i just moved to ohio about a month ago. raining and 30-40 degrees here. just haven't updated my profile yet. but yeah...it *was* great living out there, and i blew off work quite often to walk the dog down to the beach (i was bout .4 miles). and i could even see catalina from there (on a really clear day).
now i have nothing *but* diy to keep me going
Darkmobius - you should have NO problem finding surplus heatsinks and parts up the wazoo out in LA. I know I did. Wished I had a tractor trailer to haul stuff back east!!
If Rockwell still has their surplus shop - that's a best bet.
I think ALL Electronics has stuff, and I found a ton of other places... there's one in Pasadena "C&H"?? and others about. Plus places on the web. Heatsinks should be a no brainer.
_-_-bear
Yeah, and there is a big "swapmeet" once a month too... made out like a bandit there... forget the name, what they called it.
If Rockwell still has their surplus shop - that's a best bet.
I think ALL Electronics has stuff, and I found a ton of other places... there's one in Pasadena "C&H"?? and others about. Plus places on the web. Heatsinks should be a no brainer.
_-_-bear
Yeah, and there is a big "swapmeet" once a month too... made out like a bandit there... forget the name, what they called it.
bear said:
Hey Bear, that's really good news to hear.
I used to go to TRW's swap meet in El Segundo on the last Saturday of every month. Get all kinds of test equipment, etc. I think that is the one you are thinking of. They still have it like clockwork, although resellers have discovered it and the deals aren't as great as they used to be.
I'll definetly look around for these places, thanks.
Re: Rollin' your own....
Hi Darkmoebius
That's a relief
I was quoted UKP 80 but I had to buy five to get this price.
You can use another type of heatsink by changing overall size to get an equivalent heat dissipation rate. The SK158 is one of the better ones for surface area and although expensive is relatively compact.
Extruded aluminum heatsinks are primarily priced on weight although complexity of profile does have some bearing.
As Bear says stuff is available from various sources but I have to specify commercially available items.
I will add some more information here.
I now use my lathe with a handle on the spindle as I can hold the bobbin more securely.
You need to be looking at
http://www.susan-parker.co.uk/zeus-out-tx-75w.htm
Wire diameter is 1.0 mm for the EI-120 laminations.
The bolts and nuts can be steel but one uses fiber washers and insulation on the internal thread so there is no magnetic circuit.
Or one can use austinetic stainless steel.
I will reorganize the website to try to make it clearer about the correct transformer information.
Almost
All four wires are wound simultaneously.
Using the dual chamber bobbin one quad filar winds about 75 turns to get a full chamber. Then turn the bobbin around and wind an identical set of turns.
This quad filar winding is important to get the best performance from the output transformer.
I am going to make one this weekend and will put up some pictures.
Unstable economies are not good for anyone, even if there is short term gains one way or the other.
Best wishes,
Susan.
Hi Darkmoebius
darkmoebius said:Aaaaah, a decent shock every once in a while is good for the circulation
In reality, your amps are no more expensive to build than a lot of the SE & PP tube amps I was considering - especially the 300B variants. In that case, I would still have to buy the driver/output tubes and the the output and interstage transformers. Filament transformers, cockets, designers capacitors, PS filtering components, etc.
The chassis and connector costs would be the same no matter what topolgy, so in that light your amps are reasonable.
That's a relief
Although, those SK158 heatsinks are another shocker$270 EUR/$360 US.
I was quoted UKP 80 but I had to buy five to get this price.
You can use another type of heatsink by changing overall size to get an equivalent heat dissipation rate. The SK158 is one of the better ones for surface area and although expensive is relatively compact.
Extruded aluminum heatsinks are primarily priced on weight although complexity of profile does have some bearing.
As Bear says stuff is available from various sources but I have to specify commercially available items.
I will add some more information here.
I now use my lathe with a handle on the spindle as I can hold the bobbin more securely.
You need to be looking at
http://www.susan-parker.co.uk/zeus-out-tx-75w.htm
Wire diameter is 1.0 mm for the EI-120 laminations.
The bolts and nuts can be steel but one uses fiber washers and insulation on the internal thread so there is no magnetic circuit.
Or one can use austinetic stainless steel.
I will reorganize the website to try to make it clearer about the correct transformer information.
Almost
All four wires are wound simultaneously.
Using the dual chamber bobbin one quad filar winds about 75 turns to get a full chamber. Then turn the bobbin around and wind an identical set of turns.
This quad filar winding is important to get the best performance from the output transformer.
I am going to make one this weekend and will put up some pictures.
And those in charge tell us this is good for AmericansMay be good for exports, but sucks for importing.
