just to be on the safe side i would like to ask some questions about that preamp.
1, would it benefit from a couple of extra RC stages after the rectier and after the voltage stabb. ?
2, would the relay ladder circuit from P1.0 work well on the output of the BSOZ or must i recalculate the resistor values....?
3, How big a transformer will suffice? 100VA 200VA?...
4, how much the output mosfets disipate?...
1, would it benefit from a couple of extra RC stages after the rectier and after the voltage stabb. ?
2, would the relay ladder circuit from P1.0 work well on the output of the BSOZ or must i recalculate the resistor values....?
3, How big a transformer will suffice? 100VA 200VA?...
4, how much the output mosfets disipate?...
Interesting questions since I am building that circuit myself. Hopefully Nelson will stop by here later. This is what I gathered in my preliminary survey of other people experiences with this design:
1) I am sure that RC filters would improve the PS noise figure at least until the noise reaches a low plateau. Recently someone reported hum problems so additional filtering maybe worthwhile.
3) I have just ordered an 80VA 60+60 V toroidal from Plitron so that better be enough!
Nelson is using two 30VA trafos.
4) The two channels should absorb a total of 160 mA. With this current figure each mosfet in the power supply section should dissipate 20 x 0.16 = 3.2 W. The actual amplifing devices should absorb considerably less.
I hear that the cooling sinks suggested by Nelson are a tad too small for comfort and that the resistor 3 W rating is also rather bold as these res. get very very hot. I don't remember for sure but 1W dissipation seems about right for each resistor.
[Edited by grataku on 11-08-2001 at 12:36 PM]
1) I am sure that RC filters would improve the PS noise figure at least until the noise reaches a low plateau. Recently someone reported hum problems so additional filtering maybe worthwhile.
3) I have just ordered an 80VA 60+60 V toroidal from Plitron so that better be enough!
Nelson is using two 30VA trafos.
4) The two channels should absorb a total of 160 mA. With this current figure each mosfet in the power supply section should dissipate 20 x 0.16 = 3.2 W. The actual amplifing devices should absorb considerably less.
I hear that the cooling sinks suggested by Nelson are a tad too small for comfort and that the resistor 3 W rating is also rather bold as these res. get very very hot. I don't remember for sure but 1W dissipation seems about right for each resistor.
[Edited by grataku on 11-08-2001 at 12:36 PM]
Tony,
320 VA and 16000uF for the preamp: what do you use for your power amp??!!
Are these types of over contruction really required here? I mean I like preamps to look like preamps, kind of slim and cute.
Was there any particular reason (ie high ripple noise) to go to such extremes in filtering caps? Did you actually compared your version to the original Pass version?
320 VA and 16000uF for the preamp: what do you use for your power amp??!!
Are these types of over contruction really required here? I mean I like preamps to look like preamps, kind of slim and cute.
Was there any particular reason (ie high ripple noise) to go to such extremes in filtering caps? Did you actually compared your version to the original Pass version?
Dual Split Bobbins
I'm only using 24 VA for each 2 channels of my 4 channel BLS pre-amp
If N. Pass doesn't get hum from using his original design and you do - try looking over what you've done again.
Don't forget the power supply board ALONE is regulated, thus even more reducing more noise (hence, no need for out of this world large uF caps).
I'm only using 24 VA for each 2 channels of my 4 channel BLS pre-amp
If N. Pass doesn't get hum from using his original design and you do - try looking over what you've done again.
Don't forget the power supply board ALONE is regulated, thus even more reducing more noise (hence, no need for out of this world large uF caps).
Hi Grataku
I had the transformers and caps. kicking around and after posting here regarding the positive aspects of oversize power supplies I replaced the stock BOSOZ power supply with the Godzilla one and everything was just more "there".
Absolutely effortless and huge soundstage and probably a bit of placebo effect as well, you know, it's big so it's got to be good.
Tony D.
I had the transformers and caps. kicking around and after posting here regarding the positive aspects of oversize power supplies I replaced the stock BOSOZ power supply with the Godzilla one and everything was just more "there".
Absolutely effortless and huge soundstage and probably a bit of placebo effect as well, you know, it's big so it's got to be good.
Tony D.
