June 7th. "Nice testimonial from Paul. It's interesting to me that in ~70-80% of cases, my humble icOn 4Pro is used in real high-end systems where it easily competes with 3-5 even 10 x costlier passive or active preamps. It's time to make a high-end version because a lot of new clients want something fancy and heavy inside a beautiful enclosure. The R&D of the Gradient Box will hopefully go quicker but I now see clearly that it will be more complex and sophisticated than the 4Pro's. The below is only a small part of it. The full box will need sixteen of these discrete digital potentiometers.
"Due to your preview pages I get a lot of questions and suggestions, even pre-orders with wildly different requirements just as expected. It's not easy to decide what to incorporate and what to reject when my digitally controlled analogue circuits with the best possible specifications are so wildly more complex than the 50-cent carbon pots or few dozen extra lines of DSP code which are the standard solutions in similar products. There will be so much active circuitry that I'm also working on a new standalone power supply for the Gradient Box, optionally for icOn and even for future icOn tube electronics with 200V. I'll never discuss or show what's inside because I just know how each 'true' audiophile could make or buy a far better-sounding PSU in a heartbeat.
"Another nuisance I just realized today is that my favorite opamps have disappeared worldwide. The lead time now is 20-50 weeks from Mouser, Farnell, RS or elsewhere. This chip shortage seems very serious. My solution? I redesign my boards for the micro-miniature QFN package versions which are rarely used by lazy audio manufacturers but have a better supply. Farnell just sent me an email that they will ship 250 x analogue switch chips which I ordered on March 15th so 68 days ago. Here is where the project currently stands:
♦ The crossover frequencies are finalized at 26¹, 40, 53, 60, 80, 92, 106, 120, 132, 145, 160, 171, 184 and 198Hz.
♦ The input, LP and HP outputs can be unbalanced or balanced but only 3 socket pairs altogether.
♦ Mono/stereo sub mode, ±10dB gain adjustment, continuous phase adjustment (a lot of customers asked for this phase feature)
♦ Remote-controlled LF EQ with 8-10 different settings and defeat.
"If someone wants to use their sub's own low pass and other features, I'll have a Gradient Box Lite version with only the HP filter with remote control." Clearly Pál had bitten off a chunk with this project. It reiterates why nobody has attempted anything like it yet. Hopefully there'll be enough ambitious listeners recognizing its virtues to justify the extended R&D baked into it.
"The user interface, display type and menu system are still big questions. After proper configuration of all the settings, this will work as a black box until the next subwoofer or loudspeaker upgrade. So the configuration process can be very different from the icOn's normal usage. I don't think that somebody will want to experiment each day with different crossover frequencies, phase adjustments and EQ. But who knows? The remote-controlled A/B testing of different settings will be unique and provide a new way to find any optimum configuration comfortably and quickly. This is very hard or nearly impossible to do with purely manual controls. If I can find a smart workflow for the setup process that won't need a lot of buttons, I'll be happy to stay with the Apple remote."
_____________________________________
¹ In subwoofer discussions, many relate to a low-pass frequency as though it were a brick-wall filter. In reality, the transitions are more gradual. Pál's 26Hz option for example could interest owners of -3dB/40Hz speakers who want to extend their reach whilst hearing them play as low as they natively do. This could be a 3-way speaker with top-quality dedicated woofer that doesn't warrant restricting its bandwidth. The same low filter setting could see 4-way speakers good to -3dB/25Hz if their owner means to bolt on a true infra sub with response to 15Hz or lower. Now a low LP frequency is king to work such a sub across a narrow bandwidth only.
July 19th. With two Grimm SB1 subwoofers inbound for review, I checked with Pál for an update. "I'm quite devastated because nothing works as before. Now everything is slow, painful and full of mistakes. I wasn't able to realize my plans for this year mainly due to new Brexit rules, the break in trade and cooperation with the EU, the chip shortage and the pandemic. I was also very mistaken when guessing the development time of the Gradient Box. First I thought that I'd replace the attenuator board with a more sophisticated crossover board in the icOn but use everything else. Currently every part of the Gradient Box is different. Dozens of interested audiophile gave me sometimes useful advice on what functions and features they are waiting for and I've tried to find a compromise. I realized that for this much functionality, the menu will need a new OLED display because the small TFT I used before is not visible enough from a listening chair. Due to the new display, I need a bigger processor with more memory. The dozens of op amps and analogue switches need a more powerful low-noise power supply. The XLR + RCA sockets need a bigger enclosure. Nothing shares with the icOn. Because standard SOIC-packaged op amps have completely disappeared with one-year lead times, I bought rarely used QFN-packaged op amps and designed the main crossover board for that. That unfortunately was a bad decision.
"It's nearly impossible to solder these parts manually. To order a fully assembled prototype board will be very expensive, slow and difficult. From another galaxy I had to chase down a few dozen SOIC quad op amps and redesign the main board again which will arrive from China on Tuesday. Even after making hundreds of PCB designs in my life, the chance to make a perfect board with this much complexity in just one attempt isn't very high. So to make a long story short, a first functional sample won't even be ready in August. Learning from previous mistakes and given the currently unpredictable supply-chain conditions, the only accountable ETA for a fully functional Gradient Box is… hopefully later this year. This is the current reality and consequence of not conducting a proper feasibility study before I started this R&D cycle. But there's always a significant risk whenever one does make such a study. One may simply never start the project due to the huge amount of time and work involved. In the meantime I'm also busy with modifications, firmware code, enclosure design plus producing, selling and supporting icOns and designing a new high-end icOn which has a growing waiting list. At least I won't be bored."
Pál just might feel a bit better if he knew that the legendary John Curl designed the Wilson Audio Active XO yet didn't bother with remote control at all. No wonder that biting off that big chunk in the midst of all the other troubles gave the GB project of our man in Manchester indigestion. And yes, GB is shorthand also for Great Britain.
July 26th. "My weekend entertainment was assembling the GB's motherboard. Before I can try and measure it, I must write the firmware's main modules." Note the silkscreen on the board: "Gradient Box – smart active filter". That's our cue to wrap this genesis story and leave the rest for the formal review.
