DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)

[prasimix]

Thanks for this explanation, and I'd like to ask you to add it to the issue that you opened on the GitHub. I really don't know what is going here. All my testing was and still is on the Ubuntu, but my colleague Martin works with Windows and didn't report anything similar neither. The only change related to the USB port is announced in firmware v1.01 when we switched from programming to native port following the suggestion from #135. Another benefit from this is that uploading speed is increased 8 times! But, if someone for any reason wants to stick with programming port it's still possible by adding the following code into your conf_user.h file (that is mentioned in assembly instructions):

Code: [Select]

#undef CONF_SERIAL_USE_NATIVE_USB_PORT
#define CONF_SERIAL_USE_NATIVE_USB_PORT 0
EDIT: Maybe in next release we should add jumper sections for SPU and SPD signals of the ADuM that are now tied to the GND.
EDIT2: SPU and SPD are currently tied to Vdd (HIGH speed) not GND (Thanks Jleg).

[jleg]

Thanks for this explanation, and I'd like to ask you to add it to the issue that you opened on the GitHub. I

Sure, will do...

really don't know what is going here. All my testing was and still is on the Ubuntu, but my colleague Marin works with Windows and didn't report anything similar neither. The only change related to the USB port is

Well, as said - it doesn't matter if it's Windows or Linux. If your USB hardware (port on the PC which is used) is 'compatible' with the Adum, you'll never notice any problem.
If you tryout however half a dozen PCs and Laptops of different age, there's obviously some percentage which will fail. My percentage is at '66% failure rate' (3 systems)   

EDIT: Maybe in next release we should add jumper sections for SPU and SPD signals of the ADuM that are now tied to the GND.

That's exactly what i also had in mind, since i noticed in schematics that those pins are on Vdd. Unfortunately i cannot simply test this with my external Adum module, because it's not easy to open without a mess... and i'm not sure if it's a good idea for me to mess with the EEZ Psu pcb, trying to scrap some Adum pins... (careful lifting?)

The speed won't matter for Scpi imo, and a fw update taking longer is imo also not an issue, given that this may happen every few months or so. Unless you're developing, of course...

[Pjoms]

Hi all!

First I have to say that I'm really very impressed of the work that has been put into this project!   
I have assembled my kit, updated the firmware, everything seems to work just fine and I’m happy.


I was playing around with the remote programming input, and noticed a strange behave. Maybe the reason is already known and obvious for some others, but not for me at that moment... Maybe a little better now, but not really...


The setup was like this:
CH1 voltage is controlled by remote input from a sinus signal generator between about 200mV to 1.8V.
Output voltage goes between about 2-30V.
Current limit @500mA.
Psu loaded by a electronic load.

Remote input signal and psu output voltage are simultary measured by a scope.
Current measured externally by a Fluke DMM, and also displayed on the electronic load.

And here it comes:
@200mA on the Fluke and the electronic load, psu shows 70mA!
@500mA on the Fluke, psu shows 172mA
@1000mA on the Fluke, psu shows 341mA
The CC @500mA has not became active yet...

I'm now starting to scratching my head.

The cause:
I measured both remote in and psu voltage out with the scope, with the scope gnd on the probes connected to "its own" negative terminal.
One on remote input (pin 2, analog gnd on push-in), and one on CH1 negative output.

The scope gnd connected those two together making the "missing" current to go beside the current measurement through the analog gnd!
Luckily it looks like everything has survived, as I don't think the analog gnd is capable to handle that kind of current for any longer time.

Disconnecting one of the scope probe gnd solved the problem.

I don’t know if it’s possible to get rid of this issue, or if it’s just something that’s have to be beware of and kept in mind.

/Anders Westman

[prasimix]

Your observation is right. Please note the following warning stated in Building instructions in Section 1.1:

During testing and taking measurements, please note that a channel’s negative output (OUT-) is NOT at ground potential with regard to the mains earth potential in the AC wiring. If you are using a multichannel oscilloscope which does not have isolated channels (the default!) you cannot simultaneously connect the test probe ground of one channel to a PGND point and another channel to an OUT- point. This will interfere with normal operation of the current control loop (IC6A, IC7) because the current sense resistor (R63 or R65) will be shorted out; as well, there is a real risk of damage to either test probes or oscilloscope circuitry, or both. Short of exotic and expensive oscilloscope probe accessories, limiting connections to one probe at a time is wise.

This is a disadvantage of low-side current measurement, so you have to keep that in mind.

[Pjoms]

Of course you have already thought about and documented it! Great!
As you say, this is an obvious downside with low side current measure.

I was thinking if this could cause some problem in some other situations too even without a scope involved, e.g. if the psu is feeding some circuit that also generate the remote control voltage.
In a such scenario I believe it could be an easy made mistake to connect the remote input ground to the channel negative output thru the external circuit causing a major interference in the current control in the psu, and potentially high current to flow the wrong way thru the push-in connector.

The interference in the current control with a much lower current measured by the psu than actual, also cause that most of all nice protections like CC, OCP, OPP will not work as expected.

Would it be possible to put a resistor in serial with the remote input ground to increase the input resistance and limit the current thru the push-in connector to a safe level that also not interfere with the current control.

