• Disclaimer: This Vendor's Forum is a paid-for commercial area. Unlike the rest of diyAudio, the Vendor has complete control of what may or may not be posted in this forum. If you wish to discuss technical matters outside the bounds of what is permitted by the Vendor, please use the non-commercial areas of diyAudio to do so.

Spud: what, again!?!

Has anyone had any experience with vacuumtubes.net?

I have bought some tubes from them over the years. As with any vendor cleaning out their stock for cheap, expect some old boxes in good to fair condition and a few dead tubes. I did buy 1000 tubes from them about a month ago when the cheap tubes sale started. They looked OK, but I haven't had the time

Might buy a handful. At $0.25 each, why not?

I got 500 tubes for $0.35 each. Where do you find tubes for $0.25 other than a hamfest?

Question: what kind of power output would you be looking at in Push-Pull configuration?

I'll know for sure when I breadboard it up. It also depends on how hard you push the tubes. If you want a guess, I will assume class AB1 and say 8 to 10 watts with the 6GF7, 15 to 20 with the 6KY8, and maybe 25 to 30 with the 6LR8. More might be available in AB2 but that really complicates things with mosfets and negative voltages.
 
...If you want a guess, I will assume class AB1 and say 8 to 10 watts with the 6GF7, 15 to 20 with the 6KY8, and maybe 25 to 30 with the 6LR8. More might be available in AB2 but that really complicates things with mosfets and negative voltages.
If the complication is elegant and small, then sure, why not marry the new with the old for something unique.

A P-P with 20W on hand will be fine for my modest listening space. I have a Musical Fidelity A1 solid state integrated amp. It's rated for 20W per side. With my loudspeakers (Axiom M3's at 95 dB efficiency), volume at 10 o'clock is plenty to blow the room away...and cook eggs or boil water...that thing gets GD'd hot!!!

So, 10W plus is gravy if I were to go the P-P. How about the Spudd (the lower case D's are capital P's rotated 180 degrees) as a name? Ok, I'm putting the cart before the horse. Maybe I should finish my SSE first, before dreaming of my next dream.

I believe I am in for one SpudSE version as a kit as a Canada Day present (optimism).
 
I know you'd mentioned how you might get the 21V for the filaments of the 21LR8. Are you considering small, separate filament transformers (like the small Antek's that spit out 20-40V) to do the job? That might make the job of using the tubes you've discovered useful in the NovarSpud.

Originally I was planning multiple filament windings on the same toroid, but it comes down to whether it's better (cost, hassle and weight wise) to do that or to just stick to the 6LR8. I'm thinking multiple power transformers (whether permanent or swappable) is probably going against the small and simple aspects of the amp, which is one of the things that I really like about it. That decision will be left until later down the track.

I have bought some tubes from them over the years. As with any vendor cleaning out their stock for cheap, expect some old boxes in good to fair condition and a few dead tubes. I did buy 1000 tubes from them about a month ago when the cheap tubes sale started. They looked OK, but I haven't had the time

Thanks.

I got 500 tubes for $0.35 each. Where do you find tubes for $0.25 other than a hamfest?

Sorry, that was a typo, meant to say $0.75 a tube. Would be the same list you were looking at I believe (10+ @ $0.75 each, 100+ @ $0.50 each, 500+ @ $0.35 each)

I'll know for sure when I breadboard it up. It also depends on how hard you push the tubes. If you want a guess, I will assume class AB1 and say 8 to 10 watts with the 6GF7, 15 to 20 with the 6KY8, and maybe 25 to 30 with the 6LR8. More might be available in AB2 but that really complicates things with mosfets and negative voltages.

So, 10W plus is gravy if I were to go the P-P.

My thoughts exactly. Especially if it looks even in the ballpark of 25 watts for 6LR8's in PP, that's more than enough for my needs. Also, for the sake of keeping the circuit a little simpler, I don't think going to AB2 is really necessary.
Besides, I'd be more than happy with the 5ish watts of the original SE design.
But I guess a lot of this comes down to long term interest in the design. Regardless of what I'm looking for from the amp, if moving toward a few more watts would be enough to get more people interested in the PCB's, then why not?
 
I know I have a few flavors of the LR8 variety in a box some place

I will probably use a simple zener diode and mosfet regulator since it is universally available, and I have managed to blow the LR8 when used as a screen regulator. Granted I was playing my guitar through the amp, but I test all my amps with a guitar preamp since it is a worst case test...and I use the SSE and likely this board as a guitar amp.

