EEZ Bench Box 3 - Sequel to EEZ H24005

[prasimix]

Another example: on the single module I found 6 ICs rotated!

You can send me that one if you like.  Good rework practice for my new employee.   

That one was an easy fix, a matter of minute or two. I found that by simply take a look at the module. Modules that ended up in junkyard requires much more work, actually I think that only moving all parts on the bare PCB could help in most of the cases.

[Kean]

Hah!  I'm looking forward to the challenge... 

[prasimix]

I had some problems accessing the forum yesterday so I am sending an update now which is very simple: all the boxes are waiting packed on 5 pallets and it is agreed that Mouser initiates the pick up. I believe it won’t have to wait long.

[danielbriggs]

Excellent work Denis to get it through to this stage; you should be proud (+ I hope maybe the sub-contractor headache can now fade from your mind slightly) as myself and I assume 100's of others will be over the moon at the new supplies 

[Black Phoenix]

Well I've been out for a while, what a rollercoaster this adventure have been...

At least looks like this "nightmare" is almost finished. And I have to say that your communications about the problems put to shame most of the kickstarter's communications were no word is heard from the creators for months.

[Andrew McNamara]

Well I've been out for a while, what a rollercoaster this adventure have been...

At least looks like this "nightmare" is almost finished. And I have to say that your communications about the problems put to shame most of the kickstarter's communications were no word is heard from the creators for months.

I certainly appreciate the communication. Watching other crowd-funded projects, and having been involved in the manufacturing of a commercial product in the (distant) past, I think even experienced designers like prasimix underestimate the pain of outsourced manufacturing. This has been a particularly ugly example (but nowhere near the worst I have heard about). I can only think of a handful of cases where everything just "came together". Particularly when crossing time zones, country borders and legal systems... it is a huge challenge.

That one I was involved in in the past... one of our manufacturing "partners", after years of working with them (which occasionally required lawyers to get what we'd paid for), essentially just decided they would make their own version of our product. Their product wasn't better or cheaper, but they were a big company and we were small, and their brand eclipsed ours, so that was the end of that.

[prasimix]

Hi everyone, one more week passed, and I'm still waiting for action from Mouser.

[MaxZ]

Wohooo congrats to your endurance!

Since you'll probably never order anything again from that great company, what are the plans for the future? Are there any alternative manufacturers?

Best regards and thank you for keeping us up to date!
Max

[prasimix]

Wohooo congrats to your endurance!

Since you'll probably never order anything again from that great company, what are the plans for the future? Are there any alternative manufacturers?

Best regards and thank you for keeping us up to date!
Max

Thanks for asking that. Yes, of course I intend to continue. The real things have yet to happen and I somehow believe that I will be able to neutralize the damage that AssemTec Europe has done to me.
Finalization of the new three modules is underway as well as modifications of the existing ones if the CE / FCC mark is to be obtained.
So I found another company that immediately: a) offered without any obligations and free of charge a test batch of both power modules so I could check the build quality and b) they offered me to assemble all kit versions and completed devices when / if we pass CE testing. It is a local company that has been working for the Italian market for a long time. I visited them a couple of weeks ago when they gave me test modules that we activated right away without any problem. And yes: they have more than 2 people. Nothing like "Bolek & Lolek Bros". Five of them were sitting with me at the meeting. I saw with my own eyes that they have at their disposal two P&P beasts of Europlacer iineo + II.
Because I don't know when (if at all) AssemTec Europe will deliver all DCP405 modules from Poland and since I couldn't even repair a certain part of them without making significant efforts, I already ordered the first series of DCP405 modules from the new company.

[Andrew McNamara]

Thanks for asking that. Yes, of course I intend to continue.

That's good to hear - I'm looking forward to seeing where you go with the BB3. I feared this experience would put you off - even crowd-funded projects that go smoothly often burn out their creators - I don't think anyone appreciates the effort required.

[prasimix]

Thanks for asking that. Yes, of course I intend to continue.

That's good to hear - I'm looking forward to seeing where you go with the BB3. I feared this experience would put you off - even crowd-funded projects that go smoothly often burn out their creators - I don't think anyone appreciates the effort required.