Unstable economies are not good for anyone, even if there is short term gains one way or the other.
Best wishes,
Susan.
Well Done Susan!
Hi Susan,
I just discovered this thread of your various personal DIY audio deversions. Just so you are aware, if my typing looks odd I think I can be or am "dslextic" as well. I have nto had time to read all the way through the thread yet. I have made it about a 1/3 of way. I will have to catch up on the thread later. For what it is worth, I agree the transformers are the key part. I have look at your site and links a bit. I like to determine if I can wide up one of the two transformers. Heck I might, I am patient, tru to wind up the other one two!
I wanted to at least place a message her to say another silent reader that has viewed part of the thread has come out of the "box", but has lots to catch on. Though I only discovered the thread today.
All the best to your steam locomotive building project if it is not past its deliverly date yet.
Regards,
John L. Males
Willowdale, Ontario
Canada
10 December 2004 19:33
Hi Susan,
I just discovered this thread of your various personal DIY audio deversions. Just so you are aware, if my typing looks odd I think I can be or am "dslextic" as well
. I have nto had time to read all the way through the thread yet. I have made it about a 1/3 of way. I will have to catch up on the thread later. For what it is worth, I agree the transformers are the key part. I have look at your site and links a bit. I like to determine if I can wide up one of the two transformers. Heck I might, I am patient, tru to wind up the other one two!I wanted to at least place a message her to say another silent reader that has viewed part of the thread has come out of the "box", but has lots to catch on. Though I only discovered the thread today.
All the best to your steam locomotive building project if it is not past its deliverly date yet.
Regards,
John L. Males
Willowdale, Ontario
Canada
10 December 2004 19:33
Re: Re: Rollin' your own....
The UKP 270 price was from Borberly Audio , perhaps there was a slight markup in that price
You can use another type of heatsink by changing overall size to get an equivalent heat dissipation rate. The SK158 is one of the better ones for surface area and although expensive is relatively compact.
Extruded aluminum heatsinks are primarily priced on weight although complexity of profile does have some bearing.
As Bear says stuff is available from various sources but I have to specify commercially available items.
Re purchasing pre-wound input transformer bobbins:
Sounds very promising.
I have a tendency to overcomplicate things, don't I? Now, I get it - just wind all four wires until the bobbin is filled(approx 75 turns). Can't get anymore simple than that. Repeat in both chambers, then cacume impregante w/ varnish.
Thanks so much for setting me straight on this, I really appreciate it.
Susan-Parker said:
The UKP 270 price was from Borberly Audio , perhaps there was a slight markup in that price
You can use another type of heatsink by changing overall size to get an equivalent heat dissipation rate. The SK158 is one of the better ones for surface area and although expensive is relatively compact.
Extruded aluminum heatsinks are primarily priced on weight although complexity of profile does have some bearing.
As Bear says stuff is available from various sources but I have to specify commercially available items.
Re purchasing pre-wound input transformer bobbins:
Sounds very promising.
I have a tendency to overcomplicate things, don't I? Now, I get it - just wind all four wires until the bobbin is filled(approx 75 turns). Can't get anymore simple than that. Repeat in both chambers, then cacume impregante w/ varnish.
Thanks so much for setting me straight on this, I really appreciate it.
Re: Re: Re: Rollin' your own....
Hi Darkmoebius,
Okay, having had a look at their site I note that that price is for two off, so not quite so horrendous although there is still quite a mark up.
SK157 200
FISCHER Heatsink
300x200mm, black anodized, not drilled
Pair: Euro 270.00
Also it's the SK157 profile as shown in my quick 3D CAD sketch, rather than the higher SK158. However I specified the latter at 150 mm length and the SK157 at 200 mm should be about the same.
My UKP 80 each ( = about Euro 232 for a pair) is for 150 mm lengths of SK158. I was quoted UKP 47 each for the SK157 X 150mm profile.
I have done some more searching and Aavid Thermalloy
http://www.aavidthermalloy.com/
have a suitable extrusion section OS245 which looks to be roughly equivalent to the SK158:
http://www.aavidthermalloy.com/bin/...45&LengthUnits=mm&ExLength=150.00&TReff=0.210
200 mm lengths of this should do the trick
Might be cheaper or at least easier to get from the USA?
Part 0S245 (VIS#400908)
http://www.aavidthermalloy.com/company/locations.shtml
I have asked for a quote.
Yes, this is correct
It's not complicated ONCE one has got one's head around the ideas and techniques. But like anything that's different it can seem very complicated at first. Don't forget it took me two years of pretty intense development work to get to the point of being able to understand what I was doing.
