• Disclaimer: This Vendor's Forum is a paid-for commercial area. Unlike the rest of diyAudio, the Vendor has complete control of what may or may not be posted in this forum. If you wish to discuss technical matters outside the bounds of what is permitted by the Vendor, please use the non-commercial areas of diyAudio to do so.

AC internal wiring

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

I am making some new power cords (what the heck, its an experiment) from Belden 19364 cable , known for being an improvement from standard.

At the same time, it made me think about the internal wiring for AC, i am using inside the Maya amps. Maybe its not ideal...and maybe i can improve it.

25268154838_c193ba4b29_z.jpg


I run my AC from the back of amplifier where the IEC input is.
Then, run both N and F wires along the center of the chassis to the front Switch (it is a dual pole switch). Then to the transformers.

Now : should these wires be shielded ? Should they be thicker than 1.5mm2?
Should i get a relay instead for the front switch and keep AC at the back ?
Does it matter ?

Here is the real thing :
34073756713_b63d769781_h.jpg


Opinions are welcome...sorry for the toy schematic.
 
The wire size depends on the mains fuse you use.
The wire must be rated higher than the fuse rating.

Twisted pair make a good emi reduction strategy.
Twist the fuse to switch wiring.
Twist the switch to transformer wiring.
Keep the loop areas at the terminations as small as possible.
Note the open loop on the input wiring near the open loop in the mains wiring.

Using a single fuse for two transformer means upto ten times longer rupture time when abuse occurs. It is safer to fuse each transformer separately. Close rated fusing is even safer, since it is another factor of 10 faster than the big fuse you need when you don't use a soft start.

You have a lot of wiring coming into that bolted connection just along from the mains input.
What are they? There should be the PE wire and nothing else.
 
Last edited:
The wire size depends on the mains fuse you use.
The wire must be rated higher than the fuse rating.

Twisted pair make a good emi reduction strategy.
Twist the fuse to switch wiring.
Twist the switch to transformer wiring.
Keep the loop areas at the terminations as small as possible.
Note the open loop on the input wiring near the open loop in the mains wiring.

Using a single fuse for two transformer means upto ten times longer rupture time when abuse occurs. It is safer to fuse each transformer separately. Close rated fusing is even safer, since it is another factor of 10 faster than the big fuse you need when you don't use a soft start.

You have a lot of wiring coming into that bolted connection just along from the mains input.
What are they? There should be the PE wire and nothing else.

I only have 1 FUSE, and that is in the IEC connector.
It is rated at 10A, 250V slow blow.
I have a PSAUDIO PPP coming soon, that has a soft start so no worries there.

The main AC wiring to and from switch is twisted.
The bolted connection is Mains Earth. It is connected to :
- Input IEC Earth pin
- Earth pin in Maya channel 1
- Earth pin in Maya channel 2
- Earth copper shield on RCA input 1
- Earth copper shield on RCA input 2
 
I only have 1 FUSE, and that is in the IEC connector.
It is rated at 10A, 250V slow blow.
A 10A mains fuse on a 230Vac supply is enormous.
It will pass 10A indefinitely during a fault, it will pass 20A for many seconds and maybe as long as half an hour during a fault.

What VA rating are your transformers?
I have a PSAUDIO PPP coming soon, that has a soft start so no worries there.

The main AC wiring to and from switch is twisted.
The bolted connection is Mains Earth. It is connected to :
- Input IEC Earth pin
OK.
- Earth pin in Maya channel 1
- Earth pin in Maya channel 2
- Earth copper shield on RCA input 1
- Earth copper shield on RCA input 2
these 4 connections are possibly wrong.
Check with AKSA where the currents flow.
What you must NOT do is force the returning current to take a long route to return to it's source.
The return route should be alongside the Flow route for that current. None of these 4 wires has a Flow route starting at the Chassis/Enclosure.
 
