Tube DI Box

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
OK, here is a draft schematic for the tube DI box PCB. The idea is that it is flexible enough that you can build almost any version with it:


Schematic_TUBEDI_2024-05-03.png
The lower triode can be configured as either an SRPP, mu follower, White follower or single ended input stage. The top triode can be configured to complement the bottom triode configuration and in the case of the single ended input stage, the top triode can be configured as a cathodyne or a simple cathode follower output stage.

There are four connectors: HT, heaters, input and transformer. If we keep the HT on the top triode down to 250V there should be no need for heater elevation.

Cheers

Ian
 
Ian,
Been traveling a lot so where is this project at? Before I left I did a parallel 6922 loaded with a 140H/12ma (49% nickel [intact audio] running at 9.7ma or ~4.8ma each, Vdc drop 5.1V) at 80V using the NMT0572. Input supply was 8V regulated took that down to 6.3V for the filament diode->into the NMT and then grounded the -72V terminal and output on ground. I have used these in many USB dacs with 12AU7 tubes as buffers. It's best to use 3 caps and bypass each leg no matter if used or not.
Anyway the results were killer into a 5K:500 output I had which was a modified Peerless design the frequency response was dead on 10-52K -1dB. Too bad I can't get more of these as we lost Mike LaFevre (MagneQuest) last summer.
Anyway catching up on this tread.
Thanks,
Gordon
 
In summary, there are a number of alternative ways or building this. The above schematic tries to capture most of them and I am currently working on a PCB layout that will allow you to build any one of them.

From the safety/regs point of view and external SMPS seems the best option with a little dc:dc boost supply for the HT.

When the PCB design is complete I will have a bunch made. Happy to give them away to anyone interested in building a version.

Cheers

Ian
 
I believe the project is in Ian's safe hands and I for one are eagerly looking forward to seeing what designs he comes up with for this tube DI.

I appreciate that this remit could be realised using load chokes with outstanding performance, but I would be more than glad to see designs that minimise the requirements for exotic iron as these days you have to pay crazy money for what are essentially simple components. We have enough issues sourcing power and output transformers for this type of build!! 😉
 
When the PCB design is complete I will have a bunch made.
I would be interested with taking a look at your GERBER and N/C Drill files and running them through my GERBER analyzer program. I would also suggest in having your PCB-design program setup in -- MILS -- instead of "Metric", as the data ends up "making more sense" by using -- whole numbers in "Mils" -- instead of "fractional numbers" as I have seen what metric files so often seems to do.

You would need to actually open up GERBER and N/C Drill files in an ASCII word-processor, like Notepad, in order to "see" the data and how everything -- actually is formatted -- in order to understand what it is I am relating here. My work with PCB GERBER and N/C Drill files has been somewhat "different" than what is normally typical on this forum.

/
 
In summary, there are a number of alternative ways or building this. The above schematic tries to capture most of them and I am currently working on a PCB layout that will allow you to build any one of them.

From the safety/regs point of view and external SMPS seems the best option with a little dc:dc boost supply for the HT.

When the PCB design is complete I will have a bunch made. Happy to give them away to anyone interested in building a version.

Cheers

Ian
I would love to build one if there’s a spare PCB Ian.
Thanks a lot,
Silas
 
I’d love to be in the mix for this pcb please if there are enough.
Just so we are clear, a the moment it will be a one size fits all PCB. So by fitting appropriate components and links you can build a SRPP, or a mu follower, or a White follower preamp stage or a cathode follower plus preamp. I will provide a table of component values and links showing how to build each version. At the moment the component values have only been worked out for the ECC88 but I will also create a table for the 12AU7. In all cases the design will drive a VT(2291 transformer (or equivalent). In some case it can be wired 2K4:600 and in other cases it will need to be wired 9K6:600. Jack sockets and XLR connectors will be off board as will the input and output level pots.

Cheers

Ian
 
Just so we are clear, a the moment it will be a one size fits all PCB. So by fitting appropriate components and links you can build a SRPP, or a mu follower, or a White follower preamp stage or a cathode follower plus preamp. I will provide a table of component values and links showing how to build each version. At the moment the component values have only been worked out for the ECC88 but I will also create a table for the 12AU7. In all cases the design will drive a VT(2291 transformer (or equivalent). In some case it can be wired 2K4:600 and in other cases it will need to be wired 9K6:600. Jack sockets and XLR connectors will be off board as will the input and output level pots.

Cheers

Ian
Is it cheating if a use an old DI enclosure stripped of its innards and replaced with your pcb, etc….?
 
I Like this form factor for something active single tube DI thing;

https://www.reichelt.nl/nl/nl/deskt...tph-1850-150-p126234.html?&trstct=pol_8&nbc=1

Not the cheapest, but it does have a single sliding flat frontpanel that will be easy to customize.
Great as floorbox and desktop unit.