Would be the same list you were looking at I believe

Yes. I got 100 25DN6's for a 50 watt sweep tube amp. 500 6HB6's for some new low cost projects, and 100 of each of the tubes that go in a little cheap guitar amp I made for the $100 amp challenge. The 6HB6's were a wide variety of brands, some in pretty worn boxes, but they were 35 cents. A few of the 25DN6's were old and in worn boxes. All of the other stuff looked good.

I don't think going to AB2 is really necessary.

It isn't going to happen here. The idea here is for a cheap low cost amp. I should have some time tonight to fire up the 5 year old board and see what smokes. Then it's time to hack it up and see what I can make out of it.

I notified out warehouse landlord to apply our previously paid last months rent to June, so the clock is now ticking.....I have 30 days to get everything out.....This means I will be rather busy this month, leaving little time for tinkering.
 
Sorry, I didn't know you were coming. I had a collection of old radios so Sherri and I dragged them all up to the Florida Antique Wireless Guild swap meet in Stans parking lot. Sherri sold all of them by 9:30 and we had to drive back to Fort Lauderdale, so we left. We have friends in the Orlando area so we will be back up there in a few weeks. Are you local to that area?

I'm living in Tampa right now, I retire from my current job today and will be getting another one soon, so I'll end up either staying here in Tampa or moving to Dayton. About a 50% chance of either right now.
 
I fired up the board after its 5 year sleep. It is still alive and works about the same as it did then. No testing was done with the 6GF7 since I don't have any left. Got to empty the warehouse, can't keep them all.

Question: what kind of power output would you be looking at in Push-Pull configuration?

I wired the board up to regulated supplies for plate and screen and used a transformer as a phase inverter. My power output guestimates were pretty close. A screen regulator is needed for pentode operation. Otherwise too much negative bias is needed to keep the current down to a reasonable value.

The 6KY8A makes just over 20 watts when biased to the edge of dim redness in a totally dark room. 15 watts is probably a good number. B+ was 300 volts. Screen was 150 volts.

The 6LR8 hit 32 watts under the same conditions. B+ was 350 volts. Screen was 150 volts. They should be fine at 20 to 25 watts.

The 6KY8 is much smaller than the 6LR8 but the plate dissipation ratings are not too different. The 6LR8 requires much more drive than the 6KY8 and once I convert the input stages over to an LTP PI there might not be enough gain for full power from a CD player. I'll know more once I try it. Don't know when though.
 
I don't know - I would bypass all this socket brouhaha and use chassis-mounted sockets that would take some abuse - the board would be mounted via spacers somewhat below the sockets with most of the components pointed down - the most sane arrangement, and I would string wires up from the PCB to the sockets. With the right arrangement of holes on the PCB, it wouldn't matter if you used a XLU8 or an XLR8, as long as there's some sort of table that tells one how to wire a socket for a particular tube to the board.
Stan has some XXLU8/LR8 tubes in his dollar list ( I forget which), so this approach could work. I don't usually buy kits/boards anyway, so I'm just mouthing off here...
 
I wired the board up to regulated supplies for plate and screen and used a transformer as a phase inverter. My power output guestimates were pretty close. A screen regulator is needed for pentode operation. Otherwise too much negative bias is needed to keep the current down to a reasonable value.
OK...so triode and UL; great. Looking over the SimplePP schematic from another thread leads to a question about a SpudPP. Do you foresee the phase inverter tranny for Push-Pull operation OR will the SpudPP schematic look fundamentally the same as the SimplePP?

I would guess the latter case...but I suck at multiple choice exams.
The 6KY8A makes just over 20 watts when biased to the edge of dim redness in a totally dark room. 15 watts is probably a good number. B+ was 300 volts. Screen was 150 volts.

The 6LR8 hit 32 watts under the same conditions. B+ was 350 volts. Screen was 150 volts. They should be fine at 20 to 25 watts.

The 6KY8 is much smaller than the 6LR8 but the plate dissipation ratings are not too different. The 6LR8 requires much more drive than the 6KY8 and once I convert the input stages over to an LTP PI there might not be enough gain for full power from a CD player. I'll know more once I try it. Don't know when though.
The power output ratings are quite encouraging. For anyone with efficient full-range, single driver speakers or two-ways, 20W should blow away even a medium sized room.
I don't know - I would bypass all this socket brouhaha...
I tried a FreePCB layout to make a 3cm by 3cm adapter board that would hold the Duodecar socket and adapt it to the Novar. Alas, the pin circles are a bit too close to allow a circle within a circle that would avoid chassis mod and only require dropping the PCB down a few millimeters to drop in the adapter board. For any fans of the Wankel engine, I have noted that adopting a cardioid shape like the Wankel combustion chamber (i.e. two intersecting circles) with pins 4,5 on the Novar close to pins 1,12 on the Duodecar might just work. With Fenris' board service, you could procure 10 such adapter boards for $10, which would be nice and open the Tubelab Spud to the compactron twin of the 6LR8, depending on supply in your market. Even though I am spatially challenged, I will keep plugging away at that one. Keystone makes 0.040" round profile micropins that can be swaged into a board (no solder types I've seen). This could work...stay tuned.
...and use chassis-mounted sockets that would take some abuse - the board would be mounted via spacers somewhat below the sockets...I don't usually buy kits/boards anyway...
From my perspective, a board addresses some of the things a novice like me would likely miss (stray signals intermingling, wire size, component placement issues, etc.) in a point-to-point layout. Knowing Tubelab would produce a design that has been prototyped, tweaked and generally ridden hard and put away wet before it's ever up for sale is a real bonus. This will reduce the chance for a loss of enthusiasm after the build by increasing the likelihood of success for DIY'ers.