Even now when I talk to some transformer manufacturers they keep telling me how to do it "properly" with separate winding layers etc. and these are people who live and breath transformer technology on a daily basis.
If you can't vacuum impregnate then painting each layer with varnish as you go will work just as well (although perhaps be a bit sticky).
I don't use any varnish for my development work as I sometimes strip down a bobbin for reuse. However I can hear the windings singing on the test bench so for a final build transformer some form of impregnation is better.
Sound coming from a non-designated source is distortion after all.
My pleasure. It is very heartening that you are interested enough to want to build my amplifier design and I do appreciate how much effort (and money) goes into building even this relatively simple design.
And I do want you to have the best possible outcome and be happy with the results.
Best wishes,
Susan.
Hi Darkmoebius,
darkmoebius said:The UKP 270 price was from Borberly Audio , perhaps there was a slight markup in that price
Okay, having had a look at their site I note that that price is for two off, so not quite so horrendous although there is still quite a mark up.
SK157 200
FISCHER Heatsink
300x200mm, black anodized, not drilled
Pair: Euro 270.00
Also it's the SK157 profile as shown in my quick 3D CAD sketch, rather than the higher SK158. However I specified the latter at 150 mm length and the SK157 at 200 mm should be about the same.
My UKP 80 each ( = about Euro 232 for a pair) is for 150 mm lengths of SK158. I was quoted UKP 47 each for the SK157 X 150mm profile.
I have done some more searching and Aavid Thermalloy
http://www.aavidthermalloy.com/
have a suitable extrusion section OS245 which looks to be roughly equivalent to the SK158:
http://www.aavidthermalloy.com/bin/...45&LengthUnits=mm&ExLength=150.00&TReff=0.210
An externally hosted image should be here but it was not working when we last tested it.
200 mm lengths of this should do the trick
Might be cheaper or at least easier to get from the USA?
Part 0S245 (VIS#400908)
http://www.aavidthermalloy.com/company/locations.shtml
I have asked for a quote.
Yes, this is correct
It's not complicated ONCE one has got one's head around the ideas and techniques. But like anything that's different it can seem very complicated at first. Don't forget it took me two years of pretty intense development work to get to the point of being able to understand what I was doing.
Even now when I talk to some transformer manufacturers they keep telling me how to do it "properly" with separate winding layers etc. and these are people who live and breath transformer technology on a daily basis.
If you can't vacuum impregnate then painting each layer with varnish as you go will work just as well (although perhaps be a bit sticky).
I don't use any varnish for my development work as I sometimes strip down a bobbin for reuse. However I can hear the windings singing on the test bench so for a final build transformer some form of impregnation is better.
Sound coming from a non-designated source is distortion after all.
My pleasure. It is very heartening that you are interested enough to want to build my amplifier design and I do appreciate how much effort (and money) goes into building even this relatively simple design.
And I do want you to have the best possible outcome and be happy with the results.
Best wishes,
Susan.
Re: Well Done Susan!
Hi Keypunch,
Thank you for your post and you interest in my work.
Welcome
It's quite possible, there is a lot of us around and it is quite common although Dyslexia as a label covers quite a lot of ground. Thank goodness for computer and online spelling checkers!
Transformers are very underrated these days and all too often regarded as a nuscience.
The output transformers are quite easy to make and can be done by hand "in the hand" with a little care and patience.
The input transformer is not quite so easy and I would admit I haven't got round to making my own here either. This will require at least a little winding frame with a turns counter to get things even.
One way of ensuring symmetry is to wins both sides simultaneously, so then it doesn't matter if the number of turns between the inner and outer layer is different, as long as both sides are exactly symmetrical.
If one does get into winding bobbins regularly then a small lathe with a spindle handle (as one would use for manual turn thread cutting) becomes very useful.
Thank you for posting. It is very heartening to know that people such as yourself do appreciate this thread and helps keep me motivated to continue with the development and testing my end. I have spent a minimum of 40 hours a week these last couple of months supporting this one thread as well as buying components and upgrading test equipment and software. As you can no doubt imagine encouragement really means a lot to me.
Thanks. My locomotive is a long term project but I am pleased to say that my first monoflue prototype boiler is now under construction.
Meanwhile Jan-Eric in Finland:
http://members.surfeu.fi/animato/steamloco.html
... has built two locomotives using radiant burner boilers based on my design concept - although he has a tri-flue and a bi-flue versions as he has 7-1/4" gauge locomotives.
We think he must have elves helping out in his workshop
Best wishes,
Susan.
