Still4given,
I wouldn't worry too much about your transformers not being torroids.
Their main advantage is a lower spreading of magnetic field. However, field strength decreases (if I recall correctly) with distance^3, so a little more distance, and suitable orientation of the PCB's relative to the transformer should solve this problem (which is mainly a problem when you stress the transformer). Look at the transformer, and try to see the orientation of the coils. This will also be the direction of the magnetic field emmitted (which may leak from the core), going in a circle path back to the other end of the transformer coil. Worst case is to have this field going straight through the plane of your PCB, as conductors of your PCB will act as windings, causing induction error (noise, hum) voltages.
In real life, though, I have found this to be a minor problem only.
Depending on type of transformer (torroidal, EI core, UI core etc), there are differences in output impedance (voltage "stiffness" or load-independence/regulation) for a given VA rating, but as long as you use big transformers (say, 4 times the estimated wattage rating of the amp), I have found this problem to be neclectable. A healthy ammount of storage capacity lowers the PSU impedance drastically, anyway.
another thing... Use plenty of output devices. The phase of a speaker will cause losses bigger than the obvious U * I calculation, so it's a good idea to use plenty of devices (they'll run cooler, too!) for a given amp and heat sink.
Jennice
I wouldn't worry too much about your transformers not being torroids.
Their main advantage is a lower spreading of magnetic field. However, field strength decreases (if I recall correctly) with distance^3, so a little more distance, and suitable orientation of the PCB's relative to the transformer should solve this problem (which is mainly a problem when you stress the transformer). Look at the transformer, and try to see the orientation of the coils. This will also be the direction of the magnetic field emmitted (which may leak from the core), going in a circle path back to the other end of the transformer coil. Worst case is to have this field going straight through the plane of your PCB, as conductors of your PCB will act as windings, causing induction error (noise, hum) voltages.
In real life, though, I have found this to be a minor problem only.
Depending on type of transformer (torroidal, EI core, UI core etc), there are differences in output impedance (voltage "stiffness" or load-independence/regulation) for a given VA rating, but as long as you use big transformers (say, 4 times the estimated wattage rating of the amp), I have found this problem to be neclectable. A healthy ammount of storage capacity lowers the PSU impedance drastically, anyway.
another thing... Use plenty of output devices. The phase of a speaker will cause losses bigger than the obvious U * I calculation, so it's a good idea to use plenty of devices (they'll run cooler, too!) for a given amp and heat sink.
Jennice
Still4given,
Maybe you'll find the PDF on ftp://sylvest.dyndns.dk/ useful. Please let me know what you think. It's on my own server (which is down occasionally - it's experimental)), so please download it in stead of reading it directly from the server.
Jennice
Maybe you'll find the PDF on ftp://sylvest.dyndns.dk/ useful. Please let me know what you think. It's on my own server (which is down occasionally - it's experimental)), so please download it in stead of reading it directly from the server.
Jennice
OK, guys. I've purchased three of the P101 boards from ESP. Rod said it should take a couple of weeks for them to arrive. I have already ordered most of the parts for them. I will have to start thinking about a PSU. I would really like to use one of XL-280 transformers if there is a way to make it work. They have +two separate secondary outputs. Three really but the one is only like 6V or something I'm told. The one that gets hooked up for the Hafler DH200 is 65-0-65 I believe. At least that's what it measures at the Filter caps. I'm told the other set puts out 70-0-70 and is for heating up the MOSFETs. I'm not sure if it has enough amps to do anything else. I know the XL-280 uses 4 filter caps as opposed to 2 caps for the DH-200.
Anyway, do any of you know if this tranny will have enough juice to run the P101? I bought three boards planning to make this a three channel like the Bryston 5B. I want to use it for the front end of a 5.1 setup.
So what do you guys think? Dumb idea? Would three channels tax the tranny too much?
Thanks, Terry
Anyway, do any of you know if this tranny will have enough juice to run the P101? I bought three boards planning to make this a three channel like the Bryston 5B. I want to use it for the front end of a 5.1 setup.
So what do you guys think? Dumb idea? Would three channels tax the tranny too much?
