Gertboard

[Gert van Loo]

Neon22 said:

Have you looked to see if RS has the more powerful versions ?

MCP4802/MCP4812/MCP4822 or even MCP4801/4811/4821

Maybe the cost difference is in your favour ? I think pinout is same

Thanks, I'll check.

[Gert van Loo]

Yes, 4812 is available. I must be getting old not to have thought about that.

glad to help. (and now may get better code that allows for these other options )

its win win

Seriously thanks for doing this Gert. Its fantastically useful.

Do you think RS would be interested in supplying the components as a kit directly, that way it'll make life easier later on when you start producing PCBs etc (only need to sell the PCB from the shop, the rest can come direct - although may not work out too well for P+P I suppose).

Could be worth seeing if they are interested? Or Farnel/Maplin too.

Great to see the progress!

[Gert van Loo]

meltwater said:

Do you think RS would be interested in supplying the components as a kit directly, that way it'll make life easier later on when you start producing PCBs etc (only need to sell the PCB from the shop, the rest can come direct - although may not work out too well for P+P I suppose).

Could be worth seeing if they are interested? Or Farnel/Maplin too.

Great to see the progress!

That has been suggested before and I some of the mention companies seem to offer such a service. I will address that in due time.

Having supplied equipment to RS and Farnell in the past I don't think they would be that  interested in doing kits.

You'd be better off speaking with Adafruit, Sparkfun or OOmlout or someone like them.

Sparkfun will often take a project on, info here

http://www.sparkfun.com/tutorials/315

[Gert van Loo]

Gertboard rev-2 arrived today. I have not had time to test it yet. I already found the first error when trying to plug in some components: the DC power socket I have has an end pin which is slightly wider then the two side pins. Applying a vice to that pin solved that for now.

The silk screen at top and bottom is totally wrong but I blame the manufacturer for that. I send them an email and await their reply.

I also did a quick verification of my choice for 12 MHz operation at 3V3. Officially it is out-of spec. There is a graph in the data sheet which shows 8MHz at 2.7V and 16MHz at 4.5 Volt or higher. There is straight line between the 8 and 16 MHz so that line has the equation F = A * V + B.  Using the two data points provided and solving the equation (That is what you learn maths for) gives: F = 4.444 * V - 4. So for V=3.3 I get 10.666 MHz not 12 MHz. I will keep it at 12 MHz as it is a nicer number and just rely on the design margins to compensate for the extra 0.3V I would need.

Great news Gert:).

You probably don't like this question but I wanted to ask: for now, when you have ver.2 in your hands ...

1) what is approximate date when we can byu Gertboard?

2) what is your final decision about soldering? (how much will be on pcb? how much we will have to solder at our homes?)

Thanks:)

Arthur.

[Gert van Loo]

There are no plans to produce this rev-2 board completely assembled. The board has been designed for manual assembly. Project schedule:

Board test : 2 weeks. (Providing nothing comes in between)
If no redesign required: Order 500-1000 boards: production time 3 weeks (That is to keep cost down)
So in about 6 weeks you might get your (bare PCB-only) board.

You have to solder it ALL at home. But the board is now down to ~20 SMD parts. (Was ~50).

I am thinking of re-designing it for ready assembled but before I go to all that trouble I need to find a way to produce them at reasonable cost. Even if that kicks-off it will take a long time before such boards are available (Look at the time it took the foundation to get it's boards made and that was with a team of people. I am alone on this.)

Post edit: stuped tiepo's

Gert said:

There are no plans to produce this rev-2 board completely assembled. The board has been designed for manual assembly. Project schedule:

Board test : 2 weeks. (Providing nothing comes in between)
If no redesign required: Order 500-1000 boards: production time 3 weeks (That is to keep cost down)
So in about 6 weeks you might get your (bare PCB-only) board.

You have to solder it ALL at home. But the board is now down to ~20 SMD parts. (Was ~50).

I am thinking of re-designing it for ready assembled but before I go to all that trouble I need to find a way to produce them at reasonable cost. Even if that kicks-off it will take a long time before such boards are available (Look at the time it took the foundation to get it"s boards made and that was with a team of people. I am alone on this.)

Post edit: stuped tiepo's

Great answer .

So I see that at the beginning of may Gertboard can be available.

What about electronics parts? Will be in package or we have to buy it alone?

Do you have any site or blog where we can read about your works, see photos of your progress or this (this forum) is only way to read about it?