Hi Keypunch,
Thank you for your post and you interest in my work.
keypunch said:Hi Susan,
I just discovered this thread of your various personal DIY audio deversions. Just so you are aware, if my typing looks odd I think I can be or am "dslextic" as well
Welcome
It's quite possible, there is a lot of us around and it is quite common although Dyslexia as a label covers quite a lot of ground. Thank goodness for computer and online spelling checkers!
Transformers are very underrated these days and all too often regarded as a nuscience.
The output transformers are quite easy to make and can be done by hand "in the hand" with a little care and patience.
The input transformer is not quite so easy and I would admit I haven't got round to making my own here either. This will require at least a little winding frame with a turns counter to get things even.
One way of ensuring symmetry is to wins both sides simultaneously, so then it doesn't matter if the number of turns between the inner and outer layer is different, as long as both sides are exactly symmetrical.
If one does get into winding bobbins regularly then a small lathe with a spindle handle (as one would use for manual turn thread cutting) becomes very useful.
Thank you for posting. It is very heartening to know that people such as yourself do appreciate this thread and helps keep me motivated to continue with the development and testing my end. I have spent a minimum of 40 hours a week these last couple of months supporting this one thread as well as buying components and upgrading test equipment and software. As you can no doubt imagine encouragement really means a lot to me.
Thanks. My locomotive is a long term project but I am pleased to say that my first monoflue prototype boiler is now under construction.
Meanwhile Jan-Eric in Finland:
http://members.surfeu.fi/animato/steamloco.html
... has built two locomotives using radiant burner boilers based on my design concept - although he has a tri-flue and a bi-flue versions as he has 7-1/4" gauge locomotives.
We think he must have elves helping out in his workshop
Best wishes,
Susan.
H E A T S I N K S
Arrgh... it ate my post in edit...
Ok.
You can get heatsinks of any size or shape to do the job.
Avvid and Thermalloy have pages to calculate the requirements.
Add 25% for safety - it is better to be too big than too small!
It's absolutely not necessary to use the same extrustion shape or size that Susan has spec'd!!
They're available cheap surplus - sometimes it is cheap enough to buy an entire chassis or subassembly to just take the heatsinks and some random parts...
Ok, here's the home brew method I use:
(ascii representation)
|E
|E
|E
| <---devices mount here
E = smaller extrusions
| = backing plate - MUST BE FLAT
mount the smaller extrusions w/ fins VERTICALLY, 90 degrees rotated from as shown w/ heat sink goo and multiple screws, flat and tight to backing plate.
backing plate must be >0.250" *minimum* - fatter is better.
or:
]|[
]|[
_| <--devices mount here (I had to put a _ in to get the "|"plate to align... but a fat "L" on the bottom is legal too...
For a two sided set up - gets you more dissipation in less vertical space.
Derate the composite heatsink by ~15% compared to an extrusion, because of losses in the coupling between surfaces - the flatter the better the coupling.
You can see a homebrew example of this on my website in the Amplifier pages with my SE Mosfet amp... the caps on the chassis are 3" diameter, for size reference. The backing plate is 0.375" and the fins are made in a metal shop.
It runs 120watts of heat all the time for a 30 watt SE amp using 6 Hitachi Mosfets biased at ~3 amps...
_-_-bear
Arrgh... it ate my post in edit...
Ok.
You can get heatsinks of any size or shape to do the job.
Avvid and Thermalloy have pages to calculate the requirements.
Add 25% for safety - it is better to be too big than too small!
It's absolutely not necessary to use the same extrustion shape or size that Susan has spec'd!!
They're available cheap surplus - sometimes it is cheap enough to buy an entire chassis or subassembly to just take the heatsinks and some random parts...
Ok, here's the home brew method I use:
(ascii representation)
|E
|E
|E
| <---devices mount here
E = smaller extrusions
| = backing plate - MUST BE FLAT
mount the smaller extrusions w/ fins VERTICALLY, 90 degrees rotated from as shown w/ heat sink goo and multiple screws, flat and tight to backing plate.
backing plate must be >0.250" *minimum* - fatter is better.
or:
]|[
]|[
_| <--devices mount here (I had to put a _ in to get the "|"plate to align... but a fat "L" on the bottom is legal too...
For a two sided set up - gets you more dissipation in less vertical space.
Derate the composite heatsink by ~15% compared to an extrusion, because of losses in the coupling between surfaces - the flatter the better the coupling.
You can see a homebrew example of this on my website in the Amplifier pages with my SE Mosfet amp... the caps on the chassis are 3" diameter, for size reference. The backing plate is 0.375" and the fins are made in a metal shop.
It runs 120watts of heat all the time for a 30 watt SE amp using 6 Hitachi Mosfets biased at ~3 amps...
