DIY project: A handy power supply PCB for small tube amps like preamps, buffers or microphones.

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

rock soderstrom

Tour de France
Joined
Oct 14, 2009
Messages
3,577
Location
Berlin
Hi guys, here is a small handy PCB project for (not only) the self-etchers among us.

You surely know this, you want to realise a small tube amp project and once again you don't have the right power transformer at hand. Many power transformers for tube circuits only come with a 6.3VAC winding, but you want to heat your tubes with a rectified and regulated 6.3VDC. This also applies to many recycled transformers from old tube equipment such as portable R2R tape recorders or radios.

What can you do? You can try to achieve this with special low voltage drop regulators and diodes. But it's not really cool and flexible.

My solution is this circuit named TPSU_2XFIL, which was designed exactly for this purpose. With this you can heat smaller tube projects with DC voltage even if there is only one 6.3VAC
winding.

TPSU_2XFIL:

TPSU_2FIL_schematic.png

The trick is a Delon bridge circuit that acts as a voltage doubler with a downstream LM317 regulator and additional (optional) double filtering stages, which is useful for certain circuits like some tube microphones that get their biasing from the heater voltage.

Other more exotic DC target voltages for the heater are also possible, such as tubes from the P/X/U series, which brings many new possibilities.

Possible applications are mic/line preamps, buffer or headamps for microphones with rather low power requirements. I would not recommend this circuit for devices like stereo compressors or similar with many tubes. Do not use this with power amplifiers!

One should always bear in mind that voltage doubling and rectification put a corresponding load on the transformer. I would multiply the required amount of current by 3, so your transformer should be on the safe side. For example, if your tubes need 400mA, then use a transformer with a min. 1200mA 6.3VAC winding and everything is fine.

The voltage doubling produces a higher voltage which is partly converted into heat by the regulator and the filtering stages. Therefore, please consider the thermal load capacity of the resistors and the regulator. A heat sink is mandatory! The board also offers an additional series resistor in front of the regulator to distribute the thermal power loss more evenly.

The B+ section is a stone-age unregulated CRCRCRC circuit with the optional possibility of Z-diode clamping similar to the NU67 of the Neumann U67.

An LED indicates the voltage present in the circuit and, together with the R10 series resistor, serves as a bleeder and also sets the predominant current of the filter circuit when working with very low load currents (e.g. a microphone) of less than 1mA.

TPSU_2FIL_3d_rendering.png

TPSU_2FIL_layout.png

As you can see, I have not specified many values in the schematic, as this depends entirely on the application and the corresponding transformer. Some components are also optional, you can but you don't have to install them.

The general rules are - Ohm's law is your friend. Note the current capacity of your diodes and resistors and the maximum voltage of the capacitors!

Here is a concrete dimensioning example for a tube microphone with a 5654 tube. The tube requires 175mA +-15mA heating current and is represented by a 33Ohm resistor. The flowing current at FIL+ is 192mA.

At B+ the tube is simulated by a 220k resistor, 0,78mA current is flowing.

TPSU_2FIL_3d_test.jpg
TPSU_2FIL_3d_example.jpg

In this example I use a transformer with the following data: 16VA, 185V/40mA, 6.3V/1.4A.

R7,R8 and R9 are each 10K, the two series resistors for the LED2 are each 100K/3W, which corresponds to a total resistance of 50K and allows 3.6mA to flow in this branch.
Since I have a following potentiometer in my application and use additional ripple filtering in the microphone itself, the 172.5V output voltage is just right. If you wanted to supply a microphone with the more common 120V without that, I would rather use 20k (or more, do the math) series resistors and install a parallel pot to R8. You will find soldering points for the external pot wiring, as well as a place for a optional clamping zener diode on the underside of the PCB.

My example works perfectly. The 6.93VAC supplied by the transformer at the input is converted very cleanly into an adjustable 6.3VDC (or more). R4 and R5 are 5.6R resistors in my example. Nothing gets excessively hot, everything is within the green range. I can't measure any residual ripple voltage with my current possibilities.

Atached are the layouts for PCB self-etching, ready for direct printing, keep it at 100% scaling.

I have atached two complete circuits on a common Euroboard 160x100mm PCB plus a single 80x100mm version.

TPSU_2FIL_3d_rendering_euroboard.png

I hope someone can use this in the future. I would appreciate feedback. Have fun! (y)

Edit: added print ready silkscreen PDF for euro PCB
Screenshot 2024-04-12 at 22-43-45 PCB_TPSU_X2FILAMENT_003_Produktion_doppelt_2024-04-12-1.pdf.png
 

Attachments

  • PCB_TPSU_X2FILAMENT_003_Produktion_doppelt_2023-06-15-1.pdf
    233.5 KB · Views: 2
  • PCB_TPSU_X2FILAMENT_003_Produktion_single_2023-06-15.pdf
    115.1 KB · Views: 2
  • PCB_TPSU_X2FILAMENT_003_Produktion_doppelt_2024-04-12-1_Silk.pdf
    315.4 KB · Views: 0
Last edited:
Thansk for sharing!
I'm also designing a PSU board for some mics and it's interesting to see your layout choices. It's similiar to mine.
It looks, like you did not use a copper ground layer. And you split the B+ 0V and H+ 0V and connected them at one point.
Can I ask why you did so?
I've read, as so often, different opinions on 0v layout on PCB's...
 
Thansk for sharing!
Thanks for the like (y)
It looks, like you did not use a copper ground layer.
Yeah right. I don't use a dedicated ground layer, I orientate myself more towards the classic tagboard construction as found in old tube amps and I like fat traces. I am a self-etcher and usually work with single-sided PCBs.

