EEZ Bench Box 3 - Sequel to EEZ H24005

[prasimix]

I still haven't solved that dilemma. Some say it should blow inwards, others outwards. Both versions seem to have their advantages and disadvantages. I tested both and the BB3 worked well.

[Kleinstein]

For the air flow I am still missing holes in the case so the air can enter / escape. Currently there are mainly some gaps on the front connectors.

With the fan blowing in one could add some input filter if needed to keep dust out.
Blowing in could also provide a little more cooling to future  high power modules like a electronic load or a 4 quadrant regulator.

For the units sold as a kit the user seem to be free to decide.

[jhenderson0107]

prasimix -

I received and successfully assembled my BB3 containing two DCP405s.  Using the factory-supplied firmware, the unit booted and appeared operational, showing no errors.  Unfortunately after updating the firmware to v1.3.2, the unit rapidly progresses through three states (depicted by attached screenshots), then enters Standby mode.  Attempting to wake the unit from standby merely repeats the above failure sequence. 

[prasimix]

OK, and what's happens if you tap and hold touchscreen? Does it wake up or not?

[jhenderson0107]

It awakens, reboots and marches through the three states again, returning to Standby mode. 

[jhenderson0107]

The module in the first slot shows evidence of thermal stress/damage near Q13 (see attached image).  Removing that module from the chassis, the system booted and reported that the remaining module in slot 2 had failed.  Upon rebooting once again the system with just module 2 installed, the system displayed the three-screens originally posted, culminating in Standby mode again. 

[prasimix]

Thermal stress was done at the very beginning by incompetent Bolek & Lolek Enterprise. There was more modules like that, but they passed my testing and I decided to deliver them. Please remove both modules and try to reach Event log and check what kind of message it contains.

[jan28]

Yesterday my BB3 arrived and I assembled it successfully thanks to the list above that warned the IDC cable could be faulty, I payed attention to the cable and that saved the day...

The assembly was nice, instructions easy to follow. My findings:
- My IDC cable was indeed faulty: The cable was pressed in at an angle causing a misalignment (and causing a short). I managed to remove the faulty connector and assemble it (one of the clips broke off but it will hold).
- The TFT display was completely loose from the from panel. It was not damaged so fixing was easily done.
- Missed 2 black screws (ordered 3 modules so needed 2 extra for the 3th PSU)
- Had to tap 2 screw holes of the front panel (due to painting?)

Had no problems upgrading the firmware to 1.3.2: Used a raspberry pi for this. Tip for the firmware upgrade manual: The USB cable needs to be disconnected to power off the MCU and reset it to normal operations. Just flipping the power switch didn't get the MCU out of DFU mode.

I really like the BB3 and the features it has! Especially the logging features are nice for me. 

Things that would improve the BB3 for me even more are:
- There is no visual indication of the encoder step size, will it add 10mv or 1V if I turn it now?
- If I use the user SW to change the encoder step size then the encoder will not respond until the window with the new step size is closed. Can using the encoder close this window?
- It would be nice if the encoder press is just as programmable as the user switch.
- If I change the display view of a channel to Power and Current then all views with graphs change to Power and Current, except the default view (two large numbers shown) with numbers in this case I think the Power and Current should be large and measured voltage should be small where Power is?

On the hardware side: The DCP405 modules are very nice and power delivery and measurements are spot on! 

On the DCM220 I noticed 2 things:
- When powered off a voltage of 1.37V appears on the output if there is no load or just a high resistance (kilo ohms) connected. If you run a dlog on the BB3 it wil even be recorded (I also verified it with a DDM):

Here I the output was of. First it was shorted (0V), short released causing it to rise slowly to 1.37V, short applied again it drops to 0V. Again DDM on the output show exactly the same values. There also is no difference between channel 1 and channel 2.
If I use a 3k3 resistor (matching the internal resistor at the output of the DCM220) the voltage drops to 0.7V about half. The current thru the resistor doesn't decrease so it's not just a capacitor discharge it's something sourcing around 0.4mA. Not a big thing for me, was just wondering where it is coming from.

