Sure wrong
Sure is wrong, aren't they? I'm sorry I don't have a solution for you other than to use a larger enclosure with enough room for each amp module to have it's own power supply.
Sure is wrong, aren't they? I'm sorry I don't have a solution for you other than to use a larger enclosure with enough room for each amp module to have it's own power supply.
I'm using 2 boards attached to the same PSU, they are playing since hours and no problems. The only issue is the "mute" which bumps in all 4 speakers when one board is already playing and i unmute the other. not a big issue anyway.
Tone?
Leave everything turned on and stop the playback. Do you here a tone when you stick your ear right up to the woofer or tweeter? Which power supply do you use?
Leave everything turned on and stop the playback. Do you here a tone when you stick your ear right up to the woofer or tweeter? Which power supply do you use?
Usually two boards need their PGND referenced to each other by a singel connection and then there should be no noise at all...
Star ground
.
I can try one more thing. My old house has two prong electrical outlets still in the listening room. The kitchen has been updated to three prong wiring. I can try adding a wire from a third prong ground in the kitchen to the ground terminal of the power supply and tie that to the star point.
.
I believe Jack is using a 12v battery for the dsp and two amp boards.
I made a star ground at the negative power supply terminal. Still has the noise. negative wire from left board, negative wire from right board, negative from source rca cable, all connected at the same screw of the power supply terminal. left rca + to one input of the left amp, right rca + to the input of the right amp. Plays music fine except for a faint 1500Hz? tone which is not so faint in Jack's 98db efficient horn speakers.
.
I can try one more thing. My old house has two prong electrical outlets still in the listening room. The kitchen has been updated to three prong wiring. I can try adding a wire from a third prong ground in the kitchen to the ground terminal of the power supply and tie that to the star point.
.
I believe Jack is using a 12v battery for the dsp and two amp boards.
I hear a slight noise on the tweeter with volume higher than 3/4, but it happen even when the second board is disconnected.
I use this PSU: HF240W-SF-30
I have one board on the desktop (made out of wood) and the other one is in the enclosure standing on plastic feet so that the two boards are both disconnected from earth. This is a test because soon im planning to bi-amping my klipsch RF-5. Does it worth a try or bi-amping it's useless like a lot of persons says?
The board already connects the RCA ground to PGND, so your star ground is in fact a loop then!
Try just one single wire between both PGND's, I've seen this work out fine a lot of times....but not on a Sure board....
MAN! The Sure noises still continue! Their first board was terrible, this one is better, hope the next one finally got it all right! Persons having been on the train from the start have not been off that cheap afterall.....
Another set of batteries
Since you have high efficiency drivers and want to keep everything small, you could try two of the 145 watt Meanwell supplies at 24v.
.
24V DC 6A 145W Regulated Switching Power Supply New - eBay (item 160435570855 end time Jun-15-10 19:02:22 PDT)
.
In order to minimize hiss and noise, and to get the best resolution in your dsp crossover, I would experiment with reducing the gain of the the amps at the feedback resistors to a level that provides the most gain that the speaker drivers need in order to be just too loud when the dsp chips are getting nearly full analog in. Building an even gain structure at every stage will minimize noise and maximize digital resolution in the ADC. This may mean using a lower supply voltage than 24v for the amps. Maybe you should buy another set of batteries first in order to get rid of the noise by running dual mono supplies until you get the gain figured out and then determine what the max supply voltage can be before buying any supplies.
.
Do the Minidsp modules have any way to show you an input level meter?
Since you have high efficiency drivers and want to keep everything small, you could try two of the 145 watt Meanwell supplies at 24v.
.
24V DC 6A 145W Regulated Switching Power Supply New - eBay (item 160435570855 end time Jun-15-10 19:02:22 PDT)
.
In order to minimize hiss and noise, and to get the best resolution in your dsp crossover, I would experiment with reducing the gain of the the amps at the feedback resistors to a level that provides the most gain that the speaker drivers need in order to be just too loud when the dsp chips are getting nearly full analog in. Building an even gain structure at every stage will minimize noise and maximize digital resolution in the ADC. This may mean using a lower supply voltage than 24v for the amps. Maybe you should buy another set of batteries first in order to get rid of the noise by running dual mono supplies until you get the gain figured out and then determine what the max supply voltage can be before buying any supplies.
.
Do the Minidsp modules have any way to show you an input level meter?
loop
How is it a loop?
The board already connects the RCA ground to PGND, so your star ground is in fact a loop then!
