Retain space and lead length to allow you to rotate the transformers.
You may find that you get a null in hum and buzz with one particular orientation for each transformer. The orientations may be different due to the two amplifier channels not being physically symmetrical.
You may find that you get a null in hum and buzz with one particular orientation for each transformer. The orientations may be different due to the two amplifier channels not being physically symmetrical.
Andrew, Is your objection to braiding (increases inductance) based on the additional wire length that results, or is it based on the interactions of the fields themselves. In my use braiding is generally done for power (hot/com/gnd) and twisting for signal as there are only two wires. Much of my application of braiding is for cosmetic effect but I'm not adverse to eliminating it if necessary.
I used twisted pairs for two wires, twisted triplets for three wires and twisted quads for 4wires.
I never braid.
You will often see that I post twisted triplet for PSU wiring leading to the amp PCB. This is very important. The three connections to the PCB must be very close.
The combination of twisted triplet and small loop area is paramount to reducing radiated fields.
Very many amp PCBs have the +ve and -ve at opposite sides/ends of the PCB and the Power Ground cms/inches away from the output devices. This is very bad for low level stages which are susceptible to picking up radiated fields.
I never braid.
You will often see that I post twisted triplet for PSU wiring leading to the amp PCB. This is very important. The three connections to the PCB must be very close.
The combination of twisted triplet and small loop area is paramount to reducing radiated fields.
Very many amp PCBs have the +ve and -ve at opposite sides/ends of the PCB and the Power Ground cms/inches away from the output devices. This is very bad for low level stages which are susceptible to picking up radiated fields.
You should always strive for small loop area. All electricity flows around a loop. This means that every Flow wire MUST be VERY CLOSE to it's Return wire. There is a Flow and a Return. Form a big loop and you create an aerial, both a transmitting aerial and a receiving aerial. Close couple the Flow and Return and you virtually eliminate the aerial by using LOW loop area.
Twisting is one of the better ways for loose wiring to ensure low loop area.
On a PCB, using two layers that have a very thin dielectric between the parallel traces gives the least loop area. There are other advantages to using thin traces and thin dielectric in a multilayer PCB, but these are mostly in the VHF and higher rather than for audio frequencies.
BUT,
there are frequencies in an audio amplifier that are way above audio.
Some of the multilayer techniques for good VHF performance are directly transferable to audio amplifiers.
Twisting is one of the better ways for loose wiring to ensure low loop area.
On a PCB, using two layers that have a very thin dielectric between the parallel traces gives the least loop area. There are other advantages to using thin traces and thin dielectric in a multilayer PCB, but these are mostly in the VHF and higher rather than for audio frequencies.
BUT,
there are frequencies in an audio amplifier that are way above audio.
Some of the multilayer techniques for good VHF performance are directly transferable to audio amplifiers.
It depends on speakers sensitivity, mines are 89dB and that heatsink was enough
I'm sure Andrew means exaclty that,
look at magnetic field:
An externally hosted image should be here but it was not working when we last tested it.
(Image from Custom Magnetics)
In fact in my build they were oriented correctly.
This is another argument.
BTW, yes, it would be better.
Clave,
your diagram of an Rcore shows a separate primary coil from the secondary coil. This is rarely the way a good Rcore is built.
Both legs will usually have a virtually identical primary AND secondary on each of them for best Rcore performance.
The primary and the secondary may even be interleaved (layered) one over the other for best performance. Unfortunately with added interlayer capacitance.
your diagram of an Rcore shows a separate primary coil from the secondary coil. This is rarely the way a good Rcore is built.
Both legs will usually have a virtually identical primary AND secondary on each of them for best Rcore performance.
The primary and the secondary may even be interleaved (layered) one over the other for best performance. Unfortunately with added interlayer capacitance.
Last edited:
Ok. I put traffos parallel to side walls to reduce magnetic interference.
I twist, not braid wires.
Remaining question is c13 ... but I ll probably a) overhung it and attach it to spacers or b) connect it to rca directly
I twist, not braid wires.
Remaining question is c13 ... but I ll probably a) overhung it and attach it to spacers or b) connect it to rca directly
Hi Andrew,
the diagram is not mine, I've taken from Custom Magnetics web site and it's similar to Kitamura-Kiden one:
An externally hosted image should be here but it was not working when we last tested it.
(From Kitamura-Kiden web site)
The goal was to suggest best orientation.
BTW interesting info on R-Core construction, thanks 🙂
I should have more carefully worded my post to say the diagram that you posted.
Sorry, it should be intended as:
Yes, twisted would be better.
No problem at all, Andrew 🙂
Connected boards.
One is working fine.
Second is not. No click and no led.
can you please tell me what to check.
One is working fine.
Second is not. No click and no led.
can you please tell me what to check.
Hi Bob.
I left it at work as it is the only place I could use the drill
I ll come back tomorrow to do that.
I d appreciate if you give me a list of possible actions.
I ll finish holes in case and take things home.
And probably tomorrow evening can do something .
A bit disappointed though 🙁
I left it at work as it is the only place I could use the drill
I ll come back tomorrow to do that.
I d appreciate if you give me a list of possible actions.
I ll finish holes in case and take things home.
And probably tomorrow evening can do something .
A bit disappointed though 🙁
An obvious question, but...you are sure the transformer is functioning and properly connected to the board?
Don't feel bad. You have the most often experienced start up problem with all MyRefs. Had it a couple times myself. Cold solder joints, solder bridges and wrong value resistors have caused most of the problems, as I remember.
Last edited:
Yes, I used same transformer with 2 boards. One worked fine, so I connected it to another board.