I'm too young to die AND still young enough for some rock 'n' roll
This is what a roadmap might look like:

1. Passing the CE test. It is in progress, I already paid it a lot and I don't want it to fail, and surprises are possible at every step. One of such surprises is mentioned in this topic.
2. Completion of three new modules (MIO168, PREL6, SMX46)
3. Deploy existing modules to create testbed for testing power modules in production (that could be a nice "use case")
4. Development of 2Q / 4Q power module
5. Development of "slave" chassis (without TFT) for reception of up to 7 modules without Mean Well converter or up to 3 power modules (with Mean Well) + 4 other modules. The existing BB3 can become a "master" for one or more "slave" chassis
6. Further integrate the FPGA (in collaboration with the ULX3S team) and define the DIB v2.0 extension of the existing bus to accept faster modules (backward compatibility has to be maintained).
7. In parallel, work will continue on EEZ Studio.

Of course all of this could happen if we manage to establish a certain sale, which could finance Martin and me to begin with so that the above roadmap would not remain just wishful thinking. Everything we do will remain 100% open source. This is of course not a requirement for any other creator who eventually decides to work with us with an agreed revenue share or royalty fee. I’m really looking forward to collaborating with some of the creators of the already known T&M devices especially those that backed the BB3 campaign.

[exe]

Meanwhile I got three broken modules. They look quite descent, I expected something worse.

I'm on vacation next couple of weeks, so they have to stay in the drawer for some time.

[prasimix]

Meanwhile I got three broken modules. They look quite descent, I expected something worse.

I'm on vacation next couple of weeks, so they have to stay in the drawer for some time.

Yeah, but they are more like a can of worms
I would suggest that you do not take for granted that any of the components work, and I think that the greatest chance is to get something from them, that section by section is detached and soldered to the new bare PCB. In that case you should start with bias power (around IC2, IC5, IC6), then power pre-regulator (IC1) and then slowly add post-regulator. For testing pre-regulator simply connect R10 (POST_OUT input) to GND. In that case it should output around +4 V.

[Kean]

I'm definitely interested in getting a couple of the MIO168 modules when they're ready, and probably in future one of the secondary chassis for expansion.

[exe]

Yeah, but they are more like a can of worms
I would suggest that you do not take for granted that any of the components work, and I think that the greatest chance is to get something from them, that section by section is detached and soldered to the new bare PCB. In that case you should start with bias power (around IC2, IC5, IC6), then power pre-regulator (IC1) and then slowly add post-regulator. For testing pre-regulator simply connect R10 (POST_OUT input) to GND. In that case it should output around +4 V.

I trust you on this and I assume that simple problems are already eliminated. I'll proceed with re-populating components on bare pcbs. Thanks for the hints!

[prasimix]

For testing pre-regulator simply connect R10 (POST_OUT input) to GND. In that case it should output around +4 V.

I forgot to mention that you need to apply +5 V to D1 (PRE_OE line), otherwise IC1 will stay disabled.

[prasimix]

Today's update:

• A few hours ago I got a packaging list for shipping! Still, the question is why they need so long  . I believe now I just need to get a call from the courier to arrange boarding.
• As @Ranayna suggested in #283 I opened a new topic titled Building log for "EEZ BB3 Enclosure & Bare Boards".
• Yesterday I shot a short video on how to communicate with BB3 using Node-RED.
  Node-RED seems very interesting to me and has great potential for quickly creating different test scenarios and collecting data so I intend to record some more examples in the future.

[No.15]

This looks fun, sorry I did not see earlier.

I just ordered, hope it helps, looking forward to this

[prasimix]

This looks fun, sorry I did not see earlier.

I just ordered, hope it helps, looking forward to this

Thank you very much! Of course it will help!

[Ranayna]

So, my modules arrived as well Nice touch with including the desoldered LEDs
To be honest, I am now seriously thinking if I was maybe a bit over confident :p There are quite a few parts on that module that will be finnicky to solder, considering that my SMT experience is limited to SO packages and 0805...

Well, I have a decent loupe... We will see

Anyway, a quick optical check of the two modules showed something that I find odd, but I do not have much experience with bigger manufacturing batches.
What I saw is that the two PCBs are slightly different.

One PCB has two tiny holes, one beneath the power connector, the other one near the female pin header. Both very close to the board edge. That same board also has what, without magnification, looks like a tiny via. Looking at it closer reveals a tiny hole in the ground plane, where the solder resist bulges just a little bit, making it look like a via.
Additionally there is also what looks like strike damage, as if something sharp punched a small hole into the ground plane, in the area between the two heatsink screws. There is another small spot with a hole in the resist as well. Both of those could of course be transport damage.
And finally, looking at the vias of that board shows a worse consistency of the tenting.
The other board also has at least one small hole in the ground plane, at a similar location as the first one, but not on the exact same location. So whatever caused the holes seems to be somewhat random.

