200 watts Gäjnklon (Gainclone) - BPA200

[Mattyo5]

I think just from lurking here on diyaudio for the most part that there is a lot of interest in an amp like this...especially if the pcb's are avaliable ... w/ different #'s of devices... 4,6,8 lm3886 (or 75) or whatever you are using. Any way you make it avaiable, people will read it I'd definitely like to read it I hope others chime in though

-Matthew K. Olson

 

[Da5id4Vz]

Proof Reeder, uhm I mean Reader...

Proof Reeder, uhm I mean Reader...

I'd be more than happy to proof read the document. Of course I have to warn you that your English may already be better than my own.

I think that hosting on your web site may be the most convenient for both yourself and readers.

DIY maintains a Links page that is a godsend reference to many of us. I'm not sure what is involved in getting you listed there. If the moderators don’t pop in with the information (might be listed under FAQ too) I'll certainly be glad to look into it.

I'll be out of touch until next Tuesday so I cant be much help until them.

-Dave

 

[peranders]

Thomas, I suggest that you

1 do your best

2 publish the article in html on a regular homepage, any! You don't have to have a fancy URL.

3 Tell clearly at the top of the document that proof reading is wanted

4 Feedback in form of a html file or text in email is wanted.

When the article is descent move it to the magazine's website.

I have the swedish article which isn't 100% correct but a good start so I will be able to help you.

I think you should take advantage of all helpfull people here.

 

[UrSv]

Well guys, I got my PCB today from Thomas. What can I say?

It looks pretty darn good IMHO. Very high quality feel, nice planning of placement and very compact. I can also confirm that the way the PCB is designed gives the possibility of using the standard standing LM3886s if one does not have any LB14 flat mount ones. With standing chips you can use a rod, as mentioned before in this thread, of some 20 mm thickness and some 30 to whatever fits high and get excellent cooling.

Excellent implementation Thomas!

Tomorrow is building day...

 

[joensd]

I´d be interested in PCB´s.
What about a little group buy and save some bucks?
I want 2 boards so I´d need 4 more for the cheaper price.
Anybody from Europe would be sensible.

Jens

 

[peranders]

I notice that the regulators are connected directly to the main power supply voltage. Most regulators have 35 V a maximum rating so care should be taken if the voltage is 42-45 volts. This can easily be fixed with a series resistor and maybe a power zener at 33 volts. There are HV versions (60 volts) but I don't know right now if they are available in TO92. Besides the old 78/79xx new regulators have appeared (car industry).

If a series resistors are connected _maybe_ an extra decoupling is needed just before the regulator.

BTW: My headphone amp works still inspite of the +- 15V at the poor AD8610 (max +-13.6V)

 

[tlmadsen]

Regulators on high voltage

Hi Per

(all this is also in the artical, but I take it anyway)

You are right about the voltage regulators. On the paper they are only god for up to 35 V input. So as you so, one could put in a coupel of resistors to lower the voltage to them, if needed.

I should say that I have used several of DAS MODUL at high voltage level (over 40 V) and the regulators have always worked perfect. Just like your expirence with the AD8610.

A lot of new regulators can be used, but just for powering a unity- gain buffer and the Servos, I don't think that you are gainig much, if anything by going to something more exotic.......

BTW. I am now selling a "starter-pack" made of one PCB and 4 x LM3886 with a 25 % discount. This means a total price of 75 $ incl shipping to most of the world. (compared to the normal $ 100.00).

Payment is in cash or paypal (add 3 %)

Thomas

 

[Mattyo5]

Do you have pcb's for 6 and 8 lm3886 chips, or just 4? and prices? thanks!

-Matthew K. Olson

 

[tlmadsen]

Das Modul prices

Matthew

The PCB for DAS MODUL is for 4 x LM3886. If you need more than this you simply parallel as many modules at needed.

The prices are:

The DAS MODUL PCB: 45$ each incl. shipping to most of the world.
(at no, I don't make a big profit on this. High quality PCB's cost money)

The LM3886-LB14: $11 each + shipping. Normally around $10 to most of the world, no matter how maney you buy.

Starter-Pack: One DAS MODUL-PCB and 4 LM3886-LB14 incl shipping to most of the world: 75 $ (this is about 25% of).


Payment is in cach or PayPal (add 3 %).

All sale is "private to private".

Thomas

 

[elcheapo]

I want a PCB.

