Yet another F5 building log

dimkasta · 2012-05-20 1:57 am

no steel shield for them either...

dimkasta · 2012-05-20 2:01 am

Buzz what voltage will you be running your turbo? and towards what wattage?

AudioSan · 2012-05-20 2:03 am

here's a shielded one that works for standar F5
Antek - AS-4218

buzzforb · 2012-05-20 2:09 am

32V and exactly like schematic. That should yield about 6A+ Ipk, good for about 200W Class A into 6 ohms, which is my speaker impedance, and about 300W dissipation per channel. Damn yhats hot. If kept it will drive my woofers in an active system.

AudioSan · 2012-05-20 2:18 am

wow

so little over 4A bias on +/-42V rails?
that will be on the limit for the diodes to conduct?
or are you gonna use more then 4 pairs of outputs?

dimkasta · 2012-05-21 1:19 pm

OK ordered the matched fets today.

Currently working on 1+2 idea for the case, but in dual mono topology (2 monoblocks)
I love it for its price and looks, without sacrificing anything in functionality.

If I manage to steal some time from work, I will try to make a 3d model of the layout

CENTRAL · 2012-05-21 1:27 pm

I sometimes see big transformers being accused of unwanted vibration effects...

AudioSan · 2012-05-21 1:32 pm

thats often the case with cheap one's. thicker the wire, bigger the chance

dimkasta · 2012-05-21 2:05 pm

I was browsing through some transformers and (instead of going for 18V) I was thinking that I will at some point try to turbo it, so I might as well go for 24V transformers from the start (32V rails as per the pre-V1 version in the Turbo article).

Plus, in the 24V range I am finding some nice 500+ VA from mouser and Hammond with free shipping.
The Antek ones will cost almost the same including shipping....
Plus, the Hammonds go as high as 1KVA (although way too heavy) which is really tempting for tests since they cost only 10 euros more... and I agree with you guys, vibration is my concern too.
I have tried Hammonds before, and they tend to vibrate as you go up in VA on the same Voltage (120VA with 12V rails was nice and silent. 225VA was vibrating like crazy)

External PSU is quickly becoming an attractive choice too...

CENTRAL · 2012-05-21 2:11 pm

I'm feeling more & more certain this building log is going to get very interesting, very soon!!!

AudioSan · 2012-05-21 2:40 pm

dimkasta said:
I was browsing through some transformers and (instead of going for 18V) I was thinking that I will at some point try to turbo it, so I might as well go for 24V transformers from the start (32V rails as per the pre-V1 version in the Turbo article).

Plus, in the 24V range I am finding some nice 500+ VA from mouser and Hammond with free shipping.
The Antek ones will cost almost the same including shipping....
Plus, the Hammonds go as high as 1KVA (although way too heavy) which is really tempting for tests since they cost only 10 euros more... and I agree with you guys, vibration is my concern too.
I have tried Hammonds before, and they tend to vibrate as you go up in VA on the same Voltage (120VA with 12V rails was nice and silent. 225VA was vibrating like crazy)

External PSU is quickly becoming an attractive choice too...

with +/-32V rails you will need dual outputs(or much lowe bias)
and you might want to cascode the J-fets too

AudioSan · 2012-05-21 2:51 pm

maybe go streit for a turbo V1? CVillers boards with dual outputs and cascoding.
2x4-500VA 2x24V transformers and a bias of 1.9A(0.95A pr device)
and volla. you have a turbo v1

2x50W classA in 8ohm

dimkasta · 2012-05-21 3:09 pm

Dual outputs is not a requirement, as long as we do not exceed half the max dissipation rating of the mosfets (as per papa's pre-V1 notes).

5W resistors and conservative heatsinking is a requirement though, things that I was already planning anyway.
Oh and removing the current limiter might be an option too.

AudioSan · 2012-05-21 3:13 pm

over 40W dissipation pr device is a little on the hot end after my taste.
32x1.3=41.6W
also 32V for the 25V rating J-fets is also a little hi after my taste.
if you only use 1pair of outputs, you also need a thicker baseplate on your sinks.

dimkasta · 2012-05-21 3:38 pm

Thanks for the input

Well I am planning to use the heatsinks specified for the stereo version to dissipate a mono version, so I guess this will be ok.

The jfets get around 30V which is rather within papa's comfort zone, provided that we either chose a lower idss (~8mA) or use a small heatsink on them.

It s getting a bit "hairy" to be honest...
I think for now I will stick to stock F5 and after I am comfortable (since this is my first build) I will go for a V3.

Fraidy-Cat mode => ON

AudioSan · 2012-05-21 3:49 pm

so i output pr side?
then you need to mount the fet's outside the board with gate stopper and source resistor soldered to the fet

also. with a heatsink with 10mm baseplate you will have only 100mm effective raidius around your fets(maybe less). the sinks will not be able to transfer heat over a longer distance. at least not very vel

dimkasta · 2012-05-21 5:30 pm

AudioSan said:
so i output pr side?

For now I will stick to 18V toroid, 23V rails and 1,3A per channel.

Too many things can go wrong and I do not feel confident yet experimenting.

I m still considering the huge toroids though

Bksabath · 2012-05-21 7:09 pm

dimkasta said:
Guys just a question.
Your suggestions are based on experience between 1KVA and smaller ones, or is it just that you are content with the smaller ones?

The reason why I am asking is that I see you saying that the small ones should be "sufficient" or "enough" or other adjectives that tend to describe something that is just working without problems.

Does anyone have actual experience indicating that the 1KVA has no sonic difference over for example the 300VA?

The results I have so far form my phono indicate that the difference is huge with a bigger transformer.

Please have a look at this blog entry by Richard RJM
http://www.diyaudio.com/forums/blog...ower-transformer-than-you-think-you-need.html

You will see that the rectifier-diode/filter combination does not need current on a continuous basis, but rather in a spike-like manner. And the spikes can reach easily 10 times the nominal current draw. Which can easily be translated according to Andrew's calculations (minus the cooling compensation) to 2,6x10=26A or and for 2x18 secondaries to 18x2x26=936VA

I have tried/verified Richard's theory on my phono and there is a huge difference in the sound.

I see no reason why the bigger PSU should behave any differently. Especially with the huge filter caps that practically act like shorts

1 Get what you can afford
2 bigher is better
3 what if you want to go balanced tomorow?

SRMcGee · 2012-05-21 7:15 pm

buzzforb said:
32V and exactly like schematic. That should yield about 6A+ Ipk, good for about 200W Class A into 6 ohms, which is my speaker impedance, and about 300W dissipation per channel. Damn yhats hot. If kept it will drive my woofers in an active system.

buzzforb:

You are apparently going to use Antek's AN-6224. Antek also offers 800VA and 1000VA toroids with 25V secondaries. Can you help me understand why the risks associated with the extra 1 volt are more compelling than the advantages of either of the significantly larger transformer alternatives?

Regards,
Scott

buzzforb · 2012-05-21 7:25 pm

No reason. 25x1.41 will yield about 34V rails. I own 8 Antek toroidal and have found them to be very stout. Just decided 1.2kV total was enough for me. I am a little cheap too!

Side note; if this were my final build, I would go for 800VA 25V secondary

Yet another F5 building log

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Member

Banned

Member

Member