Thanks, Terry
Well, to begin with 65-0-65 is suitable for a 400W/channel amp _ approximately. I think that might be pushing the P101 beyond what it was intended to handle unless you want to experiment with Exicon "double die" MOSFETs and a really big heatsink.
On the other hand if you truly have two secondaries and not just one with a center tap you can parrallel them have have something suitable for ~100W/ch. 32.5/35 - 0 - 32.5/35
This will work
__________
( +65/70
(
(
(__________ 0
__________
( 0
(
(
(__________ -65/70
This will not allow you to parrallel
________________
( +65/70
(________________
( 0
(________________ -65/70
Assuming you can parrallel them, your thansformer will not have a 0V reference, so take a look at Rod's project on how to create one - you just need to be sure to use appropriate resistors in the case of figure 1. (I use this set up for a bench power supply to check new amps.) Figure 2 ought to work as well so ong as you use 15V zeners to power the opamps.
On the other hand if you truly have two secondaries and not just one with a center tap you can parrallel them have have something suitable for ~100W/ch. 32.5/35 - 0 - 32.5/35
This will work
__________
( +65/70
(
(
(__________ 0
__________
( 0
(
(
(__________ -65/70
This will not allow you to parrallel
________________
( +65/70
(________________
( 0
(________________ -65/70
Assuming you can parrallel them, your thansformer will not have a 0V reference, so take a look at Rod's project on how to create one - you just need to be sure to use appropriate resistors in the case of figure 1. (I use this set up for a bench power supply to check new amps.) Figure 2 ought to work as well so ong as you use 15V zeners to power the opamps.
Wow, my Haflers are just 115WPC. The use the Hitachi MOSFETs and pretty big heatsinks though. I already bought the MOSFETs for the P101. They are the 2SH1058 and 2SJ162.
The manual for the P101 has this warning;
Warning: The power supply voltage used for this amplifier is potentially lethal! With +/- 70V, the total voltage is 140V which can kill you!
Maybe I'm misrepresenting my transformer. I just looked at the recomended power supply for the P101 and it shows a 50-0-50 Transformer but after rectification it is showing +70,-70 rails with 6x4, 700uF/80V caps. My Hafler only has 65V rails with 2-10000uf/75V caps.
Maybe my tranny is too small for this amp. Any idea?
Thanks Terry
The manual for the P101 has this warning;
Warning: The power supply voltage used for this amplifier is potentially lethal! With +/- 70V, the total voltage is 140V which can kill you!
Maybe I'm misrepresenting my transformer. I just looked at the recomended power supply for the P101 and it shows a 50-0-50 Transformer but after rectification it is showing +70,-70 rails with 6x4, 700uF/80V caps. My Hafler only has 65V rails with 2-10000uf/75V caps.
Maybe my tranny is too small for this amp. Any idea?
Thanks Terry
Sam9,
How do you get away with this...!? I my book it's not possible to get +/- 32V, unless there are really 4 identical secondaries.
The only work-around is to parallel and get one 65V secondary, and then some added electronics to get this made symmetrical. Personally, I don't like that approach, especially not with these power levels, as the output impedance of the transformer is hot used to maintain the 0V reference point.
Terry,
What kind of transformer is it? (torroidal, EI core...?)
(somehow I have the feeling I've already asked this once before
)
What's the weight of it? If you take the weight for a given core type transformer, you can often get an idea of VA rating, by comparing it to those in electronic shops.
Jennice
How do you get away with this...!? I my book it's not possible to get +/- 32V, unless there are really 4 identical secondaries.
The only work-around is to parallel and get one 65V secondary, and then some added electronics to get this made symmetrical. Personally, I don't like that approach, especially not with these power levels, as the output impedance of the transformer is hot used to maintain the 0V reference point.
Terry,
What kind of transformer is it? (torroidal, EI core...?)
(somehow I have the feeling I've already asked this once before
)What's the weight of it? If you take the weight for a given core type transformer, you can often get an idea of VA rating, by comparing it to those in electronic shops.
Jennice
Transformers will work fine..
Actually, you got just about the perfect fit for your transformer. I know more about the Halfers than I do Rod's amp - but here is why I say that they will work.