Last edited:
I only have 1 FUSE, and that is in the IEC connector.
It is rated at 10A, 250V slow blow.
I have a PSAUDIO PPP coming soon, that has a soft start so no worries there.

The main AC wiring to and from switch is twisted.
The bolted connection is Mains Earth. It is connected to :
- Input IEC Earth pin
- Earth pin in Maya channel 1
- Earth pin in Maya channel 2
- Earth copper shield on RCA input 1
- Earth copper shield on RCA input 2

Hi saf,
Yes, follow Andrew's advices.
The inlet fuse in the IEC should be much smaller amperage and fast blow.
With Hugh's module construction you do need to connect the chassis earth spade to the "earth" bolt on the chassis.
There is no reason to connect the rca to the "earth" bolt.
Be sure to follow connection advices with the Belden cable, especially with regards the drain wire. I suspect you are already aware of that.
Steve
 
Hi Filipe,

It is easy to confuse this earthing and it's crucial to get it right. I've spent years trying to perfect a repeatable earth system on my amps and I'm pretty sure this is the best I've done yet.

Here are the connections I suggest:

1. RCA input HOT to module, and BRAID to Sig Earth on the module too, NOT to the chassis/star earth on the enclosure.
2. IEC power earth to chassis/star earth on the enclosure.
3. Power earth wire from modules to chassis/star earth on the enclosure.

If you have an electrostatic wire on the toroids, you may also connect them to chassis/star earth on the enclosure.

Do NOT connect RCA signal input GROUND to the chassis/star. This will create hum and noise.

Be sure that the ground of the RCA does NOT connect to the enclosure on the enclosure panel. This will also create hum and noise.

Twist ALL AC lines tightly so that they do not radiate AC noise into the modules.

A standard fast blow 10A on the IEC input socket will accommodate two 220/240Vac 300VA toroids with ease. At full operation, the amp will be pulling just under 3A from the mains, but mostly it's well under an amp in use. A 10A fast blow is good protection in this situation.

This regime works very well and does not require a slow turn on circuit. The 300VA toroid current surge at switchoff is around 11 times the max current (more than 13A EACH) but it persists for only two cycles, which is just 40mS, and this is not sufficient to blow the 10A fast blow fuse. The actual peak is only a few mS in fact, it is not consistent along the 40mS turn on event.

Thanks to AndrewT, Stvnharr and Scott Joplin.


Cheers,

Hugh
 
Last edited:
Ok about the twisting, i will check.
Nobody recommended to use shielded AC twisted wire from IEC to switch. I would assume this shielding would benefit then ac noise redutction.

This leeds me to explain the 4 connections that i referred:

- Earth pin in Maya channel 1
- Earth pin in Maya channel 2
These EARTH connections are on the Maya modules and Hugh clearly says to connect them to Star Earth. 100% sure here.

- Earth copper shield on RCA input 1
- Earth copper shield on RCA input 2~
I do not use a coax cable for RCA input wire. I use a shielded twisted pair.
So i have 2 conductors, one for HOT and one for NEUTRAL. Plus i have the copper shielding that i connect to EARTH GROUND at the source end (wire from rca input end). This shielded copper is NOT connected to anything besides the start earth bolt. This is purely for shielding ac noise and i confirmed it WORKS.

Thus, thinking of doing the same approach on the AC wire input. Shield the twisted pair and connect this to star earth...

By the way the Belden 19364 is a very well known and respectable AC wire.
 
Hi Filipe,

Here's a quick reply for you!

You could use shield AC twisted wire from IEC to Switch (and from switch to transformers as well!) but I do not believe that it will reduce the noise because it's low frequency. If it were 1kHz it would be a big issue.

When you examine the Earth spade on the Maya, you will see that they are connected to the 0V rail reference on the pcb using two 3A back to back diodes. Put another way, current will only flow through this wire to star/chassis earth on the enclosure when a DC figure is outside the insulating window of -0.6V to +0.6V. It will never pass AC signal from the amp 0V reference down to star/chassis/power earth. This is quite intentional, because if you put in a 10R 2W resistor in parallel with the BtB diodes you actually bring mains ground into the amp 0V reference, and it can cause hum. Not always, but often. This system avoids the issue whilst affording protection in any DC arising in the amp (or even at mains ground!).