Will my 1010VA Transformer work?
I ordered 2 of the
1010VA 37.5V transformer from Apex jr
The nice guy there said that he measured the output of
the 4 taps at 37.5 volts under NO load.
My dream SOZ would be about 25 watts/ch, which
would use a transformer rated at 32 volts
My thought was that the output voltage would sag down to about that much if rated as manufacturers rate them (under load-right?
Also, the outputs aren't necessarily center tapped.
But- couldn't I just parallel 2 to get more amps, the connect the black wires to create a common. Then wouldn't the remaining wires be plus and minus supplys? I should wait for their arrival, but I need to know if I've screwed up ordering them.
I'm pretty ignorant obviously, but I've gotten responses from
Nelson Pass the various times I've written him. This time I'm trying to avoid bothering him again!
One more vote for Nelson being a great guy!!!
I hope useful info for other SOZ builders:
Apex jr has a bunch of 8 ohm 50 watt alu power resistors
at $2 ea
I ordered 32!
also 6 1 ohm 50 watters
Maybe not non-inductive-but what can you do?
I can't afford to pay $6 apiece for them!
I got six pretty huge heatsinks per channel from B.G. Micro:
http://bgm.bgmicro.com/prodinfo.asp...3&prodid=ACS1341&page=1&cri=heat+sink&stype=3
They are each 4" x 10.5" x 1" and only 3.95 ea.
I got a bunch of power supply caps from these guys:
http://hometown.aol.com/brigar2/capacitors/capacitors.html
30,000 mfd at 63 volts
If this doesn't take you to the right place, just select computer grade capacitors. These say McIntosh on them-At least they were designed for power amps!
I have no connection to these guys but I've searched the web pretty hard and these seem to be the best deals.
Any help on the transformers?
Mark Cronander
I ordered 2 of the
1010VA 37.5V transformer from Apex jr
The nice guy there said that he measured the output of
the 4 taps at 37.5 volts under NO load.
My dream SOZ would be about 25 watts/ch, which
would use a transformer rated at 32 volts
My thought was that the output voltage would sag down to about that much if rated as manufacturers rate them (under load-right?
Also, the outputs aren't necessarily center tapped.
But- couldn't I just parallel 2 to get more amps, the connect the black wires to create a common. Then wouldn't the remaining wires be plus and minus supplys? I should wait for their arrival, but I need to know if I've screwed up ordering them.
I'm pretty ignorant obviously, but I've gotten responses from
Nelson Pass the various times I've written him. This time I'm trying to avoid bothering him again!
One more vote for Nelson being a great guy!!!
I hope useful info for other SOZ builders:
Apex jr has a bunch of 8 ohm 50 watt alu power resistors
at $2 ea
I ordered 32!
also 6 1 ohm 50 watters
Maybe not non-inductive-but what can you do?
I can't afford to pay $6 apiece for them!
I got six pretty huge heatsinks per channel from B.G. Micro:
http://bgm.bgmicro.com/prodinfo.asp...3&prodid=ACS1341&page=1&cri=heat+sink&stype=3
They are each 4" x 10.5" x 1" and only 3.95 ea.
I got a bunch of power supply caps from these guys:
http://hometown.aol.com/brigar2/capacitors/capacitors.html
30,000 mfd at 63 volts
If this doesn't take you to the right place, just select computer grade capacitors. These say McIntosh on them-At least they were designed for power amps!
I have no connection to these guys but I've searched the web pretty hard and these seem to be the best deals.
Any help on the transformers?
Mark Cronander
I think your post is for SoZ (not related to Bride of SOZ pre-amp).
Anyways, if you're looking 25 watt SoZ - 6 of those heatsinks you've ordered is quite a bit on the lean side. Imagine that almost all 1000 watts (1 channel) is being dissipated into 6 of those heatsinks.
I'm building up a collection of heatsinks myself below:
http://www.geocities.com/super_bq/Heatsinks.html
Getting a transformer that is center-tapped or not is based upon the type of power supply configuration you plan to use. You don't have to do "dual-complimentary" that uses the center-tap (like I did on my SoZ). Full-wave rectification should do the job - only disadvantage is the possibility of more mechanical hum from the transformer.
