markk
Member
So, this thread pulls in a lot of disparate elements, but, bear with me.
So, getting back into audio after a long layoff, I decided to pull out a set of Peerless 830500 subs that I had NOS for, like, 15 years...sigh. Not saying you should buy these, but using these turns out to be a great refresher. They happen to be the same subs that are used in SL's Thor. And that's educational because this is an extraordinarily well documented project with a spreadsheet that is still applicable today. And if you want to understand exactly how excursion limits, box size, amplifier voltage and current all interact on a deeper level than most discussions, it's worth a read.
Would I buy these today? Well, no, but they're free, in my garage, so... And long ago I split half a pallet of finished sub cabinets from some manufacturer going out of business in SoCal with Charlie Laub. On pair was a set of mid-century modern walnut cabinets and the cutout was a perfect fit for the 830500, so just drop in. Mid century modern is back...
So, I just drop in the 830500's into the finished cabinets along with a little polyfill and measure away. Unfortunately, the resultant F3 of the box and driver combo is around 47hz. Not ideal without eq. So, this is where the idea of using a miniDSP to eq the subs. Since the sub out are RCA, I'm stuck with analog in. I picked up the miniDSP 8x12 V2.0 because I can use two channels to eq these two subs and feed it into a buckeye stereo hypex NC502MP, and can eq the front three before feeding them into another Hypex with 3 more of the miniDSP channels (part two of the saga...) Also this unit can be upgraded to DIRAC. I mean, so can the Denon, but what happens when that is EOL'd?
So, running Clio out into the buckeye and doing nearfield measurements, all is going well. See the graph below.
Yes, I know, the LF rolloff isn't ideal. Will fix... Nonetheless, with about 8dB of gain, I can get a reasonably flat response to 30hz. So, finish programming the miniDSP, unplug Clio and plug the Denon 3800 sub 1 and 2 outs into the miniDSP, and the outputs into the Hypex, fire everything up and, well, a ground loop hum to make you go insane!
So, this is interesting. In a bad way... I mean it's loud. Sweet 60 hz hum and all it's progeny appear...
Clio driving the miniDSP into the Hypex does not cause ground loop. But, Denon driving the miniDSP into the Hypex causes hum. So, what is really going on?
The key to life is having measurement equipment. Otherwise you just randomly switch cables, plugs, electricians, bybee filters...
So, disconnect the RCA sub out of the Denon and measure. Well, that is the green curve. Some noise, but in the range of -90dbV. Seems about right for the Denon. Plug in the miniDSP input to the Denon sub output and measuring the low level sub 1 out gives the aqua curve. Note the miniDSP is NOT on, not plugged into 12 volts.
The mere fact of plugging the Denon sub out into the miniDSP is causing the Denon to generate all that 60 cycle and related products. The miniDSP is off, so the entire RCA loop is acting a bit like an antenna and likely basking in the warm EMF of the Denon's power supply unit and the Denon is happy to feed the miniDSP with almost 80 millivolts self generated noise courtesy of said antenna.
So, who's at fault here, the Denon or the miniDSP? Well, so, it's a bit like Jonny Depp and Amber Heard. They're both lacking... to put it nicely. The miniDSP input is not robust enough to reject the spurious voltage in the way that the Hypex can when it's directly connected to the Denon. Having said that, the Denon is clearly where the spurious voltage is being generated, so Denon takes a hit here to...
So what's the answer? Well part of the problem is nothing is grounded anymore. The Denon is isolated. The miniDSP is isolated. The Hypex is isolated. Turns out, tying the RCA shield to earth ground works fine. This is the yellow curve. There is "ground loop" so to speak between the Denon and the miniDSP, and forcing the shield to ground seems to clear up the hum quite nicely...
Note I am NOT recommending you always blindly tie your shield to earth ground. In my universe of well designed products, all grounds would be same ground and all shields would be grounded. But that is not the way every electronic device is designed these days. So, as they say, this post is only for entertainment, seek the advice of a qualified electrician, analog systems design engineer (er, are there any of those left...?), or shaman..
In my childhood I worked on a TV set where the chassis was HOT. I mean, electrically hot. Like, touch the chassis and you die hot. Black wire hot... Yes, there was a warning label, but really? Yeah, back in the 70's you could design a patently dangerous product, stick a warning label on, and well, you're ok...
PS the purple curve is an attempt to tie the chassis ground of the Denon, miniDSP and RCA shield together. Marginally better, but not the correct solution.
Hope you found that interesting. Especially if you have something in the miniDSP family and are experiencing hum...
