Preparing EEZ H24005 for crowdfunding

[ddavidebor]

I can't seem to be able to order one from crowdsupply, are they available to be bought?

[LapTop006]

Ok, some debugging later and I found what the fan issue was with my unit.

The fan was wired for the standard PC pinout (pin 2 hot, pin 1 cold, pin 3 sense), but the PSU has pin 1 hot. This particular fan will not run in reverse, once I swapped the hot & cold in the fan plug it started spinning.

[mattlm]

Ok, some debugging later and I found what the fan issue was with my unit.

The fan was wired for the standard PC pinout (pin 2 hot, pin 1 cold, pin 3 sense), but the PSU has pin 1 hot. This particular fan will not run in reverse, once I swapped the hot & cold in the fan plug it started spinning.

Thanks for your debugging efforts, I haven't got back to my unit just yet, but I'm assuming that it probably has the same issue!

So it passed the self test with this fix?

[prasimix]

I did notice a bug with the power button down in the right hand corner for the power options. It beeps when you select it, then pops up the power menu with the various options to select.  If you hit the power button again, the menu disappears and there is no beep from hitting the power button. So it just missing the beep sound.

Please create for this one a new issue on GitHub here that Martin can follow up.

I would like to see more graphing options on the device LCD.   But there is limited space left on the Arduino.

You mean more "graphing" options a top of existing YT-view? (you can toggle main values on main page with gauge icon in the top right corner).

I would also like see dedicated battery charging options and again graphs to show the charging cycle details. It might be just easier to code a front end python application for battery charging functions than using precise space on the Arduino.  Just some things I was thinking about while I was playing around with it.

This is reported as #158. Without additional ADC inputs for monitoring single cell we cannot provide charging of new types of batteries. But, for old one such as Pb, NiCd and NiMH it could be usable. We need some guidelines for all charging phases. Even fast charging could be possible if 10K NTC input for battery temp monitoring is used.

[prasimix]

Ethernet patch cable: I was not sure on the orientation of the plug on the front panel. Until I discovered the marked dot on the circuit boards, so I made sure the colored cables matched them

Please take another look on Fig. 24 (bottom right corner).

[prasimix]

Also, it would be nice to have PDF's of the schematic available as a non-Eagle user to help troubleshoot it.

You can download it from here (it's available on the bottom of Technology overview page too).

[prasimix]

I can't seem to be able to order one from crowdsupply, are they available to be bought?

No it's not available. Crowd Supply announced that they are going to take care of second batch. For anything above that feel free to contact me by PM or email.

[prasimix]

Ok, some debugging later and I found what the fan issue was with my unit.

The fan was wired for the standard PC pinout (pin 2 hot, pin 1 cold, pin 3 sense), but the PSU has pin 1 hot. This particular fan will not run in reverse, once I swapped the hot & cold in the fan plug it started spinning.

Good, that's why is important to check ALL wiring (that include also e.g. 10-pin IDC cable [A1] orientation!) before you start assembling (as suggested in Before you start section).

[Strada916]

I was just playing the other day and found. In parallel mode ie 40V 10A the OPP trips at 155W even if the value is set to 310W. Is this a known bug?

Sent from my SM-A520K using Tapatalk

[prasimix]

Hm, it could be connected to #62, but it's "masked" with OCP trip when OCP is enabled. I tried my PSU (with firmware v1.01) and didn't find anything strange. Vout is set to 33 V, and Iout to 8 A.



Please let me know what is your output settings? For example you cannot set Vout to 40 V and Iout to 10 A because that is 400 W and fimware will report an error.

[Strada916]

This is the unit power up with 310W dialled in. I get the error message no problem.ie can not set beyond 310W.

 First photo is No load. CH2 shows it's in CC and the current shows it's set to 5A, although set to 10A.

Second photo with a load it will go up to 10A, as per setting. However shows only 5A no load but delivered 10A. As per setting.
Is this normal?

Sent from my SM-A520K using Tapatalk

[prasimix]

Second photo with a load it will go up to 10A, as per setting. However shows only 5A no load but delivered 10A. As per setting.
Is this normal?

Yes, that's normal, of better to say works as expected . Please note that new set value (5.001 A) is displayed in different color. That's indication that firmware is doing output balancing because no load is connected. It's a trick needed because channels even nicely calibrated do not have exact output voltage: a few mV is enough that one channel start to sinking current/power due to that difference. And channel indeed can do that up to ~300 mA since it has downprogrammer circuit. Q7 has very limited heatsink and its dissipation is limited by firmware to 1 W. If dissipation became more then 1 W for 5 or more seconds, firmware will start to limit a current and you can see that on the display. It's a bit confusing, and maybe we would do it in other way, but so far it's tested and works just fine.
Similarly, when outputs are coupled in series, you can see that set voltage is reducing but this time with load connected.