Geertings ,

Arthur

[Gert van Loo]

I will post schematics & parts list. I am not planning to make order lists for the various vendors. Not because I am lazy but because it takes a lot of time and anybody can do it once the schematics & parts list are out. (Also any list will get reactions of 'I know where to get it cheaper') I have too much to do already. e.g. I already spend many, many hours on the board assembly manual.

I have no blog, no facebook or twitter account. No wish for any of that and too busy with developing all kinds of HW. I was planning to make a picture this weekend and show it, but the silkscreen (production??) error make it look awful.

component kits for the gertboard or any other board relating to the raspberry pi should be discussed with various retailers/supplyers to help bring the costs down… a plan that would work well I think…

edit: as in I would go to my local supplyer and ask if they would help with making component kits.

Gert said:

I will post schematics & parts list. I am not planning to make order lists for the various vendors. Not because I am lazy but because it takes a lot of time and anybody can do it once the schematics & parts list are out. (Also any list will get reactions of 'I know where to get it cheaper') I have too much to do already. e.g. I already spend many, many hours on the board assembly manual.

I have no blog, no facebook or twitter account. No wish for any of that and too busy with developing all kinds of HW. I was planning to make a picture this weekend and show it, but the silkscreen (production??) error make it look awful.

OK, I see that you are a very busy man - I understand you very good.

I think many of ppl will want to buy your PCB and if you will publish parts list we can start to collect items .

If there is no official Gertboard site then I hope you will post here an announcement about PCB and a list (or it will be on main raspberrypi.org site).

Thank you!

There are services, who do all the work for small batches and it is not too expensive:

http://www.seeedstudio.com/propagate/

and there are other services to be sure that you really have the required number of buyers to be profitable, if it gets expensive:

http://www.indiegogo.com

Maybe we should let Gert to prepare a sellable PCB (+elements list) ... and then help him to extends project into kit (to solder at home), then semi-soldered kit and so on...

http://www.indiegogo.com looks very interesting, I can be first for Gertboard PCB .

[Gert van Loo]

Status update: One Gertboard rev2 has been assembled and I just spend some time repairing the SD card on the damaged Beta board. It's alive! Just some pictures of the SD-soldering the damaged GPIO connector (Which I still have to bend back) and the assembled rev-2 board.

[Gert van Loo]

Basic buffer I/O seems OK but I have not yet tested every of the 12 buffers in each direction. Switches seem to work too. I just managed to get the two-channel Digital to Analogue converter working. See scope picture below.

kudos on getting the beta board working again

how long (as an estimate) till the boards are ready for public consumption?

[Gert van Loo]

If will be about 4 weeks after I finished fully testing the curren tboard. That is moslty to keep the price down. If you want PCBs sooner the manufacturer price goes up rapidly up to £50 per PCB for 2 working days.

I will also need some time to finish the manual.

p.s. ADC is also working!

ouch, worth the wait then

I am registering my interest now and ready to pre-order. 

Who will be the distributor?

[Gert van Loo]

These are bare PCBs only thus much simpler to produce or sell. I talked to Liz and they will probably go trough the raspberrypi.com website.

raspberrypi.com has been down from just before the r-pi launch. Hopefully, this means it is going to be back and functioning soon. Is the plan still early-mid April availability for these PCBs?

Is there anything the community could help you with? Given the interest in the r-pi, I hope the Gertboard release is not a replay, albeit at a smaller scale, of the r-pi day. My hunch, only a hunch, is that the Gertboard will get the level of interest from hacker/techie types that liz,eben and co were originally expecting for the first run of the r-pi. Perhaps using eBay in conjunction with raspberrypi.com will alleviate the load.

If there is only 5% of people interested in a Gertboard  per million RPi sold…. hackers/techies/schools/robotics programs/ DIY’s/mad scientists/and of course --- Dr. Evil!

That will be a lot of boards…

[Gert van Loo]

I expect more like 1%. With 10.00 boards out that would be 100 boards.

I have the ADC and the DAC working. The six open collectors (ULN2803) work.

The only thing which still needs testing is the ATmega chip. At the moment I am stuck with that. SPI  interface works and I can read the fuses and blow them. But as soon as I program the debug-wire fuse the system dies. I checked the reset/DBG pin and I can't see anything wrong with it. Correct 10K pull up to 3V3 in parallel with the diode. If I use 5k to pull the pin down it goes to half voltage so there really seems to be a 10K pull up only on that. (So I did not solder the diode in the wrong way). I don't dare program that fuse so for now I am loading test images using the SPI interface. I just ordered 10 more chips to have some back-up as I am on my last still-working chip. I assume I can revive them as soon as I know what is wrong.