_-_-bear
Re: H E A T S I N K S
Dear Bear,
Thank you for your helpful post.
You are quite correct and the heatsinks I am showing are only an example of the size and capacity required.
Heatsinks can be quite different sizes and can be bought very cheeply from surplus or eBay as you say.
However I do feel that I should give some sort of definitive starting point and representation that people can use to get an idea about what they are looking for.
The CAD 3D sketch I have done is only one way of implimenting the amplifier - however it is not just an intelectual exercise as this is what I am intending to build myself. All the main componets are sized accuratly and based on stock items. Thus it is a "real" design that others may base their own amplifiers on should they so wish.
For those of us who have been making a living in electronics for may years a lot of this stuff is very simple and second nature. However there are others who are not so experienced and I try to give enough detail in the hope that everyone can understand the technicalities of what I am talking about and give the best chance of sucess.
Many thanks.
Best wishes,
Susan.
Dear Bear,
Thank you for your helpful post.
bear said:
You are quite correct and the heatsinks I am showing are only an example of the size and capacity required.
Heatsinks can be quite different sizes and can be bought very cheeply from surplus or eBay as you say.
However I do feel that I should give some sort of definitive starting point and representation that people can use to get an idea about what they are looking for.
The CAD 3D sketch I have done is only one way of implimenting the amplifier - however it is not just an intelectual exercise as this is what I am intending to build myself. All the main componets are sized accuratly and based on stock items. Thus it is a "real" design that others may base their own amplifiers on should they so wish.
For those of us who have been making a living in electronics for may years a lot of this stuff is very simple and second nature. However there are others who are not so experienced and I try to give enough detail in the hope that everyone can understand the technicalities of what I am talking about and give the best chance of sucess.
Many thanks.
Best wishes,
Susan.
Found great heatsinks for $15 US ea......
Following Bear's suggestions, I started searching around for surplus heatsinks with sufficient surface area and dissipation capacity.
After wading through endless CPU heatsink listings on eBay, I came across one guy with quite a few cases left of NOS(circa 1997) Wakefield high density heatsinks (pictured below) that could be perfect for Susan's amp(or any project), so I ordered 6 just in case I cannot find anything better.
They are black annodized aluminum measuring 5.875"(149mm) x 12"(305mm) with 16 fins at a depth of 2.875"(73mm). The heatsink plate is 1/8"(3.2mm) thick and total weight is 5 lbs(2.3Kg)
My tentative plan is to use them exactly as Susan has in her new rendering(two cut to 8" side-by-side)and add the two remaining 4" pieces side-by-side against the flatback side of the 8" pieces above the amp chasis(like below). The mosfets will be mounted just inside the top of the chasiss. 1,113 sq. inches(718,063 sq. mm) of heatsink surface should be enough.
[|]
[|]
[|]___[||]__
[|...............|
[|_________|
These are probably only worthwhile for those living in the USA because of shipping weight. The guy's instructions are to use the "Buy it now" price of $15 and then tell him how many pieces you really want in the winner's message. He will get back to you with the total price plus shipping.
Following Bear's suggestions, I started searching around for surplus heatsinks with sufficient surface area and dissipation capacity.
After wading through endless CPU heatsink listings on eBay, I came across one guy with quite a few cases left of NOS(circa 1997) Wakefield high density heatsinks (pictured below) that could be perfect for Susan's amp(or any project), so I ordered 6 just in case I cannot find anything better.
They are black annodized aluminum measuring 5.875"(149mm) x 12"(305mm) with 16 fins at a depth of 2.875"(73mm). The heatsink plate is 1/8"(3.2mm) thick and total weight is 5 lbs(2.3Kg)
My tentative plan is to use them exactly as Susan has in her new rendering(two cut to 8" side-by-side)and add the two remaining 4" pieces side-by-side against the flatback side of the 8" pieces above the amp chasis(like below). The mosfets will be mounted just inside the top of the chasiss. 1,113 sq. inches(718,063 sq. mm) of heatsink surface should be enough.
[|]
[|]
[|]___[||]__
[|...............|
[|_________|
These are probably only worthwhile for those living in the USA because of shipping weight. The guy's instructions are to use the "Buy it now" price of $15 and then tell him how many pieces you really want in the winner's message. He will get back to you with the total price plus shipping.
Website Updates
Dear All,
Pages updated this evening are:
http://www.susan-parker.co.uk/zeus-out-tx-75w.htm
and
http://www.susan-parker.co.uk/zeus-toroid-amp-1.htm
Note that the FFT 1 kHz signal levels are not at 0 dB but at -9 and -20 dB.
Best wishes,
Susan.
