Tube mic PSU for a tube DI

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Chrisfromthepast

Well-known member
Joined
Mar 5, 2018
Messages
265
I’ve been wanting to experiment with some tube DI topologies, with the end goal of tracking bass guitar.
I went a little royermod crazy in 2020, and I currently have more tube mic PSUs than I ever use at one time.
If Im careful about current consumption and the pinout, is there any technical or historical reason that I can’t use one of these microphone PSU’s to power a tube DI?

The one I’m setting my sights on has 12v heater and 140v b+. A single 12ax7 would be small and simple enough for me to shove into an old DI outfitted with a 5 pin xlr.
Is this project a go?
Why haven’t I been able to find a similar example?
 
Chrisfromthepast said:
The one I’m setting my sights on has 12v heater and 140v b+. A single 12ax7 would be small and simple enough for me to shove into an old DI outfitted with a 5 pin xlr.
Is this project a go?
Why haven’t I been able to find a similar example?

I'm sorry but I don't seem to understand the second part of your post.

If your PSU's can supply the power requirements the DI circuit you are using needs, then don't see any problem with that.
 
Is your PSU regulated? Can it handle the power consumption of the DI circuit?
If it's not regulated, you probably need to change some resistors in the PSU to set the voltages for the DI as the power consumption is probably different than your mic. So the voltages will probably rise or fall depending on the power need of the DI.
 
A tube DI is almost the same as microphone gain stage.

1meg input instead of many megs of input resistance.
 
Thanks for the encouragement!

If im careful to use the same type of tube, I’ll automatically obey the heater current requirements of the PSU.
Thats fairly easy.
Once I change the 1 gigaohm resistor from grid 1 to ground, does anything else automaticall need to change in the circuit?
Im a bit of a tube dunce, and I haven’t found the magic article or tutorial that makes me feel confident about choosing the right resistor values.

If I choose a PSU designed for a EF-86 in a triode cathode follower configuration, and simply swap the 1G to a 1M, I have a good starting point?
All of these royermod based circuits look dead simple, but are confusing to tweak because they elegantly rely on stray capacitance and internal resistances to lower the parts count.

The capsule polarization is also part of the trick too, right?
There will be DC on the grid so I need a coupling capacitor so that I don’t put DC on whatever I plug into the 1/4 jack.
.1u? Im starting to see why this isn’t a common sister project to the tube mic builds
 
Personally I would advise you to collect already made Tube DI circuits and check if their PSU requirements fit the bill than trying to adapt a mic circuit to use as a DI

My 2 cents

There’s some Tube DI circuites here in the forum.
Search Aguilar Tube DI, Demeter Tube DI, New York Dave one bottle
 
I've build this circuit, with very nice EQ as well... transformer balanced line out.

http://www.frontiernet.net/~jff/SonOfAlembic/SonOfAlembicF2B.html

Great on Rhodes, bass and guitar.
 
Yeah the Alembic is the way to go. I've built a few for customers and everyone loved them. Great for recording and/or live use. Works great with guitars too.
 
Chrisfromthepast said:
Thanks for the encouragement!

If im careful to use the same type of tube, I’ll automatically obey the heater current requirements of the PSU.
Thats fairly easy.
Once I change the 1 gigaohm resistor from grid 1 to ground, does anything else automatically need to change in the circuit?
On the face of it there is nothing else you need to do. However, the input signal from a guitar is likely to be at least 10 times greater than the signal from a mic capsule. Depending on the circuit used, this may or may not cause the tube to seriously distort. To avoid this you might need to change the bias point but if the design is a cathode follower then it will almost certainly be OK as it is.

The other item that may struggle is the output transformer, again because it will have been designed to handle microphone levels rather than guitar levels. If this does turn out to be a problem the the only solution is a bigger one.

As others have pointed out, tube DI circuits tend to use a different topology, but i see no reason not to try your idea.

Cheers

Ian
 
As always noise is added to a thread

For the first question
If the B+ and heater supplies can supply enough current without overheating should answer your question

It is funny tube microphones tend to be two basic circuits
Plate out triode gain stage cap to transformer
or cathode follower cap to transformer
Nothing special except for the grid resistance you need to keep in mind.

If the input is too high a level and you want 1meg input use two 470Ks in series and connect the grid to the center node or use a low noise 1 meg potentiometer at the input etc.

I don't see a different topology in tube DIs to a "standard" tube microphone circuit  it is just a voltage and/or current gain stage
 
Gus said:
I don't see a different topology in tube DIs to a "standard" tube microphone circuit  it is just a voltage and/or current gain stage
Tube microphone circuit tend to use a capacitor coupled output transformer, whether driven from the plate ot the cathode. This avoids dc current flowing in the transformer primary. Tube DI boxes tend to use a transformer as a plate load, passing dc.

Cheers

Ian
 
ruffrecords said:
As others have pointed out, tube DI circuits tend to use a different topology, but i see no reason not to try your idea.

Cool! Thanks Ian!

This is kind of an bonus project, that I was hoping to build from parts on hand. I can try it out as soon as I sneak by the lighting department for a 5pin male xlr.

The royermod2 project is designed for 5840 tubes and one version of the power supply I built has a capacitively coupled unbalanced output.
It works great on a most of the prosumer gear, like focusrite scarlet, but doesn’t drive every transformer based mic preamp well.
With that limitation understood, I hopefully can sidestep the transformer issue.
True, I won’t get as good as a sound as this gem:
https://groupdiy.com/index.php?topic=61139.80
I have been drooling over the alembic build lately too.

Both of these are on my list, and I have found enclosures and a few random parts for each, but I tend to have too many kettles cooking at once...

470k resistor divider is a great tip, thanks!

