Cheap power supplies
-
- proppa neck!
- Posts: 537
- Joined: Mon Sep 02, 2019 11:06 am
Cheap power supplies
Got a 48v switch mode power here but it’s suffering from RF getting into it when the amp is powered up. Iv used these cheap power supply’s before and they’ve worked fine.
When the amp starts up it makes something on the supply buzz and then it shuts down completely after about 10 seconds.
When powering up a non RF device, it works at full load.
What is it that it’s interfering with ?
When the amp starts up it makes something on the supply buzz and then it shuts down completely after about 10 seconds.
When powering up a non RF device, it works at full load.
What is it that it’s interfering with ?
-
- proppa neck!
- Posts: 2778
- Joined: Tue Apr 05, 2016 1:23 am
Re: Cheap power supplies
How are you filtering the output of the PSU? You need to prevent RF getting back up the power rails into the supply - I used a mixture of small value capacitors from rail to ground at strategic points, and inductors in series. Some of the "Coutant" branded PSUs used to suffer if RF got onto the "sense" inputs of the PSU - these were designed so that you could monitor the actual voltage getting to the board at the board, so that the supply would adjust itself to give precisely the set voltage at the board. I found that these would turn the output voltage down radically if RF got in up the "sense" wires!
It's all in the filtering and in the layout!
It's all in the filtering and in the layout!
"Why is my rig humming?"
"Because it doesn't know the words!"![Wink ;)](./images/smilies/icon_e_wink.gif)
"Because it doesn't know the words!"
![Wink ;)](./images/smilies/icon_e_wink.gif)
-
- proppa neck!
- Posts: 537
- Joined: Mon Sep 02, 2019 11:06 am
Re: Cheap power supplies
How would I go about filtering it? At the moment I just use a wire and ferrite bead but it’s not doing anything. I can’t solder caps to ground from a wire. Would using some coax work to block rf jumping on the cable ?
-
- proppa neck!
- Posts: 2778
- Joined: Tue Apr 05, 2016 1:23 am
Re: Cheap power supplies
Coax might help - as long as it's not a resonant length! Make sure that the supply is filtered when it reaches the board. I usually use 1n ceramic, 100n polyester, 10µ electrolytic in parallel - these should strip most of the RF off the cable. I also add a ferrite bead or two (sometimes with the lead passed through a couple of times) - everything helps!
One thing that frequently catches out PA constructors is when there is some resonance at a couple of MHz going on in the supply (or in the collector / drain circuit of the power devices) - this can lead to huge spurs either side of your main carrier (a frequent problem with A****f PAs) which will usually appear on the most inconvenient frequency....
Back in the 80s, one of Dawson's LWR rigs caused the DTI guys to work on a Christmas Day when he managed to get a spur on to the input frequency of a local Ambulance repeater! This is why - if you're building anything more that a "bug" - you really do need a good Spectrum Analyser.......
One thing that frequently catches out PA constructors is when there is some resonance at a couple of MHz going on in the supply (or in the collector / drain circuit of the power devices) - this can lead to huge spurs either side of your main carrier (a frequent problem with A****f PAs) which will usually appear on the most inconvenient frequency....
Back in the 80s, one of Dawson's LWR rigs caused the DTI guys to work on a Christmas Day when he managed to get a spur on to the input frequency of a local Ambulance repeater! This is why - if you're building anything more that a "bug" - you really do need a good Spectrum Analyser.......
"Why is my rig humming?"
"Because it doesn't know the words!"![Wink ;)](./images/smilies/icon_e_wink.gif)
"Because it doesn't know the words!"
![Wink ;)](./images/smilies/icon_e_wink.gif)
-
- big in da game.. trust
- Posts: 95
- Joined: Sun Jan 17, 2016 1:49 pm
- Location: Netherlands
Re: Cheap power supplies
I recently experienced the same thing: A decent (older but not crappy) industrial switched power supply caused many spurs on the dial.
RFhead told me i should use torroids / Ferrite cores on the output of the supply.
This helped really good.
RFhead told me i should use torroids / Ferrite cores on the output of the supply.
This helped really good.
- sinus trouble
- proppa neck!
- Posts: 1410
- Joined: Fri Aug 22, 2014 11:34 pm
Re: Cheap power supplies
Another possibility to the problem could be ground loops?
A bundle of ground connections (Negative) creates differential paths for energy to flow throughout!
If i was building a rig? I would have a single DC ground point from PSU to the chassis or case!
The individual PCBs would then be bolted to the chassis and fed with positive supply as required! Apart from signal Ground? No separate DC Ground is needed!
A bundle of ground connections (Negative) creates differential paths for energy to flow throughout!
If i was building a rig? I would have a single DC ground point from PSU to the chassis or case!
The individual PCBs would then be bolted to the chassis and fed with positive supply as required! Apart from signal Ground? No separate DC Ground is needed!