Dear All,
Pages updated this evening are:
http://www.susan-parker.co.uk/zeus-out-tx-75w.htm
and
http://www.susan-parker.co.uk/zeus-toroid-amp-1.htm
Note that the FFT 1 kHz signal levels are not at 0 dB but at -9 and -20 dB.
Best wishes,
Susan.
Zeus crashes 0.01% THD+N barrier
Dear All,
Using the new split chamber bobbin I have the following test results:
Output Transformer 4:1 mode, 8 ohm load.
Vac %THD+N Watts
1.52 0.0141 0.3
1.92 0.0115 0.46
2.00 0.0098 0.5
2.28 0.0089 0.65
2.56 0.0082 0.8
2.71 0.0073 0.9
2.87 0.0070 1.0
3.04 0.0067 1.2
3.41 0.0062 1.5
3.61 0.0058 1.6
3.83 0.0056 1.8
4.05 0.0058 2.0
4.29 0.0058 2.3
4.55 0.0059 2.6
4.82 0.0058 2.9
5.01 0.0059 3.1
5.41 0.0061 3.7
5.72 0.0063 4.1
6.06 0.0066 4.5
Output Impedance = 0.25 ohms.
See:
http://www.susan-parker.co.uk/zeus-test-2sk1529-fft-1khz.htm
for graphs etc.
N.B. The Emu1212 + TPA line driver loopback is at 0.005% so some improvement in measuring setup is required. PC based test systems are okay but do have their limitations and I will have to start "upgrading" my soundcard's analog side sometime soon.
Best wishes,
Susan.
Dear All,
Using the new split chamber bobbin I have the following test results:
Output Transformer 4:1 mode, 8 ohm load.
Vac %THD+N Watts
1.52 0.0141 0.3
1.92 0.0115 0.46
2.00 0.0098 0.5
2.28 0.0089 0.65
2.56 0.0082 0.8
2.71 0.0073 0.9
2.87 0.0070 1.0
3.04 0.0067 1.2
3.41 0.0062 1.5
3.61 0.0058 1.6
3.83 0.0056 1.8
4.05 0.0058 2.0
4.29 0.0058 2.3
4.55 0.0059 2.6
4.82 0.0058 2.9
5.01 0.0059 3.1
5.41 0.0061 3.7
5.72 0.0063 4.1
6.06 0.0066 4.5
Output Impedance = 0.25 ohms.
See:
http://www.susan-parker.co.uk/zeus-test-2sk1529-fft-1khz.htm
for graphs etc.
N.B. The Emu1212 + TPA line driver loopback is at 0.005% so some improvement in measuring setup is required. PC based test systems are okay but do have their limitations and I will have to start "upgrading" my soundcard's analog side sometime soon.
Best wishes,
Susan.
Re: Re: Found great heatsinks for $15 US ea......
Yep, I'm going to cut two 12" heatsinks per monoblock down to 8" and then mount those side-by-side to a 8" x 12" copper plate (>0.25" thick per Bear's advice"). Then, I will mount the 4" x 6" remnants side-by-side on the frontside of the plate above the top of the chassis. The mosfets will be mounted half way up the plate, just inside the top of the chassis.
Let's see if this works:
[ ]= heatsinks
|| = copper plate
[||]
[||]
[||]
[||]____{||}_____
[||<---mosfets---|
[||---------------|
[||_____________|
Susan-Parker said:
Yep, I'm going to cut two 12" heatsinks per monoblock down to 8" and then mount those side-by-side to a 8" x 12" copper plate (>0.25" thick per Bear's advice"). Then, I will mount the 4" x 6" remnants side-by-side on the frontside of the plate above the top of the chassis. The mosfets will be mounted half way up the plate, just inside the top of the chassis.
Let's see if this works:
[ ]= heatsinks
|| = copper plate
[||]
[||]
[||]
[||]____{||}_____
[||<---mosfets---|
[||---------------|
[||_____________|
Re: Zeus crashes 0.01% THD+N barrier
funny !!
I like it, but i have to mutch noise, 100 Hz on my (Monito-)Speaker.
So i need change the setup !
You know, i use this setup: http://bauteile-fuer-die-elektronik.de/projekte/Zeus/project/index.php
All the best and i wish you good look !
- uwe
Susan-Parker said:
funny !!
I like it, but i have to mutch noise, 100 Hz on my (Monito-)Speaker.
So i need change the setup !
You know, i use this setup: http://bauteile-fuer-die-elektronik.de/projekte/Zeus/project/index.php
All the best and i wish you good look !
- uwe
Re: Re: Zeus crashes 0.01% THD+N barrier
Hi Uwe,
Good to hear from you, I trust you are keeping well.