- When comparing the measured currents from the DCM220 they seemed to be off with pretty significant values (up to 5-10%). After some playing around I found that this happens if the voltage set value is increased, when decreasing the set voltage (and with a resistive load also decreasing the current) the current reading is spot on again. In other words: there is some kind of hysteresis in the current reading.
To illustrate this, I created a small stair case like voltage list:

Running this list with a 14 ohm resistive load gives:

The shown current is at 10V AFTER 8V step. Measuring 650mA, DDM says 700mA

The shown current is at 10V AFTER 12V step. Measuring 700mA, DDM also says 700mA

Why is this happening?

PS. Did I already mention I really like the logging and graphing features?   

[prasimix]

Massive thanks @Jan28 for your feedback. So far, I have opened two new issues: #81 and #82 (need to discuss that with Martin).

The DCM220 is designed as an additional module for powering non-demanding loads (electric drill, battery charging, etc.) and your observations regarding the DCM220 are correct: it cannot lower the voltage below about 1.4 V without additional load. I don't know if that can be done better, perhaps used method to set voltage (affecting directly FB voltage divider with programming voltage) is not optimal. At least we can move up minimum set voltage to 1.5 V (calibrated).

Both current setting and monitoring comes directly from LT3763 and it is all but accurate. In particular, it becomes inaccurate for currents less than 150 mA. The manufacturer himself declares current accuracy of 6%.

[jan28]

Thanks for your fast answer. I'll have a look in the LT3763 datasheet.

If the values are not accurate (which is fine for this modules use case) I would suggest to display them with less digits. In case of the DCM220 current: Always show it as A instead of mA
E.g.: 0.65 A ( or even 0.6A) instead of 650mA. The last one implies an accuracy which is not present. 6% of my 700mA example is 42mA showing a value between 0.66 and 0.74 A OR 658mA and 742mA.

[jhenderson0107]

prasimix -

The log file is attached.

[prasimix]

Thanks a lot Jim. I'll discuss this with Martin tomorrow. So, if I understood correctly problems started with the latest firmware? Did you tried to go back to v1.0 that was originally loaded?

[Andrew McNamara]

Please remove both modules and try to reach Event log and check what kind of message it contains.

I'm assuming it goes into standby because the self-test of the plugged-in modules failed? Would it be possible to just power down the modules and leave the MCU running so an error can be displayed (edit: maybe this is #33)?

As an aside, I suspect another way to get to the log would be to pop the SD card and read it with another device?

[jhenderson0107]

prasimix -

Reverting the firmware back to v1.0.0 restores original functionality.  Sporadically, both DCP405 modules will pass initial self-test and can generate accurate voltages in CV mode when running v1.0.0 FW.  Frequently however, the second module will fail self-test.  See attached log.txt 

Andrew -
I did retrieve the log.txt from the micro SD card using a Windows box and posted it as an attachment earlier in this thread. 

[fuzzoli]

Hi Denis,

First power-up looks great!  The case color is fantastic!  I've been putting my BB3 through it's paces and unfortunately hit a snag.  In testing out any coupled mode (series, parallel, etc.), each DCP405 switches into CC mode and the voltage drops to several hundred millivolts.   It sure seems like there's a short, but now neither DCP405 works even in uncoupled mode.  The CC LED on one module flickers as well.

Swapping #1 for #2 does not help, and leaving only one module populated still yields the CC / low voltage issue.

I have upgraded to the latest firmware.

Tomorrow I'm going go take the BB3 apart completely and look for shorts.  In the meantime, any thoughts on to what might be wrong?

Great job to you and Martin!

Thanks,

-Frank

[prasimix]

Hi Denis,

First power-up looks great!  The case color is fantastic!  I've been putting my BB3 through it's paces and unfortunately hit a snag.  In testing out any coupled mode (series, parallel, etc.), each DCP405 switches into CC mode and the voltage drops to several hundred millivolts.   It sure seems like there's a short, but now neither DCP405 works even in uncoupled mode.  The CC LED on one module flickers as well.

Swapping #1 for #2 does not help, and leaving only one module populated still yields the CC / low voltage issue.

I have upgraded to the latest firmware.

Tomorrow I'm going go take the BB3 apart completely and look for shorts.  In the meantime, any thoughts on to what might be wrong?

Great job to you and Martin!

Thanks,

-Frank

Hi Frank, just to be sure: you cannot get more then several hundred millivolts when both voltage and current are set to non-zero values?

EDIT: before doing anything please check if all fuses are fine.