So, getting back into audio after a long layoff, I decided to pull out a set of Peerless 830500 subs that I had NOS for, like, 15 years...sigh. Not saying you should buy these, but using these turns out to be a great refresher. They happen to be the same subs that are used in SL's Thor. And that's educational because this is an extraordinarily well documented project with a spreadsheet that is still applicable today. And if you want to understand exactly how excursion limits, box size, amplifier voltage and current all interact on a deeper level than most discussions, it's worth a read.
Would I buy these today? Well, no, but they're free, in my garage, so... And long ago I split half a pallet of finished sub cabinets from some manufacturer going out of business in SoCal with Charlie Laub. On pair was a set of mid-century modern walnut cabinets and the cutout was a perfect fit for the 830500, so just drop in. Mid century modern is back...
So, I just drop in the 830500's into the finished cabinets along with a little polyfill and measure away. Unfortunately, the resultant F3 of the box and driver combo is around 47hz. Not ideal without eq. So, this is where the idea of using a miniDSP to eq the subs. Since the sub out are RCA, I'm stuck with analog in. I picked up the miniDSP 8x12 V2.0 because I can use two channels to eq these two subs and feed it into a buckeye stereo hypex NC502MP, and can eq the front three before feeding them into another Hypex with 3 more of the miniDSP channels (part two of the saga...) Also this unit can be upgraded to DIRAC. I mean, so can the Denon, but what happens when that is EOL'd?
So, running Clio out into the buckeye and doing nearfield measurements, all is going well. See the graph below.
Yes, I know, the LF rolloff isn't ideal. Will fix... Nonetheless, with about 8dB of gain, I can get a reasonably flat response to 30hz. So, finish programming the miniDSP, unplug Clio and plug the Denon 3800 sub 1 and 2 outs into the miniDSP, and the outputs into the Hypex, fire everything up and, well, a ground loop hum to make you go insane!
So, this is interesting. In a bad way... I mean it's loud. Sweet 60 hz hum and all it's progeny appear...
Clio driving the miniDSP into the Hypex does not cause ground loop. But, Denon driving the miniDSP into the Hypex causes hum. So, what is really going on?
The key to life is having measurement equipment. Otherwise you just randomly switch cables, plugs, electricians, bybee filters...
So, disconnect the RCA sub out of the Denon and measure. Well, that is the green curve. Some noise, but in the range of -90dbV. Seems about right for the Denon. Plug in the miniDSP input to the Denon sub output and measuring the low level sub 1 out gives the aqua curve. Note the miniDSP is NOT on, not plugged into 12 volts.
The mere fact of plugging the Denon sub out into the miniDSP is causing the Denon to generate all that 60 cycle and related products. The miniDSP is off, so the entire RCA loop is acting a bit like an antenna and likely basking in the warm EMF of the Denon's power supply unit and the Denon is happy to feed the miniDSP with almost 80 millivolts self generated noise courtesy of said antenna.
So, who's at fault here, the Denon or the miniDSP? Well, so, it's a bit like Jonny Depp and Amber Heard. They're both lacking... to put it nicely. The miniDSP input is not robust enough to reject the spurious voltage in the way that the Hypex can when it's directly connected to the Denon. Having said that, the Denon is clearly where the spurious voltage is being generated, so Denon takes a hit here to...
So what's the answer? Well part of the problem is nothing is grounded anymore. The Denon is isolated. The miniDSP is isolated. The Hypex is isolated. Turns out, tying the RCA shield to earth ground works fine. This is the yellow curve. There is "ground loop" so to speak between the Denon and the miniDSP, and forcing the shield to ground seems to clear up the hum quite nicely...
Note I am NOT recommending you always blindly tie your shield to earth ground. In my universe of well designed products, all grounds would be same ground and all shields would be grounded. But that is not the way every electronic device is designed these days. So, as they say, this post is only for entertainment, seek the advice of a qualified electrician, analog systems design engineer (er, are there any of those left...?), or shaman..
In my childhood I worked on a TV set where the chassis was HOT. I mean, electrically hot. Like, touch the chassis and you die hot. Black wire hot... Yes, there was a warning label, but really? Yeah, back in the 70's you could design a patently dangerous product, stick a warning label on, and well, you're ok...
PS the purple curve is an attempt to tie the chassis ground of the Denon, miniDSP and RCA shield together. Marginally better, but not the correct solution.
Hope you found that interesting. Especially if you have something in the miniDSP family and are experiencing hum...