In doubt now as on second look it seems the front panel is fixed to the extrusion parts :unsure:
FYI:
1715604157015.png

[it seems the front panel is fixed to the extrusion parts] -- Here is what the -- TOP-EXTRUSION ONLY -- piece looks like:

1715604329156.png


>> TOP-EXTRUSION -- LEFT-SIDE VIEW:

1715604655929.png

[it does have a single sliding flat front-panel that will be easy to customize] -- The main "front-panel" area should still be rather easy to customize, despite how "gnarly" the ends may look!!! How this extruded enclosure fits together has nothing to do with anybody being able to drill a bunch of hole cutouts into it in order to make it be what you want it to be.

>> BOTTOM-EXTRUSION ONLY:

1715605513865.png

>> BOTTOM-EXTRUSION LEFT-SIDE View:
1715605632207.png

-----------------------↑↑↑↑↑↑↑↑↑↑↑↑↑↑↑↑----PCB FITS WITHIN THIS AREA----↑↑↑↑↑↑↑↑↑↑↑↑↑↑↑↑↑↑-------------------

>> The PCB size to fit this enclosure would be 4.55" X 5.85" (115.87mm X 148.59mm).

>> ENTIRE EXTRUDED ENCLOSURE SIDE-VIEW:


1715607175279.png

Just in case anyone was wondering.....

/
 
Last edited:
Hi Ian,
Should there be any boards spare I would love to put one together!
Thanks for another interesting thread.
 
Ian,
Thanks for all your hard work on this. The 6922/ECC88/6DJ8 is the best bet for this kind of project. The 12AU7A is a great tube but with high Rp and output Z it would require a higher primary impedance. I tried using the one I have at 15K:500 for the 12AU7A and it doesn't work well in SRPP. If you go up in Z on the primary you of course start to loose top end. Also there is a difference in filament and pinout which is not hard to get around with some headers.
If you like I could lay this out in Altium pretty quickly. I have everything in my data base for everything tube.
If you still want to use the 12V switcher, how about using one of these switching LM317 type modules to get the 6.3V out for the filament? Quick and easy supply for that and most of these run pretty high frequency.
Another option would be the Hammond 261D6 which has international primary, 250VCT and 6.3@1A which would make a pretty easy supply for a single tube DI like this.
Anyway happy to help out!
Thanks,
Gordon
 
Ian,
Thanks for all your hard work on this. The 6922/ECC88/6DJ8 is the best bet for this kind of project. The 12AU7A is a great tube but with high Rp and output Z it would require a higher primary impedance. I tried using the one I have at 15K:500 for the 12AU7A and it doesn't work well in SRPP. If you go up in Z on the primary you of course start to loose top end. Also there is a difference in filament and pinout which is not hard to get around with some headers.
I agree. I only included the 12AU7 because it has convenient 12V heaters. In all other respects the ECC88 is a far superior tube. If only there was a 12V heater version. Actually there is; it is the 8416 but unfortunately they are hard to find and rather expensive:

https://www.ebay.co.uk/sch/i.html?_from=R40&_nkw=8416+tube&_sacat=0&_odkw=8416+vacuum+tube&_osacat=0

Cheapest I have found is just under $40 at Tube Depot

An alternative might be the 12BH7 but I am not sure this is any better than a 12AU7 in terms of distortion and drive capability. A better choice might be the ECC99 which has a reasonable mu and a high gm (meaning a low rp). I have used these in a headphones amp in the past. The only downside is its high heater current
If you like I could lay this out in Altium pretty quickly. I have everything in my data base for everything tube.
If you still want to use the 12V switcher, how about using one of these switching LM317 type modules to get the 6.3V out for the filament? Quick and easy supply for that and most of these run pretty high frequency.

Do you mean something like the LM2678-ADJ ? Looks easy to use and operates at 260KHz.
Another option would be the Hammond 261D6 which has international primary, 250VCT and 6.3@1A which would make a pretty easy supply for a single tube DI like this.

We might be able to get away with the smaller 262B6. Heater current is spot on and with a voltage doubler we should be able to get 250V at 10mA or so for the HT
Anyway happy to help out!
Thanks,
Gordon
THanks for the offer. I will certainly bear you in mind.

Cheers

Ian
 
Ian,
There are a number of the switching LM317 alternatives at DigiKey. They are listed under Linear Regulator Replacements. I have never used these but have drawn up schematics using those.
I am not a big fan of doublers, the 261D6 would give you raw about 184Vdc in the cases I have used them on. Especially because I would probably DC regulate the 6922 heater as I have a boat load of those NJM2396 6.3V 1.5A units here. The lower current usage on the 6.3 will boost the HV a little. I would probably go with a PI filter into maybe a high voltage regulator and run the 6922 at 5-6ma SRPP and boost the filament up like 40-50v off ground.

I am waiting right now for parts so maybe I come up with something pretty quickly.

What output are you thinking of using?
Thanks,
Gordon
 
Last edited:
Back
Top