And hey, with the current exchange rates US$ to CDN$, even though Canada just stopped minting pennies, at least my Maple Leaf 2 cents is about the same as two Lincoln's.
 
I would guess the latter case...but I suck at multiple choice exams.

E) None of the above.

In order to maintain electrical compatibility with the dual SE option a single driver stage must remain connected to each output stage. There is exactly one way to do this. Currently there is a resistor and a capacitor to ground on the cathode of each input tube. You add a jumper so that the two cathodes can be tied together, then add a CCS to ground. This creates an LTP. For the SE version you place the R's and C's (or LED's) and leave out the CCS and jumper. For the P-P version you install the CCS and jumper, leave out the R's and C's. This also requires a capacitor in series with the input, which is replaced with a jumper in the SE version.

I have modified SSE boards in this manner to achieve 60+ watts, but it requires a bit more work since the CCS plate loads must be removed.

I don't know - I would bypass all this socket brouhaha and use chassis-mounted sockets that would take some abuse

If I wanted to use chassis mounted sockets, then I would just PTP wire the entire amp. The circuit is pretty simple.

a board addresses some of the things a novice like me would likely miss

That is really the only reason for doing this board. There have been several people succesfully build an SSE using PTP wiring from my schematic. There have been a few PTP SSE builds that didn't work including one guy in England who got rather annoyed when I couldn't help him fix the hum in an amp I didn't build, can not see, and had no clue how he laid it out or wired it.

Keystone makes 0.040" round profile micropins that can be swaged into a board (no solder types I've seen).

I have made adapters before. You want pins that have to be staked into the board. If you rely on solder alone, the joint will eventually get brittle and go intermittent with repeated use.
 
E) None of the above...
E ?!?!?! What happened to C and D that I didn't even think of?

Oh, man. I just flashed back 26 years to my physics 101 mid-term exam where the professors created a multi-choosey exam; 12 questions with most pre-loaded with "common mistakes" that the great unwashed will often make in their leap to an answer. In many cases, my fellow students were beaten by statistical "monkey" that would randomly choose and get more than 25% correct OR the one guy (this is true) had a total brain fart before the exam, created a pattern and hit that "one in a million shot" and got 12 out of 12. Lucky S.O.B.
I have made adapters before. You want pins that have to be staked into the board. If you rely on solder alone, the joint will eventually get brittle and go intermittent with repeated use.
The ones I was looking at are mechanically swaged into vias in the board of a specific hole diameter (a few thou over). So, these would be mechanically held on both sides of the board. There are also the stake type/press fit, but I think I'm happier DIY'ing a swage fit than thinking about the possible creaking and cracking sounds of a pressfit going awry if the board is not supported enough from behind...I suppose some kind of mini-arbor press thingy might do the trick. First things first though, can Novar/Duodecar circles intermingle in a nearly aligned fashion. Then, I split the atom at my kitchen sink and cold fusion will soon follow.
 
I have made adapters before...
The trick for the 6LU8/6LR8 adapter appears to be that although the guts of these tubes are the same, the pinouts look like pick-up-sticks.

In trying to keep the filament pins pointing in the same direction (Novar 4,5 to the Duodecar 1,12), I am seeing some sort of weird Kama Sutra type intermeshing circles that could work without too much offset and the trace lengths are not too long from the grid stoppers and cathode resistor for the pentode section and the bias diode for the triode section.

Still might work if I can find an ideal angle for clocking the Novar tube pin circle that would align with the socket below. Making progress anyway. It's raining here and it feels a bit like The Shining...all duodecars and no play makes Jack a novar boy...
 
