New Posts  All Forums:Forum Nav:

DIT tube buffer? - Page 3

post #31 of 43
11/28/06 at 2:28am
Oh well, I just have to make one more post.

I designed this buffer a few years ago and haven't looked at it since. A fresh look is always a good thing to clear the cobwebs from your thinking. In this case, for some strange reason, I used a depletion fet for the plate bootstrap device.

Now a depletion fet is the simplest thing to use for the CCS but the bootstrap fet can easily be any one of the ubiquitous high voltage enhancement fets. Why I didn't see this before is one of those "how could I have been so dumb?" moments.

Anyway, here's a new schematic using an irf820 for the bootstrap fet. What can I say? Thanks to j4cbo for making me think about this agin.
Reply
post #32 of 43
11/28/06 at 8:23am
I saw that the DN2540N5 is a non-stocked part, with 50 piece minimum (though not crazy at $1.54 ea) at Mouser. So seeing the IRF820 in it's place makes this more appealing to me. I have some of those on hand already, and the DN3545N3 is readily available.
Reply
post #33 of 43
11/28/06 at 11:09am
The DN2540N5 is in stock in single quantities; it's the N3 that's not available right now, so you can still get it if the irf820 is less convenient.
Reply
post #34 of 43
11/28/06 at 11:22am
Quote:
Originally Posted by runeight View Post
The output impedance of the buffer at low freqs depends on the size of the output cap. With the 4u7 cap that is on the schematic the Zo will go from about 135R at 1kHz to about 350R at 100Hz to about 1000R at 10Hz.
Ok, I think I follow this. Excuse my ignorance, but is 4u7 the same as 4uF?
Quote:
Originally Posted by runeight View Post
This to say that if you follow the rule that the input impedance of the amp should be at least 10x the output impedance of the source, you're probably in ok shape with a 10k amp. A 25k Zi amp is much better. Anything lower than 10k would begin to stress the buffer design.
Could I go with a smaller output cap if I planned on only using the CD player with my amp? Say 2.2uF?
Quote:
Originally Posted by runeight View Post
Does this help?
Very much so, thanks.
Quote:
Originally Posted by runeight View Post
BTW, if you need gain in the output stage of the CD player and you want to use tubes, there are ways to do that. How much gain does your SS output stage make?
Not sure, but given the short cable runs and the amount of gain in my amplifier, I don't think it will be a problem.
Reply
post #35 of 43
11/28/06 at 11:36am
Quote:
Originally Posted by Alfiax View Post
The DN2540N5 is in stock in single quantities; it's the N3 that's not available right now, so you can still get it if the irf820 is less convenient.
Ahh... the "G" suffix threw off my search. N5 is non-stock, but the N5-G is stocked.
Reply
post #36 of 43
11/28/06 at 1:54pm
Philodox,

4u7 is the same as 4.7uF

Yes, you can make the cap smaller as the Zi of the amp gets higher. You'll just get a slightly faster roll off at low freqs if you do. You could start with 2u2 and then if you don't like it add another set in parallel to get 4u4.

Regarding the gain, it could be a problem, depending.

The SS output for the CD player might have to take a low signal DAC output and amplify it before sending it on. If you are planning to hook the tube buffers directly to the DAC, this might not work depending on what the signal level is there. Also, some DAC outputs are in the form of current rather than voltage output and the amplifying stage needs to do an I-V transformation. Not hard, but one needs to be aware of it if this is so.

So, is there any way you can determine what the signal level is where you are planning to connect the buffers?

Gents, thanks for the work on the dn2540. Even though it is available in qty 1, there is really no reason to use it. The buffer can use almost any >=300V enhancement fet and this simply makes parts sourcing easier. In fact, many folks may have these lying around in their parts bins.

I'm posting another schematic for completeness just to show the power switch, fuse connections.
Reply
post #37 of 43
11/28/06 at 3:28pm
Just wanted to clarify that the DN3545 on the cathode remains the TO-92 and not the SOT package..dB
Reply
post #38 of 43
11/28/06 at 3:34pm
Correct, the dn3545 in the CCS only dissipates about 100mW. TO92 should be fine for this. a small finned heatsink wouldn't hurt, but probably isn't necessary.
Reply
post #39 of 43
11/28/06 at 6:51pm
BTW gentlemen, the Hammond 262B6 transformer will cover both the 120V and the 6.3V secondaries for this amp. So you only need one transformer.

See: http://www.hammondmfg.com/261.htm
Reply
post #40 of 43
11/29/06 at 2:38am
The 262B6 looks like a handy solution... Unfortunately none of the typical hammond distributors seem to have it in stock. The 262E6 seems to be in stock at mouser but is way overspecced for the task; something probably would have to be done to keep the filament voltage down...
Reply
post #41 of 43
11/29/06 at 2:53am
Yes, I see what you mean. The usual suspects don't carry inventory, although they would probably special order it upon request.

The 262e6 would be too big really, but it could be used if, as you say, the filament voltage is taken care of.

An alternative is to just use two small transformers. Triad makes several that would work here. I was just trying to keep things as simple as possible.

Of course, this really isn't useful until someone decides to build one of these buggers.
Reply
post #42 of 43
12/1/06 at 11:22am
Just to let you all know that Newark doesn't carry stock of the 262b6 but if ordered , it ships directly from the factory with a 2 week lag time. They stock the IRF820 too but not the DN3545 ( but mouser is back ordered on the depletion mosfets so it seems that more than one supplier is going to have to be used. ) I have put a BOM together if anyone is interested..dB
Reply
post #43 of 43
2/9/11 at 9:24am

I have the fortune to obtain the tube buffer built by dBel84 (runeight's design). I am hearing an improvement in my system: Gamma2 Dac --> tube buffer --> M3 amp --> DT770 / 32 ohm. The tube buffer extend the soundstage, fills up the mids and make the highs smoother; without the tube buffer the combo was a bit aggressive with dance / eletronica type of music. I am not a DIY guy, but I think this tube buffer sounds really good and specs well at the same time. Thanks for the great job,dBel84 and runeight! This tube buffer is so good that I am thinking to get another one......but I do not know who else can build one.

Reply
New Posts  All Forums:Forum Nav:
  Return Home