Re: Re: My K-12M's doing the same thing
if you bring the current draw for the power section into line, the power trannie will run considerably cooler. radiated heat from the tubes is the least of your worries. it's core overheating that causes trannies to fail most quickly.
it's not so beefy so...reduce current draw, reduce heat.
here's why it pays for the newbie to get the good literature. one aspect of design will naturally lead you to another. each solution leads to more questions.
as usual..."everything is everything" hahha so when the next question comes up...your trusty Radiotron has a detailed discussion of the whys and wherefores...
GG said:
if you bring the current draw for the power section into line, the power trannie will run considerably cooler. radiated heat from the tubes is the least of your worries. it's core overheating that causes trannies to fail most quickly.
it's not so beefy so...reduce current draw, reduce heat.
here's why it pays for the newbie to get the good literature. one aspect of design will naturally lead you to another. each solution leads to more questions.
as usual..."everything is everything" hahha so when the next question comes up...your trusty Radiotron has a detailed discussion of the whys and wherefores...
I've been pondering the differences (besides the tube pinout) between the K-12m and the version AES sells which uses 11BM8 tubes instead of 11MS8 tubes. I really got the feeling based on my measurments at how hot the 11MS8s were running (all of them exceeding the specificed plate dissipation, and a visible glow to prove it!) that this kit was really not designed around the 11MS8 specificially - it just happened to be close enough and S5 Electronics has a stash of 'em.
After playing around a bit on the NJ7P Tube Database, the 6GV8 caught my attention. It has the same pinout as the 11MS8, similar (at least, they seem similar enough to me) characteristics, and a 7 watt plate dissipation like the 11BM8. Even better, I can get them locally for $3.50 each + tax. It looked like the only reason these tubes aren't plug-and-play is that the filament voltage is half that of the 11MS8.
I decied to take the plunge and buy 8 of whatever brands I could scrounge up locally of 6GV8 tubes (figured I'd save on the matching fee and just match 'em myself by current draw if they worked), and mod the filament supply so I could switch between running all four filaments in parallel (stock) or parallel filament pairs in series with each other to drop the voltage down to about 6.1v (worked!). I realize I don't have any experience with tubes and it would be pure dumb luck if these tubes worked at all, leave alone sounding good. I had $29 of NOS tubes on the line and what I figured to be the potential of letting the magic smoke out of my output transformers. It's a risk I was willing to take in the name of science. Okay, that and since I got my Jolida, the K-12m has just become something for me to tinker with.
The 6GV8 bounty included 6 RCA tubes, consisting of a pair made in Great Britian (Mullards!?), another pair made in Italy, and a final pair made in Holland. The remaining two tubes were Sylvania, made in India. I decided to go with the Great Britian and Italian pairs to start. Guess what - I'm glad to say the tube substitution experiment worked!
To my ears, they actually sound much clearer high end than the overheated 11MS8 tubes, and there doesn't seem to be any noises or osciliations present. Bass is still rolled off and flabby as it is with the 11MS8 tubes, I think the output transformers will have to take the blame on that. Gain seems slightly higher, and the headroom about the same, at least at the volume levels I was able to test. I was going to really crank this thing to see where distortion kicks in but the roomies are asleep and won't appreciate a racket at this hour. In leu of a "full blast test", I'm going to provide the bias measurements I took from each tube.
Sylvania / RCA 6GV8 tubes:
Cathode v = 195.0
Sylvania
Tube 1: 21.7ma * 195.0 = 4.2315w
Tube 2: 41.3ma * 195.0 = 8.0535w
Cathode v = 195.2
RCA (Holland)
Tube 3: 33.3ma * 195.2 = 6.50016w
Tube 4: 29.6ma * 195.2 = 5.77792w
RCA 6GV8 tubes:
Cathode v = 198.2
RCA (Italy)
Tube 1: 30.3ma * 198.2 = 6.00546w
Tube 2: 30.1ma * 198.2 = 5.96582w
Cathode v = 198.0
RCA (Great Britain)
Tube 3: 31.7ma * 198.0 = 6.2766w
Tube 4: 30.9ma * 198.0 = 6.1182w
Okay, it's still not biased to 70%, but it's no longer roasting the tubes full tilt. The best part: Cheaper tubes and no more glowing plates!
I'm going to follow-up to this post with some pictures.
After playing around a bit on the NJ7P Tube Database, the 6GV8 caught my attention. It has the same pinout as the 11MS8, similar (at least, they seem similar enough to me) characteristics, and a 7 watt plate dissipation like the 11BM8. Even better, I can get them locally for $3.50 each + tax. It looked like the only reason these tubes aren't plug-and-play is that the filament voltage is half that of the 11MS8.