Anyways, if you're looking 25 watt SoZ - 6 of those heatsinks you've ordered is quite a bit on the lean side. Imagine that almost all 1000 watts (1 channel) is being dissipated into 6 of those heatsinks.
I'm building up a collection of heatsinks myself below:
http://www.geocities.com/super_bq/Heatsinks.html
Getting a transformer that is center-tapped or not is based upon the type of power supply configuration you plan to use. You don't have to do "dual-complimentary" that uses the center-tap (like I did on my SoZ). Full-wave rectification should do the job - only disadvantage is the possibility of more mechanical hum from the transformer.
Power trans blues
Thanks for your response.
You are right- this is a SOZ question
Maybe I'll be demoted to an even lower title, my excuse is that it was my first post!
From what I read on the Pass site, the 25 watt monoblock
dissapates more like 500 watts. The 1kva trans is to avoid ripple.
Maybe I misread it , but I've misread it about 50 times if so.....
Those transformers are looking less desireable all the time.
I should have ordered Plitrons in the right size for twice the price.
I assumed that the 4 taps each had a plus and minus, which I thought could be configured into a plus minus configuration
Now I realize I have no idea what I'm getting!!
Any Idea how much the voltage will sag from 37 when the trans is loaded with 500watts?
Mark
Thanks for your response.
You are right- this is a SOZ question
Maybe I'll be demoted to an even lower title, my excuse is that it was my first post!
From what I read on the Pass site, the 25 watt monoblock
dissapates more like 500 watts. The 1kva trans is to avoid ripple.
Maybe I misread it , but I've misread it about 50 times if so.....
Those transformers are looking less desireable all the time.
I should have ordered Plitrons in the right size for twice the price.
I assumed that the 4 taps each had a plus and minus, which I thought could be configured into a plus minus configuration
Now I realize I have no idea what I'm getting!!
Any Idea how much the voltage will sag from 37 when the trans is loaded with 500watts?
Mark
Mark,
Some thoughts:
--Those transformers have four separate secondary windings. You'll be running two pairs of wires in parallel for each rail. Be careful hooking them up; make sure you don't get them backwards.
--The 37.5VAC from those transformers will give you a rail of approximately 51VDC (unloaded) assuming a cap input filter, after taking rectifier losses into account.
--No, it won't sag more than a few volts...or rather, by the time it does the transformer will be getting very, very hot indeed. You don't want it to sag that much, trust me. Count on, say, 48-50VDC rails under a reasonable load.
--A SOZ running on 48V rails is not a reasonable load. It's a current hog travelling incognito. A transformer to give +-48VDC rails will have to be rated much higher than 1010VA to withstand the SOZ circuit.
--But...there's a possibility that things might just work out. If, instead of a cap input filter, you were to use a choke input, the (loaded) DC rail would be more on the order of 35VDC, which is closer to what you had intended. A fringe benefit is that choke input filters deliver proportionately more current than cap input filters.
--Note that the caps in a choke input filter will rise to the full 51V unloaded, so buy 63V caps just to be sure you're prepared for any contingency.
--You're still going to need a bodacious amount of heatsinking, no matter what. (If all else fails, you can water cool the thing.)
--Now your only problem is to track down large chokes...or wind your own. (Don't worry, they're easy.) Take this slow and careful and you'll do just fine.
Grey
Some thoughts:
--Those transformers have four separate secondary windings. You'll be running two pairs of wires in parallel for each rail. Be careful hooking them up; make sure you don't get them backwards.
--The 37.5VAC from those transformers will give you a rail of approximately 51VDC (unloaded) assuming a cap input filter, after taking rectifier losses into account.
--No, it won't sag more than a few volts...or rather, by the time it does the transformer will be getting very, very hot indeed. You don't want it to sag that much, trust me. Count on, say, 48-50VDC rails under a reasonable load.
--A SOZ running on 48V rails is not a reasonable load. It's a current hog travelling incognito. A transformer to give +-48VDC rails will have to be rated much higher than 1010VA to withstand the SOZ circuit.
--But...there's a possibility that things might just work out. If, instead of a cap input filter, you were to use a choke input, the (loaded) DC rail would be more on the order of 35VDC, which is closer to what you had intended. A fringe benefit is that choke input filters deliver proportionately more current than cap input filters.
