM and ADC to the Pi. So, yes you could drive motors/servos, dim LEDs, etc. And on the input side, you can use LDRs, pots, hall effects etc.

So, it's a simple project to add a PCA9685 and 2x ADS1015 on top of a Pi so you can easily access them via I2C. It also level-shifts the comms. That's it in a nutshell. I think I mention Arduino too much in the campaign but it is not an Arduino at all, it is so I don't NEED to use an Arduino with a Pi (which I've done many times).

Sorry for my poor details, I was really nervous telling you lot about it. I love this place but it's a scary place to float your ideas/projects! I want honest and expert feedback though, any advice/tips/constructive abuse is genuinely appreciated.

John

[Jelby-John]

new banner makes much more sense.

Yes, thanks for the suggestion, very useful.  Others close to me have told me that AFTER you. So thanks.

Promoting this has been difficult. I got plenty of backing when Kickstarter were featuring it as a new project but since then, I've resorted to Twitter and making YouTube videos to get the word out. I hate marketing/selling.

A guy I met at a RaspberryJam (maker-type get-together) is going to take a board and some flyers to a faire in Madrid at the weekend. And I'm hoping the video series (and relentless tweets about it) will get some interest. There are plenty of Pi users but I need ones who do electronics too and that seems to be a small percentage. If anyone has any tips on promoting, please let me know. Please!

I sent a board to Dave about 3 weeks ago which I hope got there in one piece. He always has so much stacked up on the shelf though, I can't see my little parcel!
I also sent boards to the Raspberry Pi Foundation (since I follow their Hat spec) and Adafruit (because I like them).

It's aimed at a beginner level (it's not really aimed at all tbh), but if anyone is interested in the faire project: video playlist

I really appreciate the input, thanks. 

[Jelby-John]

Big thanks to Dave for mentioning it on his latest mailbag! I'm not sure if he was calling the 'Hat' name itself stupid or my board name, either way he is right I guess.

Two things to note:
1) It is open-source or at least I think it is. I have uploaded the design files and I have no intention of suing anyone... I have not registered it anywhere as open-source but other than that, I think I'm just missing the little logo.
2) I have (preliminarily) decided to drop the two DACs from the final board (if it's successful). Too niche for multi-purpose use IMO, am I wrong?

THANKS DAVE! <- If 'new' David sees this, please pass on my thanks! I don't imagine the big guy has time to follow the CF threads that don't rely on pseudo-science!

[Jelby-John]

Thanks for the input mojo-chan. Firstly, no I haven't posted on the Pi forums because ... well, my excuse sucks so I'll just take the advice and post in the best topic I can find (no Crowd-Funded one like EEVBlog).

And I know what you mean about the negativity, I have learnt a few things about myself since launching the campaign but the main one is that I am really bad at selling stuff. In engineering, I trade in honesty/'whole-picture' and it's foreign for me to focus on the good and omit the bad. It feels inherently dishonest.

And with that honesty... I'm pretty certain that it's simply too expensive. I think it needs to be around the 15GBP/22USD mark to get enough people wanting it. I have had nearly 800 views of my KS video but only 29 backers. That suggests to me that there is interest but they are being put off by something. Perhaps by my negative language (I don't deny it's an issue), but more I think it's the price.  For the record though, I've left very little profit margin. This was never going to make me much money because I couldn't stuff and reflow the boards myself so would have to pay a good chunk to get them 100% fabbed for me.

Obviously I still hope the campaign succeeds and I have 100% confidence in my ability to delivery on time if it does so, but I fear I'll have to return to the drawing board and work out how to get the price down if I want to relaunch.

I really appreciate the input, this has been a great learning process and an eye-opener on many levels.
Thanks, John

Oh and I'm working on a little YouTube project of modding an RC car - ripping out the board and replacing it with a Pi + Hatalogico + perfboard w/ H-bridges on it. Will control via Bluetooth and a WiiMote I think. I'd like to add more to the car to use the ADCs and more PWMs but time is against me. The campaign will be a week from ending by the time I get these onto YouTube. 

[dadler]

I am one of your backers and think this is a great idea.

I do think you are too self-deprecating. Even the title of this topic "Another Taspberry Pi Hat on Kickstarter" implies that this board is nothing special, when it is very useful. The price is a little high, but certainly reasonable (in my opinion). But this is the sort of thing solved through economies of scale. I think people will back you because nothing else like this really exists (and they want it to exist), not because the price is "right". You are in a tough situation, though, due to the extremely low price of the Pi--I think there is some curious consumer psychology at play here. You get an entire working computer for $35, so an add-on that costs about the same will need to be perceived as providing about the same value as the $35 base product. I think your product does, as my time is valuable and your product makes the Pi much easier to use. I think you should drive this point forward as a primary selling point, as it is the main reason I backed your campaign (the first thing I have ever backed).

On a side note, I personally would use the DACs (for audio synthesis), but I don't know that I would use them at the same time as the ADC/PWM channels. I think it was wise to remove them. In the future, perhaps you can offer two boards, one DAC/ADC focused (audio/data logging/other analog), and one PWM/digital focused (for robotics, copters, other digital stuff). Just a thought.

