Omni Box - One Box to Rule Them All

[vigilante397]

Started this in the SMD thread and due to popular request (literally one request) am giving it its own thread.

I had been talking to a guitarist friend a while ago and we talked about how big of a pain it is to bring a huge amp to gigs, and how hard it can be to get a good recorded sound with an amp with microphones, and also how inconvenient it is to get the urge to jam at 2AM when wife and kids are sleeping. So he jokingly told me that as an engineer I should solve all of those problems with a single box that should fit in a 1590B. A few months later, I'm working on a prototype.

The last time I had a really cool idea I got super excited and rushed into ordering boards with a semi-verified design, so now I have about 20 useless PCBs lying around (I use them as nerdy bookmarks). So this time I broke the design up into modules that I'm building and testing individually before I throw everything onto one board and send it to the fab house.

The design for this box has the following "modules":

- Alembic F2B preamp running a 6N17B-VR tube
- Speaker Emulator; I've auditioned a few but for now I've settled on Lart's Simple CabSim v2
- DI transformer for balanced XLR out
- Headphone amp; started with an LM386 design I wasn't thrilled with, scrapped it and auditioning some different ones
- Nixie-style SMPS for 200V running on a 9V wall-wart

The preamp module is complete and I'm thrilled with the sound. The more I play with it the less I think I need the master volume control, so I will probably be getting rid of that. The preamp is pretty clean, so even with the gain cranked it doesn't get very dirty, so what I had labelled as gain will work just fine as a volume control. I used RG's schematic for the preamp:



And this is what the stuffed and working preamp board looks like:





And a quick video:

I have the SMPS board etched and 99% stuffed, but it looks like my bag of SMD 555 timers has disappeared, so replacements have been ordered and will be here tomorrow. I also etched out the speaker emulator board last night (I was able to get the layout down to 1" x 1" with 0805 components ) and the parts for that should be here early next week. I finished the initial headphone amp last night, but I wasn't thrilled with the results. It wasn't hard to make it sound good, but it was really easy to make it sound bad. So I have another design that I will be trying probably next week.

Once I have all the modules verified individually I'll link them all together human centipede style and test them as a unit, and assuming there are no problems there I'll smash all the designs together into one board and send it out for fabrication. If anyone is really interested I can sell a few boards as I'll be ordering at least 10 and I very seriously doubt I'll have people lining up to buy finished units from me.

Anyway, sorry for the long read, and I expect to keep updating this until the project is done.

[bluebunny]

Started this in the SMD thread and due to popular request (literally one request) am giving it its own thread.

Yay!!      (Guilty as charged...)

Anyway, sorry for the long read, and I expect to keep updating this until the project is done.

Popcorn on standby.   

[deadastronaut]

yep, got my popcorn too....

[vigilante397]

Started this in the SMD thread and due to popular request (literally one request) am giving it its own thread.

Yay!!      (Guilty as charged...)

Did you notice how I said you were popular?

[Marcos - Munky]

This board is just beautiful! Question: I heard subminis have a short lifespan, is that correct?

[vigilante397]

I personally haven't heard that, but I imagine it depends on the tube and how you treat them. Many are rated for 150V on the plates but we tend to run them at 200V or so, they probably don't love that. But that being said a lot (most?) of the subminis on the market are surplus military tubes, and I know military spec tubes tend to last a while as the manufacturing tolerances are tighter, the glass is thicker, and they just seem to do well. All that being said though, while they aren't generally as easy to replace (though some people do work up a socketing solution) it wouldn't be the end of the world to replace one even if it was soldered in.

[anotherjim]

Good choices.

BTW, If made with 12AX7, that F2-B scheme has pins 6&8 swapped!

[Marcos - Munky]

Thanks for your answer. I asked more because they're way hard to find here in Brazil. I happen to have a spare one and plan on building the B2F some day, but I'll just hate if the tube dies and I can't find an replacement.

[vigilante397]

BTW, If made with 12AX7, that F2-B scheme has pins 6&8 swapped!

Hah! Didn't even notice, thanks for pointing that out. Fortunately I went off the parts (plate/grid/cathode) instead of the numbers, but that could definitely have someone new to tubes wondering what they did wrong. I'll post my drawings later on, probably when I get everything finalized.

[anotherjim]

I made it with the swap. It didn't kill the tube, but neither did it function   

I treat the experience as a swipe upside the head from RG for committing the sin of assumption

Funny how, since Alembic stuck a schematic inside the lids, I cannot find a definitive rendition anywhere on the webs. Many other versions increase the mids pot value, although I'm fairly sure that the original was 10k.

[slashandburn]

Brave man trying to fit all that in 1590B.  I'm watching with intruige. 

How come you ditched the 386 as a headphone amp module? Lacking for clean's? Otherwise I've found it it to be pretty usuable and fun. On top of the low parts count it could probably also double as a Purple Plexi/Sonic Titan style distortion module. 

[anotherjim]

It depends on the type of phones. Average consumer models are 30ohm and easily driven by low voltage circuits. Proper studio headphones have tougher requirements and higher impedances. 80ohm? 150ohm?

[vigilante397]

How come you ditched the 386 as a headphone amp module? Lacking for clean's?

