This I have been experimenting - call it a hunch - on the effects of bypassing electrolytic capcitors (Nichicon FW and KW) with 0.1 uF TDK ceramics (Mouser 810-FK28X7S2A104K) with the diamond buffer circuit used in both my B-board preamp and Sapphire headphone amplifier.

This being a mod, I had to solder the caps to the underside of the boards, attached to the leads of the Nichicon 100uF electrolyic capacitors.

I used four ceramics per channel, one per active device in the diamond buffer if you like.

I did several other changes on the B-board at the same time, so it wasn't obvious until I modded the Sapphire in the same way what was the result of the bypassing. Anyway, with both the improvement was immediate and dramatic: any sense of "transistor-like" treble glare is completely quenched. The whole top end takes a step backwards, not in extension, but in prominence.

Less audiophile detail, more swinging mojo.

Update...

B-board vs. 0247.

Comparison of the noise baselines, measured at the circuit output using a NI USB-6215 DAQ. Unloaded for the preamps, and with a 6 ohm load for the 0347 amplifier.

• B-board: -139 [300Hz-100kHz] 0 dB gain (-139 - 0 = -139 dB, 110 nV sqrtHz input referred). The actual B-board output noise is below this measurement threshold.
• 0247: -124 dB [300Hz-100kHz] 14 dB gain (-124 - 14 = -139 dB, 110 nV sqrtHz input referred).
• 0347: -109 dB [300Hz-100kHz] 31 dB gain (-109 - 31 = -140 dB, 100 nV sqrtHz input referred).

See the attached plot for the FFT data. Note the peak at 28 Hz is an artifact of the measurement apparatus.

By way of comparison, a typical audio opamp has an input referred voltage noise figure of 3-8 nV sqrtHz (-170 ~ -160 dB) and can be expected to return this datasheet specification in most well-designed circuits. In other words the output noise is going to be about -160 dB + the circuit gain.

The 0247...

Nor the remake, the original. Good film. Story shakes out something like "Twelve Angry Men in the desert": Put together a small, random group of people and pressure them to complete a task. In the case of "The Flight of the Pheonix" this is to make an airworthy plane (this one) from the crashed remains of another (this one).

So. We start with my old red Gainclone case, and a pair of these buffer boards , and a Takman resistor, 24 position stepped attenuator from eBay, unassembled, and start working to transform something old into something new.

Here's my LM3875 gainclone. Served me well, but it is time to bid adieu! (at least to the guts):


Opened up, we see the circuit board, such that it is, and my home-built 11 position attenuator:
...

The original stereo VSPS project is now quite mature. I did a couple of minor tweaks on the last re-spin of the layout, removing the pads for the old Black Gate coupling caps (long gone), adding thermal isolation, and making room for a pair of optional ceramic bypass caps next to the op amp.

The new boards are matt black. Quite cool. I have a small stack if you need any.

All the info you need to build one is attached.

Once in a while I get emails asking after the X-reg evaluation boards. These are handy little 5x8 cm test boards for the X-reg voltage regulator - or they would be, if I had ever bothered to get a set made.

Since the circuit is built into the VSPS300 and Phonoclone 3, it wasn't really a big priority.

Anyhow, yesterday I re-drew the evaluation boards and I'll be getting a batch made in the near future, along with some of the stereo VSPS that people were also asking after recently. They won't exactly fly off the shelves, but I started to feel bad for those few people who actually wanted them.

I did get a couple of responses to my invitation. Well, two, actually.

So congratulations! Free boards will be heading your way in about a month, and I'll throw in most if not all the parts, too.

First was simonov's entry. He's clearly done this several times before. His confident layout ticks all the right buttons: ground plane (check), thermal isolation (check), clean, geometric layout (check). One jumper required, but that's a very minor offence.



While I allowed modifications of the circuit, simonov went and redesigned pretty much the whole thing. CCS replaces the source resistor, current limiter and capacitance multiplier blocks added. The BOM was starting to spiral, and no longer uses my standard parts set so I'd have difficulty supplying kits based on his design. His modifications, although certainly improvements, ended up counting against him....

I tell people: "Buy a nice, heavy power transformer. It will sound better."

They are skeptical, because the circuit only draws a couple of watts, and less than 100 mA current.

The image below shows how the power transformer, and rectifier diodes, actually work much harder than you would estimate from looking at the output power.

It shows a zener regulated supply with a load drawing 100 mA at 20 V. That's 2 W.

As a result of the capacitor input filter directly after the diodes, however, the diodes and transformer do not conduct current all the time, but instead for just a couple of milliseconds twice every cycle of the AC wave. They have to supply all the output current in just that short space of time. As you can see in the simulation, the diodes are pushing peak currents well in excess of 1A or 10x the output current. This is a typical "normal" power supply with a initial ripple ratio of a modest 1/40, things...

I just remembered I hadn't announced it here on the blog. Time to change that.

I'm holding a little contest, to encourage people to get into circuit board design, and to showcase the talent on tap here at diyaudio.

The idea is pretty simple. I hand out a circuit schematic, you design a circuit board based on the schematic. It's all done in Eagle CAD software, which is free to download and available on Mac/Windows/Linux platforms.

The circuit is a headphone amplifier of my design, and fairly simple. It's easy enough to come up with a workable layout, the trick is to come up with a really good one, that is small, convenient, clear, logical, good looking and that follows sound electrical design practice.

Winner will get a pair of boards, complete with all components, built by me. Runners up get sent pairs of the winning board design.

Deadline is the end of the month, so get on it!

All the info is here....

The B-board line buffer is back from the fab and running on the bench.

Cute little board. Diamond buffer with BD-135/136 transistors and a built in Z-reg. Designed to buffer the output of op amp circuits to help drive cables and otherwise isolate the op amp feedback loop from the big bad outside world. It can also be used as a unity gain preamplifier, or, with by changing a couple of resistors and adding small heatsinks, to drive headphones.

Development thread here.

UPDATE: Also finished a second set of boards, configured as a headphone driver (5x bias current, heatsinks).

This is a followup of sorts to the X-reg, though there is nothing original about the circuit this time around.

It's just a Zener voltage regulator with a series pass transistor. I lifted this particular configuration from the Pionner C-21 preamplifier and re-tuned it for op amp applications. My main interest here is trying to make a small and convenient board layout.

I've used this circuit block already in the Sapphire amp and come away impressed.

The output is about 1 V less than the Zener voltage, and the input voltage should be about 3-6 V above the Zener reference voltage. I'm working here with 17 V unregulated supply and 12 V Zeners, but the values can be reconfigured easily enough for any output from 4-24 V

This is a low current circuit. If you are just powering a few op amps, no heatsinks are needed. Above 25 mA small heatsinks are a good idea. The circuit is not designed for output currents above 100 mA.