Commercial Gainclone kit- building instructions

[kierownik]

I meant the electrolytic next to the chip

[Peter Daniel]

You might consider BG N 100/50 for mid/high channels, they are still available from Percy.

Otherwise I would probably go with regular sizes.

[kierownik]

Thanks Peter

[martinbls]

Hello to everybody!

I just ordered a LM3875 Classic Kit from Peter and I still have some questions. I would be happy if some of you guys could tell me about your thoughts and experiences!

1) First and most important: The kit comes without any coupling capacitor. I plan to test some nos dacs in the near future. Some of these have output coupling caps, but some are DC coupled, without any output coupling cap at all. I'm not sure wheather some of these devices could damage my amp or my speakers because of this, so I thought about using a cap (BG N 4,7uF/50V) instead of R1 to be on the safe side. But, what do you think? May it cause problems using the LM3875 kit without input-coupling cap in combination with a source that has no output-coupling cap? (Everybody says something like this: "If you don't use a input-coupling cap you have to make sure that there is no DC passing through the amp". - But: How can I be absolutely sure about that, especially when using different source devices from time to time?) Of course, I would like to go without coupling cap on the LM-kit.

2) I got a nice enclosure for the amp to hands, but it is rather small. A 250VA or even 300VA toroid won't fit in, but a 160VA would. Is there anybody who uses a 160VA for this kit? What are your thoughts about that? Of course I could buy another enclosure and go for a larger trans, but I like small hifi components and therefore this case would be great.

3) I have heard that the small 10uF Panasonic caps on the rectifier board are not really necessary. Would it probably be better to omit them? Side note: I plan to build a stereo setup with one trans and one rectifier board only.

4) Do you have an advise for a proper heatsink, expressed in K/W per channel? Would something around 2 K/W per channel be sufficient?

That's it for the moment, thanks a lot!

martin

[raypalmer]

I use a 140VA transformer in a stereo setup(my enclosures are tiny), it works superbly well! I have buckets more power than needed.

As for heatsinking, I just used a chunk of aluminum about 8cm by 8cm by 15mm or so. Peter's design is extremely versatile.

There's my 140 in the wee enclosure

[AndrewT]

Measure the output offset of the source equipment.
If it never varies from <=+-10mVdc, then you can assume it has a DC blocking cap inside or that they have adopted some form of DC servo to reduce the offset.

If you choose to DC couple the source equipment to the power amps, then I recommend that you fit DC detection to the amplifier output and this can trigger input mute and/or output isolation. You can also fit a DC servo to reduce the excessive and variable residual output offset that is very likely to exist in a DC coupled amplifier.

[canadino]

Without source connected I had a channel DC offset from 4mV (pot minimum) to 46mV (pot maximum), the other channel from 44mV to 120mV.
With source connected (Linn mimik II) I had a channel DC offset from 13mV (pot minimum) to 57mV (pot maximum), the other channel from 80mV to 230mV.
After I install coupling caps (photo) and the DC offset go crazy: one channel measure 82mV (pot minimum) 87 (pot max), other channel 56mV (pot minimum) and pot max jumping values from 200mV and 2V.
Do you think the installation of ci caps is correct? I really don't understand...

[Peter Daniel]

Originally posted by canadino
Without source connected I had a channel DC offset from 4mV (pot minimum) to 46mV (pot maximum), the other channel from 44mV to 120mV.
With source connected (Linn mimik II) I had a channel DC offset from 13mV (pot minimum) to 57mV (pot maximum), the other channel from 80mV to 230mV.
After I install coupling caps (photo) and the DC offset go crazy: one channel measure 82mV (pot minimum) 87 (pot max), other channel 56mV (pot minimum) and pot max jumping values from 200mV and 2V.
Do you think the installation of ci caps is correct? I really don't understand...

It is perfectly fine for the first channel. I posted DC offset measurements for the batch of chips here: http://www.diyaudio.com/forums/showt...77#post1524877 The offset depends on input impedance and will vary from approx 0 to 70mV, that's for R2 (shunt resistance) being 15K maximum. When you install coupling cap, your pot is no longer in parallel with R2 and R2 being 22k will result in even higher offset so your 82mV reading for the first channel is fine and that's what I measured in the original Gaincard which also uses coupling cap.

But there is something wrong with your other channel, either the chip is defective or you mixed up connections or parts placement.

[martinbls]

Hi Peter!

This is (once again...) a question about the small Panasonic FC's used on the rectifier board. On page 1 of this thread you said: "...I supply 10uF Panasonics instead, and they can be used to sort of "improve" the sound of larger caps, as it is commonly practised by other manufacturers. I noticed that in most cases such additional caps only spoil things and it's best to test both setups and choose better sounding option." From the pictures on this page I was able to see that you don't use them yourself on the rectifier board.

I have 6 pieces of Black Gate N 4,7/50 lying around, brand new ones. As said before, I want to use two of them as input coupling cap instead of R1, so 4 pieces are left over. What about using these 4 BGs instead of the 10uF FC's on the rectifier board?

The question is: How does the sound change when using these BG's instead of using NO caps at all on the rectifier board?

martin

[Peter Daniel]

Originally posted by martinbls
The question is: How does the sound change when using these BG's instead of using NO caps at all on the rectifier board?

The question is, how hard is it to try and see for yourself?

For tests like that, I usually install some sort of sockets and plug the caps in an and out to estimate the difference. Here's an example of such approach when I was modding Bidat: http://www.diyaudio.com/forums/showt...425#post903425

Generally, you will notice slight sonic influence of BG N caps when added in parallel with main filter caps. I already forgot how they sound, but I once expressed my thoughts on a subject here:

http://www.diyaudio.com/forums/showt...431#post574431

http://www.diyaudio.com/forums/showt...230#post581230