Please would you tell , where did you got that GC PCB ?

I already wrote to HIFI.DK magazine but also for redundancy I post here same question.

If I see a GC at reasonable price, I'll buy it othewrise I'll quit :=)

IMHO too much mysteria about Sakura's controversies.

Bests

 

[peranders]

Just get in contact with Mr Madsen, above.

 

[carlosfm]

Re: Das Modul

Under "work" this top-profile get surprisingly hot, which just tells you how bad cooling you get if you are just using the hole.
Thomas

Heat goes up.
I would definitely not use the chips this way.
The top-profile will be , and the heatsink doesn't get at all.

Excellent work on the PCB, Thomas.
But the servos are not for me.

 

[tlmadsen]

top-profil and servos

To carlosfm

You are right about your observation about the heat, but the top-profile also gets hot if you use the "standing-up" approch. In both cases the heatsink too get hot. I think it is therefore resonable to conclude that the top-profile does have a (positive)funktion.

Thanks for the nice words on the PCB.

I am looking at something without the servos. Persoanlly I like them, but they take up a lot of expensive PCB-area.

Regards

Thomas

 

[UrSv]

Re: Re: Das Modul

Heat goes up.
I would definitely not use the chips this way.
The top-profile will be , and the heatsink doesn't get at all.

Excellent work on the PCB, Thomas.
But the servos are not for me.

Sure you're not confusing heat conduction in the metal pieces involved with hot air rising? I'd be surprised to see heat having any tendency to travel upwards rather then sideways in a piece of aluminium.

 

[carlosfm]

Re: Re: Re: Das Modul

Sure you're not confusing heat conduction in the metal pieces involved with hot air rising? I'd be surprised to see heat having any tendency to travel upwards rather then sideways in a piece of aluminium.

The position of that piece of aluminium is important.
It's just that it dissipates better on a vertical position than horizontal.
In a horizontal position, with the chip on top, heat tends to stick near the chip and doesn't go away that easily.
It comes back to the chip.

 

[peranders]

I think the main problem is to unsure heat conduction from the chip out to the heatsink. The best solution is to attach the LM's directly to the heatsink but this can be impractical for some. My pcb is done for this. For me t's even more important because I want to use the TF type.

An externally hosted image should be here but it was not working when we last tested it.

 

[Jax]

The important thermal interface is from the heatsink to ambient air. This works best with vertical fins since air tend to raise when heated.

The position of the chip is totally irrelevant.

I can see an advantage with horizontally mounted chips if you squeeze the chips between an U-formed clamp and an L-profile in contact with a vertical main heatsink. The U will help dissipate heat but most of the heat is sucked out from the L to the main heatsink.

Heat goes from warm to cold in a thermal conductor, not back to the warm part.

 

[peranders]

Mr Madsen has already stated why he did his arrangement and it's only important if the amp will go in heavy duty, otherwise I gather that his cooling arrangement is sufficient.

 

[macboy]

It's just that it dissipates better on a vertical position than horizontal.
In a horizontal position, with the chip on top, heat tends to stick near the chip and doesn't go away that easily.
It comes back to the chip.

Heat conduction through a chunk of metal has nothing to do with the orientation with respect to the direction of gravity. Up or down doesn't matter. Heat convection (into the air) certainly is affected by direction if you are relying on natural airflow rather than forced air (a fan), because hot air will rise. "Heat" does not rise; heat will conduct equally in any/all directions through a heat conductor such as aluminum.

I think the main problem is to unsure heat conduction from the chip out to the heatsink. The best solution is to attach the LM's directly to the heatsink but this can be impractical for some. My pcb is done for this. For me t's even more important because I want to use the TF type.

You are absolutely correct, that you should mount the devices directly to the heatsink (no silicone pad or mica insulator, just thermal grease/compound) in order to get the best heat conduction from the device into the heatsink. But using the TF type device defeats this entirely; the package's thermal resistance is much higher than the metal tab on the the normal device. Even if mounted directly to the heatsink you will not be assured of better conduction with the TF device that the normal device used a good quality and properly installed isolator between the device and the heatsink. The best solution is the normal device mounted to the heatsink with a good quality thermal compound. The only difficulty is that you must now electrically isolate the heatsink from the chassis.

 

[peranders]

macboy, I agree that the TF is worse but I imagine that for home environment it will work. If it doesn't work good it's only to change to the "T" type.