A simple AM band portable radio will answer the questions for you... The cheaper the radio, the better this works... just tune it off station and place it right next to an SMPS... Hear anything?
If there was any leaked RF going on the radio would sound off like a gieger counter as it picked up harmonics of the chopper's frequencies.
Hi please note that an updated version of this PSU exists. It has an extra vertical PCB. What caught my eye was the little clearance between 230V mains voltage tracks and GND. I just wonder how safe this device is. It may be due to the picture. Is that OK with your version of this device?
Often PE is a disregarded item with cheap PSU's from China. It either is not connected (!!!) or it is connected inadequately so it can be improved in many cases. This is the price we pay for cheap 😀
Also some sellers offer the device with a better Talema transformer for 230V mains operation. The often used BingZi types are wound for 220V Chinese mains voltage. I tried a few times to convince BingZi to adapt to 230V but no replies from them. Not surprising....
Free ship New Version 25W Full Division MOS Linear Power Supply HIFI LPS PSU-in Amplifier from Consumer Electronics on AliExpress
Last edited:
My schedule and workbench are both open so I opened it up for a peek. Turns out I have the updated version. There was a long delay before mine was shipped so perhaps they ran out of the $58 one and sent me the $68 version.
I could be unhappy about not getting exactly what I ordered... But I'm not. 🙂
Attachments
This is probably off topic (or isn't it as the OT was looking for a good linear PSU?) but the not connected PE pin worries me. Also the very small clearance between mains voltage carrying PCB tracks.
For which output voltage is your sample set? I see 2 x 9V windings.
Please note and possibly solve the following points:
- 2 x 110V (or 220V) primaries BingZi! Could be replaced for better 2 x 115V (or 230V) Talema. Does the BingZi hum?
- not enough clearance around mains voltage carrying PCB tracks (at least I would not dare to do it like this with 230V AC mains voltage)
- 2SJ407 is used instead of the much lower RdsON P-MOSFET IRF9Z34N as written on the PCB.
- connect the PE pin to the casing by means of sturdy flexible green/yellow wire soldered or crimped to a cable shoe (eyelet) fastened with M3 bolt, nut and a washer. You could use the nut nearby but make sure the eyelet can not make a short circuit.
- the CD11k cap is made by Foshan. Maybe a well known brand and type could be used.
It is a discrete LDO with very low noise. I know these perform well but as you can see there are certainly some issues. I had ordered one but I never received it.
For which output voltage is your sample set? I see 2 x 9V windings.
Please note and possibly solve the following points:
- 2 x 110V (or 220V) primaries BingZi! Could be replaced for better 2 x 115V (or 230V) Talema. Does the BingZi hum?
- not enough clearance around mains voltage carrying PCB tracks (at least I would not dare to do it like this with 230V AC mains voltage)
- 2SJ407 is used instead of the much lower RdsON P-MOSFET IRF9Z34N as written on the PCB.
- connect the PE pin to the casing by means of sturdy flexible green/yellow wire soldered or crimped to a cable shoe (eyelet) fastened with M3 bolt, nut and a washer. You could use the nut nearby but make sure the eyelet can not make a short circuit.
- the CD11k cap is made by Foshan. Maybe a well known brand and type could be used.
It is a discrete LDO with very low noise. I know these perform well but as you can see there are certainly some issues. I had ordered one but I never received it.
Last edited:
This device does not seems to be built according to the 2014/35/UE requirements (Low voltage directive), to me. It would be funny to see what happens to that naked 250V mains fuse that sits over a ground plane silkscreen, when the ESD test probe does its job on the mains input.
Last edited:
My guess it that it is worth trying to replace the IEC inlet for a version that takes wires. Then connect the wires to the transformer and mains switch and cut the PCB traces. Challenge will be to wire the switch safely. Maybe replacing the mains switch alltogether for a panel mount version will do the job.
As mine did not arrive I decided to draw my own version. If time permits of course.
As mine did not arrive I decided to draw my own version. If time permits of course.
JP when you commented (#83) on the mains traces being so near the ground plane it seemed to me you felt the ground pin was connected.
In your above post you seem to feel it's NOT connected.
I assure you it is.
All 3 pins of the IEC socket are on the bottom. The ground trace is bonded to the case via that visible bolt.
In my case I'm on 120v so I'm not concerned about this spacing.
In your above post you seem to feel it's NOT connected.
I assure you it is.
All 3 pins of the IEC socket are on the bottom. The ground trace is bonded to the case via that visible bolt.
In my case I'm on 120v so I'm not concerned about this spacing.
AFAIK, don't trust me on this, PE clearances can be minimal as long as any fault current clears the fuse rather than the PE. The internal fuse is going to violate that one without thinking about it so the unit will be wholly reliant on the fuse in the leads plug. A little bit more thought, an internal fuse holder or clearance on the board for the internal fuse end caps... 2mm ?, and it might have made more sense.
In the UK I think it is 6mm from mains. So in USA it will be around 3mm.
The problem is worst when mains gets on say a high impedance audio input where it doesn't blow the fuse.
Trace clearance should be at least 0.4mm according to ANSI/IPC-2221, buy UL requires 1mm on this specific instance (I may be wrong, I am not an expert on UL). This is the minimum clearance required; but I always see more on good quality devices. I have repared boards with 1mm clearance that experienced arcing, maybe due to random mains voltage spikes.
When ground is connected trough a single bolt, it must have a fastener to avoid unscrewing (two nuts, as example); the board coating is not enough to pass dielectric strength tests if a naked conductor (the fuse) is directly over it. This is why I have doubt about the CE mark conformity. It is only a rough guess, a closer examination would be needed to come at a conclusion, of course.
When ground is connected trough a single bolt, it must have a fastener to avoid unscrewing (two nuts, as example); the board coating is not enough to pass dielectric strength tests if a naked conductor (the fuse) is directly over it. This is why I have doubt about the CE mark conformity. It is only a rough guess, a closer examination would be needed to come at a conclusion, of course.
The PCB is coated: spacing could be less. We can assume that the PCB CAD software does apply the correct rules. The issue is: the PCB has been designed for a fuse holder, that has been removed on manufacturing. And the PCB work was maybe rushed a bit: manual routing on the top layer would have been useful here.