I assume that boards like these would be fully electrically tested? Otherwise, should whatever caused the holes have hit a trace, might that be a reason why the modules do not work?
I will try to capture a proper image of those holes. If I manage to do that I will post them here this evening.

[exe]

One PCB has two tiny holes, one beneath the power connector, the other one near the female pin header.

That could be tooling holes. At least jlcpcb does that.

I'm too was a bit overconfident with my skills to repair the modules, but at least I don't need to buy BOM for them .

[exe]

2 @prasimix: how safe is it to power up the non-working modules?

[prasimix]

2 @prasimix: how safe is it to power up the non-working modules?

Reasonable safe, but ONLY with current limit! My first step in testing (after visual control which itself can results in finding multiple non-sense and quality issues) is to apply 40 V to DC power input with current limit set to 30 mA. Module don't need to be connected to the backplane and has to draw no more then 20 mA.

Hey, but wait, shouldn't we start talking about this here? 

[jan28]

This is what a roadmap might look like:

1. Passing the CE test. It is in progress, I already paid it a lot and I don't want it to fail, and surprises are possible at every step. One of such surprises is mentioned in this topic.
2. Completion of three new modules (MIO168, PREL6, SMX46)
3. Deploy existing modules to create testbed for testing power modules in production (that could be a nice "use case")
4. Development of 2Q / 4Q power module
5. Development of "slave" chassis (without TFT) for reception of up to 7 modules without Mean Well converter or up to 3 power modules (with Mean Well) + 4 other modules. The existing BB3 can become a "master" for one or more "slave" chassis
6. Further integrate the FPGA (in collaboration with the ULX3S team) and define the DIB v2.0 extension of the existing bus to accept faster modules (backward compatibility has to be maintained).
7. In parallel, work will continue on EEZ Studio.

Nice you're sharing your roadmap ideas! 

Personally I'm very interested in the MIO168 module. The hardware looks capable to support a lot of different use cases: My primary use would be for it to simulate MCU outputs during the 'tweaking phase' of the hardware parts of my builds/experiments. Currently I find myself using my (fancy) signal generator way too often for generating DC voltages or PWM signals (simulating pins of the MCU of the project). The MIO168 has a powerful MCU of its own. That got me wondering: What will be the sample rates for setting/reading the outputs/inputs? And how to avoid skew or frequency differences with other modules/chassis: Most modules have there own clocks and I didn't see a synchronisation/clock/heartbeat pin in the current DIB specs?

Other module ideas I've been dreaming about in anticipation of receiving the EEZ BB3:
- A USB-PD module: More and more USB-PD is being used to deliver serious power to devices and I think it would be nice to be capable of setting PD limits/options and measuring power usages and read error messages easily. This might not need to be a separate DIB module: Maybe it's also possible to use a DCP405 module (power, sense and analog control) and use the UART of the chassis to talk to a external PCB with the USB-PD hardware incl. a small MCU. Software on the BB3 will rerad/set the USB-PD parameters. Or add this to existing module as an extra port and/or piggyboard.       
- A development module: I think there are several people interested in building specific modules. Having a module that has just the optocouplers, EEPROM and breaks out the (isolated) SPI,I2C pins to the front. This would allow for fast and safe tweaking up a prototype module or adapting existing hardware designs to the BB3 without the need for a BB3 specific module fist or the risk of blowing up other parts of the precious BB3... 
- A straight forward DC load module.

Of course I need a slave chassis with al these nice modules to come....

Keep up your very nice work and transparent communication!

[prasimix]

I assume that boards like these would be fully electrically tested? Otherwise, should whatever caused the holes have hit a trace, might that be a reason why the modules do not work?

Nothing about AssemTec Europe is business as usual and i can't tell if the modules have been tested or not. Even if I ask them something like that, it doesn't mean much to me anymore: I've heard too many of their fairy tales. One think is for sure: the basic visual inspection (with naked eyes!) wasn't performed.

I will try to capture a proper image of those holes. If I manage to do that I will post them here this evening.

Thanks in advance for the effort. I am interested in what it looks like and possibly add it to the existing documentation about this adventure.