Unbalance sharing in output mosfets

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Look at output ground. Same noise there (as I suspect)?

If so, that is calling for a healthy mains filter and other common-mode choke action (at the output, too). A 1k||10n connection of secondary-side ground to PE may also be of use. Common-mode noise in SMPS' and associated EMI is nasty stuff, and with some bad luck a redesign may be needed to get things in check. Bruno Putzeys says his target levels are 20, preferably 30 dB below FCC Class B requirements to have RF-sensitive devices covered. And those aren't necessarily easy to hit to begin with.

If the noise is appearing across the output, that would point to an RF sensitivity or PSRR problem in the actual amplifier circuit.
 
Here's a picture of the Oscope connected to output of amp
** I short cut the input of amp in this test**
Channel 2
Volts/Div = 50mv
Time/Div = 5 us
probe attenuation is on 1X

kz63_56877.jpg
 
Hi Guys

SMPS is about the worst to use for a linear amplifier. The output impedance is high and variable which is exactly the opposite of what is required. The LC or RC filters are the simplest approach to try initially.

The better approach is to add active hum filters on the rails as these will have the low and constant impedance needed for the amp.

In either approach, the extra filters provide a point of decoupling the ground nodes to isolate the noisey switching currents from the clean audio currents. The output caps in both approaches must be quite large in value to affect the impedance change required.

Have fun
 
Hi Guys

SMPS is about the worst to use for a linear amplifier. The output impedance is high and variable which is exactly the opposite of what is required. The LC or RC filters are the simplest approach to try initially.

The better approach is to add active hum filters on the rails as these will have the low and constant impedance needed for the amp.

In either approach, the extra filters provide a point of decoupling the ground nodes to isolate the noisey switching currents from the clean audio currents. The output caps in both approaches must be quite large in value to affect the impedance change required.

Have fun
Oversimplification and generalizing is no good. Unregulated hardswitched SMPS has pretty constant output impedance and it could easily have better regulation than a linear PSU. Active filters could be used with an SMPS as well.

The output impedance is high and variable which is exactly the opposite of what is required. The LC or RC filters are the simplest approach to try initially.
Adding RC filters make PSU impedance higher at low frequencies.
 
Hi Guys

My experience and that of others I know makes if very clear that SMPS is entirely the wrong choice for support of a linear amplifier. Feel free to disagree.

In those cases where a SMPS "had to be " used, the active stages were added between it and the load and everything was fine - except that SMPS life is miserable compared to a properly rated linear supply. There is quite a bit of "broadcast quality" gear that has redundant SMPSs where that was never a requirement with linear supplies. Obviously economics drives SPMS design more than performance does.

Have fun
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.