[drteming]

I got my Little Dot Mk II several years ago. After playing around with it for a while, and totally screwing up the PCB with lifted pads and such ('cause I didn't know what I was doing), I put it away on the shelf. In the meantime, I got into building tube guitar amplifiers, then tube hi-fi amps, and recently, OTL headphone amps. Having learned quite a bit about hollow state electronics, I decided to see what I can do with the long neglected Little Dot. Cracking the little guy open, I forgot the horrors that I had inflicted on the PCB. I salvaged the power transformer and whatever components, and traced out the circuit:

It's actually a really neat little circuit. The gain switches are variable negative feedback switches, with resistances of 470k, 149.9k, 82.5k, and 59.9k. The EF95 is wired as a triode. In the manual, the output sections were described as having a SRPP topology. Actually, it's a self biasing White cathode follower, which is actually better at driving low impedance loads such as headphones.

I made some modifications to the circuit:

First, I moved the input coupling capacitor to before the volume potentiometer. A 100k stepped attenuator was swapped in for the cheesy 10k pot. I set the negative feedback resistor to a fixed 100k as I had the switches set to 82.5k, but had wanted just a bit more gain, but not as much as 149.4k. I added a 100 ohm resistor between the screen and the plate to prevent oscillations. I also increased the film bypass capacitor on the output coupling cap (220uF) to 2.2uF from 0.68uF, increased the bypass caps on the power supply to 0.1uF from 0.047uF, and added a virtual center tap to the heaters of the output tubes. The original heater elevation voltage was about 5V, and I upped it to about 28V. I also beefed up the bridge rectifier, using 4 UF5408 diodes.

The chassis drilled and painted. The orientation has been changed. It's a Hammond 10" x 6" x 2" aluminum chassis:

Power supply built on a turret board:

Power supply installed, along with the rest of the power circuit, and the filaments:

More pictures to come as I progress on the build.

 

[drteming]

I forgot to mention, with pins 2 and 7 of the input stage tied together, the following list of tubes can be swapped in: 6AJ5, 6BH6, 6AK5(5654), 6AG5(6186), 6AU6, 6BC5, 6CB6, 6DE6, 6AH6, 6DK6, and 6EW6 (listed in order of gain from low to high). There are probably other tubes that can work, but these are what I have in my stash.

 

[HiFan]

That's a lot of work (looks beautiful) here. It seems like a good chance to do some mod on it

Isn't 330uf as incoming filter too small? My 2 cents.

 

[drteming]

Isn't 330uf as incoming filter too small? My 2 cents.

Well, there's the input cap, followed by 2 RC filters before power is taken off. The total capacitance is 990uF. For a low powered tube amp, this should be plenty. I'll figure out the ripple voltage after I finish the amp and measure the power draw.

I decided to add back in the variable negative feedback in light of rolling in higher gain tubes.

Mostly done. Here's a shot before the coupling caps so the layout can be clearly seen:

 

[drteming]

All done.

I revised the negative feedback scheme with a set 220k resistor and switched in 330k and 100k, for paralleled resistance of 132k and 68.8k. With only the 220k resistor there is an attenuation of -3.4dB, with 132K, -5.4dB, and 68.8k, -9.2dB. I have nicely matched Voskhod 6J1P-EV and Novosibirsk 6N6P in the amp, and I had forgotten how nice it sounds.

One final change, I upped the film cap on the output to 4.7uF.

The ripple voltage to the 6N6P was 0.27mV and 3uV to the front section, so there's plenty of filtering in the power supply. No audible hum or noises.

The total resting power draw for the amp was only about 36mA. An interesting build can be done with a Hammond 269EX power transformer and a 6X4 rectifier tube with appropriate dropping resistors in the RC filtering circuit.

 

[kostas6a3]

Nice work. Congs.

 

[wantech]

will you use 6H30 ?

 

[drteming]

will you use 6H30 ?

Nope. I'm not paying that kind of money for basically no difference in sound. The output section is a White cathode follower, basically there to match the output impedance to headphones. The 6H30 (or 6N30, if you will) tube earned it's reputation as a driver for power amplifiers, a completely different role.

 

[wantech]

what will the sound changes if I remove the 220k feedback resistor ? thanks

 

[drteming]

what will the sound changes if I remove the 220k feedback resistor ? thanks

Without the 220k NFB resistor, you will have an additional 3.4dB of gain, at a cost of slightly decreased bandwidth, slight increase in harmonic distortion, and a more tubey sound.

 

[wantech]

thanks

 

[Datun Walnut]

This is seriously nice. I love my LD mk2 but always felt extra space for decent caps and a decent pot and a meatier psu. Mines developed an intermittent fault so the time is right to think about this rebuild. I'm going to start planning mine. I think Octal bases for the back line as I love my 6SN7GTs. Thanks for this

 

[lupoal]

@drteming

very nice job!!

I have an LD MK2 too and I'm thinking to remove the bottle neck (components quality) copying your idea... but, as I'm quite e rookie ... do you have a list, like an excel file for example, with all component required and info about them (for example... resistors tollerances, wattage of used resistors in power supply stage , etc. etc.) ?

 

[Datun Walnut]

@drteming

very nice job!!

I have an LD MK2 too and I'm thinking to remove the bottle neck (components quality) copying your idea... but, as I'm quite e rookie ... do you have a list, like an excel file for example, with all component required and info about them (for example... resistors tollerances, wattage of used resistors in power supply stage , etc. etc.) ?

That would be handy. I'm starting to gather components for my build. I reckon there's no harm in going tighter on the tolerances and higher on the wattages. And metal film resistors where you can. The best quality caps I can afford will be a must

 

[drteming]

Doh! I cleaned up my Mouser project list last week, and deleted a bunch of them, this one included. However, all the components are in the schematic above. Starting with the power supply, the 2 180 Ohm resistors are Ohmite 45F180E, wirewound, rated at 5 W. The filtering caps are United Chemicon KXJ, rated at 250V. The two resistors that form the voltage divider for the heater center tap can be 1 W whatever. The bypass capacitor is 10uF, 100V; I used what I had on hand which was a Vishay. All the resistors in the amp are Vishay RN60 or CCF60. CMF60 will also do. The RN60 are military rated for 1/4 W, but for civilian use, they are actually 1 W resistors. As for the coupling caps, any polypropylene cap of a reputable brand should do. I would use caps rated for at least 250V. The output coupling electrolytics are Nichicon UPW, chosen for their low impedance nature.

One note: Because I'm using a stepped attenuator, I was able to move the input coupling capacitor to before the attenuator. If you are using a regular potentiometer, I'd keep it on the wiper to avoid any DC from the grid of the 6AK5. It probably won't matter at first, but eventually, the pot will become scratchy from the DC.