[Strada916]

Second photo with a load it will go up to 10A, as per setting. However shows only 5A no load but delivered 10A. As per setting.
Is this normal?

Yes, that's normal, of better to say works as expected . Please note that new set value (5.001 A) is displayed in different color. That's indication that firmware is doing output balancing because no load is connected. It's a trick needed because channels even nicely calibrated do not have exact output voltage: a few mV is enough that one channel start to sinking current/power due to that difference. And channel indeed can do that up to ~300 mA since it has downprogrammer circuit. Q7 has very limited heatsink and its dissipation is limited by firmware to 1 W. If dissipation became more then 1 W for 5 or more seconds, firmware will start to limit a current and you can see that on the display. It's a bit confusing, and maybe we would do it in other way, but so far it's tested and works just fine.
Similarly, when outputs are coupled in series, you can see that set voltage is reducing but this time with load connected.

Ok that clears things.
Thank you. Blowing resistor left right and centre. LoL. Fun Fun Fun.

Thank you Prasimix.

Sent from my SM-A520K using Tapatalk

[prasimix]

I forgot to mention that you can see how output coupling works in this video (that is an older firmware but the principle is remaining the same):

]

[Strada916]

I forgot to mention that you can see how output coupling works in this video (that is an older firmware but the principle is remaining the same):

]

Thank you.

Sent from my SM-A520K using Tapatalk

[Dwaine]

This PSU is absolutely fantastic!!!!!  I never knew having a PSU with a great display and real working Ethernet could be so much fun.   

[ChrisG]

Very nice PSU. Putting it together was pretty straightforward with some small exceptions on crew-hole alignment. Programming to v1.01 went pretty smooth (via my iMac). During calibration it's not immediately clear that only entering the SET button the change is applied on the outputs. (I've blown also one fuse moving from 450mA to the 2.4250A calibration mode..... ) PLEASE MAKE SURE YOU CHANGE YOUR DMM inputs then. Not in possession of a 6.5+ high accuracy DMM I've used my trusted BM869s to do this. I've noticed that Channel 1 calibration values for mV is deviating a lot from Channel 2. Lastly after calibration both channels deviate +1 and -1 mV on the volt range upon to 10V or so. Meaning the same DMM 896s measures then a -1 mV on Channel 1 and a +1mV on Channel 2. I've recalibrated 3 times but with same results. Anybody any suggestion or comment what I'm doing wrong or perhaps a confirmation that this is all well within spec perhaps?

[prasimix]

It's fine that you have a +/-1 mV after calibration. Used ADC has only 15-bit resolution that gives you over a full range (40 V) max. precision of 40 / 32768 = 1.2 mV what is not enough for mV precision. Anyway the planned resolution for this PSU was just 10 mV.

Also please note that mV precision does not make a lot of sense when load is connected and without carefully wiring of remote sensing inputs. The internal resistance between point where voltage sense is connected and output terminals is above 20 miliohms. Therefore voltage drop on e.g. 2.5 A will be 50 mV. I've reported that as #67, and maybe we could add that output "adjustment" as suggested.

[ChrisG]

Thanks Prasimix! Clear explanation. I was not having any load attached (apart from the DMM 869s). I'll read note #67 too now. Again many thanks for providing such a nice and affordable PSU.

[prasimix]

Just a quick note: all leftovers gone except 6 sets of bare PCBs (r5B12 that is used for this group buy). Please contact me by PM or email if you eventually need them (spare parts, building from scratch, whatever).

[all_repair]

I can't seem to be able to order one from crowdsupply, are they available to be bought?

No it's not available. Crowd Supply announced that they are going to take care of second batch. For anything above that feel free to contact me by PM or email.

Thinking to get a level C, but no update on crowdsupply so far.  Any news? 

[prasimix]

I can't seem to be able to order one from crowdsupply, are they available to be bought?

No it's not available. Crowd Supply announced that they are going to take care of second batch. For anything above that feel free to contact me by PM or email.

Thinking to get a level C, but no update on crowdsupply so far.  Any news?

Not really, and don't have any clue what Crowd Supply is going to do.

[Dwaine]

I must be doing something wrong with the firmware updating.   When I download the scpi-parser zipfile from GITHUB and try to load it in Arduino IDE.   It keeps saying that the library is invalid?   

I'm grabbing it from here...
https://github.com/eez-open/scpi-parser

I even tried just a git clone of it and copy it into the libraries folder for the Arduino IDE.   The ethernet library is perfectly fine downloading the zipfile and importing.

[prasimix]

Are you using v1.02 that is the latest released version?

EDIT: please note that SCPI parser is not integral part of the firmware (check this folder).

[Dwaine]

Did a git pull from master.  I fixed the problem.  I copied from the libraries folder to the Arduino library folder.  It worked after that.   I was trying to download the zip files and put them into the Arduino library folder.
The firmware from the git pull is v1.1