Classic amp topologies made with submini tubes

[Kevin Mitchell]

I haven't even come close to finishing my submini blackface style amp   

Lookin' great Thomas!

[thomasha]

finally made a video for this one, is there someone still interested in these small amps?

[Kipper4]

That's great Thomas.
Yes I'm still intrested, I will eventually get around to doing one.
Don't give up the good work mate.

[Metaldestroyer]

Alright, here it is in its open, unboxed form. Only took 3 years to make.

It isn't quite a classic amp topology but it is based on a Top Hat Club Deluxe which is based on an AC30 Top Boost channel. I used 6n17b for the preamp, 6n16b for the PI and 6p30b for the output stage. The whole thing runs off of a 12v DC adapter which then goes through 2 SMPS: one MAX1771 for the HV and one TPS564201 for the heaters. I'll post a full schematic when I get around to drawing one; the one I worked off of to build it has some mistakes and I changed quite a few values in order to get it working.

When I first plugged it in, it sounded grainy and was barely audible. So I went back in and replaced the PI circuit with the one Thomasha used for his JCM800, which didn't solve the problem. Then I noticed I forgot to ground the OT secondary, which made the whole amp stop making sound. Then I figured out I had flipped the secondary common and one of the primary leads  . It's a wonder the thing output any sound at all, but somehow I didn't kill the tubes or the transformer so it's fine I guess. Now all that's left is to take the whole thing apart and put it in the box.

[tommycataus]

I'm definitely still interested in this subject. Metaldestroyer, that looks and sounds awesome! I'd love to glance at a schematic if/when you have time to draw one.

The problem I am having is mainly due to the fact that I'm new to building tube amps. Other than specific transformers that are named, I don't understand what to look for. I do understand the risks and have worked with high voltage circuits before.

I'm guessing the SMPS circuits negate the need for a power transformer, but what about the output transformer? What do I look for in the primary and secondary coils? Am I limited to only a few options or are there myriad choices out there? I'm keen to learn more about this component in particular, but Google falls a little short on the detail. Thanks in advance.

[jubal81]

That TPS564201 looks like a great chip. Looking forward to seeing the schem. Great work!

[thomasha]

Nice build, hoping to see some pictures too, it already looks interesting with those chicken head knobs.

You could also tell us more, if there were any problems with your build

I wanted for some time, to make something fail proof, so it would be easier to other members to build, but the SMPS is always the most problematic part and introduces a lot of noise if not tamed adequately.
If you managed to get around this would love to see your layout and which inductor  and mosfet you're using with the MAX1771.
I found some good inductors some time ago, but during me testing there was also a lot of them that just vibrated too much, and produced this coil squeal sound.
Another member showed me a guy at ebay selling smd SMPS that could deliver up to 20W at 200v. Maybe that could make the whole process easier.

You had luck with your build problems, the tubes and OT are more forgiving than the SMPS chip, I think I lost the track of how much I already burned just because one filter cap had a bad solder joint.

keep the good work going!

[Metaldestroyer]

I'll just drop this off here 



I'll do a full write up with schematics, stories, pictures, etc. when exams stop

[thomasha]

wow that looks awesome, need to make some pictures like that.

Some time ago I finished another build, that was more like a remake of an older amplifier.
The superfly using russian tubes never had enough gain to me, (at lower volume settings...).
So I changed it to something in the line of the jjs amplifier with only one triode in SE.
The circuit and the looks where both night train inspired, and I named it the "Tiny Nightingale"

Here some pictures:






Cheers,
Thomas

[tonyharker]

Can you make your pictures larger please?

[EBK]

Can you make your pictures larger please?

If you take out the "m" at the end of his links, you can see the larger version.

[Ripthorn]

Maybe 5 years ago I made a 2W plexi/JCM800.  It was based around a 1959 super lead, but when I saw how similar it was to the JCM800, I worked up a relay switching system that would allow me to be in plexi mode with either input 1 or inputs 1 and 2 jumped and then I could flick a switch (or use my foot switch) and go into a JCM800.  It sounds quite good, but I don't use it a ton.  I'm thinking of building something else into that chassis.

[thomasha]

Hi, just adding another build!

This time using the 5672 direct heated tube in push-pull at only 87v.
The preamp still uses the 6N16B and 6N17B russian tubes. In the future I plan to use direct heated tubes for all the amp to go low filament current and all battery if possible....

For now I tried a AC30/liverpool mixed schematic for the preamp and I loved the sound of it!
It can blast 0.5W on a 8 ohm load when fully distorted!

The schematic follows>


As can be seen the last stage is biased using the 44 ohms and the 62 ohms resistors, which gives a grid bias of about -7.5v.
Using the 22.5k transformer from Hammond it only goes to class AB in with the maximum voltage swing.
Although there is a layout for it, to make it faster I used a perf board:



Some measurements:

Here is a measurement of the signal at the speaker terminals (8 ohms)

This build also required the traditional small combo, as used in my previous builds. This time I wen't with a semi finished box, cut it in half, remade the finger joints in the bottom and it ended up looking like this:






To keep it simple I just used some wax this time,
I liked it better than the shiny finish of my first builds (less finger prints).