1. The Hafler DH200 rails are 63v. Meaning +/- 63v using the center tap as 0. Otherwise expressed as 63-0-63. This is after rectification with the associated diode voltage drops.
From a current point of view, they are sized for the 200 watts per side at 8ohms. And it does that realistically, so I am guessing that it is rated around 600va.
2. Rod's project 101 has max voltage rails of 70.
3. So the transformers for the 200 will work - for two channels, not three.
4 You said you wanted to build 3 channels. If you have three of those trannies, and can afford to use them... and the extra cost of the caps, build it monoblock style - one power supply per channel. This will result in by far the best sound quality.
Actually, you got just about the perfect fit for your transformer. I know more about the Halfers than I do Rod's amp - but here is why I say that they will work.
1. The Hafler DH200 rails are 63v. Meaning +/- 63v using the center tap as 0. Otherwise expressed as 63-0-63. This is after rectification with the associated diode voltage drops.
From a current point of view, they are sized for the 200 watts per side at 8ohms. And it does that realistically, so I am guessing that it is rated around 600va.
2. Rod's project 101 has max voltage rails of 70.
3. So the transformers for the 200 will work - for two channels, not three.
4 You said you wanted to build 3 channels. If you have three of those trannies, and can afford to use them... and the extra cost of the caps, build it monoblock style - one power supply per channel. This will result in by far the best sound quality.
Jennice,
I'm sure it must be EI. It weighs 15lbs. I'll see if I can compare it.
Sawzall,
Thanks for the info. I only have two extra trannies. If you say it would be better to use three trannies, maybe I'll look into getting three toroids instead. I would like this amp to sound it's best if I can. Not sure how much real estate I'll need for all those trannies though.
Those boards won't be here for two weeks so that gives me a bit of time to figure things out. What size toroids would I need if I only power one board with each?
Thanks, Terry
I'm sure it must be EI. It weighs 15lbs. I'll see if I can compare it.
Sawzall,
Thanks for the info. I only have two extra trannies. If you say it would be better to use three trannies, maybe I'll look into getting three toroids instead. I would like this amp to sound it's best if I can. Not sure how much real estate I'll need for all those trannies though.
Those boards won't be here for two weeks so that gives me a bit of time to figure things out. What size toroids would I need if I only power one board with each?
Thanks, Terry
still4given said:
It is a EI.
As for the proper size transformer if you wish to run it off or one or two transformers.
One transformer - 3 channels at 200 watts each into 8 ohms. If you worked in a perfect world, you could simply add them up. But it does not. I would not attempt it with a transformer rated for less than 1000 va.
Two transformer. In this case you would get a transformer that did *not* use a center tap, but would give you the required 65 to 70 volts using the two secondary leads. I would look to use a trans rated at 600 va
Actually, as I am sitting here thinking perhaps the is another alternative using the two old Hafler trans.- let me check and I will get back to you.
In the meantime, go look for the Duncan tranformer website. He has a powersupply simulation tool there that is really helpful when building a supply.
Hi,
use your 2 transformers. If you have 4 wires on the secondary i.e. 2 @40vac.
Wire the two outputs of each transformer in parallel, that will give you double the current rating and a single rail of 63v
Use 2 bridge rectifiers and after smoothing connect a plus to a minus & you'll end up with 63-0-63 and a total power rating of 1200va.
By the way +-63v will probably give between 150 to 170w into 8r and times 3 gives a total output of about 480w, then times 1.5 for efficiency of output gives 720w as your psu requirement. You will have almost double this meaning you could rate your output stage to run continuously into 5r or pushing it 4r. All 3 channels will almost certainly not run max power at the same time & then not for very long & I think if you tripled up the rectifiers and caps you will find that the pair of transformers will easily cope with 3 channel 300w into 4r
regards Andrew T.
use your 2 transformers. If you have 4 wires on the secondary i.e. 2 @40vac.
Wire the two outputs of each transformer in parallel, that will give you double the current rating and a single rail of 63v
Use 2 bridge rectifiers and after smoothing connect a plus to a minus & you'll end up with 63-0-63 and a total power rating of 1200va.