There are certainly many good reasons for a ground layer in PCB design, but I have had very good experiences with my variant so far and will stick with it for that kind of stuff.
And you split the B+ 0V and H+ 0V and connected them at one point.
Can I ask why you did so?
I am not an expert on grounding. My way of doing it is based on reading, studying other designs and my own experience.

I usually always connect the grounds externally to a central ground point, but keep the grounds separate on the board to be able to elevate the heater if the circuit requires it because Ufk could be exceeded. That is foreseeably not the case here.

Therefore, I have already connected the grounds on the PCB, which has the advantage that one screw terminal can go to the XLR 5pin input and the other to the central PE ground point of the housing. With microphones, I don't separate the grounds from B+ and the filament, it's all the same.

I always take the grounds at the output of the PSUs to be as far away as possible from the "dirty" rectifier diodes and large filter capacitors.
 
Last edited:
I am probably the slowest metal worker and mechanic on this planet.

If someone had to pay me for that, I'd be rich or unemployed.

To the point, my practical TPSU_X project is progressing at snail's pace. I just upgraded and lit up my voltmeter.

20230624_220552.jpg20230626_164506.jpg
20230625_204449.jpg20230625_200614.jpg20230627_154404.jpg20230627_154331.jpg20230627_160942.jpg20230627_161036.jpg
 
Last edited:
I don't like blue LED's.

I took out the LED's from my China VU's and put in a little pilot licht (the ones that look like a fuse)

It just looks so much better compared to a couple of LED's
 
I don't like blue LED's.

I took out the LED's from my China VU's and put in a little pilot licht (the ones that look like a fuse)

It just looks so much better compared to a couple of LED's
I'm not the biggest fan of blue LEDs either, but I didn't have any others in 3mm size. The space was very tight, I had to remove some material, now it barely fits.

Do you have a source for the bulbs? How big are they?

PS: The LEDs look a bit better in real life than on the photos.
 
lampie.jpg
Here it is inserted in a 35mm VU
It actually has it's connections on one side, the VU was actually designed to hold a real pilot light, but that never made it.

I'm pretty sure I got these from Farnell as I have an account there and get most of my stuff there.
 
I would separate the grounds of the filament voltage and the HT voltage.
(You can always connect them on the terminal block.)
Because some microphones want a negative heater voltage. (+ grounded).
For example the U67.
 
I would separate the grounds of the filament voltage and the HT voltage.
(You can always connect them on the terminal block.)
Because some microphones want a negative heater voltage. (+ grounded).
For example the U67.
I know that's a good point and I had thought about it, but then decided against it. The bridge can be Dremel´ed away if necessary, that was my planned emergency exit in case I wanted to separate the grounds after all.
 
I am probably the slowest metal worker and mechanic on this planet.

If someone had to pay me for that, I'd be rich or unemployed.

To the point, my practical TPSU_X project is progressing at snail's pace. I just upgraded and lit up my voltmeter.

View attachment 110796View attachment 110797
View attachment 110798View attachment 110799View attachment 110800View attachment 110801View attachment 110802View attachment 110803
Hi,

It seems green, blue and red are the standard colors for voltmeters and stand for different voltage levels.
Blue generally stands for 30-600V, right?
Linda
 
Hi,

It seems green, blue and red are the standard colors for voltmeters and stand for different voltage levels.
Blue generally stands for 30-600V, right?
Linda
Hi Linda, sorry I don't understand your post. Are you talking about the color of the lighting? As far as I know, there is no "color code". In fact, the vast majority of analog (moving coil) voltmeters are unlit.

Or did I misunderstand you?

Cheers!
 
Hi Rock,

Thanks for your reply. I'm sorry for my unprofessional question. 🤝
Linda
 

Attachments

  • 20.png
    20.png
    32.4 KB · Views: 0
NO NEED TO BE SORRY!!! << Completely "unprofessional" questions are always welcomed on this forum so us "Old Farts" on here can pass on whatever it is that we may know about the question asked and further educate the "newbie"!!! Then.....one day.....you will end up "just like us" after we have become "worm-food" and are "pushing daisies" somewhere!!! So.....ASK AWAY!!! YAY!!!
Thank you very much! 1688531181868.png
Linda
 
[the whole Gerber topic is new to me] -- Whenever it is that you decide to post your GERBER and N/C Drill data files, if you would want to send them to me first to check them over before you do post them, I will be happy to give them a "once over" for you!!!
Thank you for the offer! (y) It's not a priority for me at the moment. I first have to revise my component library a little and, above all, test it. It's good to know that we have expert knowledge on this subject here in the forum.

Another topic, is there a rule or source of information regarding the distance on the PCB for tracks with high voltages, as is the case with tube circuits, for example?

How close can a ground plane and say a 200V DC trace get to each other before it becomes critical in terms of electrical flashover? Is there any reliable information on this?
 
"Aluminum is just shiny wood." 🤪

Drilling large holes is not so easy for the average DIYer.

I've been milling it for a while now. Goes quite well. :cool: Don't take your best router bits for it. The super cheap ones from ebay are just fine for this sacrifice.

20230708_172512.jpg20230708_172804.jpg20230708_173927.jpg20230708_173927.jpg20230708_173906.jpg

The usual cheap router bits can not plunge into the material, you need to pre-drill a starting hole.

Do it only at your own risk! Probably a stupid idea, so wear personal safety equipment!
 

Attachments

  • 20230708_172804.jpg
    20230708_172804.jpg
    584.6 KB · Views: 0
  • 20230708_173702.jpg
    20230708_173702.jpg
    579.7 KB · Views: 0
  • 20230708_173927.jpg
    20230708_173927.jpg
    818 KB · Views: 1
Last edited:

Latest posts

Back
Top