[prasimix]

My first thought is that as far as channel output are floating (i.e. minus is not connected to Earth) and input signal ground is not earthed that nothing unexpected can happen. But, I have to think again about it .

[s8548a]

Hi,

I have got a notification that my kit just has been arrived at local post office with a whopping $195 customs duty!

btw the original packing has gone it seems and the box looks very small.

can some one post a picture of the package when it arrived and how much weighs it?

Thanks.

[garnix]

I was surprised as well how small the package was.

[s8548a]

Thanks @garnix.

The box resembles like as you have shown.
i will get it on monday.

[Aigor]

Another unit wen well no problem at all amazing project

[iainwhite]

I built my psu up this week, and also made up a 1 Ohm power resistor to use for calibration (using a scrap AMD CPU heatsink).

Very pleased with it.

[biot]

Another happy EEZ H24005 user checking in here. Great job, prasimix!

[mkudlacek]

Hello Prasimix,
 I'm sorry I have to report a faulty PSU. After putting everything together I flipped the front panel switch and nothing happend. After double (more like triple) checking nothing seemed to be out of order. So I started to trace power on aux PSU board and it seems I have faulty Vigortronix AC/DC module. The 230V is on the input pins, but the outputs are nothing like 12V a 5V. Check the screenshots from my oscilloscope. The first one with big wave is from 12V pins, the second one is from 5V pins. Both measurements are taken with the front PCB completely disconnected. What should I do?

[prasimix]

Did you measure that with or without Arduino shield board connected?

[mkudlacek]

Without Arduino and front PCB connected. Only the DC power supplies were connected to the AUX PSU.

[prasimix]

Connected AC/DC modules are not import for this. I could try to contact Vigortronix (manufacturer of TR1) and see if they have any idea what's going on.

[mkudlacek]

I can unsolder the Vigortronix module and test it outside of the AUX PSU. Thus we can determine whether the error is in the module itself or some combination error with the rest of the board. Will I be able to do it with ordinary microsoldering gun?

[s8548a]

Hello Prasimix,

I just finished assembling of my PSU kit and everything went smooth.
a BIG thanks to you for this great work done 

successfully updated the firmware to latest version and finished the calibration.
btw am not satisfied with the calibration, may be because of my DMM's quality( only the 1 ohm resistor is enough right? i have read something about 3.3 ohm resistor somewhere).

1-am not able to establish a serial communication with PC for getting a debug trace(*IDN returns nothing), only the programming port is working, native port is code10 as mentioned earlier in this thread.
   and when i connected the usb cable directly to DUE's port, the PSU is always ON irrespective of the standby switch position, is that normal?

2- and just noticed that when the outputs are OFF, there is around -400mV present on the outputs and if i changed the current to non zero, it reduces to -50mV, is my PSU is ok?

[prasimix]

I can unsolder the Vigortronix module and test it outside of the AUX PSU. Thus we can determine whether the error is in the module itself or some combination error with the rest of the board. Will I be able to do it with ordinary microsoldering gun?

Sure, especially with some help of desoldering braid.

[prasimix]

successfully updated the firmware to latest version and finished the calibration.
btw am not satisfied with the calibration, may be because of my DMM's quality( only the 1 ohm resistor is enough right? i have read something about 3.3 ohm resistor somewhere).

Everything below approx. 4 ohm should be ok. What kind of problem your have?

1-am not able to establish a serial communication with PC for getting a debug trace(*IDN returns nothing), only the programming port is working, native port is code10 as mentioned earlier in this thread.
   and when i connected the usb cable directly to DUE's port, the PSU is always ON irrespective of the standby switch position, is that normal?

Yes, that is normal since Arduino Due is in that case powered from PC (over USB cable).

2- and just noticed that when the outputs are OFF, there is around -400mV present on the outputs and if i changed the current to non zero, it reduces to -50mV, is my PSU is ok?

Yes, it is.

[mkudlacek]

I can unsolder the Vigortronix module and test it outside of the AUX PSU. Thus we can determine whether the error is in the module itself or some combination error with the rest of the board. Will I be able to do it with ordinary microsoldering gun?

Sure, especially with some help of desoldering braid.

I can't get it out, even with the desoldering braid. I ordered a rework station, I'll post a result as soon as I'll get it, hopefully this week.

[prasimix]

Hm, it shouldn't be so difficult, but ok I cannot judge of your (de)soldering skills. I did it few times in past with no hassle, just ordinary soldering iron, desoldering braid and eventually a few drops of flux.

[mkudlacek]

Ok, so it took a little bit of persuasion, but the module is out. I hooked it up to scope, but the result is the same. No 5V and strange 11-18V wave on the 12V output. Hopefully the 18V didn't do anything bad to the rest of the power supply.

[prasimix]

Thanks, I'm going to contact Vigortronix tomorrow.

[reportingsjr]

Wow, I'm really sad I missed the crowdfunding of this! What an amazingly well done piece of equipment.

If anyone who has one wants to sell, send me a PM!!!!!

Prasimix, is there a list to join just in case you do end up deciding to do another run?