I'm planning to make some videos soon, to show how it sounds.
Not as loud as the previous builds (what I was looking for! I haven't added a master volume in this one), but as good sounding.
The advantages are:
-less filament current than using 3 or 4 russian tubes. (500mA against 800mA)
-push-pull output stage which gives (IMO) a better tone.
-lower wattage (around 200mW ), TV output level in the small speaker or even higher if using proper 12" speakers.

As I said before, If I could do it only using direct heated tubes with series filaments this could have an even lower filament current, which would allow using some 3.7v Li-Ion batteries (3x to have almost 12v).

Cheers

[Ripthorn]

That is so amazingly cool!  I'm working on a little headphone amp with integrated cab sim and overdrive that will get a dovetailed walnut head cab, but that little Vox is just so cool!

[thomasha]

I made a quick video clean/dirty with the volume maxed,
to add the distortion I increase the guitar's volume.

Trebble was like 3 o'clock, while bass was at 9 o'clock.

[vigilante397]

I know this thread is semi-old, but I just happened upon a stash of 6N16, 6N17, and 6P30, and I'm commenting here to remind myself to do something cool with them

[bluebunny]

I know this thread is semi-old . . .

This thread is never too old for stuff like this:

. . . but I just happened upon a stash of 6N16, 6N17, and 6P30, and I'm commenting here to remind myself to do something cool with them

I wonder if any of Rick's old submini projects would suit these, perhaps with a bit of tweaking?  I've built a few and really like them.

[vigilante397]

The problem I seem to be having is that half the links in here are dead and the other half are blocked by the firewall at work

I'm thinking I want to make a little high-gain amp because that's the main tone I don't really have covered in my current amp collection. I was hoping Metaldestroyer would post details on his amp, but he seems to have vanished. Oh well, to the breadboard then

[Metaldestroyer]

Sorry I disappeared for a year. I decided the best place to put the write-up would be my own site and it took me until now to get around to it. Here it is everyone:

http://projectstudioproject.com/2018/05/30/subminiature-tube-amplifier/

[thomasha]

Hi guys, I already posted over madbeans, but maybe there is someone interested in this one here too.

I started another subminiature amplifier where the main idea was to use the pencil tubes I got over ebay. I have some old 5678 and 5672s that are already used, but still working. Someone probably took them from a military radio or something.

This time I aimed at a mesa boogie kind of distortion, similar to what was done with the DR Boogei pedal. Actually I started working on the preamp, using the 5678 in triode mode.

I started from this schematic, since it has some voltages.

It was originally posted here: https://music-electronics-forum.com/showthread.php?t=40525

I checked the load lines just to find out that most of the stages are center biased, with exception of the cold clipper with the 39k resistor. My first attempt already sounded pretty nice:

But it required a proper design.

For the 5678 the load lines at 45v or 60V look like this:

Here the values


Since the 5678 and the 5672 have directly heated cathodes, the cathode follower would be really difficult to implement without a proper transformer. Here I wanted something that could run from 9V and use as little current as possible. In a series string the filaments only require 50mA!

So for the cathode follower stage I used a lower gain/ higher current stage, with a 20k plate resistor. Not sure if this is the best approach. What would you do?

The important values at the table are the cathode voltages, since I will have to bias each stage at that value. In the meantime I tried to make a spice model for the 5678 tube to simulate it, and compare with the simulation of the original preamp. Therefore I used the paint_kit tool. My lines look like this:



And the schematic:


I was very confident as I compared both sine waves, since both have a similar shape. I could get it even closer by adjusting the tone.
The only difference is the voltage swing, since in the simulation I was testing a new voltage divider before the last 5678 tube, where the cathode follower, with unity gain, should be.



For the output stage I tested some different configurations, but ended up using the 5672 tube. I even used the specified values at the datasheet, with 65V and a bias of -6.5V.

Since all the filaments are in series I decoupled the signal to ground at each filament and used fixed bias as the difference between the cathode voltage and the grid leak resistor of the previous stage. For example, the seconda stage has it's cathode at 2.5V (negative side of the filament), while the grid is at 1.4V approx., resulting in a bias of -1.1V. I Used a trimpot to adjust the voltages and also decoupled every trimpot, so that the signal sees the trimpot as ground. The extra current from the anode is really low in comparison to the filament current (0.1mA vs. 50mA), so that no parallel resistor is required.

For the output stage, on the other hand, there should be at least 3.3mA, so that I added a 1.5k resistor to ground (filament is at 7.5V). The final schematic looks like this:


So the good news is that it sounds awesome, and is really quiet. The bad news is that the 5678 is quite microphonic. I did all my testing over the table, and as I put the board over my 1x12" cab it had some ringing. I could use the 5672 in place of the 5678, but the filaments must be reversed, or I build the amp inside a small cab where I could use some rubber feet to dampen the vibration. Actually it's not that bad, since it's not that loud, and I can always put it on the nearest unplugged amp.

I have some 5672 that I used in other projects, which I could remove just to test the circuit and make a proper amp, without the microphonics. What do you think?