Shielded twisted pair for the input wire is very similar to balanced input and is good thinking. I do use good quality coax, it is my preference, but twisted, shielded is probably slightly better. Your shield and the neutral wire should both be connected to OUTER on the RCA plug, and at the other end to the neutral pin on the Maya module, as marked. The hot signal wire is then connected, as you have done it, from center point on RCA socket and at the other end, to the HOT pin on the Maya module.

I don't think you can do it any better, and it should be very, very quiet. By separating the two AC power supplies on the modules, with connection of ground only at the DC level (that is, no center tap on the transformer) hugely improves the S/N ratio of the amplifier because connecting a transformer CT secondary point to the 0V reference on an amplifier invites parasitic noise from the transformer directly into the modules. I really do not like using a full wave rectifier across positive and negative rails with ground to CT; it's the usual technique, but it's noisy, particularly if transformers are built with the CT slightly to one side of the 0V midpoint of the secondary.

How is the performance now it's completely run in, Filipe? Have you had a few audiophile friends have a good listen? What do they think?

Merry Christmas to you and your family!

Hugh
 
Hi Hugh, Hi everyone.

I am a new member and first of all a lot of health in the new year.
I'm interested in this post even though I do not know the Maya module.

Hugh, Hi Hugh, Hi everyone.

I am a new diyAudio Member and first of all I wish a lot of health in the new year.
I'm interested in this post even though I do not know the Maya module.
Hugh, I assume that the maya module is mono (not stereo), I am attaching the basic cabling scheme, it is well understood? It means that Maya has a separate signal ground and power ground?
An externally hosted image should be here but it was not working when we last tested it.


Hugh, where I can find detailed information about the Maya module? Is selling and has a final price?

Thanks

Jan
 
You are showing close coupled/twisted wire pairs for all the inter-module connections. Well done. So many Members here get this completely wrong.

You are showing a weird arrangement for the two solid "EARTH" wires.
What are you trying to achieve?

You have two broken lines from transformer to Safety Earth.
What are these?
 
Hi Andrew,
I assume that the Maya modules have separate signal ground and power ground (green wires), but I'm not sure it must be confirmed by Hugh.
If this is the case, it's perfect for minimal noise and hum.

Broken lines (green) - this is the case if the transformers have grounded shielding.

Thank for your remarks

Jan
 
Hi Andrew,
I assume that the Maya modules have separate signal ground and power ground (green wires), but I'm not sure it must be confirmed by Hugh.
If this is the case, it's perfect for minimal noise and hum.

Broken lines (green) - this is the case if the transformers have grounded shielding.

Thank for your remarks

Jan
If each Power Amplifier has separated Signal Return and Speaker Return, then a voltage referencing connection needs to be made between the two, either on board or across the board. Definitely not to a remote Chassis connection.

If however that remote chassis connection is a Safety connection, then it should not be a long wire across the chassis. It can be a short wire to the chassis adjacent to the PCB terminal.

Grounded shielding sounds like an inter-winding screen.
If you have an inter-winding screen, then the effectiveness of the screen (it is a capacitive filter) depends on the impedance to the Enclosure/Chassis. A low impedance allows the screen to be more effective at higher frequencies. Again a long wire to a remote Chassis connection is not the best way to achieve low impedance. A short wire/strip/braid to the enclosure, adjacent to where the wire comes out of the transformer insulation is better than a long wire trailing across the Chassis.
 
If each Power Amplifier has separated Signal Return and Speaker Return, then a voltage referencing connection needs to be made between the two, either on board or across the board. Definitely not to a remote Chassis connection.

