Pasi P said:Here is my amp with T-net caps. Wires are messy yet.
Grounding middle of the caps with 6mm2 wire. All other wires 2,5mm2. Suggestions to improve something in lay-out?
Love all the attention to detail and expert machine work. Nice layout with short paths. I certainly can appreciate how much time and effort went into this project.
Now undo all the transformer leads and twist them in pairs. Do the output wires as twisted pairs too. This will look better and sound better due to lower noise and cross talk.
Roger
This will look better and sound better due to lower noise and cross talk.
How can output wires add to crosstalk? Only if they are near the open ends of the input wires(unbalanced) or input electronics. Proper earthing eliminates most crosstalk over the supply lines.
Pasi P,
I see you have 2 modules next to each other, is this a stereo set or bridge? The output coil will stray on the input opamp underneath the other module and causes crosstalk. I had this experience with my first build amp. Look on the scope and channel output. better turn 1 amp 180 degrees so the 2 output coils are facing eachother or the 2 input amps.
I see you have 2 modules next to each other, is this a stereo set or bridge? The output coil will stray on the input opamp underneath the other module and causes crosstalk. I had this experience with my first build amp. Look on the scope and channel output. better turn 1 amp 180 degrees so the 2 output coils are facing eachother or the 2 input amps.
Hi Bert,
It is 4-channel amp for my Linkwitz Orions M&T.
Interesting finding. First time i heard that modules can do somethig bad to eachother when near each other.
It is 4-channel amp for my Linkwitz Orions M&T.
Interesting finding. First time i heard that modules can do somethig bad to eachother when near each other.
after some extensive "cap hunting" i've found a dealer of bc components mkp379's (the ones bruno was talking about for the output filter cap). They are somewhat smaller than the mkp479's, which perform pretty good imho, and have similar specs. Not only that, they are box-type and not dipped epoxy.
As the specs are pretty similar i wouldn;t expect a big difference but has anyone tried these already?
how much of a difference does an external supply for the low voltage section make?
As the specs are pretty similar i wouldn;t expect a big difference but has anyone tried these already?
how much of a difference does an external supply for the low voltage section make?
Hi Matjans,
Do they have the caps in quantities?
I'd expect the benefits seen from having an isolated input supply could vary widely.It would be far more significant if you undersized your transformer and therefor didn't have the best load regulation possible,and would also depend on whatever else you did with your input stage to perhaps improve PSRR or otherwise.
Regards,
Chris
Do they have the caps in quantities?
I'd expect the benefits seen from having an isolated input supply could vary widely.It would be far more significant if you undersized your transformer and therefor didn't have the best load regulation possible,and would also depend on whatever else you did with your input stage to perhaps improve PSRR or otherwise.
Regards,
Chris
Bgt said:
How can output wires add to crosstalk? Only if they are near the open ends of the input wires(unbalanced) or input electronics. Proper earthing eliminates most crosstalk over the supply lines.
We are talking HF cross talk here. 2 bare wires running near each other will act like a transformer and actually couple together quite effectively at high frequencies. I think since there is a good bit of residual switching frequency left in the output they would couple more than you would expect. Remember this energy is rather large as we are talking outputs here. This can also find its way into the inputs and be a further cause of degradation.
Twisting them together takes very little effort and costs nothing extra. It is an improvement so why not do it?
Roger
It is an improvement so why not do it?
You're absolutely right but when I measure the crosstalk between outputs it is very low already even with untwisted output wires. I always want to have at least 90db at 15kc separation/crosstalk. If I don't get that there must be something wrong. But I will try this with the next amp and compare.
i got the caps from ecomal ( www.ecomal.com ), a european vishay-importer. They were the only one in the netherlands i tried that could get their hands on the mkp379's.
Hi Matjans, do you already have them? How many did you order? Tried them already?matjans said:i got the caps from ecomal ( www.ecomal.com ), a european vishay-importer. They were the only one in the netherlands i tried that could get their hands on the mkp379's.
yes i already have them, I ordered 5, haven;t tried them out yet, i'm pretty happy with the mkp479's.
Bgt said:Pasi P,
The output coil will stray on the input opamp underneath the other module and causes crosstalk. I had this experience with my first build amp. Look on the scope and channel output. better turn 1 amp 180 degrees so the 2 output coils are facing eachother or the 2 input amps.
Okay, I turn other module of channel upside down (coil not anymore near other module op-amp). I also twisted speaker cables same time, as Roger recommend.
And.. What a improvement!
Do not know wich one does the change or both.. I havent investigated.
Changing T-net caps is nothing compared with this!
How it is possible? Can there be so much crosstalk with old lay-out? (adjacent modules are in same channel; mid+tweeter amp). Or can it be twisted speaker cables, can not believe.
Anyway, if somebody has same kind of lay-out than mine (picture in Post #401 this thread), i very strongly recommend do the same. And twist speaker cables also..
Hi Pasi,
It's from twisting the wires, unless you're hearing whistling, the module is shielded and very well bypassed, with great care taken in selection of the filter core for lowest radiation possible.
Goes to show you proper techniques aren't are worth the effort especially for the little they cost, and a little research in this area can go a long way.
Regards,
Chris
It's from twisting the wires, unless you're hearing whistling, the module is shielded and very well bypassed, with great care taken in selection of the filter core for lowest radiation possible.
Goes to show you proper techniques aren't are worth the effort especially for the little they cost, and a little research in this area can go a long way.
Regards,
Chris
"It's in the details"
Hi guys,
Yes, to quote Bruno “it's in the details”. I have been astounded to hear the difference that replacing 1” of wire caused. A lot of little improvements like this cost nothing but some time. I know we can get very impatient to hear our latest wonders but one has to remember that a well constructed piece will last many years so is worth our best effort. For instance I am now upgrading the volume control in a custom preamp I did over 8 years ago. It still is his baby and working very well. For projects that can give years of enjoyment isn’t the extra care justified?
Roger
Hi guys,
Yes, to quote Bruno “it's in the details”. I have been astounded to hear the difference that replacing 1” of wire caused. A lot of little improvements like this cost nothing but some time. I know we can get very impatient to hear our latest wonders but one has to remember that a well constructed piece will last many years so is worth our best effort. For instance I am now upgrading the volume control in a custom preamp I did over 8 years ago. It still is his baby and working very well. For projects that can give years of enjoyment isn’t the extra care justified?
Roger
the module is shielded and very well bypassed, with great care taken in selection of the filter core for lowest radiation possible.
This is absolutely not the case. Try it and put the probe on the output of the amp. with the preamp next to the output coil of the other amp. you inject some frequency in. Now you will see a lot of crosstalk at the higher frequencies. Ofcourse load both modules with 8 ohm resistors and put a 600 ohms resistor on the input of the amp. you measure.
- Status
- Not open for further replies.
- Home
- Amplifiers
- Class D
- Hotrodding the UCD modules