Try just one single wire between both PGND's, I've seen this work out fine a lot of times....but not on a Sure board....
MAN! The Sure noises still continue! Their first board was terrible, this one is better, hope the next one finally got it all right! Persons having been on the train from the start have not been off that cheap afterall.....
How is it a loop?
Actually, I have.
Since I have received a 4*100W board as 2*100W (4-2X100 as you proposed) few days ago it is too early for any comparison.
I've accommodated a discrete linear supply from previous projects and changed input caps so far.
I am waiting for the Meanwell smps and Wurth coils you've suggested.
My take on 3886 is very serious one: 300VA trafo with 4 duals, separate discrete voltage regulators for each channel, boutique parts...
The DIY preamp is equally ambitious project.
We'll see. I'll let you know.
Baki
Ok.
Ok.
But:
Either you are connecting source GND (which should go to AGND) to PGND or you (and then there's a real loop) connect the source GND to both (because there is already one single AGND-PGND connection on the board). Either case can result in noise, read the application notes.
So what was AGND doing?
Source GND shouldn't be at PGND at all, and certainly not bolted to any casing... Not that I care, but it's not really a good test then is it?
What struck me as well is that the Sure board uses a resistor of zero ohms to connect AGND and PGND. Even though this works according to Tripath application notes I noticed most designs use a small ferrite choke. I've connected whole bunches of amps together many times and used very sensitive PA speakers and never got such noise problems, so I may be able to help analize the problem, that's all...
I'm sorry, but I see a lot of work being done like the sorcerer's apprentice, no scientific approach whatsoever. Guys spending hours, perhaps even days listening to the sound of inductors. Even though it can be quite entertaining, technically it's not really getting anywhere.
Source GND shouldn't be at PGND at all, and certainly not bolted to any casing... Not that I care, but it's not really a good test then is it?
What struck me as well is that the Sure board uses a resistor of zero ohms to connect AGND and PGND. Even though this works according to Tripath application notes I noticed most designs use a small ferrite choke. I've connected whole bunches of amps together many times and used very sensitive PA speakers and never got such noise problems, so I may be able to help analize the problem, that's all...
I'm sorry, but I see a lot of work being done like the sorcerer's apprentice, no scientific approach whatsoever. Guys spending hours, perhaps even days listening to the sound of inductors. Even though it can be quite entertaining, technically it's not really getting anywhere.
Last edited:
I am using IBM Stinkpad 19V 3.4A PS, what can I improve there? Amp is totally silent with it and I am not sure I hear any distortion any more.
I am thinking into finally building an 2-way active speaker with linkwitz-riley active crossover in front of sure board. Anyone tried it already (I confess I did not go over all 170 pages in this thread)?
Too small
.
The silver fan is also a known culprit for bad sound. Stop it momentarily while the amp is playing to hear it's effect. You will want to change it for a Zalman heat sink.
.
If you go active, you can sometimes get a decent crossover using textbook filters with something like a DCX2496 preceded by a DEQ2496 and adjusting the EQ by ear but you will eventually want to put together a measurement rig with ARTA to develop a better crossover.
For one thing that power supply is a little too small. That is only 65 watts and you are trying to power a 200 watt amp. Also, I have read many posts by people using various laptop power supplies that were disappointed with the results and then found a big improvement when changing to the 350 watt MeanWell. The feedback resistors in the amp are set to give the best sound at 32v so you may buy the 36v supply and turn down the front panel adjuster to 32. This voltage level will allow you to achieve the rated power. Less volts at 19v, less power. You will also want a seperate power supply for each amp board if you use more than one. This will prevent the odd noises we have been talking about and allow you to position the amps close to the left and right speakers in order to minimize the speaker wires for another improvement in sound.
.
The silver fan is also a known culprit for bad sound. Stop it momentarily while the amp is playing to hear it's effect. You will want to change it for a Zalman heat sink.
.
If you go active, you can sometimes get a decent crossover using textbook filters with something like a DCX2496 preceded by a DEQ2496 and adjusting the EQ by ear but you will eventually want to put together a measurement rig with ARTA to develop a better crossover.
I am not playing even remotely close to 200W at all, and my mistake - it's 20V 4.5A, so 90W. I don't think I am playing more than 10 watts into 8ohm, but ok - I can give it a try with more muscle. I can also slip it into a GC, with two 220VA torroids and 35V on rails. Any benefits from running on heavy unregulated torroid PS?
Anyway, if I eliminate passive crossover losses, I am guessing that even 90W going directly to speaker drivers should be loud enough.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.