I was hoping there might be someone in Michigan that might be close to the seller and be able to pickup and avoid a double shipping bill. I'll contact the seller and find out what the breakdown is for discounts.
I am glad to see one of the international group buy folks had received there boards. Hoorah for the postal service -- so far--.
The guy on ebay with the heatsinks seems to be a really nice guy from my involvement with him. He had mentioned a larger more to our liking extrusion he was working on. If someone would contact him I too would be interested in a few sinks. fThough I am not sure that a group buy would be necessary for this. Direct sale might be just fine.
Tad
The guy on ebay with the heatsinks seems to be a really nice guy from my involvement with him. He had mentioned a larger more to our liking extrusion he was working on. If someone would contact him I too would be interested in a few sinks. fThough I am not sure that a group buy would be necessary for this. Direct sale might be just fine.
Tad
Thanks to Tryonziess bring up this GB.
I got it yesterday and feel what it so successful masterpiece.
Starting to collect components but a question about the PWD-GND.
Where is the suitable location on PCB run the thick wire to the center tap of the main capacitor. (I intend to use 2 more largest capacitors per channel out of the PCB)
I am not sure it is the hole underneath of the fuse holder ?
One more question about the output DC offset control. Why this circuit can't find DC offset control ?
I got it yesterday and feel what it so successful masterpiece.
Starting to collect components but a question about the PWD-GND.
Where is the suitable location on PCB run the thick wire to the center tap of the main capacitor. (I intend to use 2 more largest capacitors per channel out of the PCB)
I am not sure it is the hole underneath of the fuse holder ?
One more question about the output DC offset control. Why this circuit can't find DC offset control ?
The hole in the center of the fuse holder is for a #6 screw to mount the fuse holder.
There is a nice ground point near the inputs - that is the signal ground, also connect the ground at the output to your star ground.
There is a nice ground point near the inputs - that is the signal ground, also connect the ground at the output to your star ground.
I got a reply from the guy on ebay selling the heatsinks, the breakdown is $24 each up to 10ft(10 peices), $1 per foot disount for every 10ft ordered thereafter. Unless there is a person to locally pickup the heatsinks, any discount gets eaten up on re-shipping for a group buy.
However, which type of bipolar capacitors is most suitable for C7, C12 ? What do you plan to use ? 😕
I use Nichicon - http://www.mouser.com/Search/Produc...PDvirtualkey64700000virtualkey647-UVP1C471MPD as long as you have a film bypass, it doesn't really matter, IMHO.
Nichicon Muse are 12 mm diameter. The footprint on the board is 10 mm. You could make them fit by mounting them sloghtly above the board. I'd fill the gap below them with hot melt if you do that.
Nichicon Muse are 12 mm diameter. The footprint on the board is 10 mm. You could make them fit by mounting them sloghtly above the board. I'd fill the gap below them with hot melt if you do that.
I received boards
thx
which output transistors to chose - original MJL 3281 / 1302 or new NJW 0281 / 0302
what Is best sounding option
are there any pos If chose MJL 4281 / 4302
I don't need more than 150 W per channel
thx
which output transistors to chose - original MJL 3281 / 1302 or new NJW 0281 / 0302
what Is best sounding option
are there any pos If chose MJL 4281 / 4302
I don't need more than 150 W per channel
The NJWs are supposed to be better matched between N and P although the spec sheet calls out the same matching numbers. Theoretically they should sound better. There is also a NJW3281/1302 pair that has a bit larger SOA
The MJL3281/1302 has a larger SOA than the NJW0281, so the amp will be more rugged. I haven't compared the SOA, but would expect it to be similar.
The NJW0281/0302 have slightly lower input capacitance, which could improve the upper end compared to the MJW3281/1302. In this amp, the upper limit is driven by the driver feedback loop, so I suspect that you are unlikely to hear a difference.
The M4281/4302 are really only needed if you are pushing the rail voltages and/or want to drive low impedance or very reactive loads. I've heard that they sound a bit better than the MJL3281/1302 though I don't have any experience.
You didn't mention your planned rail voltage, but from your 150W requirement, I guess you're shooting for ~60V. The NJWs would be fine (unless you're going crazy low impedance). Use the NJW3281/1302 at $1.70 ea. for a bit more SOA margin than the NJW0281/0302 at $1.45 ea.
EDIT: just to add to analysis paralysis, take a look at FJL4315/4215 available at Mouser. Similar SOA to the NJW0281 at the same price. 😉
The MJL3281/1302 has a larger SOA than the NJW0281, so the amp will be more rugged. I haven't compared the SOA, but would expect it to be similar.
The NJW0281/0302 have slightly lower input capacitance, which could improve the upper end compared to the MJW3281/1302. In this amp, the upper limit is driven by the driver feedback loop, so I suspect that you are unlikely to hear a difference.
The M4281/4302 are really only needed if you are pushing the rail voltages and/or want to drive low impedance or very reactive loads. I've heard that they sound a bit better than the MJL3281/1302 though I don't have any experience.
You didn't mention your planned rail voltage, but from your 150W requirement, I guess you're shooting for ~60V. The NJWs would be fine (unless you're going crazy low impedance). Use the NJW3281/1302 at $1.70 ea. for a bit more SOA margin than the NJW0281/0302 at $1.45 ea.