I am as stupid as I look! ![Neutral :|](./images/smilies/icon_neutral.gif)
![Neutral :|](./images/smilies/icon_neutral.gif)
-
- proppa neck!
- Posts: 2778
- Joined: Tue Apr 05, 2016 1:23 am
Re: Cheap power supplies
"Star Point" earthing is the way to go. I usually put a stud through the case near to the supply smoothing capacitors, and earth everything radially back to this central point. The trick is to eliminate circulating earth currents - remember that every metal path that conducts electricity has its own resistance, and the further the flow, the higher the resistance. Internal coaxial cables are always earthed at just one end, so the screen is at the same potential throughout.
Remember also that you can get capacitive coupling between PCB tracks and the case, or between cables. Sometimes simply re-routing a lead can eliminate hum problems - it's sometimes down to trial-and-error. Some commercial equipment manufacturers have cable routing clips specifically designed to hold wires in particular positions. I remember in one instance that moving a regulator sub-PCB just a couple of centimetres completely eliminated hum!
When frequencies get higher - UHF and above - layout becomes even more significant.
Remember also that you can get capacitive coupling between PCB tracks and the case, or between cables. Sometimes simply re-routing a lead can eliminate hum problems - it's sometimes down to trial-and-error. Some commercial equipment manufacturers have cable routing clips specifically designed to hold wires in particular positions. I remember in one instance that moving a regulator sub-PCB just a couple of centimetres completely eliminated hum!
When frequencies get higher - UHF and above - layout becomes even more significant.
"Why is my rig humming?"
"Because it doesn't know the words!"![Wink ;)](./images/smilies/icon_e_wink.gif)
"Because it doesn't know the words!"
![Wink ;)](./images/smilies/icon_e_wink.gif)
- sinus trouble
- proppa neck!
- Posts: 1410
- Joined: Fri Aug 22, 2014 11:34 pm
Re: Cheap power supplies
Spot on Albert!Albert H wrote: ↑Tue May 24, 2022 5:06 pm "Star Point" earthing is the way to go. I usually put a stud through the case near to the supply smoothing capacitors, and earth everything radially back to this central point. The trick is to eliminate circulating earth currents - remember that every metal path that conducts electricity has its own resistance, and the further the flow, the higher the resistance. Internal coaxial cables are always earthed at just one end, so the screen is at the same potential throughout.
Remember also that you can get capacitive coupling between PCB tracks and the case, or between cables. Sometimes simply re-routing a lead can eliminate hum problems - it's sometimes down to trial-and-error. Some commercial equipment manufacturers have cable routing clips specifically designed to hold wires in particular positions. I remember in one instance that moving a regulator sub-PCB just a couple of centimetres completely eliminated hum!
When frequencies get higher - UHF and above - layout becomes even more significant.
![Smile :)](./images/smilies/icon_e_smile.gif)
Power cables can effectively become antennas under the right (Or wrong) conditions! which can disrupt the workings of a switch mode PSU!
RF Power amplifiers switch huge amounts of voltage and current at very high frequencies! It is enevitable that it will effect your DC connections!
Comparing this to a purely resistive load is madness! Not even on the same level!
I am as stupid as I look! ![Neutral :|](./images/smilies/icon_neutral.gif)
![Neutral :|](./images/smilies/icon_neutral.gif)
-
- proppa neck!
- Posts: 537
- Joined: Mon Sep 02, 2019 11:06 am
Re: Cheap power supplies
most switch modes are ok. its just the cheaper ones that suffer with the interfearance. I couldnt get it filtered out so just had to buy a new one. I bought a slighty more expensive one and works fine. just got the 1 ferrite on the wire coming from the amp and no hums/interfearance at all.
- rigmo
- tower block dreamin
- Posts: 462
- Joined: Wed Jan 30, 2019 9:35 pm
Re: Cheap power supplies
Few of them I make some modification.. exp. add some capacitance on primary.. better rf filtering.. etc and also on primary plus tvs diode.
- sinus trouble
- proppa neck!
- Posts: 1410
- Joined: Fri Aug 22, 2014 11:34 pm
Re: Cheap power supplies
I am pleased to hear that you have sorted it XXLXXL wrote: ↑Thu May 26, 2022 12:37 am most switch modes are ok. its just the cheaper ones that suffer with the interfearance. I couldnt get it filtered out so just had to buy a new one. I bought a slighty more expensive one and works fine. just got the 1 ferrite on the wire coming from the amp and no hums/interfearance at all.
![Smile :)](./images/smilies/icon_e_smile.gif)
Based on the vague info that you gave us? I would suspect that "Buzzing" noise came from the high frequency transformer within the PSU!
When a PSU surges! The primary driver tends to slump in frequency! This in turn is picked up by the feedback circuitry and attempts to correct the situation!
Failure to maintain the initial surge causes the PSU to shut down!
Thats my guess on what could be happening?
I am as stupid as I look! ![Neutral :|](./images/smilies/icon_neutral.gif)
![Neutral :|](./images/smilies/icon_neutral.gif)