Bijou All Tube Futterman Headphone Amplifier - Page 158

chobint

The screw terminals are great for swapping out components when you're experimenting. No soldering needed & much time saved. When the final config is reached (if ever), components can be soldered into place permanently. But I'd still use gold plated posts (standoffs) for connections to the boards to facilitate any maintenance that may be required.

I've been looking all over the place for these. Do you have a link to a vendor?

Thanks!

If I can recall, you are the 3rd Bijou builder (after me and chobint) who uses film caps at the output of your amp.
However sound improvement is a very subjective matter, your impressions is quite consistent with mine.



As for bypass cap C6 (parallel to main output cap C5), I just omitted it in my build since it seems to be redundant and doesn’t add any thing to the excellent sound quality of the 330µF MKP film output cap. I can imagine that the bypass cap C6 can bring some improvement when put it in parallel to electrolytic caps.

You know, gents, there is a way to eliminate the O/P and futterman feedback caps in the Bijou. But, it requires a split HV supply and is more dangerous to your headphones than having the protection of the O/P cap if anything goes wrong.

If any of you are high risk builders, I can show you how to do this.

You mean something like ±125V configuration?
However I prefer to play safe, I would love to know what you have in mind.

Ferrari, you might actually be able to do this. I have already tested in in a slightly different amp and the O/P DC is very stable if the PS is stable.



Will need +/-15V for the opamp. Or maybe even +/-5V.

The PS complexity is one of the main reasons I didn't do this for the original Bijou. We have gone from one B+ supply to 5 supplies including the opamp. It's just more build complexity.

But, the two caps associated with the output are now gone.

holland, I may not have explained myself very well in describing these. I'm referring to small gold plated steel or brass pins about 1 to 1.5 cm long with a shoulder half way along their length. You solder one end into the component or wire connection hole on the PC board & the external component or connecting wire to the projecting end. It eliminates having to remove the boards when changing components, but still requires a soldered connection.

I obtain them from a Jaycar, an electronics parts supplier near where I live. If you are in Australia, you should be able to source them from your local Jaycar, DS Electronics or Altronics. If they are what you require, PM me with your details & I'll send you a packet.

ferrari, your Bijou build looks stunning. I wish I had the patience & flair to create such an aesthetic piece.

Regarding the OP & bypass caps, I noticed the biggest difference (improvement) when the stock 470uF electro was bypassed with the Jupiter 1uF. The move to the Mundorf 200uF Tubecap was an improvement, for certain, but not to the same extent. I was very happy with the sound using the good bypass & electro OP cap, but I had already purchased the Mundorfs & was keen to try them. But I won't be going back, as the sound is now very much the way I like it - full & extended in frequency range, but delicate & 3D-like at the same time. It is, however, very critical of source material. My impression is that there's a synergy of the components in the build as it now exists. The challenge I have is to put the whole plot together in a more rational, organised & aesthetic build on a couple of new chassis.

I will disconnect the Jupiter bypass cap & report back with some impressions.

Nice Alex! That would definitely be interesting to hear vs the stock PS configuration.

Well I pulled the trigger on a slightly longer mini-lathe the Harbor-Freight 7" x 12". Eventhough I called the order in last night, it won't be processed until Monday. It looks similar to the other...........

my Mundorf output film caps are still sitting on my workbench, waiting for a Bijou chassis to be built. the amp circuitry will be P2P. no plans to go balanced at this point. i have to finish some projects for a couple of friends first: a preamp/CK2III with an off-board PS and LM7805 regulator to power an on-board gamma-1 DAC; a P2P Gainclone; and a High Octane tube guitar amp.

rolotube,
it's nice to see another Bijou build.

V-Div,
i find the lay-out and chassis build to be the most time consuming and challenging part of DIY. it's quite rewarding to have an aesthetically pleasing and good sounding piece of DIY sitting on the hi-fi rack, and well worth the sweat.

I've been looking at the schematic, and it looks like a very cool design. What is the most probable failure mode - the ECC99 going south or the opamp losing control?

Would an e12 monitoring the output provide enough safety?

Looks like a good p2p candidate....

An e12 will be necessary on the O/P to prevent damage from DC offset problems. But the e12, because it is electro-mechanical is not fast enough to handle certain kinds of failures. Like the ones you're asking about.

If the opamp goes while the amp is working the O/P will slew to one of the rails pretty quickly. And because the voltages are so high the e12 may not cut off in time.

Tube failure of the kind that would matter here is rare. A direct plate to cathode short or arc would put HV on the headphones almost immediately. The e12 won't even see this type of transient. Protecting against it is like protecting against an asteroid hitting the earth. It happens so rarely that it may not be worth the trouble but when it does happen dinosaurs die.

Nonetheless, I have built this O/P stage and I know that the opamp will zero the DC offset and keep it under 10mV if the split PS is regulated.

So the only way to prevent catastrophic failure will be to have a current limited PS so that even if there is a transient it's not large enough to damage headphones. This is actually possible with a shunt regulated PS.

This all can be done if someone really wants to get rid of those O/P caps.

And I thought I was crazy with two electrostat amps, and an idht dynamic amp...

Ron, I looked at that lathe at HF yesterday (they were having a big sale), as it was right next to the dovetail jig (my real target). It was probably the nicest machine in the store (ok, not saying a whole lot for HF), but I was a bit tempted myself. You will do well with that.

You are right, it’s indeed a trade-off between OCL and an increasing complexity of the PS.
The low symmetrical PS (±5V… ±15V) for the DC servo circuit should not be a problem, a simple PS based on 78xx/79xx should be sufficient here.
More critical is the symmetrical ±125V, this circuit should be very solide in order to guarantee a stable DC offset at the output.

For the moment I would love to let my Bijou as it is, since I listen to it every day and it sounds great (CAN'T miss it).
But I still have a couple of PCB’s in my drawer, and I will certainly try your proposed schematic out as soon as I have time.
If it turns out as what we are thinking, then we have a winner: a Bijou with the best output cap up to now (OCL & OTL).
If it doesn’t for whatever reason, no problem at all… we all know that’s the nature of R&D.

As I understand it, a fundamental principle in the the initial design of the Bijou was simplicity. The idea was to use just one PS, the trade-off being a requirement for an OP capacitor. Whilst it may not be the most elegant solution, the simplicity of this is what attracted me in the first place. I can appreciate the virtues of a symmetrical PS, but loosing the OP cap in favour of greater complication takes the design outside its original brief, IMO. That being said, it would still be a very interesting exercise. But as reliable as modern parts are, I'm not sure I'd want to risk my expensive HD650/K701's trying it. With my luck, Murphy's Law would be likely to intercede & if just one critical component went south, that would be the end of them.

There are plenty of HQ caps available, both electro & film, that will do an admirable job of coupling the output to the headphones. Part of the interest & fun of DIY is trying different combinations to see what works best for individual builds. IMO, this is a "no risk" solution that has the potential to be near as damn-it in terms of sound quality to a no-OP cap design, if the components are chosen carefully.

I'm sure its possible, but I find it very difficult to imagine how the sound of the Bijou could be greatly improved beyond application of the mods that have been suggested & use of top shelf parts. I think its evolved to be hard up against the vertical part of the curve of the law of diminishing returns, where an exponential amount of time, effort & money would need to invested to extract a very small amount of gain.

Just my thoughts.