If I found a suitable output transformer, I would find one that could handle dc on the primary. That seems to be the standard way most guitar amp output stages, and tube DI’s do it.

Sidebar: is DC on the primary  how the silk knob is still implemented by neve licensed gear?
 
I built rev 1, and its a bit lackluster.
I took A DI I haven’t used in years, replaced the 3 pin xlr with a 5 pin and removed the little tiny transformer.
Transformer gone, I had plenty of room for a subminiature tube.
I chose a 5840 workalike with a very long part number.
The pinout was available here:
https://groupdiy.com/index.php?topic=75320.0

To start as simple as possible, I just added what I had available from the parts bin
22n into 4M4.
I built my power supply without an output transformer, so this project is really limited to bedroom sized studios because cable runs must be short.
I understand that having an unbalanced output doesn’t truly fulfill the requirements of a DI.

Here’s what I built:
https://imgur.com/gallery/jWeKA48

https://i.imgur.com/Xc7HWn7.jpg

The pad switches on the DI still work, but are not useful for guitar or bass.
With no pads the output is pretty weak.
I compared the output to a microphone designed for this psu, and the mic was easily 15db hotter. I wasn’t expecting that result.

Bass guitar sounded fine, but it was nothing to write home about. Not having any gain, and with the weak unbalanced output this box just didn’t do much for me.
I don’t see a use case.
Rev2 is going to need an output transformer.


 
I am not surprised the output is very low. From the schematic you posted it looks like the plate of the tube goes straight to the HT. There is no plate resistor for the signal to develop across.

Cheers

Ian
 
Ian look again, it's a cathode follower.


Chris, depending on what you load the output with on your unbalanced connection, that could reduce the level since the circuit is not up to the task to drive it.
 
Chrisfromthepast said:
I built rev 1, and its a bit lackluster.
I took A DI I haven’t used in years, replaced the 3 pin xlr with a 5 pin and removed the little tiny transformer.
Transformer gone, I had plenty of room for a subminiature tube.
I chose a 5840 workalike with a very long part number.
The pinout was available here:
https://groupdiy.com/index.php?topic=75320.0

To start as simple as possible, I just added what I had available from the parts bin
22n into 4M4.
I built my power supply without an output transformer, so this project is really limited to bedroom sized studios because cable runs must be short.
I understand that having an unbalanced output doesn’t truly fulfill the requirements of a DI.

Here’s what I built:
https://imgur.com/gallery/jWeKA48

https://i.imgur.com/Xc7HWn7.jpg

The pad switches on the DI still work, but are not useful for guitar or bass.
With no pads the output is pretty weak.
I compared the output to a microphone designed for this psu, and the mic was easily 15db hotter. I wasn’t expecting that result.

Bass guitar sounded fine, but it was nothing to write home about. Not having any gain, and with the weak unbalanced output this box just didn’t do much for me.
I don’t see a use case.
Rev2 is going to need an output transformer.

That circuit is not good for a DI
it uses contact bias and leakage to both charge the capsule and bias the CF
Adding an input resistor to ground(move the resistor you added) and a coupling cap to the grid still might have issues

If you want to use a cathode follower (CF)look at other circuits to copy or read up on CFs
If you look at an AKG C60 CF you will need a cap and another input resistor to ground it has better defined bias

Do you want voltage gain and then reduce the output to get some distortion OR do you want as clean as possible?
If you want as clean as possible build a solid state DI.
 
ruffrecords said:
Doh! What an idiot I am.

Cheers

Ian

Never, Ian!  I was confused by the drawing as well and thrown off by the handwritten numbers by the XLR pinout.

Question to the O/P: Is there enough room inside the PSU to fit a submini tube and a shielded transformer?  It seems to me the extra cable and box could be eliminated.
 
mjrippe said:
Never, Ian!  I was confused by the drawing as well and thrown off by the handwritten numbers by the XLR pinout.

Question to the O/P: Is there enough room inside the PSU to fit a submini tube and a shielded transformer?  It seems to me the extra cable and box could be eliminated.

Thats a good idea about eliminating the box. The PSU is for 2 mics and its pretty full aready. A submini tube and 3 resistors per channel is about all I could fit.
My apologies everyone, this is a bit sloppy. I was really hoping for a free lunch.
I appreciate all the input.
I turned your ideas for the next version into a proper schematic and will lay it out the PCB and release the gerbers in gratitude when we get it finalized.

That aside, Im a few resistors away from a textbook cathode follower circuit.
Inside the PSU, there is 39k from cathode to ground. From what I can tell, Im just missing 1 resistor, and I've misplaced 1 resistor
In the article this PSU was built from, David Royer shows a more typical common cathode arrangement, and then suggests you build his stripped down version for his particular application.

https://web.archive.org/web/20090901000000/http://www.diyfactory.com/projects/royerproject/royermod_2.pdf

It seems like he is taking advantage of a corner case thats too advanced for me.

Back to a typical common cathode. Rg, needs to connect to the the 39k Rk that is inside the PSU.
Additionally, I need R bias  between that node and the cathode. Is 1k a fine starting point?
I've attached a revised version that looks alot like the c60 schematic now.

Gus, I have plenty of clean options, so Id prefer something with gain and some output attenuation. I haven’t found a load line calculator for these tubes, so I can’t fake my way through this one. In any case, gain would require another stage, or a common cathode configuration, right?
Should I just try the 1k bias resistor and leave the 4M4 Rg and 39k Rk?
B+is 95 loaded.
If I understand cathode followers correctly, the output signal will be roughly the same amplitude as the input, but lower impedance and reverse polarity, and no change in resistor values will give me gain of more than 1.

Thanks for your patience!
Chris
 

Attachments

  • 5840 DI.pdf
    25.2 KB · Views: 32

Latest posts

Back
Top