Bijou All Tube Futterman Headphone Amplifier - Page 4

her shann the 370DAX has the same secondary windings as the 270DAX except for the CT on the heaters and a Bias Tap. If you will only be using your transformer for the Bijou you can cut the heater CT (not the HV CT!!) and the Bias Tap and wire the 370 identically to the 270. If you don't want to cut, just tape them off with good insulating tape or heatshrink. Follow the Hammond wiring color scheme and you should have no trouble.

It's pretty easy to put a DPDT switch on the primary of the transformer so you don't have to actually rewire the circuit each time. However, you have to be really careful about flipping the switch to the wrong setting.

Chokes - chokes are great in tube PSs. I've used many of them myself. But be careful to think through the DC resistance of the choke. The two resistors R7 and R8 are calculated to drop a certain range of voltage so the the input to the to the regulator is high enough to keep the regulator regulating at 250V. The Hammond 158L has an R of 411 ohms. Depending on exactly how you use the choke this has to be thought through.

Actually they make special switches for voltage selection (do a search on Digikey for voltage select and choose switches when it comes back). These usually are marked 110/220 or whatever and require a screwdriver to switch them. Sample Digikey PN CKC3002-ND for example.

I'm using the recommended 270 DAX.

That choke has a DCR of 411R. It is replacing a resistor of 360R. The maximum current of both the rectifier and the transformer is 90mA, which means that the choke would drop, at worst, an additional ~5V. Differences in wall voltage will account for much bigger swings.

Now, the bigger issue is that the current rating of the choke is only 75mA, and while Hammond rates their chokes pretty conservatively, that might not be enough. If that's the case, the 158M has a current rating of 100mA with a H of 10H and a DCR of about 200R, so it might be a better choice.

And if it does run at close to 90mA, you might try an EZ81 instead of the EZ80.

Running a few comparison sims shows this:

1. With the resistors only:

279V input to regulator
140uV of 120Hz ripple

2. With a single choke and R8, R7, C5 removed:

292V input to regulator
500uV of 120Hz ripple

The regulator has enough rejection to deal with the additional ripple in the case of a stand alone choke.

The original PS heatsinking would not have been able to handle the extra power dissipation on the pass device (running at 42V and 55mA with the choke). But the new production boards have bigger heatsinks to increase flexibility for diy'ing the PS just for things like this.

3. With R8 replaced with the choke and R7 reduced to 240R and with C4 and C5 reduced to 47u you'll about the same result as case #2 above.

The maximum draw comes from using 6N6Pi tubes. It should be around 62-65mA. This is close to the max on this choke, and it will warm up some, but my guess is that it will work out alright.

I'm not sure, however, that you will hear much difference in the audio with the choke hiding behind a sand regulator. I guess one of you guys will try it and let us know.

Hmmm..... to choke or not to choke. ;-)

LOL. Couldn't have said it better.

But, this is an interesting idea. I decided to add a section to the Cavalli Audio website describing one way to do the choke in the PS. It's in the tweaks section. I may add some other ways to do it also, to give some choices.

Although the original design expressly avoided big and heavy chokes, thanks for dsavitsk for suggesting the idea.

Just wondering about PS caps / ripples and such:

1) If PS caps are there to reduce the ripples - I suppose the bigger the better (more expensive)? Is there concern that certain caps are slow to react and unable to react to transients?

2) As far as I can understand - chokes are there to reduce ripples but they drop the voltage - this allows smaller / better caps?

3) Films caps are better than electrolytics as output / coupling caps - is that the same in the power supplies -> since no audio signal passes through them.

4) Terminologies - what's coupling caps / what's decoupling caps.

5) I see that some designs that has line suppression caps (small 0.0033uF example) across the AC terminals - is that to suppress EMF / filter out HF spikes from switching supplies noise. Any advantage or disadvantage to include in these amp designs?

thanks

her shann

1) Electrolytics have an internal inductance and resistance (actually all capacitors do, but electrolytics have more). The higher the frequency the more the inductance influences the behavior, reducing its behavior as a capacitor. The inductances are small and these frequencies are generally well above the audio band. But, as you note, it's the transients that matter. When a fast transient response is demanded from the amp this, in turn, demands a fast transient response from the PS. Large electrolytics will not respond as well as, say, good film caps. This is why many PS designers parallel film caps with the electrolytics or try to eliminate the electrolytics altogether.

2) Chokes are very effective at filtering out the AC (including a lot of the lineside noise). The more effective the choke is, the smaller the caps can be.

3) Film caps are always better than electrolytics, but they can't be made with large enough capacitance to be practical where large capacitances are needed, such as in power supplies. But, as noted in 1, the PS sees transients too and having film caps in the PS can be a benefit even if no "signal" passes through the PS.

4) Coupling caps couple the signal from one place to another, usually between one audio stage and the next. There are two coupling caps in the Bijou between the phase splitter and the output triodes. Decoupling caps do just the opposite. They placed in the circuit to prevent the signal from getting from one place to another. Most of the time decoupling caps are located on the PS rails so that when one stage causes the PS voltage to fluctuate the other stages are screened from these fluctuations. This keeps the stages from interacting with each other through the PS.

5) Yes, those caps are to remove AC line noise. There is no disadvantage that I am aware of to including them in the build.

Anyone got their boards yet? Can't wait to get started. I will be using 370DAX/120VAC. Ordered 6n6p from Ukraine, Russian 71-74 military stock. I have Russian 6DJ8 in hand, marked "made in U.S.S.R." National brand. Might use high-end Japanese 9 pin sockets, $8.95 each(angela.com). What is the volume pot brand to use in this project?

Alps RK27 for the inputs and even for the NFB if you really wanted. Panasonic EVJ series is also good.

I just payed for my kit last night. I haven't picked up tubes or a traffo yet, though.

Anyone heard of RK27 50k being used for the volume pot? Runeight?

There are only a few Bijous around at this moment, until more of you get your boards/kits. I'm not aware that any of them are using this pot.

However, and good quality 50k or 100k will be suitable for the volume pot and should give you no problems.

Does anyone have a source for the 370DAX?



The website states:
The Bijou is equipped with a variable NFB circuit. This helps the amplifier handle the range of headphone impedance from 32R to 300R.

How will it go with 25 Ohm headphone such as the Denon D2000? Will I need to make any modifications?