[Jelby-John]

Thanks to both,

daddler: Massive thanks for the support on Kickstarter, it's very much appreciated. And yes, you're right, this is a terrible thread subject! I refer you to my earlier commenting about me sucking! As for the price, I think you hit the nail on the head. I don't think my board is super expensive, it just looks it compared to the Pi. The 22USD price-point I mentioned in the last message is IMO a very attractive price but the only way I can achieve that is to make them myself in quantities over 500 (and at 500 there is no real profit).
In terms of different versions of the board, I know what you mean but I was trying to include only the most multi-purpose functions and leave room for expansion. Therefore, to get DAC focused board you mentioned, you could make a little board with a couple of MCP4725s on to sit on the expansion port. If you build a similar board with a H-Bridges and 9DOF, you have a handy robot platform. I SHOULD have made all of this clear in the campaign and provided example expansion boards. But I didn't.  Either not enough hours in the day or I'm too impatient.

mojo-chan: I will be posting on the Pi forums later today (with much more positive language) and will keep my fingers crossed. As for posting on here, it's probably the most useful thing I did. I sent the board to the mailbag as promotion but I wasn't expecting backers from this thread, I hoped for constructive feedback and advice and I got that in spades.  Wise heads live here. 

I'll post back on the Pi forums outcome as I figure this thread could be useful for others launching a Pi-related crowd-funded project. A learn-by-my-mistakes guide.

[dadler]

I was trying to include only the most multi-purpose functions and leave room for expansion. Therefore, to get DAC focused board you mentioned, you could make a little board with a couple of MCP4725s on to sit on the expansion port. If you build a similar board with a H-Bridges and 9DOF, you have a handy robot platform. I SHOULD have made all of this clear in the campaign and provided example expansion boards.

Ahh, I hadn't explicitly thought of it that way. I guess that is the trick-you need to do all the thinking and explain options/approaches in the campaign. Like a bullet point list of Hatalogico+DAC=example audio synthesis project, Hatalogico+CAN/I2C transceiver=embedded automotive logger, etc. Give people lots of ideas that may pique interest in current or new projects.

[Jelby-John]

Hi daddler,

I forgot to also explain that the expansion can either be a small board to sit within the white-dotted area of the board, or it could be the full size and utilise the full dual-row male pins like this video: https://youtu.be/TbgEcstdr0s?t=14s << This video only uses two of the three dual-row as I didn't need the communications for this small project.

The small breakout area just has 5V, Gnd and 5V I2C pair. I thought this would be useful for things that don't require any extra connections: RTC, Temperature/Barometer/Altimeter, 9DoF, etc. But if you need several of those things, then a full-size custom board on top would be a good way to go.

I'm trying to make a list like you suggest and will update the Kickstarter page and use it on the RPi forums when I'm happy with the wording.

Thanks for the input. 

[Jelby-John]

Well, I posted to the Raspberry Pi forums but they immediately moved the post into the 'For Sale' topic where it received scant attention. 107 views to-date (https://www.raspberrypi.org/forums/viewtopic.php?f=93&t=108381&p=745651).

I was however, contacted by LetsMakeRobots.com who are setting up a marketplace and actively seeking sellers. They have organised a few reviewers for the board and would like it featured in the store if the campaign is successful. Robotics is a natural fit so am quite pleased about it but I think it could be too late for this campaign. I may have to relaunch immediately if the reviews are good.

Quick question: I've been told that - especially in tech - the last 7 days is when most pledges are made on crowd-funding platforms. Does anyone know if this is a myth (or better, have some stats to view)? But either way, I'm not pinning my hopes on finding the last 80% in the last week!

[Phoenix]

a PCA9685

What made you choose this IC? Seems a bit limiting to have a 1kHz PWM frequency (maybe a few kHz with external oscillator) if you're intending it for small motor control.

[Jelby-John]

Pheonix,

Six reasons:
1) Price, in bulk I could get them for under 1 GBP.
2) Widely-used with software ready-to-go.
3) Ok, 2 reasons.

I knew you could add an external oscillator but I did not allow for an unpopulated footprint on the first prototype (there is a TH-pin for it though). To be honest, you're the first to question me on it and frankly, I hadn't really thought enough about it.

I hadn't intended it for use with small motors any more than any other use, I was trying to keep it as flexible as possible. But I do absolutely want it to work well with small motors. I actually thought the variable frequency was ideal for this. I appreciate that there might be some whine at this range but switching beyond 20kHz is beyond many/most components surely? Isn't it? That seems to be very fast switching but I could be - and frequently am - wrong.

Is there a chip you'd suggest as a possible replacement? I've got Prototype 2 complete/drawn but am not sure I will actually get it made as I've already thought of a couple of things for prototype 3. Suggestions are very welcome.

Thanks for the question, 3 weeks of answering folk's questions and I hadn't heard that one before. Refreshing. 
John

[Phoenix]

I hadn't intended it for use with small motors any more than any other use, I was trying to keep it as flexible as possible. But I do absolutely want it to work well with small motors. I actually thought the variable frequency was ideal for this. I appreciate that there might be some whine at this range but switching beyond 20kHz is beyond many/most components surely? Isn't it? That seems to be very fast switching but I could be - and frequently am - wrong.

Perhaps I'm interpriting the purpose incorrectly. It seems that many electronic speed controllers want a PWM input in the range offered by the PCA9685 - so it's just fine. The speed controllers will then generate their own PWM signal (or whatever is required for the motor) and frequencies of 100s of kHz here are not unreasonable.

My mind was thinking of using the hat to PWM the motor directly, probably not its intended use.