The trick for the 6LU8/6LR8...the pinouts look like pick-up-sticks.
OK...have an adapter if anyone is interested. The attached ZIP file holds a FreePCB layout file for a two-layer board 5mm by 5mm (can be considerably smaller if needed) that allows the soldering in of a Duodecar socket and press-fit or swage micro pins for the Novar.
In trying to keep the filament pins pointing in the same direction (Novar 4,5 to the Duodecar 1,12)...Still might work if I can find an ideal angle for clocking the Novar tube pin circle that would align with the socket below...
The ideal angle appears to be minus 5 degrees or so for Pin 1 of the Novar at 3 o'clock (i.e. "zero degrees" in trig parlance).

The pin circles interlace. The tricky bit is that the Novar has a tube pin circle of 17.45mm and the Duodecar 19.1mm. I found Duodecar sockets on the web with an advertised PC mount pin circle of 20mm to 22mm and used that as a basis. Clearly, if a wider footprint for the PC mount side of the Duodecar were available from a reliable source, the B9E Novar circle of pins would be much, much easier to fit within the Duodecar.

On to something else as useless. Good practice in FreePCB and high school trigonometry though.
 

Attachments

  • Duodecar2NovarAdapter.zip
    3.4 KB · Views: 61
OK...have an adapter if anyone is interested...
Oh...just realized from an earlier post we might also want to know the transcription of Duodecar to Novar pinouts 6LU8 to 6LR8:

Novar B9E (pin#) Function Duodecar B12C (pin#)
1 Triode Cathode 11
2 Pentode Grid 6 & 7
3 Pentode Cathode 9
4 Filament A 1
5 Filament B 12
6 Pentode Plate 4
7 Pentode Screen 8
8 Triode Plate 2
9 Triode Grid 10
-- Not Connected 3 & 5

Trust I got this right from the respective data sheets. I suppose short runs of insulated wire to a chassis mount socket would permit the use of a B12C socket with the Novar based PCBoard.
 
Socket Adapter

To all:

Attached is a better version. I found Peter Millett's Duodecar (6LU8) socket source (among a few), which has a PC mount pin circle of 25.4mm (1.0 inches). With the 17.45mm specification for the B9E socket pin circle for the Novar (6LR8). This is a slightly more elegant arrangement than my earlier post/attachment.

Nice to have the option for 9 or 12 pin for the Spud so one can choose based on available stocks from your preferred suppliers.

Working it the other way (9 to 12 pin adapter) is a much bigger challenge, in my estimation.

Time for a Pikesville and water...TTFN
 

Attachments

  • Duodecar2Novar_Adapter_r0.zip
    6.3 KB · Views: 64
E) None of the above.

In order to maintain electrical compatibility with the dual SE option a single driver stage must remain connected to each output stage. There is exactly one way to do this. Currently there is a resistor and a capacitor to ground on the cathode of each input tube. You add a jumper so that the two cathodes can be tied together, then add a CCS to ground. This creates an LTP...
Since I am trying to learn about the circuit designs along the way, are you talking AC coupled or DC coupled Long Tail Pair as described here: link

So, IF I have deciphered the DC coupled LTP correctly, the mod to be made would use the Triode sections of both 6LR8 tubes to create the differential pair phase inverter and voltage gain stage that would pass onto the two pentodes in the output stage in a push pull setup.

More coffee this AM might make this even clearer...or muddier if I am off base on my decryption of the circuit. Why oh why can't this be a fuel or ignition map from a powertrain control module? THAT I understand.
 
If this board goes ahead, when may be available?

It looks like some version of this board will happen. When, is a hard question to answer. I have a full time engineering job that consumes most of my time. Tubelab is a part time venture that barely covers its expenses. To fix this, I have decided to eliminate the warehouse rent by closing out the warehouse. I have already informed the landlord that June is my last month. This gives me 25 more days to remove all my stuff. I will be very busy moving stuff until July.

After that I need to fix by messed up website, re-do some of the instructions for the SSE to match the new board revision that is coming.

I still need to finish the design work and testing on the Spud SE board, re-do the layout, get some boards made, build, test and document them, and create a set of build instructions.

The reality of all of this work says that a fully finsihed board with a parts kit and a proper set of instructions is still several months away if no major interruptions come my way. A few boards may be available to early adopters sooner. These are for the porpose of finding potential problems, and verifying the instructions and parts kits.

are you talking AC coupled or DC coupled Long Tail Pair......use the Triode sections of both 6LR8 tubes to create the differential pair phase inverter and voltage gain stage

The concept is to use the two triode sections to create an LTP similar to the one shown in your link. Exactly how it will be biased hasn't been determined yet. The DC coupled variant has the grids at DC ground and a negative voltage on the CCS feeding the cathodes. This requires a negative voltage source. The AC coupled variant has the cathode CCS tied to ground and the grids fed from a posithve voltage source. This requires a second set of coupling capacitors in the signal path. I will breadboard both variants and test them, but I would prefer not to add more caps.