Intrest Check: 120w @ 8ohm Tube Buffered LM3875 Amplifier

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have came up with this pcb design for a tube buffered LM3875 amplifer, it puts out just over 120w into a pair of 8 ohm speakers (stereo).

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


Click here for full size image (high res)

The PCB is not currently finished, but i am wondering if anyone is intrested? The pcb itself should not run more then $15 each, i am hoping.

Basicly, it is a standard gainclone fed by a tual triode tube, such as the ECC88. 3 Seperate power supplys are used.

1. Main Gainclone Powersupply (400va transformer)
2. Tube power supply (40va transformer)
3 Tube heater and digital power supply

What you can not yet see on the board, is a 555 timer delay circuit, that will connect to a external relay to give the tube time to heat up before the 2 main power supplys are switched on.

Some additional tydying and more sensible routing needs to be done on the board, but i would be delighted if anyone is intrested in some sort of group buy.

Please note - as the pcb is not yet finished there will be some random pads / incorrect routing that i have not yet took care of.

Also - Im aware of the 6v running thru the AC.

This board is about 2/3 of a A4 page wide.

Thanks
 
Ok so you mean that you are going to have 25v or less on the secondary of the transformer right? I was going to say you are going to burn up the chips....

Can you do a schematic? From what I can tell you might have (a few) extra things that you may not need. If you are driving it with tubes simplicity is going to give the richest sound, I am pretty sure.

Also the spaces on what I think, but am not sure, are capacitors that carry signal might be a little close together for a modification I have recently tried and found amazing results with. All it requires is that you have capacitors for the output (or any carrying information, if you have more, or REALLY want to make some sweet sounding stuff with the effort involved) have isolated cans (the metal around the electrolytic element inside).

If you are not star grounding everything via p2p, which from what I can tell you are not, a modification to the board would be to have a ground PCB hole next to each capacitor to run a solid piece of copper around the capacitor (jacket removed) and then soldered into ground.

If you do not believe me that this can make a difference like between Teflon V caps and some harsh no name electrolytics, try it in something and be amazed.

There is no real reason to do this to non-electrolytics because there is no induction that is measurable to the plastic or paper etc.

The voltage does not seem to matter as long as you have solid wire pressing or some thing that holds the grounding wire on. The amperage is next to nothing. Also for the sake of electromagnetic field of induction it is important to do a loop around the capacitor. The same wire can be used for both output caps for the one side.

Oh and unless you are going to have two transformers for the supply to the LM3875s I doubt having two sets of capacitors and diodes is going to do any good. You could save money and cut to one, or put in some bypass capability so people would have the option to use two transformers and make it dual mono.
 
Destroyer OS. said:
Ok so you mean that you are going to have 25v or less on the secondary of the transformer right? I was going to say you are going to burn up the chips....

Can you do a schematic? From what I can tell you might have (a few) extra things that you may not need. If you are driving it with tubes simplicity is going to give the richest sound, I am pretty sure.

Also the spaces on what I think, but am not sure, are capacitors that carry signal might be a little close together for a modification I have recently tried and found amazing results with. All it requires is that you have capacitors for the output (or any carrying information, if you have more, or REALLY want to make some sweet sounding stuff with the effort involved) have isolated cans (the metal around the electrolytic element inside).

If you are not star grounding everything via p2p, which from what I can tell you are not, a modification to the board would be to have a ground PCB hole next to each capacitor to run a solid piece of copper around the capacitor (jacket removed) and then soldered into ground.

If you do not believe me that this can make a difference like between Teflon V caps and some harsh no name electrolytics, try it in something and be amazed.

There is no real reason to do this to non-electrolytics because there is no induction that is measurable to the plastic or paper etc.

The voltage does not seem to matter as long as you have solid wire pressing or some thing that holds the grounding wire on. The amperage is next to nothing. Also for the sake of electromagnetic field of induction it is important to do a loop around the capacitor. The same wire can be used for both output caps for the one side.

Oh and unless you are going to have two transformers for the supply to the LM3875s I doubt having two sets of capacitors and diodes is going to do any good. You could save money and cut to one, or put in some bypass capability so people would have the option to use two transformers and make it dual mono.

