Hi everyone, this my first original design that I believe is worth publishing. I have recently bought ten 6J1 tubes and wanted to make something with them, so I tried with this. I am very happy with the result, this sounds nice with both guitar and bass, and the sound can be set from very clean to gentle overdrive.
(https://s26.postimg.cc/74bs2q0lh/vintage_voyage_schematic.jpg) (https://postimg.cc/image/74bs2q0lh/)
If anyone is interested I will put part list here, as well as PCB.
I'm intrigued by the Mode control. What does that do? Is it sort of like a triode/pentode blend? How would you describe the difference it makes to he sound?
Yes, it is actually triode/pentode blend. Triode sounds very clean, while pentode is a lot louder and breaks up with overdrive easily.
QuoteYes, it is actually triode/pentode blend. Triode sounds very clean, while pentode is a lot louder and breaks up with overdrive easily.
You have been busy.
Cool idea.
It's not an original idea - found it on web, but it gave me an opportunity to have two-in-one pedal, using cheap pentode tubes and a pot :) . I hope I will have time to box it during this week, and then record some demo sounds and prepare PCB files for sharing. But if anyone's anxious to start building before that, have in mind that R14 can be used to "dose" overdrive. I've chosen 180k with idea not to have too much of it, but increasing this value will cetainly add more. I have also added 10k between treble W lug and gain CW lug, just in case.
HEJ :)
Any News here?
Not yet - I noticed that Gain and Treble pots produce scratching when I turn them, so I replaced them, but scratching is still here. I also tried different value pots (in case the same value pots are all faulty), but it is there again. So I presume there is some problem in my design, I will try to figure out what.
That would indicate DC on the pots. Either multiple caps leaking (C8 and C4, maybe C5), or grid leakage from the second EF95. Measure with your meter, you'll figure it out.
I measure 0.55V on gain 2&3, as well as on treble 2&3. What should I do? Put blocking cap between gain pod and tube 2?
Thanks thermionix, putting 1u between gain lug and 2nd tube's grid fixed the problem! The question is: why is there some grid leakage? Bad tube, or bad design? Shall I leave that 1u cap on my PCB?
Could the second stage anode be idling at too low DC volts compared to the screen volts? You have larger plate load resistor there and maybe the screen supply needs to be from a voltage divider rather than a dropper to hold its voltage lower?
Indeed, but I also have 2M2 on stage 2 screen, so I though it will be enough. I have measured DC voltage at stage 1 grid right now, and it is 0.64V. Inaudible since I don't have a pot there, but it seems that I have a problem there too!
Could it just be the condition of the tubes? I have a 2-channel F2B clone in the works and out of the 4 grids, just one has a few 100mv on it (which I'm not worried about). As you may know, all of those triodes have very similar DC conditions. The tubes are old British Mullard ECC83 and I've given them regulated DC heater supply I measure at 6.4v.
I've been reading....
http://www.tubebooks.org/Books/Atwood/Tomer%201960%20Getting%20the%20Most%20Out%20of%20Vacuum%20Tubes.pdf
...and from that I gather that something as simple as a grid wire too close to the hot cathode, can of itself make the grid go a little positive (as it emits some electrons making it so).
Quote from: rankot on April 22, 2018, 04:08:30 PM
Thanks thermionix, putting 1u between gain lug and 2nd tube's grid fixed the problem! The question is: why is there some grid leakage? Bad tube, or bad design? Shall I leave that 1u cap on my PCB?
Some tubes have grid leakage, as Jim says. Your grid still needs a DC path to ground, which your gain pot was doing before you put the cap in. I would remove the cap and find a tube that doesn't have the grid leakage.
Or I can add 1M grid leak between grid and that cap, so in case I get another leaky tube it will still work fine?
However, I hope it is just leaky tube, not leaky design, because I like the sound.
If you are measuring positive voltage on a control grid, it is likely that the EF95 / 6AK5 tube is gassy. In some designs, tubes are self-biased by using a large resistor of around 10 megohms from grid to ground. The grid intercepts electrons given off by the cathode and is biased negative by about a volt. This is how the 12SQ7 / 12AV6 / 12AT6 triodes were biased in the classic All-American 5 radios (named because they had 5 tubes) that was the majority of tube radios built between 1939 and 1970.
If the bias is positive, the electrons are hitting gas atoms in the tube and knocking an electron off, making them positive ions. Since the cathodes in this design are self-biased, they are at a positive voltage with respect to the grid, so the positive ions head for the nearest more negative electrode, which is the grid, and charge it positive. There doesn't appear to be a design fault that would cause a positive grid although a leaky coupling capacitor could do it as well for the second tube. Check by disconnecting the coupling capacitor and if the grid voltage becomes zero or negative, replace the capacitor.
I've measured DC again, and this is what I got:
-0.63V on first tube's grid
-0.86V on second
So both DC voltages are negative (no AC signal in when measuring).
Maybe they were negative in previous measuring, but I didn't pay attention to sign.
I already put 1u blocking capacitor between stage 2 grid and volume pot's lug, so there is some DC on grid (-0.86V), but not on the other side of capacitor. My conclusion is that capacitors between stage 1 and 2 are OK (all brand new, Wima MKS), and I believe that those are just normal biasing voltages? False alarm due to wrong measurement?
That sounds like what you would expect.
But the fact is that scratching stopped after I've installed that 1u capacitor between gain pot and grid stopper, so most probably the second tube is leaky. Never mind, this cap doesn't affect the sound, so I will leave it there.
So, this is the final version:
(https://s26.postimg.cc/vijr8mnud/vintage_voyage_schematic_F2.jpg) (https://postimg.cc/image/vijr8mnud/)
Few remarks:
- C8 is not necessary, but I like to use it there because it blocks HV DC to the tonestack, so all the other caps between that one and stage two don't have to be high voltage rated.
- C7 and C14 are there to provide space for different sized output caps, actually only one of them is needed (I use 47n and it's quite fine).
- L1/L2/L4 If you can find 150uH/3A inductor, use it; if not you can use two 330uH/1A (they are smaller).
- L3 is low current rated small inductor, to provide some filtering.
- U1 is NE555.
- I have read somewhere that R25 should be 1 or 2W rated, but I don't think it's really important.
- C15/C18/C21 may not be necessary, I left them there just in case, they are ment to filter radio frequencies.
- D3/D4 can probalby be left out, they protect MOSFETS from static electricity.
- Two LEDs because I wanted to have them under tube sockets.
- If you can buy 7.5R/2W resistor to drop voltage for heater you can use it, my local store sells only 15R/1W, so I used two in parallel.
- B+ shall be 120V±10%.
- all capacitors rated 400V can be 250V, I used 400's because of local availability.
If I forgot something, please ask.
I will soon have some PCBs for sale, so if anyone is interested, PM me and include e-mail address in message, because I couldn't answer to some PM's in the past, due to full mailboxes of members who contacted me.
Any chance for sound samples?
T.
Soon, I hope! :D
I used boost converter found here: https://threeneurons.wordpress.com/nixie-power-supply/ (https://threeneurons.wordpress.com/nixie-power-supply/). Mine is running at 31kHz, but it maybe better to run it on higher frequency? I don't know, I'll try it.
I have changed few things due to suggestions by Dexter in a Facebook group: Ct6 is now 680p instead of 2n2, and L1 is 47uH (L2/L4 are 100uH if you use them). Also, with this settings, boost converter works more stable and smaller size drum-shaped inductors can be used.