EDIT2: please also check are both DCP modules calibrated, and calibration is enabled:

[prasimix]

prasimix -

Reverting the firmware back to v1.0.0 restores original functionality.  Sporadically, both DCP405 modules will pass initial self-test and can generate accurate voltages in CV mode when running v1.0.0 FW.  Frequently however, the second module will fail self-test.  See attached log.txt 

Andrew -
I did retrieve the log.txt from the micro SD card using a Windows box and posted it as an attachment earlier in this thread.

I can see multiple issues here that could belongs to MCU and DCP modules. As an next step, please try to remove both DCP modules and check if Missing media is still reporting. This is not firmware related issue, it's probably result of modules stress during transport.

[Andrew McNamara]

Tomorrow I'm going go take the BB3 apart completely and look for shorts.  In the meantime, any thoughts on to what might be wrong?

Maybe check the modules are inserted correctly into the backplane (not off-by-one, etc), and check the relays on the backplane?

[prasimix]

Tomorrow I'm going go take the BB3 apart completely and look for shorts.  In the meantime, any thoughts on to what might be wrong?

Maybe check the modules are inserted correctly into the backplane (not off-by-one, etc), and check the relays on the backplane?

Yes, another test is to plug one module at the time in the third slot where module output is not going thru wiring for coupling in series or parallel.

[prasimix]

I'm assuming it goes into standby because the self-test of the plugged-in modules failed? Would it be possible to just power down the modules and leave the MCU running so an error can be displayed (edit: maybe this is #33)?

That's right, I assigned #33 to next release to simplify debugging task in case like this.

[jhenderson0107]

prasimix -

Removing the modules, the MCU powers up normally, reporting no errors.  Media is recognized (confirmed via System Information screen). 

Installing either module individually or both simultaneously, module self-test will fail frequently (~50%).  Both modules fail sporadically and with identical symptoms.  Typically, only a single module fails during a boot session.  Rarely, both modules will pass self-test.  If a module passes self-test, it can successfully drive 1A @ 20V into a Maynuo 9811 active load.  If a module fails self-test, it cannot be used unless the system is power-cycled and the module happens to self-test successfully. 

The latest log is attached. 

[prasimix]

Thanks for update. Both DCP modules are ready for repairment . We have at least two options: that you try to take a look and find some obvious issue or that we arrange return and replacement. 

[knotlogic]

So, as usual, I'm late to the thread.  I squeaked into the Crowd Supply campaign shortly before it closed and have been meaning to chime in with words of support, but it's been a rough year for me so far.  Maybe not as rough as some others here... but still.

So a big 'Thank You!' to prasimix and the team for all the updates and an extremely well run campaign.  It would have been impressive if it had run problem free, but that you had to suffer all those setbacks makes it all the more awesome.

My unit shipped pretty quickly.  I got an e-mail shipment notification on the 15th, though I'm 12 hours ahead so it was actually sent on the 14th.  And it said the unit had actually shipped on the 13th.... I received it on the 16th.  So 3 days to ship, but that's been my experience with Mouser for all my orders from them.

Sadly it wasn't till Sunday (18th) that I had time to open the box.  Not enough time to assemble it, but enough time to check through everything.  Looks like nothing is missing or damaged, and no components falling off, so that's a good start.  I'm hoping to have time to start building it this coming weekend.  The only thing is that possibly the screws holding the LCD onto the front panel look a tarnished/rusty, but others have already seen the same.

[Ranayna]

Something odd is going on with the second channel of my DCM220.

Without a load, just a DMM, once i get to roughly 11 Volts, i get a voltage fluctuating between the set voltage and about 10.5 volt. When the voltage drops, the channel also switches to CC, resulting in a red/green flashing of the LED. This fluctuation is visible on screen and also on the DMM. If i crank the voltage up to 19 volts or hight, this seems to stop.

The first channel does not show that behaviour.

Firmware of the BB3 is still on 1.0.0, of the DCM220 it is 0.5.

[prasimix]

The thing is that I've managed to repair at least two dozen of DCM220 modules. Most of them had problem with ch2 and I rectify that at the end by replacing IC13 (LT3763) . Perhaps that is one of that modules that passed my testing but not survive shipping. So please check if you can replace it, or we arrange return and replacement.