I think a schematic would clear a lot up for you. I have two sets of capacitors and diodes, becuase the set on the left is to power the high voltage tube supply. The power supply in the middle is just a standard split rail power supply, i do not understand the problem here. The power supply for the chips, after rectification will be at about 35v if a 30v transformer is used. If you are talking about the 2 sets of capacitors and diodes in the centre power supply, well one bank of capacitors are for + and the other for - voltage.

No capacitors carry signal, except the large coupling film caps you can see on the board.

Find the schematic below, sorry for the mess. The square box is the tube, i didnt fancy making a nice symbol for it. Tube datasheet is here.

http://lib.store.yahoo.net/lib/thetubestore/JJ-E88CC.pdf

The amplifier is infact very simple as you can see.

Click here for schematic


May aswell add the latest board, still unfinished.

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

Click here for full size

I hope this clears things up for you.
 
Is it possible to do for LM3886?

I ask this because Tube + LM3875 sound and power output is approached by some LM1875 designs, and with very little effort.

I think that LM3886, which actually does need a buffer, is also appealing to more people because it can shove today's inefficient speakers appropriately. ;) The LM3886 can also withstand 4 ohm loads (typical of MTM speakers) with grace.

AND you can interpose a delay with LM3886 inbuilt mute circuit--so you can omit the relay. ;)
 
danielwritesbac said:
Is it possible to do for LM3886?

I ask this because Tube + LM3875 sound and power output is approached by some LM1875 designs, and with very little effort.

I think that LM3886, which actually does need a buffer, is also appealing to more people because it can shove today's inefficient speakers appropriately. ;) The LM3886 can also withstand 4 ohm loads (typical of MTM speakers) with grace.

AND you can interpose a delay with LM3886 inbuilt mute circuit--so you can omit the relay. ;)


Hmm, it is possible to redisign for the 3386, but im wondering if it is actually worth doing. I allready have a couple of 3386 monoblocks, so i did actually want to try the 3875 for a change. I will have a look at a few 3386 schematics and give it a think.
 
The 3875 can handle 4ohm just fine so long as you keep it within the power parameters, read the PDF on it.

R103 if I am not mistaken is not necessary at all since you have capacitors inbetween the tube and the 3875, they should, if they are the right size, block everything, if that is why you have the resistors in there. Plus I think the input schematic wise would run to the +

Forgive me for not being an electrical engineer but I can tell you for the raw tonality that you want with having tubes invovled simple is best with the 3875, just use the right capacitors and resistors.

After going back and looking over what I have done with my amp I realized that my modification with the capacitors worked for the incoming to chip power capacitors. It might just be a lot of filtration at that point. I sure wonder what it could do to output capacitors etc.

Here is how simple signal needs to be, minus R1 so long as the source is safe,

http://www.diyaudio.com/forums/attachment.php?s=&postid=1508778&stamp=1210693738
 
Destroyer OS. said:
The 3875 can handle 4ohm just fine so long as you keep it within the power parameters, read the PDF on it.

R103 if I am not mistaken is not necessary at all since you have capacitors inbetween the tube and the 3875, they should, if they are the right size, block everything, if that is why you have the resistors in there. Plus I think the input schematic wise would run to the +

Forgive me for not being an electrical engineer but I can tell you for the raw tonality that you want with having tubes invovled simple is best with the 3875, just use the right capacitors and resistors.

After going back and looking over what I have done with my amp I realized that my modification with the capacitors worked for the incoming to chip power capacitors. It might just be a lot of filtration at that point. I sure wonder what it could do to output capacitors etc.

Here is how simple signal needs to be, minus R1 so long as the source is safe,

http://www.diyaudio.com/forums/attachment.php?s=&postid=1508778&stamp=1210693738


Thankyou for the comments, i have removed a couple of resistors from the schematic and used the values in your drawing. I have emmited R1 from the equation, as it is directly after a coupling cap like you said.

The 'final' board layout is below, pending minor changes.

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


Click here for full size image

Star grounding is now being used throughout the board, to try and keep ground loops to a minimum. I am waiting for the quote to come back from my boardhouse, but anyone who is intrested in this final design please let me know.

Please note, part labeling is not complete as you can see, dont worry i will clear up the mess :)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.