--Note that the caps in a choke input filter will rise to the full 51V unloaded, so buy 63V caps just to be sure you're prepared for any contingency.
--You're still going to need a bodacious amount of heatsinking, no matter what. (If all else fails, you can water cool the thing.)
--Now your only problem is to track down large chokes...or wind your own. (Don't worry, they're easy.) Take this slow and careful and you'll do just fine.
Grey
1010 Transformer angst
Well, thanks for the kindly advice.
I was able to determine from the Pass web pages that the voltage was too high. I just had the ignorant hope that the fact that the transformer was measured without load was significant.
I have now learned that rectification raises the voltage. sometimes learning can be painful! To Nelson Pass's credit, he states that the transformer should be rated 20% less than the rail voltage desired.
A choke would certainly put me in trendy company.
After hearing about the results Musical Fidelity is getting with this approach, I was tempted to ask NP about incorporating one.
I have always suspected that some of the "tube magic" was due to things other than tubes- I think even Mr. Pass has theorized that tube circuits are generally required to be simple and that this simplicity might be one of the major tube "advantages". Now the choke regulation is looking to be significant also.
So- two more questions
1. (you knew this was coming!!!) what sort of specs would a choke of this type have?-back to cruising the web!
Do they have to be wound on a core? I suspect there is more to it than just putting the choke in(where?) What else do I need?
2. I have already ordered the the 120,000 mfd of caps/ch mentioned in my original post above. BUT I also have 2 10 amp variacs.
Can I just use one on each amp and keep the AC voltage turned down to about 70 volts to the trans? (match both channels with a voltmeter)I think the toroid transformer would still handle the same current-right?
3. I could order the plitron toroids- but there's that pesky divorce possibility!
Thanks, Mark
Well, thanks for the kindly advice.
I was able to determine from the Pass web pages that the voltage was too high. I just had the ignorant hope that the fact that the transformer was measured without load was significant.
I have now learned that rectification raises the voltage. sometimes learning can be painful! To Nelson Pass's credit, he states that the transformer should be rated 20% less than the rail voltage desired.
A choke would certainly put me in trendy company.
After hearing about the results Musical Fidelity is getting with this approach, I was tempted to ask NP about incorporating one.
I have always suspected that some of the "tube magic" was due to things other than tubes- I think even Mr. Pass has theorized that tube circuits are generally required to be simple and that this simplicity might be one of the major tube "advantages". Now the choke regulation is looking to be significant also.
So- two more questions
1. (you knew this was coming!!!) what sort of specs would a choke of this type have?-back to cruising the web!
Do they have to be wound on a core? I suspect there is more to it than just putting the choke in(where?) What else do I need?
2. I have already ordered the the 120,000 mfd of caps/ch mentioned in my original post above. BUT I also have 2 10 amp variacs.
Can I just use one on each amp and keep the AC voltage turned down to about 70 volts to the trans? (match both channels with a voltmeter)I think the toroid transformer would still handle the same current-right?
3. I could order the plitron toroids- but there's that pesky divorce possibility!
Thanks, Mark
I've been very happy with the ERSE inductors from
Zalytron (www.Zalytron.com) which can take a lot
of DC current before saturation.
Zalytron (www.Zalytron.com) which can take a lot
of DC current before saturation.
Inductors
Check out the supersize inductors i've made for my SoZ
http://www.geocities.com/super_bq/soz3.html
Future upgrade would be to move the amp to run at an even higher output wattage for running less efficient speakers. The amp is definitly over heatsinked so with the right transformer, I estimate it could easily do a total of 80 watts output maybe 100 watts.
Power turn on surge hasn't been too much of a problem for me right now. However, i'm curious what a 4KVA toroidal could pull from 240 VAC mainlines? I know 2KVA seems more of a problem to those living in countries with 120 VAC power.
Check out the supersize inductors i've made for my SoZ
http://www.geocities.com/super_bq/soz3.html
Future upgrade would be to move the amp to run at an even higher output wattage for running less efficient speakers. The amp is definitly over heatsinked so with the right transformer, I estimate it could easily do a total of 80 watts output maybe 100 watts.
