audio-talk

Hi, I'd like to put a question to the knowledgable if I may.

The never ending Transcendent Beast OTL project is moving glacially forward and I've run into something which looks a bit odd to me.

As it's a copyrighted design I can't post the schematic, but the circuit signal grounds all tie back to a short copper ground bus on the main circuit board as a star ground, including the negative speaker terminal.

However, I cannot find anything in the assembly manual, or evidence in the pictures, of a direct connection between the signal ground and the chassis / mains safety earth, which is present on the Son of Beast amplifier.

I have posted a question on the Transcendent forums about this but haven't had a response from Bruce about it - the post has been up for 5 days so it looks like I'm not going to get one. I have had a response from another user who confirms what I'm seeing and that I've not missed anything.

I suspect that in order to solve any potential ground loop problems, the idea is that a 'normal' RCA input connection from a preamp where the interconnect shield is connected to earth, will ground the signal ground bus via the interconnect shield.

However, I'm going to be using line input transformers at the Beast end to isolate and unbalance the incoming line signal from my AMB preamp - these ones : http://www.jensen-transformers.com/wp-c ... 1p4-11.pdf , and I'm following the recommended connection with a ground lift switch.

The result of this, as far as I can see, will be a floating signal ground if I build the amp as instructed. From an amplifier operation perspective, I think the amp should still work as the ground point will still be a valid return path. I do have a safety concern about this though, what if a fault condition were to place several hundred volts on the ground bus? With no route back to safety earth, the speaker connectors could remain at that potential?!

With an isolated floating +/- input signal, it seems to me best practice to connect that star ground directly to chassis / mains safety earth. Would the collective agree with that approach?

A 10 chm 1 watt resistor between the chassis/PE and signal ground/star. Nelson Pass uses an NTC thermistor in a lot of his designs.

Interesting idea about the thermistors, as it happens I have two spare CL-60's. They were originally specified for the inrush current limiting but I've since decided to replace them with an active soft-start circuit per Rod Elliott, which will also more elegantly integrate the remote 12V power-up / power-down triggering.

So as of now they are surplus to requirements, but I guess they could be used for this connection, with the resistance reducing in the event of a fault.

Hang on though - is that 10R resistance even needed in this case?

I'm using input transformers, one of their benefits is the breaking up of ground loops. The incoming signal hot and cold phase will be on pins 2 and 3 of the XLR, connected to the floating primary, with pin 1 (shield) connected to chassis through a parallel resistor/cap and ground lift switch, per the Jensen recommendation.

So even with just a straight piece of wire between the star ground point and chassis, I can't see a route around which a ground loop would develop. It would also be safer.

Granted the situation would be different if I had a conventional arrangement of unbalanced RCA socket with shield connected to signal ground and no transformer-coupling.

SteveH wrote: ↑Fri Feb 23, 2018 7:13 pm A 10 Ohm 1 watt resistor between the chassis/PE and signal ground/star. Nelson Pass uses an NTC thermistor in a lot of his designs.

That's what I would do too.

Hang on though - is that 10R resistance even needed in this case?

Won't hurt, but should stop a hum loop, and you've had bother with that too?

Mike H wrote: ↑Sat Feb 24, 2018 10:31 am

SteveH wrote: ↑Fri Feb 23, 2018 7:13 pm A 10 Ohm 1 watt resistor between the chassis/PE and signal ground/star. Nelson Pass uses an NTC thermistor in a lot of his designs.

That's what I would do too.

Hang on though - is that 10R resistance even needed in this case?

Won't hurt, but should stop a hum loop, and you've had bother with that too?

Hi Mike - yes I have, which is one of the reasons I chose to use input transformers on the new amps, to solve that problem.

If the speakers were connected to a low voltage solid state amp or transformer coupled I wouldn't worry, but these are directly connected to the output stage supplied by +/-170V, and the more I think about this the more I feel that 0V needs as unencumbered a connection to safety earth as the chassis.

Appreciate the input - it certainly confirmed my main question, that I do need to reference 0V to mains earth somehow. I think I'll try the straight piece of wire first, then in the event I run into problems I'll take it from there.

Thanks!

I have only skim read this but don't most manufacturers recommend that the one and only signal to mains safety earth connection should be in the source component (or one of them) ?

My system has no signal to mains safety earth connection at all at the moment. It is completely floating.

Stu

karatestu wrote: ↑Sat Mar 10, 2018 1:28 pm I have only skim read this but don't most manufacturers recommend that the one and only signal to mains safety earth connection should be in the source component (or one of them) ?

My system has no signal to mains safety earth connection at all at the moment. It is completely floating.

Stu

Signal - ground ?

Ov point to mains earth do you mean ?

pre65 wrote: ↑Sat Mar 10, 2018 1:46 pm

karatestu wrote: ↑Sat Mar 10, 2018 1:28 pm I have only skim read this but don't most manufacturers recommend that the one and only signal to mains safety earth connection should be in the source component (or one of them) ?

My system has no signal to mains safety earth connection at all at the moment. It is completely floating.

Stu

Signal - ground ?

Ov point to mains earth do you mean ?

Yes that is what i meant.

karatestu wrote: ↑Sat Mar 10, 2018 1:28 pm I have only skim read this but don't most manufacturers recommend that the one and only signal to mains safety earth connection should be in the source component (or one of them) ?

My system has no signal to mains safety earth connection at all at the moment. It is completely floating.

Stu

didn't Steve(Cresswell) have an issue with this sometime back when he plugged a grounded(safety) pre-amp into a floating power amp that had a fault...My memory may be iffy here but If I recall correctly there were fireworks.

All my valve amps are signal ground to chassis safety...My fet amps are sig grd to chassis safety via thermistor. I have to mention that I'm not the greatest authority on this.

ed wrote: ↑Sat Mar 10, 2018 2:47 pm

karatestu wrote: ↑Sat Mar 10, 2018 1:28 pm I have only skim read this but don't most manufacturers recommend that the one and only signal to mains safety earth connection should be in the source component (or one of them) ?

My system has no signal to mains safety earth connection at all at the moment. It is completely floating.

Stu

didn't Steve(Cresswell) have an issue with this sometime back when he plugged a grounded(safety) pre-amp into a floating power amp that had a fault...My memory may be iffy here but If I recall correctly there were fireworks.

All my valve amps are signal ground to chassis safety...My fet amps are sig grd to chassis safety via thermistor. I have to mention that I'm not the greatest authority on this.

Yes,
I was thrown onto my back by a high voltage zap as I plugged in the interconnect. Highly unpleasant and I thought from then on, never again.

Decision cemented then, on that basis - 0v to safety with a straight piece of wire, no resistors, no thermistors. The galvanic isolation gives me absolutely no reason not to.

Jesus.......
Fukin armature's

Nanook of the North wrote: ↑Sat Apr 14, 2018 9:43 pm Jesus.......
Fukin armature's

Hi Jammy