By the way +-63v will probably give between 150 to 170w into 8r and times 3 gives a total output of about 480w, then times 1.5 for efficiency of output gives 720w as your psu requirement. You will have almost double this meaning you could rate your output stage to run continuously into 5r or pushing it 4r. All 3 channels will almost certainly not run max power at the same time & then not for very long & I think if you tripled up the rectifiers and caps you will find that the pair of transformers will easily cope with 3 channel 300w into 4r
regards Andrew T.
15lbs of transformer sounds like 500 - 600VA to me.
In reality, AndrewT's idea should work, as long as you don't push t
he PSU too hard. The channels won't run full power constantly like in class A, but still keep in mind to de-rate the VA rating to (roughly) 60% (so I've been told by the manufacturer) due to the higher losses / saturation, when using it with a rectifier. The RMS power is still below the VA rating, but the peak currents are much higher, as it will be charging the capacitors in shorter bursts of high current. This give a magnetizing curve in the core, which is far from the indended shape when used as AC source.
Jennice
In reality, AndrewT's idea should work, as long as you don't push t
he PSU too hard. The channels won't run full power constantly like in class A, but still keep in mind to de-rate the VA rating to (roughly) 60% (so I've been told by the manufacturer) due to the higher losses / saturation, when using it with a rectifier. The RMS power is still below the VA rating, but the peak currents are much higher, as it will be charging the capacitors in shorter bursts of high current. This give a magnetizing curve in the core, which is far from the indended shape when used as AC source.
Jennice
Thanks guys for the info.
I think, if I need more than one transformer, I'd rather not use the Hafler trannies, For one thing, just the trannies alone would weigh 30lbs. Another thing is that, as I have them right now, the ones not being used are the TT-XL280E, and the old tranny out of my P230. The one from the P230 only has 3 secondary wires. The TT-XL280E has 7 secondary wires. I think I would be better off buying one large XFR rather than trying to rig these up for this.
Would I be better off getting one 1000va tranny or three smaller ones?
And by the way, where is a good place to buy Toroidal trannies here in the US?
I have purchased the PCB from Rod for a soft-start system. Do I need three seperate ones if I go with three trannies?
I hope you guys don't mind all these questions. I'm just trying to measure twice and cut once.
Blessings, Terry
I think, if I need more than one transformer, I'd rather not use the Hafler trannies, For one thing, just the trannies alone would weigh 30lbs. Another thing is that, as I have them right now, the ones not being used are the TT-XL280E, and the old tranny out of my P230. The one from the P230 only has 3 secondary wires. The TT-XL280E has 7 secondary wires. I think I would be better off buying one large XFR rather than trying to rig these up for this.
Would I be better off getting one 1000va tranny or three smaller ones?
And by the way, where is a good place to buy Toroidal trannies here in the US?
I have purchased the PCB from Rod for a soft-start system. Do I need three seperate ones if I go with three trannies?
I hope you guys don't mind all these questions. I'm just trying to measure twice and cut once.
Blessings, Terry
still4given said:
And...?
My last project used two 7.5 kG transformers in one box... ended up at 30 Kg.
An odd number of wires means that at least one has a center-tap. Not necesarily a bad thing...
For three seperate channels, I'd go for three transformers. (Good channel seperation) If it were one tri-amp'ed speaker, I'd go for one big (low output resistance)
I have purchased the PCB from Rod for a soft-start system. Do I need three seperate ones if I go with three trannies?
I hope you guys don't mind all these questions. I'm just trying to measure twice and cut once.
Depends... What's the rating of his circuit? (can you tell a project number, if you don't have the specs at hand?
Questions is what this forum is all about. (In my oppinion, anyway).