Exactly that. The overdriven 386 was unpleasant to my ears, and it was inconvenient to have to turn the headphone amp down when the preamp was up and turn it up when the preamp is down. I'm hoping to find a setup (possibly unrealistic) where I can find a set-and-forget level on the headphone amp without having undesirable clipping when I adjust the preamp level.

I don't have a specific goal for impedance to drive, I just have a few different sets of headphones, and my very scientific testing involves trying different headphones and measuring their "sounding good" level.

[vigilante397]

UPDATE (now with schematics!)

I was out of town for the weekend but I came back to a mailbox full of components, so I finished soldering up the PSU last night and I can now add that to the list of working parts. I'll probably pull a couple bulk caps off the output for the final as it's still quiet with just one, but I figured I had room so why not?







I would post the schematic for the whole design, but there's really no point in doing that until I've confirmed everything works. So here's my drawing of the schematic. I'm planning on pulling off the master volume as I don't find it useful and just throwing in a pull-down resistor there like the original schematic.



And here's the PSU. You can see it's just the standard nixie-style PSU with no changes, plus a 7806 regulator with a diode between middle pin and ground to get it up to 6.3V. Unfortunately I forgot to look into SMD 7806's and 1N5817's so those were through-hole. The advantage there of course is that I can put components on both sides of a single-sided PCB as long as some are TH and some are SMD



I have boards milled and ready to solder for the speaker sim and the new headphone amp I'm trying, hopefully I'll get time this evening after work to get them soldered up and tested. Due to the size of the DI transformer I'm thinking of making a daughter board so I can mount it sideways and attach it directly to the chassis. The transformers I'm trying out are 10k:600 ohm from GGG. I have a couple on hand already and I plan to try them out as soon as I get the speaker sim working, at which point I can get some decent recordings for what the DI output will sound like. Feel free to refill your popcorn as necessary

[Ben N]

That is bloody brilliant! And inspiring. I wonder if I can do this in thru-hole and without cab sim and the power amp, with either the nixie or a backwards wallwart tranny for power. OK, maybe not in a 1590B, but still. Sounds incredible.

[vigilante397]

That is bloody brilliant! And inspiring. I wonder if I can do this in thru-hole and without cab sim and the power amp, with either the nixie or a backwards wallwart tranny for power. OK, maybe not in a 1590B, but still. Sounds incredible.

Thanks It should absolutely be doable even in a 1590B as long as you're using a submini tube and the nixie supply. I was thinking about trying a through-hole layout of just the preamp and the power supply, maybe I'll throw that together. There are tons of things you can really only do in bigger boxes, I just personally think 1590B's are the most aesthetically pleasing so I try to do everything in them.

Well I really hope the cabsim I'm trying out tonight sounds good, because I already got excited and added it to my final board layout I had an idea on how to fix the 386 headphone amp so I'll also return to that tonight. The big thing I like about the 386 is low parts count, so if I can get that doing what I want I should be able to fit it on here without changing the size of the board. Here's a quick shot of the layout (red is top, blue is bottom) and a 3D model of most of the parts on the board. I decided to get rid of the master volume pot and make the bright switch (which I had previously ignored) board mounted. The final size of the board is 53mm x 48.5mm (2.1" x 1.9").

EDIT: just realized I used the wrong pins on the bright switch so it would work backwards. Whoops. I can fix that.

[Ben N]

Is there any issue with venting heat?

[vigilante397]

Is there any issue with venting heat?

Not so far, but I also haven't run it in a box yet. I'm only running a single tube, so the 6V regulator, which is usually the hottest component (at least in my tube designs) is only feeding about 300mA and isn't heating up terribly. The tube gets a little warm, but the main reason I wanted it on the same side as the pots is so I can vent the top of the box over the tube.

Speaking of which, just finished the through-hole layout. If I can find some time I'll see if I can get a board milled out this week to verify the layout, but I thought it went pretty well. Single-sided with just one jumper, 16mm board-mounted pots, and still kept it to 2.1" x 2.6" I didn't want to use the 400V polyester caps I usually put in amps, so I used WIMA 250V caps on the high-voltage, which saved a bit of space as they're only a little bigger than the caps I usually use for builds.

[Ben N]

You work fast, Man!
I like the little daughterboard for the tube, but I think I would try to put an 8-pin IC socket there and socket the little bugger.

[vigilante397]

I like the little daughterboard for the tube, but I think I would try to put an 8-pin IC socket there and socket the little bugger.

I see a lot of people doing that and I definitely understand the logic behind it. The only reason (literally just one reason) I don't socket them is for space. I love the 1590B form factor and I decided a while ago I didn't want tubes poking out the top. I've also seen a lot of people just bend the legs to get them to lay flat, but that makes me nervous because it becomes very easy for legs to short into each other. I worked out a socket for my little horizontal mount daughterboards, but again it just added too much height and made it hard to fit everything in the box.

I do have a pair each of 6N16 and 6N17 that I keep on some vero with headers so I can plug them into IC sockets, but I use them exclusively for breadboarding now.

You work fast, Man!

And yup, it's a pretty slow day at work. Sometimes I get bored with the PCBs I get paid to design so on my breaks I like to do designs like this to relax Plus it's good problem-solving practice and makes me better at my job