amp15 picture thread

[mikechw]

Here is a bottom picture of AMP15:


1/4 W Gate metal film resistors measured 3.5ohm, I couldn't get 1/8W resistors.
With 1N4001G diode.


Still haven't try with both channel running...
I now need to resolder the 2 bulk caps because it is getting loose...

[more10]

quote:Originally posted by mikechw

The input cap is a Wima MKP10 2.2uF 160V.

Why did you change input cap? This is higher voltage of course, but that cannot be the reason.

Now I realize the uses of all those holes where the input cap is. You can put any sized cap there :-)

[V-bro]

Ok as promised:

































The green PCB models are prototype boards. I've also been experimenting on some of the other boards so there may be changes compared to what's in the stock kits...

[mikechw]

quote:Originally posted by more10

quote:Originally posted by mikechw

The input cap is a Wima MKP10 2.2uF 160V.

Why did you change input cap? This is higher voltage of course, but that cannot be the reason.

Now I realize the uses of all those holes where the input cap is. You can put any sized cap there :-)

Polypropene is slightly better than Polyester.
But it would be a little larger and it may pick up unwanted noises.

[wadoka]

Yes, the first set of pics are my Amp15. Note that the power transfer wires are NOT fitted at this point, look further down for amps that have them.

Please forgive my lack of knowledge, what are the advantages of running faster BBM settings? I guess all you Truepathers know, but I've yet to build one

[mikechw]

A faster BBM means a shorter dead time ===> less distortion...

I have been running it for more than a few hours / single channel.

V-bro measured ~45mA for VN10 80ns both channel.
I think with 40ns BBM, current draw may be doubled, so its close to limit of 100mA for the 7810.

I will solder both input channels tomorrow and test it with both channels running at high volume...

[mikechw]

Actually the idle switching frequency is around 600KHz, according to AN16 application note.

40ns Break Before Make setting ===> 1 / 40ns ===> 25MHz
80ns ===> 1 / 80ns ===> 12.5MHz

Considering 25Mhz is ~40 times faster than the idle switching frequency, and all these high frequency will be filtered out by the output filter, smoothed out to form the actual sound wave. I think the quality of the output filter should be more important factor affecting the sound quality.

[mikechw]

AMP15 put in case:


Power led added on the +ve rail with a 10K resistor.


It is powering two KEF 6ohm speakers.

[wadoka]

Nice!
Do you mind if I ask where the case came from and what input caps are you using?

Matt

[mikechw]

I bought the case and the softstart module from diyclub.biz

http://diyclub.biz/catalog/product_i...roducts_id=599

It doesn't include power switch, rca sockets, speaker terminals and IEC plug.
I bought those from a local electronic shop.

There are six hinges that connect the side,top,bottom,rear and front panels.
After I put all the screws in, there are 0.5mm gap between the panels on one side and no gap on the other side.
So the quality is not very good, but you will only notice it if you look closely.

There were some pre-drilled holes on the bottom panel but it was for another amp.
I drilled 2 holes for the 2 mosfet and one for the transformer.
I mounted the 2 transistor on the top of the board and the 2 diodes on the bottom.
The 2 diodes are just touching the bottom panel without screws.
( I was lazy to measure and drill 2 more holes )

The soft start unit and the board is stick to the bottom panel through double side 3M tapes.

I have to put tapes on the top of the transformer because it was very close to the top panel. Transformer is 6 cm in height and the internal height of the case is 6.3cm.

Input capacitors are MKP10 2.2uF 160V.

After running it for over 2 hours, the amplifier does get warm.
I then measured DC offset from both speaker terminals and it was 6mV.
I turned it off for one minute, then power it on again and DC offset is around 1mV.
I think I need to retune the DC offset resistor.