Power turn on surge hasn't been too much of a problem for me right now. However, i'm curious what a 4KVA toroidal could pull from 240 VAC mainlines? I know 2KVA seems more of a problem to those living in countries with 120 VAC power.
whoeee!
a post from Jolly St. Nelson on Christmas eve! Are the Erse inductors being proposed for the chokes?
My resistors and transformers arrived today but I'm not allowed to look at them till manana
Even if the variacs save me this time (do they?) I'm still pretty interested in the choke thang.They Are the design flavor of the month, after all. AND they would allow me to run the amp without the variacs if I put in a soft start gizmo. I hope even a choke approach uses some of these caps I bought! Any disadvatages to the chokes? (other than even more money no doubt!)
Mark
a post from Jolly St. Nelson on Christmas eve! Are the Erse inductors being proposed for the chokes?
My resistors and transformers arrived today but I'm not allowed to look at them till manana
Even if the variacs save me this time (do they?) I'm still pretty interested in the choke thang.They Are the design flavor of the month, after all. AND they would allow me to run the amp without the variacs if I put in a soft start gizmo. I hope even a choke approach uses some of these caps I bought! Any disadvatages to the chokes? (other than even more money no doubt!)
Mark
Choke supply
Mark,
Well, I guess my 15 minutes of fame are over. I just assumed that Nelson Pass and Grey Rollins had dropped everything (including Christmas) to work out the details of a choke regulated power supply!
I assumed that the choke that Grey mentioned was different than the pi filter supply - is it? or were you just suggesting a higher value for the inductors. I have no fear of winding them, actually I like mindless work like that!
I'll fire it up with the variac, which will let me determine how much power the assembly can take, then maybe order Plitrons at that voltage later.
another option: I have heard of people unwinding transformers a bit to lower the volts. My assumption is that the amperes the transformer can handle remains the same, so the VA rating of the
transformer goes down. Thus, my 1kva trans becomes a .8 kva if I reduce the secondary to 80%? I assume I am still ahead because there is less voltage pushing current into the amp?
Actually, I am well on my way. The resistors I got are perfect, and mounting all 39 of them on my heatsinks will keep me busy.
To sum up:
Grey, what is this choke supply? values?
Will the variac solve the problem for the moment?
Anyone have some matched MOSFETS to sell?
gotta go,
Mark
Mark,
Well, I guess my 15 minutes of fame are over. I just assumed that Nelson Pass and Grey Rollins had dropped everything (including Christmas) to work out the details of a choke regulated power supply!
I assumed that the choke that Grey mentioned was different than the pi filter supply - is it? or were you just suggesting a higher value for the inductors. I have no fear of winding them, actually I like mindless work like that!
I'll fire it up with the variac, which will let me determine how much power the assembly can take, then maybe order Plitrons at that voltage later.
another option: I have heard of people unwinding transformers a bit to lower the volts. My assumption is that the amperes the transformer can handle remains the same, so the VA rating of the
transformer goes down. Thus, my 1kva trans becomes a .8 kva if I reduce the secondary to 80%? I assume I am still ahead because there is less voltage pushing current into the amp?
Actually, I am well on my way. The resistors I got are perfect, and mounting all 39 of them on my heatsinks will keep me busy.
To sum up:
Grey, what is this choke supply? values?
Will the variac solve the problem for the moment?
Anyone have some matched MOSFETS to sell?
gotta go,
Mark
Mark,
A PI filter is rectifier->cap->inductor->cap, but there are other options. You can use an inductor as the first element: rectifier->inductor->cap, which is what I was referring to. It's called an "L" filter, by the way, since it looks like an upside-down, backwards capital L. Yes, you can keep on chaining these things out into the distance for as long as your wallet and patience hold out, but after about two or three stages you're well into diminishing returns.
Values? Whatever you can find to throw at it. Bigger is better in all cases. I looked at the inductors Nelson mentioned at one point, but can't remember what values they had. Hit their website and look things over.
The formula for reactance of an inductor (aka choke) is X= 2*PI*F*L where
X= reactance--kinda like resistance, but it varies with frequency--this is in ohms
F= frequency
L= inductance of the inductor in Henries (not millihenries or whatever, be sure to get the decimal in the right place)
We could get into things like the Q or stray capacitance of the inductor, but let's leave that for the time being. To a first approximation, the one formula above will do the job.