How is the sound for you (apart from the hum)?
Okay.
I have found that on the bench I need at least 50 cm separation from the power transformer to the input transformer to get mains pickup down to an acceptable level.
You can see the 50 Hz and it's harmonics down at the -120 dB level in the residual level graph on the most recent set of tests.
http://www.susan-parker.co.uk/zeus-test-2sk1529-fft-1khz.htm
"Residual noise for above tests (signal = mute)." i.e. the third graph down, as I am now doing 43 Hz to 96 kHz FFTs.
The upshot of this is that although toroids have much closer fields than EI transformers they still have some and at the levels and sensitivities we are using it can be apparent.
Hence the new Zeus 75 design has the input transformer in a steel chassis and places the power supply in it's own separate steel box. The field thus has to get through two magnetic screens to the input transformer and if I am going to be really pedantic I can have the input transformer inside it's own Mumetal can as well.
Secondly I propose that you try cross coupling the two output transformers using the 230 volt windings. This will link the two and hopefully help balance the low frequency pickup.
At the moment you are running two single ended zeus output stages and these won't necessarily cancel out 100% of the supply ripple or induced currents.
I tried using a 500 VA toroid as a proper output transformer and did get a quite reasonable bandwidth with the top end -3 dB at 20 kHz.
http://www.susan-parker.co.uk/zeus-toroid-amp-1.htm
(Added to the front of this page.)
A 330 VA transformer might go a bit higher, but one has to watch the bias currents as the windings get quite thin.
This page:
http://www.hardwarebook.net/table/AWG.html
is really useful as it gives average currents for each wire size (in mm and AWG). So for example max average current for a 1.0 mm wire is 2.36 amps.
With the test setup and 2:1 transformer I have about 2.5 amps peak at 50 watts into 8 ohms, so this is about right (as I am not going to be running 50 watts continuously in my living room!).
This gives a good guide for the bias current setting as well.
BTW. I noticed your picture has changed. It looks too good to be a live webcam. I envy you your view (in a very positive manner) as all I get to see when I look out of my windows are houses and gray rooftops stretching off to the horizon.
Best wishes,
Susan.
Hi Uwe,
Good to hear from you, I trust you are keeping well.
dx.master said:
How is the sound for you (apart from the hum)?
Okay.
I have found that on the bench I need at least 50 cm separation from the power transformer to the input transformer to get mains pickup down to an acceptable level.
You can see the 50 Hz and it's harmonics down at the -120 dB level in the residual level graph on the most recent set of tests.
http://www.susan-parker.co.uk/zeus-test-2sk1529-fft-1khz.htm
"Residual noise for above tests (signal = mute)." i.e. the third graph down, as I am now doing 43 Hz to 96 kHz FFTs.
The upshot of this is that although toroids have much closer fields than EI transformers they still have some and at the levels and sensitivities we are using it can be apparent.
Hence the new Zeus 75 design has the input transformer in a steel chassis and places the power supply in it's own separate steel box. The field thus has to get through two magnetic screens to the input transformer and if I am going to be really pedantic I can have the input transformer inside it's own Mumetal can as well.
Secondly I propose that you try cross coupling the two output transformers using the 230 volt windings. This will link the two and hopefully help balance the low frequency pickup.
At the moment you are running two single ended zeus output stages and these won't necessarily cancel out 100% of the supply ripple or induced currents.
I tried using a 500 VA toroid as a proper output transformer and did get a quite reasonable bandwidth with the top end -3 dB at 20 kHz.
http://www.susan-parker.co.uk/zeus-toroid-amp-1.htm
(Added to the front of this page.)
A 330 VA transformer might go a bit higher, but one has to watch the bias currents as the windings get quite thin.
This page:
http://www.hardwarebook.net/table/AWG.html
is really useful as it gives average currents for each wire size (in mm and AWG). So for example max average current for a 1.0 mm wire is 2.36 amps.
With the test setup and 2:1 transformer I have about 2.5 amps peak at 50 watts into 8 ohms, so this is about right (as I am not going to be running 50 watts continuously in my living room!).
This gives a good guide for the bias current setting as well.
BTW. I noticed your picture has changed. It looks too good to be a live webcam. I envy you your view (in a very positive manner) as all I get to see when I look out of my windows are houses and gray rooftops stretching off to the horizon.
Best wishes,
Susan.
Re: Re: Re: Zeus crashes 0.01% THD+N barrier
Yes, we have today +6°C and some sunshine, now it's raining ..
I like the Sound !