If however that remote chassis connection is a Safety connection, then it should not be a long wire across the chassis. It can be a short wire to the chassis adjacent to the PCB terminal.

Grounded shielding sounds like an inter-winding screen.
If you have an inter-winding screen, then the effectiveness of the screen (it is a capacitive filter) depends on the impedance to the Enclosure/Chassis. A low impedance allows the screen to be more effective at higher frequencies. Again a long wire to a remote Chassis connection is not the best way to achieve low impedance. A short wire/strip/braid to the enclosure, adjacent to where the wire comes out of the transformer insulation is better than a long wire trailing across the Chassis.

I totally agree with everything..... yes it's an inter-windig screen (excuse my English) ..... the attached schematic is just a principle and it is not a real amplifier. The real construction (design) would be other .....

Attached is a picture of ASPEN Maya module with connecting points .... but it will probably not allow very short connections.

dsf3185.jpg

Thanks

Jan
 
Hi Jan,

Welcome to the forum! You have done a very good analysis of the pcb...... bravo!

The signal and speaker ground are linked with a 10R resistor on the pcb.

The chassis is isolated from the speaker ground with two back to back high current diodes. This for normal purposes the signal and speaker ground are isolated from the chassis and IEC ground unless there is a voltage appears between them of +/-0.6V. This would then ground the module to the chassis in the event of a failure, without any risk of electrical damage or breach of a safety.

There is a DC voltage protection on the speaker hot wire as well. Should there be more than DC 2V plus or minus an electronic relay disconnects it with a few milliseconds. This is an optodrive on two series high current mosfets. It also waits to connect the speaker at switch on so the amp has settled.

Hugh
 
Hi Jan,

Welcome to the forum! You have done a very good analysis of the pcb...... bravo!

The signal and speaker ground are linked with a 10R resistor on the pcb.

The chassis is isolated from the speaker ground with two back to back high current diodes. This for normal purposes the signal and speaker ground are isolated from the chassis and IEC ground unless there is a voltage appears between them of +/-0.6V. This would then ground the module to the chassis in the event of a failure, without any risk of electrical damage or breach of a safety.

There is a DC voltage protection on the speaker hot wire as well. Should there be more than DC 2V plus or minus an electronic relay disconnects it with a few milliseconds. This is an optodrive on two series high current mosfets. It also waits to connect the speaker at switch on so the amp has settled.

Hugh

Thanks Hugh,

Perfect, you also offered the world a great solution to the hum and noise problems. It can not connect incorrectly.
It sounds like a simple task, but it certainly needed a lot of effort and time.

All the modules I know need different bizarre solutions to remove hum and noise and the result is often not satisfactory, incorrect cable connection will make hum and noise.

:up:

P. S.
Hugh, where we find technical information and price? I did not find it on your web site.
 
Jan,

Thank you for your analysis. :happy1: Your diagram is perfect with the one exception that secondary #1 AC is connected via two spades on one end of the module and the secondary #2 AC is connected on the other end of the module. You have the connections correct; just the point at which they link up on the module. Because I was bringing AC into a amp pcb, I decided to install the power supplies, two of them positive and negative, at each end of the amp to minimise interference. It has worked out well and is utterly silent. I have found that good quality toroids are important too; one Maya did not present with silent background until I replaced the cheaper chinese toroids with high quality versions, then it was silent as the grave.

Not shown is the signal ground. I have the speaker 0V ground midpoint on the pcb, but the signal ground, the feedback shunt cap and the servo reference cap are all connected to the midpoint 0V spade via a 10R smd resistor. This is very important to ground lift the very sensitive input grounds.

I have been remiss with my website for years and hoping to fix it early this year. I have attached a pdf brochure of the Maya which describes most of the technical features, at least enough to get an idea that it is rather unusual and and evolution rather than a revolution.

Thank you for your continuing interest in Aspen!

Cheers,

Hugh
 

Attachments

  • MAYA200 BROCHURE.pdf
    654.9 KB · Views: 181
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.