You'll be looking for an inductor with as low a DC resistance as you can get (larger gauge wire is the ticket, here), but with as high a reactance as possible at whatever frequency you're trying to filter. In this case, you're looking to knock out stuff in the 100Hz (if your AC line frequency is 50Hz) or 120Hz (if your AC is 60Hz) and above in multiples or either 50 or 60Hz.
Sit down with a calculator (or spreadsheet) and work it through. It'll soon start to come clear after you've fiddled it through.
Bear in mind that even a little bit of inductance helps, particularly with very large caps.
Grey
A PI filter is rectifier->cap->inductor->cap, but there are other options. You can use an inductor as the first element: rectifier->inductor->cap, which is what I was referring to. It's called an "L" filter, by the way, since it looks like an upside-down, backwards capital L. Yes, you can keep on chaining these things out into the distance for as long as your wallet and patience hold out, but after about two or three stages you're well into diminishing returns.
Values? Whatever you can find to throw at it. Bigger is better in all cases. I looked at the inductors Nelson mentioned at one point, but can't remember what values they had. Hit their website and look things over.
The formula for reactance of an inductor (aka choke) is X= 2*PI*F*L where
X= reactance--kinda like resistance, but it varies with frequency--this is in ohms
F= frequency
L= inductance of the inductor in Henries (not millihenries or whatever, be sure to get the decimal in the right place)
We could get into things like the Q or stray capacitance of the inductor, but let's leave that for the time being. To a first approximation, the one formula above will do the job.
You'll be looking for an inductor with as low a DC resistance as you can get (larger gauge wire is the ticket, here), but with as high a reactance as possible at whatever frequency you're trying to filter. In this case, you're looking to knock out stuff in the 100Hz (if your AC line frequency is 50Hz) or 120Hz (if your AC is 60Hz) and above in multiples or either 50 or 60Hz.
Sit down with a calculator (or spreadsheet) and work it through. It'll soon start to come clear after you've fiddled it through.
Bear in mind that even a little bit of inductance helps, particularly with very large caps.
Grey
Well Nelson, you are an immense fish (metaphorically - I don't know what you look like!)
in a pond that is perhaps smaller than we imagine.
For this reason, I think you still have quite a bit more mileage
on the fame o meter. Try to stay out of People magazine!
I suppose you have a better perspective in that you have to
market to this puddle.
Grey Rollins:
I have now learned a bunch, thanks to you both.
I promise to order a copy of Art of Electronics as soon as my funds recover from all the Zen stuff (ironic, huh?)
I now know that the pi filter WAS a choke supply.
I know where to get the coils (those Erse coils ARE a good value- cheaper than many, before considering performance)
I had always planned for this supply as Nelson has it shown with the pi filter on his site. It sure fits the amp design conceptually:
bigger is better!
Here are a few more questions :
the (square) rectifier bridge has one pin that is different next to the cut off corner. Which connection is this? the one to the ground?
If I parallel inductors, the current capacity clearly would go up.
What happens to the inductance- the same? Double?
No rush, still gotta track down the transistors
mark
in a pond that is perhaps smaller than we imagine.
For this reason, I think you still have quite a bit more mileage
on the fame o meter. Try to stay out of People magazine!
I suppose you have a better perspective in that you have to
market to this puddle.
Grey Rollins:
I have now learned a bunch, thanks to you both.
I promise to order a copy of Art of Electronics as soon as my funds recover from all the Zen stuff (ironic, huh?)
I now know that the pi filter WAS a choke supply.
I know where to get the coils (those Erse coils ARE a good value- cheaper than many, before considering performance)
I had always planned for this supply as Nelson has it shown with the pi filter on his site. It sure fits the amp design conceptually:
bigger is better!
Here are a few more questions :
the (square) rectifier bridge has one pin that is different next to the cut off corner. Which connection is this? the one to the ground?
If I parallel inductors, the current capacity clearly would go up.
What happens to the inductance- the same? Double?
No rush, still gotta track down the transistors
mark
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.