Today i orderd other one Input Tr1 :
- http://bauteile-fuer-die-elektronik...=/012-20041108/01-TR1_Input_Transformator.jpg
with a dealer near by me.
So i can start make a second channel
.
All other construction unit, like the output Tr (Tr2):
- http://bauteile-fuer-die-elektronik...howimg.php?file=/013-20041218/01-20041218.jpg
the Heatsink with mosfet IRF530N:
- http://bauteile-fuer-die-elektronik...ile=/011-20041102/05-heatsink_with_mosfet.jpg
the Power Supply Tr (Tr3) and the power supply :
- http://bauteile-fuer-die-elektronik...mg.php?file=/011-20041102/02-power_supply.jpg
ans so on where HERE !
I will test, if i can separate the power transformer more to the input transformer.
You can see i try my best: on the left is the TR3 and on the right the TR1.
- http://bauteile-fuer-die-elektronik...howimg.php?file=/012-20041108/04-Zeus_Amp.jpg
Yes i understand.
Yes it's usefull for me.
Yes ?!
I use three digital images an stitch them together with a imagetool.
Then some more strip and resize -> the picture is ready !
I like your project 'Zeus' and I hope we can find all a nice and sounding well solution with our construction unit we have !
A nice weekend - uwe.
Yes, we have today +6°C and some sunshine, now it's raining ..
I like the Sound !
Today i orderd other one Input Tr1 :
- http://bauteile-fuer-die-elektronik...=/012-20041108/01-TR1_Input_Transformator.jpg
with a dealer near by me.
So i can start make a second channel
.All other construction unit, like the output Tr (Tr2):
- http://bauteile-fuer-die-elektronik...howimg.php?file=/013-20041218/01-20041218.jpg
the Heatsink with mosfet IRF530N:
- http://bauteile-fuer-die-elektronik...ile=/011-20041102/05-heatsink_with_mosfet.jpg
the Power Supply Tr (Tr3) and the power supply :
- http://bauteile-fuer-die-elektronik...mg.php?file=/011-20041102/02-power_supply.jpg
ans so on where HERE !
I will test, if i can separate the power transformer more to the input transformer.
You can see i try my best: on the left is the TR3 and on the right the TR1.
- http://bauteile-fuer-die-elektronik...howimg.php?file=/012-20041108/04-Zeus_Amp.jpg
Yes i understand.
Yes it's usefull for me.
Yes ?!
I use three digital images an stitch them together with a imagetool.
Then some more strip and resize -> the picture is ready !
I like your project 'Zeus' and I hope we can find all a nice and sounding well solution with our construction unit we have !
A nice weekend - uwe.
Re: Re: Re: Found great heatsinks for $15 US ea......
Copper = bad.
Copper and aluminum are not friends. You will get horrendous electrolytic corrosion effects.
Go with aluminum plate. Cheaper and easier to machine, easier to find.
As mentioned, the *flatter* both surfaces are the better the heat transfer.
If you use Plastic pak mosfets you can mount flat to the plate at the bottom.
If you use TO3 style (metal case) mosfets you'll need to add an "L" bracket, (thicker is better) machined to accept the pinout for the Mosfet mounting holes and pins. (Use a commercially drilled heatsink as ur template - hint)
Unless, of course, you go through the plate with the mosfets.
In which case there is little benefit to putting heatsinks BELOW the point at which the mosfeters are mounted... heat likes to rise.
Leave some meat below for thermal spreading, and put the heat fins above...
Use as many screws as possible to hold the the pieces together, with heatsink compound inbetween.
Derate according to the calcs on the heatsink mfrs sites...
Figure on 4x the nominal ouput power to be used in heat, and ur good to go...
_-_-bear
darkmoebius said:
Copper = bad.
Copper and aluminum are not friends. You will get horrendous electrolytic corrosion effects.
Go with aluminum plate. Cheaper and easier to machine, easier to find.
As mentioned, the *flatter* both surfaces are the better the heat transfer.
If you use Plastic pak mosfets you can mount flat to the plate at the bottom.
If you use TO3 style (metal case) mosfets you'll need to add an "L" bracket, (thicker is better) machined to accept the pinout for the Mosfet mounting holes and pins. (Use a commercially drilled heatsink as ur template - hint)
Unless, of course, you go through the plate with the mosfets.
In which case there is little benefit to putting heatsinks BELOW the point at which the mosfeters are mounted... heat likes to rise.
Leave some meat below for thermal spreading, and put the heat fins above...
Use as many screws as possible to hold the the pieces together, with heatsink compound inbetween.
Derate according to the calcs on the heatsink mfrs sites...
Figure on 4x the nominal ouput power to be used in heat, and ur good to go...
_-_-bear