Jennice
"where is a good place to buy Toroidal trannies here in the US?"
http://partsexpress.com/
has Avel trannies. For Avel items not on the web page contact:
Beverly
Special Order Rep
800.338.0531 Xt179
bevr@partsexpress.com
Also
http://avellindberg.com/
http://www.victoriamagnetics.com/
http://www.belfuse.com/signaltransformer/
http://www.toroid.com/
http://www.plitron.com/ in Canada, but good
Prosit
http://partsexpress.com/
has Avel trannies. For Avel items not on the web page contact:
Beverly
Special Order Rep
800.338.0531 Xt179
bevr@partsexpress.com
Also
http://avellindberg.com/
http://www.victoriamagnetics.com/
http://www.belfuse.com/signaltransformer/
http://www.toroid.com/
http://www.plitron.com/ in Canada, but good
Prosit
I know 30 Kg is a lot... especially when it doesn't have any carrying handles. 
The sides are all heat sink. I bought some industrial 420mm wide x 200 mm high heat sinks. I don't know what you'd call them, but the extrusion profile looks like a huge comb for your hair. And this extruded stuff is where you carry them. This calls for protective gloves - otherwise you'll cut yourself. I'm speaking from experience.
Then again, I believe in it as a sort of anti-theft feature.
It's got two (conservatively rated) 500 VA transformers, 8 x 15.000uF/63V capacitors, and 24 power transistors in total (200W Motorola types).
Overkill? Maybe, but it has a reason:
Each transformer supplies two power amp stages. Indivitually, they're rated at (only?) 80W @ 8ohm, but they're dessigned to be stable at 4 ohm bridged! (although this would stress the PSU at high levels). At 8 ohm bridged they're about 300 W stereo, full time (4 channels bridged in pairs). Then again, my B&W speakers don't like that kind of power, and it only plays modeately louder than the 80W before my speakers start do dislike it. However, when used at 80W/channel, it has a lot of energy just waiting to be released!
At the moment I'm making two stereo blocks for my brother, based on the same circuit, but rated 150W @ 8 ohm. These will not be designed to be bridged. Cramming it all into one box was a mistake. Too much stuff in too little space. Things started to affect each other. Just not adviseable to be repeated.
The new ones will have 550VA UI-core transformers, and 100.000uF total per stereo amp. That should be enough - even for my brother who's going to have them. Even though, he's concerned: Are you sure they won't run out of juice? Can they be bridged, just to make sure I got enough power? What a nutcase!
The design criteria was the use of 100mm high heat sinks, so I had the manufacturer find out how much power a tansformer for this could have. The transformer is just about 90mm high, so this is going to be fun...
As for your soft-start, I'll get back to you on that...
Jennice
The sides are all heat sink. I bought some industrial 420mm wide x 200 mm high heat sinks. I don't know what you'd call them, but the extrusion profile looks like a huge comb for your hair. And this extruded stuff is where you carry them. This calls for protective gloves - otherwise you'll cut yourself. I'm speaking from experience.
Then again, I believe in it as a sort of anti-theft feature.
It's got two (conservatively rated) 500 VA transformers, 8 x 15.000uF/63V capacitors, and 24 power transistors in total (200W Motorola types).
Overkill? Maybe, but it has a reason:
Each transformer supplies two power amp stages. Indivitually, they're rated at (only?) 80W @ 8ohm, but they're dessigned to be stable at 4 ohm bridged! (although this would stress the PSU at high levels). At 8 ohm bridged they're about 300 W stereo, full time (4 channels bridged in pairs). Then again, my B&W speakers don't like that kind of power, and it only plays modeately louder than the 80W before my speakers start do dislike it. However, when used at 80W/channel, it has a lot of energy just waiting to be released!
At the moment I'm making two stereo blocks for my brother, based on the same circuit, but rated 150W @ 8 ohm. These will not be designed to be bridged. Cramming it all into one box was a mistake. Too much stuff in too little space. Things started to affect each other. Just not adviseable to be repeated.
The new ones will have 550VA UI-core transformers, and 100.000uF total per stereo amp. That should be enough - even for my brother who's going to have them. Even though, he's concerned: Are you sure they won't run out of juice? Can they be bridged, just to make sure I got enough power? What a nutcase!
The design criteria was the use of 100mm high heat sinks, so I had the manufacturer find out how much power a tansformer for this could have. The transformer is just about 90mm high, so this is going to be fun...
As for your soft-start, I'll get back to you on that...
Jennice
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.