EDIT: just to add to analysis paralysis, take a look at FJL4315/4215 available at Mouser. Similar SOA to the NJW0281 at the same price. 😉
thx for yor reply
yes NJW 3281 / 1302 has to be best bargain but It Is TO-3P package not TO-264
can It be used with this pcb
yes NJW 3281 / 1302 has to be best bargain but It Is TO-3P package not TO-264
can It be used with this pcb
The lead spacing is the same and the T0-3P package is ~4 mm NARROWER than the TO264. Yes, it works fine.
Data sheets are your friends 😉
Data sheets are your friends 😉
Rod Elliot swears by the 4281 output devices. He has made several strong comments in favor of these particular devices. I believe that Jens is also a proponent of the mjl4281 as I am.
They are expensive at the onset, however an equivalent amplifier of this caliber would quite possibly set you back a few thousand dollars. And It is really only a few dollars more in the overall scheme. The chassis and power supply capacitors will be the main cost factor in this amp.
You probably will never HEAR the difference but you will have a stronger output section with more rugged devices. Shop around it pays off try Avnet.com. -- a little hint from Bob -- nice folks.
And the DIY community really does have some deep knowledge.
Tad
They are expensive at the onset, however an equivalent amplifier of this caliber would quite possibly set you back a few thousand dollars. And It is really only a few dollars more in the overall scheme. The chassis and power supply capacitors will be the main cost factor in this amp.
You probably will never HEAR the difference but you will have a stronger output section with more rugged devices. Shop around it pays off try Avnet.com. -- a little hint from Bob -- nice folks.
And the DIY community really does have some deep knowledge.
Tad
R3 and R22 calculation
I have an Avel D4062 Toroid that I totally forgot. The specs are:
primary: 2x 0-120
secondary: 2 x 0-40
530VAC
After the drop in the bridge rectifier and the CL3, I guess I will be getting somewhere at 54VDC. Use the formula in the BOM, R3 and R8 = (54-40)/8.2*1000 = 1.7K
Is that correct?
Thanks!!
I have an Avel D4062 Toroid that I totally forgot. The specs are:
primary: 2x 0-120
secondary: 2 x 0-40
530VAC
After the drop in the bridge rectifier and the CL3, I guess I will be getting somewhere at 54VDC. Use the formula in the BOM, R3 and R8 = (54-40)/8.2*1000 = 1.7K
Is that correct?
Thanks!!
Hi,
my 230:40Vac transformer gives +-58Vdc from 240Vac.
The amp puts out 170W into 8r0 just before clipping.
my 230:40Vac transformer gives +-58Vdc from 240Vac.
The amp puts out 170W into 8r0 just before clipping.
so far best to use output devices with our pcb theoreticaly are
1 MJL 4281 / 4302
2 NJW 3281 / 1302
3 MJL 3281 / 1302
my choice are NJW 3281 / 1302 but Digikey has not got In stock
Mouser has them but they use UPS and that will go too expensive for me
does anybody want to share a buy from Mouser and ship to me
I will buy 25 pcs pairs
1 MJL 4281 / 4302
2 NJW 3281 / 1302
3 MJL 3281 / 1302
my choice are NJW 3281 / 1302 but Digikey has not got In stock
Mouser has them but they use UPS and that will go too expensive for me
does anybody want to share a buy from Mouser and ship to me
I will buy 25 pcs pairs
Yeah, my main is usually around 120-123VAC most of the time. Since this toroid is rated for 120 primary, I kind of don't expect too much on the upside of the secondary. Since my speakers are 4 ohm, I probably need a hotter bias more than the voltage. Do you think using 0.33R for the emitter will overload this toroid? Also, does 1.7K sound right for R3 and R8?
Thanks!!
Thanks!!
In this amp the emitter resistors have nothing to do with setting bias.
Five 0R47 emitter resistors in parallel give a DC resistance of <0R1. However, this resistance is inside the feedback loop so the effective output impedance is much lower. I don't know if anyone has measured the output impedance of this version, but I am sure that is signifcantly lower than 0R1.
IIRC Self suggests that "optimum" bias for lowest distortion means about .25V across the emitter resistors with bipolar outputs. That's 0.5A with 0R47 resistors, or .75A with 0R33. At 60V rails that's 300W or 420W dissipation for 10 output devices.
Guess we're not going there. But, I think this means that 0R47 is a better choice from a distortion perspective.
Pass' Aleph series use the source resistors in the active current sources to set the bias.
Five 0R47 emitter resistors in parallel give a DC resistance of <0R1. However, this resistance is inside the feedback loop so the effective output impedance is much lower. I don't know if anyone has measured the output impedance of this version, but I am sure that is signifcantly lower than 0R1.
IIRC Self suggests that "optimum" bias for lowest distortion means about .25V across the emitter resistors with bipolar outputs. That's 0.5A with 0R47 resistors, or .75A with 0R33. At 60V rails that's 300W or 420W dissipation for 10 output devices.
Guess we're not going there. But, I think this means that 0R47 is a better choice from a distortion perspective.
Pass' Aleph series use the source resistors in the active current sources to set the bias.