F5 Turbo Circuit Boards

[lhquam]

Originally Posted by Horio

Your option 1 is very similar to what I am wanting to build (except I would be using 8 single FET output boards per channel). I've realized that I don't need to cascode as the rails on the balanced build are going to be half that of an equivalent power single-end amp. I'm thinking of shooting for +/-24V rails which should be well within the safe limit for the JFETs.

I have ideas for 3 versions of the balanced F5T:

1. Balanced stereo at 46W/channel using 2 x 1/2 E007 heatsink with fan as shown in F5 power amplifier.
2. Balanced stereo at 100W/channel using 2 x E007 heatsinks fin-to-fin with 2 fans.
3. Balanced monoblock at 200W/channel using 2 x E007 heatsinks fin-to-fin with 2 fans.

[sekess]

Hey Ihquam,
I was thinking about up a similar forced air build. Where did you get the heatsink simulation software from?
Thanks,
Steve

Originally Posted by lhquam

Here is a heatsink simulation based on using 1-1/2 HeatsinkUSA E007.
Unfortunately, HeatsinkUSA doesn't make anything wider than 256 mm, so I cut one E007 in half and weld it to a full E007 to make a 36 fin 384 wide. The resulting heatsink is 384 mm (W) x 130 mm (L) x 74 mm (H), with a baseplate thickness of 9.5 mm and has 36 fins. Too bad that such a heatsink isn't available off-the-shelf.

The simulation assumes 8 MOSFETs dissipating 22 watts each spaced according to the F5-Turbo-v2/3 layout. I also assume FORCED CONVECTION. The image shows the heat rise using .693 m/s air flow rate. Looks pretty good.

I imagine assembling two such heatsinks with the fins facing each other and using 3x 120mm low-noise fans in the same manner I used with the F5 shown HERE: F5 power amplifier

[sippy]

Toecutter,
I was intending the V3 build as a long term project as the build cost is looking some what.......large.
I dont know what kinda beta testing time frame you would be looking at, but I'm willing to beta a V2 build, if you want to talk more I'll shoot you a pm with my email addy.

[Tea-Bag]

Prototyping doesn't necessarily mean doing everything all nice and tidy.
If you can get the front end, and maybe a pair of outputs working on a channel with a MUR3020, I say it's good to go.
Some of it is about component stuffing to make sure things in reality put together fairly easily and precisely.

[sippy]

Aye Tea-Bag (love the id btw (being a yorkshire man and all)).......I've done proto work on 2 and 4 stroke porting, but its been a while since I've built an amp. V2 would allow me to help proto and if it works as a single channel, i'd have a ps only needing a second channel to be added.
My offer still stands.

[chalkandtalk]

Ihquam:

I too aim for 2 balanced monoblocks, and 100W per channel. I've looked at your heatsink tunnel with fan and it's a cracking idea. So I'm going to emulate that. Have you calculated any bias and voltage ratings?

Chalk

[SRMcGee]

Folks:

As excited as I am about this project, I confess that my confusion about how to heatsink a 100w/ch F5T is leading me to second-guess reserving V3 Output Boards; perhaps using the single FET output boards and spreading the heat across a larger physical space would be a more stable solution (but who needs a pair of HiFi2000 5U 500mm monoblock amps?). I've been spending a lot of time trying to identify adequate cost-effective solutions but am hampered by massive ignorance. My hope is that those of you who know what you are doing can provide some heatsinking guidance before Andy closes his pending group buy.

Regards,
Scott

[UKToecutter]

I wonder how many people will actually bias to Class A @ 100W???

[Horio]

I'm thinking of shooting for around 100W into 8 ohms Class A with my balanced mono-blocks. Going balanced makes it a bit easier.

[rdk845]

Originally Posted by Horio

I'm thinking of shooting for around 100W into 8 ohms Class A with my balanced mono-blocks. Going balanced makes it a bit easier.

Hi Horio,

I'm probably wrong here, but at given heatsink capacity and given transformer VA rating, wouldn't single ended better in getting higher class A power, while balanced better in getting better overall power? Even though you get to use lower rail voltage at given overall taget power with balanced so higher bias current per device at given heatsink and transformer capacity, in SE mode 4 pairs of devices are in pararell adding up the current, while in balance mode there are only 2 of 2 pair devices in parallel to add the current against speakers? I'd appreciate your correction if I'm wrong.

Thanks!