I have a box with two TPA3255 amp modules in it (I don't think it actually matters what kind of amps they are, but oh well). A single 48 VDC single rail SMPS powers both of them.
Right now, the power inlet module has its ground tab connected directly to a chassis grounding bolt very, very securely. I also have a ground loop breaker, with its safety ground connection made also very securely to chassis. Its HRG (high-resistance ground) side is tied to the 0 VDC output of the SMPS.
My SMPS has, on its mains input side, L-N-G/E. My question is, assuming I want to minimize any ground loop issues if using single-ended interconnects, do I want to attach G to chassis safety ground, or to HRG, or not connected at all? Keeping in mind the 0 VDC is tied to HRG and the chassis is very well connected to safety earth.
I believe, but am not certain, that SMPS inlet G is not tied directly to ("common with") its 0 VDC output. I don't know what, if any, capacitors are placed between input & output sides.
Peace and good health to all, and thanks in advance.
Right now, the power inlet module has its ground tab connected directly to a chassis grounding bolt very, very securely. I also have a ground loop breaker, with its safety ground connection made also very securely to chassis. Its HRG (high-resistance ground) side is tied to the 0 VDC output of the SMPS.
My SMPS has, on its mains input side, L-N-G/E. My question is, assuming I want to minimize any ground loop issues if using single-ended interconnects, do I want to attach G to chassis safety ground, or to HRG, or not connected at all? Keeping in mind the 0 VDC is tied to HRG and the chassis is very well connected to safety earth.
I believe, but am not certain, that SMPS inlet G is not tied directly to ("common with") its 0 VDC output. I don't know what, if any, capacitors are placed between input & output sides.
Peace and good health to all, and thanks in advance.
Have you consulted a moldymeter about this yet? I take it you don't have a datasheet for the SMPS?
If everything is as you say, G would go to chassis/safety ground. With 0 V going to HRG, RCA inputs could even remain connected to chassis, even if it's not recommended as this quickly creates external ground loops e.g. with PC sources. So isolated jacks and small RF bypass caps (maybe 1 nF) to chassis it is then. (You may try using a balanced input stage as suggested by Bruno Putzeys, though providing its supplies may prove inconvenient. I guess you could go single supply and regulate +48 down to +30 or something, we're not talking a massive current drain.)
That's about the best you can do with chassis construction as it is.
I confess it took me embarrassingly much longer than I wanted to understand your comment about the moldymeter... 🙂 No, I do not have a datasheet for the SMPS.
I do currently have SMPS inlet G going to safety ground, so that's nice. Could you elaborate a bit on the RF bypass caps? Connecting between the shield tab on the isolated RCA jacks and to chassis right beside the jack? If I have multiple jacks should I do 1 chassis bypass/decoupling point per jack, or can I "bus" the chassis side of all the bypass caps to a single chassis point?
I do currently have SMPS inlet G going to safety ground, so that's nice. Could you elaborate a bit on the RF bypass caps? Connecting between the shield tab on the isolated RCA jacks and to chassis right beside the jack? If I have multiple jacks should I do 1 chassis bypass/decoupling point per jack, or can I "bus" the chassis side of all the bypass caps to a single chassis point?
EEVBlog Dave is pretty much etched into my brain at this point, sorry.
Shield tab to chassis, correct (1-10 nF typ). One per jack would generally give lower inductance which is critical at RF, though combining two in a pair may not be much worse. I guess you could use braid or copper foil if you did need to "bus" things.
