I need some advice on selecting a ferrite to filter out noise spikes from a switching power supply that is being fed into a radio. I am an electronics hobbyist with spotty knowledge, mostly from the late 70’s to early 80’s (TTL, CMOS, op-amps), but unfortunately minimal RF.
If this is not the appropriate forum or site to post this question, please let me know.
We have several fairly old devices that include VHF radio receivers (along with some digital decoder stuff that is not relevant to this issue). The devices previously had simple, clean linear power supplies running off 240VAC, but were just retrofitted to run off 48VDC. The retrofit included a 48VDC to 12VDC switching voltage regulator. It is a P7812 by CUI INC, which now feeds the original 7805 linear regulator. The original transformer and FWB were simply removed, replaced with the P7812.
The problem is the RF stage is fed upstream of the 7805, so it sees the noise spikes from the P7812. I’m not sure of the exact amount, but I believe the RF IC is being fed 100mV spikes, give or take. (The 12VDC has up to 200mV spikes, but there is a voltage divider that drops the RF IC’s power to 6.2VDC.) Post retrofit, the devices appear to have a significant sensitivity loss – they are no longer able to decode broadcasts as well as they used to.
From what I have read, a ferrite could help significantly, but there are so many variables in selecting ferrites, I would only be guessing. The power is quite low, varying from 75mA to 140mA.
I understand ferrites can be configured to filter out either common mode or differential mode noise. Since I am connecting the scope across the power supply output where it is fed into the device, I assume this is differential mode noise.
I have attached a photo of the scope trace. In retrospect, I probably should has zoomed into the spike more to determine its rise time, but again I am out of my element – I am not that familiar with working scopes. If it is needed, I can return to the lap for additional testing.
I am confident the noise source is the P7812, as when we replaced it with the lab power supply the spikes disappeared.
Thank you in advance,
parkerea
If this is not the appropriate forum or site to post this question, please let me know.
We have several fairly old devices that include VHF radio receivers (along with some digital decoder stuff that is not relevant to this issue). The devices previously had simple, clean linear power supplies running off 240VAC, but were just retrofitted to run off 48VDC. The retrofit included a 48VDC to 12VDC switching voltage regulator. It is a P7812 by CUI INC, which now feeds the original 7805 linear regulator. The original transformer and FWB were simply removed, replaced with the P7812.
The problem is the RF stage is fed upstream of the 7805, so it sees the noise spikes from the P7812. I’m not sure of the exact amount, but I believe the RF IC is being fed 100mV spikes, give or take. (The 12VDC has up to 200mV spikes, but there is a voltage divider that drops the RF IC’s power to 6.2VDC.) Post retrofit, the devices appear to have a significant sensitivity loss – they are no longer able to decode broadcasts as well as they used to.
From what I have read, a ferrite could help significantly, but there are so many variables in selecting ferrites, I would only be guessing. The power is quite low, varying from 75mA to 140mA.
I understand ferrites can be configured to filter out either common mode or differential mode noise. Since I am connecting the scope across the power supply output where it is fed into the device, I assume this is differential mode noise.
I have attached a photo of the scope trace. In retrospect, I probably should has zoomed into the spike more to determine its rise time, but again I am out of my element – I am not that familiar with working scopes. If it is needed, I can return to the lap for additional testing.
I am confident the noise source is the P7812, as when we replaced it with the lab power supply the spikes disappeared.
Thank you in advance,
parkerea
