Faulty Connexelectronic SMPS2000R [Resolved]

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi all,
I ordered two SMPS2000R power supplies to power my Tripath TA0105A amps. One works OK but the other was dead on arrival. I have tried everything to get help from Cristi, but to no avail. I've even recorded videos of my attempts to debug the circuit and posted it on YouTube. Cristi tells me that video sites are blocked where he comes from, so I hosted the video on my site. Apparently that didn't work either...so I cut the video into 5 and emailed them, but that didn't work either...

Next was Skype - but you both have to be on at the same time for that to work...now all attempts to communicate via email have gone unanswered.

Two questions:
  1. Has anyone else had this problem with Connexelectronic?
  2. Any ideas how to debug the SMPS2000R?!

Thanks!
Gareth
 
Video here: Testing the Connex Electronic SMPS2000R - YouTube

This is what's happening:
  1. Checked all fuses on board with continuity meter - all fine.
  2. Connected the board to the mains (240V) using the Live, Neutral and Earth wires with 13amp fuse.
  3. Turned on power from mains and POWER LED turned on
  4. Measured +ve rail to GND: 0V
  5. Measured -ve rail to GND: 0V
  6. Measured +ve AUX to AUX GND: 0V
  7. Measured -ve AUX to AUX GND: 0V

I was asked to check pads 1, 2 and 3 from the attachment which measured as follows:
  1. 2.7V
  2. 10.56V
  3. 12.57V

Apparently, they should have measured:
  1. 12.5 - 15V
  2. 11 - 13.5V
  3. 16 - 22V

Thanks for any help!!!
Gareth
 

Attachments

  • SMPS2000R Debug.GIF
    SMPS2000R Debug.GIF
    185.9 KB · Views: 686
Basically you just use the multimeter to trace the circuit, the path the current flows from input to output to see where the values aren't right anymore.

There seems to be a typical line filter, then at bridge rectifier input you should still have near 240VAC, then on the output you'd have about 335V DC, then tracing over to one or more switching transistors you'd have their base driven by the control circuit, and an emitter and collector voltage levels (or mosfet gate, drain and source pins). If no voltage is coming out of the transistors and they are driven at their base still, suspect transistor fault. If there is no base drive, begin tracing the control circuit. Locate the control IC and its datasheet or make a schematic of a discrete design to better understand what to measure in it.

If on the other hand you still have a good voltage level going into the transformer, measure at the transformer output (secondary) windings too.

It helps that you have one that works ok because you can compare readings between the two. Just be careful, playing with line powered power supplies is potentially dangerous. Some would advise hooking this up with a light bulb in series on the AC input to limit current.
 
Last edited:
Hi all,
I ordered two SMPS2000R power supplies to power my Tripath TA0105A amps. One works OK but the other was dead on arrival. I have tried everything to get help from Cristi, but to no avail. I've even recorded videos of my attempts to debug the circuit and posted it on YouTube. Cristi tells me that video sites are blocked where he comes from, so I hosted the video on my site. Apparently that didn't work either...so I cut the video into 5 and emailed them, but that didn't work either...

Next was Skype - but you both have to be on at the same time for that to work...now all attempts to communicate via email have gone unanswered.Two questions:
  1. Has anyone else had this problem with Connexelectronic?
  2. Any ideas how to debug the SMPS2000R?!

Thanks!
Gareth



Emails to them requesting info has gone unanswered, i never buy from those who don't respond , if before sales are poor and marginal, imagine after sale... :mad:
 
is it just a 2 layer pcb...if so , then there is some hope of finding the fault.......but really , we need a schematic....

Do you know what PWM controller it is using, -if you can tell us this we can tell you what to expect on its pins........check the pwm controller is receiving power...............do the obvious first...check for any missing components etc......is you ntc faile dopen circuit.......are the diodes of your diode bridge still diodes...etc etc

it looks like an unusual smps....states it can deliver 2600W from 115VAC........the rms mains input current woul dbe LITERALLY ENORMOUS because it doesnt have a PFC stage..........such enormous current could well result in resonations at power up.....eg resonations in the emi filter etc etc...have you checked the white "bathtub" resistor...is that open circuit?

Also, is the black thing at the input a relay...if so, check it operates...it looks like an inrush relay......................also, trace all power rails.....what about the startup circuit, is that working ok...?

Ask connexelectric if the designer still works for them......its likely they got in a designer, who designed the product then just left....this is very much less than ideal.
 
Last edited:
OK i saw your youtube video......so you say on the bad board the ntc is NOT heat-pasted to the heatsink?

I see you not use isolation transformer, so i am reluctant to ask you to follow live line through to the rectifier....and then check bus, cuzz i dont want you accidentally slipping and zapping yourself.....but anyway , to properly test you really need to power down, then solder wires to the test voltage points...then coroc clip these to your DVM probes, but this will take time as i know.

Can you show us the board underside?

You need to go round-the-pins on the pwm controller, but you cant really do that with that dvm as you'll keep slipping off.
 
you don't need schematic (unless it still can be posted). Such supply are really not hard to understand, there is no weird element used that could not be understood. All you need is to check the voltages, if those are ok, check the waveforms and at this point you can already say where the problem is...

you just have to pull out the scope and use it, nothing to it
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.