I decied to take the plunge and buy 8 of whatever brands I could scrounge up locally of 6GV8 tubes (figured I'd save on the matching fee and just match 'em myself by current draw if they worked), and mod the filament supply so I could switch between running all four filaments in parallel (stock) or parallel filament pairs in series with each other to drop the voltage down to about 6.1v (worked!). I realize I don't have any experience with tubes and it would be pure dumb luck if these tubes worked at all, leave alone sounding good. I had $29 of NOS tubes on the line and what I figured to be the potential of letting the magic smoke out of my output transformers. It's a risk I was willing to take in the name of science. Okay, that and since I got my Jolida, the K-12m has just become something for me to tinker with.
The 6GV8 bounty included 6 RCA tubes, consisting of a pair made in Great Britian (Mullards!?), another pair made in Italy, and a final pair made in Holland. The remaining two tubes were Sylvania, made in India. I decided to go with the Great Britian and Italian pairs to start. Guess what - I'm glad to say the tube substitution experiment worked!
To my ears, they actually sound much clearer high end than the overheated 11MS8 tubes, and there doesn't seem to be any noises or osciliations present. Bass is still rolled off and flabby as it is with the 11MS8 tubes, I think the output transformers will have to take the blame on that. Gain seems slightly higher, and the headroom about the same, at least at the volume levels I was able to test. I was going to really crank this thing to see where distortion kicks in but the roomies are asleep and won't appreciate a racket at this hour. In leu of a "full blast test", I'm going to provide the bias measurements I took from each tube.
Sylvania / RCA 6GV8 tubes:
Cathode v = 195.0
Sylvania
Tube 1: 21.7ma * 195.0 = 4.2315w
Tube 2: 41.3ma * 195.0 = 8.0535w
Cathode v = 195.2
RCA (Holland)
Tube 3: 33.3ma * 195.2 = 6.50016w
Tube 4: 29.6ma * 195.2 = 5.77792w
RCA 6GV8 tubes:
Cathode v = 198.2
RCA (Italy)
Tube 1: 30.3ma * 198.2 = 6.00546w
Tube 2: 30.1ma * 198.2 = 5.96582w
Cathode v = 198.0
RCA (Great Britain)
Tube 3: 31.7ma * 198.0 = 6.2766w
Tube 4: 30.9ma * 198.0 = 6.1182w
Okay, it's still not biased to 70%, but it's no longer roasting the tubes full tilt. The best part: Cheaper tubes and no more glowing plates!
I'm going to follow-up to this post with some pictures.
Powercntrl said:
I decied to take the plunge and buy 8 of whatever brands I could scrounge up locally of 6GV8 tubes (figured I'd save on the matching fee and just match 'em myself by current draw if they worked), and mod the filament supply so I could switch between running all four filaments in parallel (stock) or parallel filament pairs in series with each other to drop the voltage down to about 6.1v (worked!). I realize I don't have any experience with tubes and it would be pure dumb luck if these tubes worked at all, leave alone sounding good. I had $29 of NOS tubes on the line and what I figured to be the potential of letting the magic smoke out of my output transformers. It's a risk I was willing to take in the name of science. Okay, that and since I got my Jolida, the K-12m has just become something for me to tinker with.
Okay, it's still not biased to 70%, but it's no longer roasting the tubes full tilt. The best part: Cheaper tubes and no more glowing plates!
haha i predict that any minute now, you'll be tweaking that jolida.
not sure where you got the 70% bias figure, but do bear in mind that it isn't a critical number, as operating points are a continuum, and the tube will function, but function differently (i.e. different power out, distortion spec and sound of course) at different points along the scale. the only hard and fast rule is the design maximum dissipation, and as you can see there are some well regarded commercial products that ignore this too!
i have found that it pays to question authority when it comes to commercial products, which are often designed for expediency, cheapness, easy parts availability, etc. rather than optimal operation. hell there's some MUCH pricier commercial offerings that have cheesy choices.
note that the database you are using lists an even lower suggested plate voltage and higher current for this new tube you are using, even though the design maximum plate voltage is the same. given the extra watt of dissipation, the designers chose to go lower voltage, higher current. in fact, this particular manual doesn't even rate this tube for class AB operation.
hmmm.
now, if you want to address the bass shy/flabbiness issue, which sounds like it's next on the list...your first step would be to get your hands on a scope and see what this amp is doing in the low frequencies. where does the waveform fall apart?
also, a suggestion re: layout...
you got a neat box, my man, but...that power trannie is awful close to the input section of your amp. now i am a fan of radical layouts, but there are some things you caint risk.
but of course your trusty radiotron has a chapter on layout, placement, chassis issues etc hehe
Powercntrl said:I really got the feeling based on my measurments at how hot the 11MS8s were running ... that this kit was really not designed around the 11MS8 specificially - it just happened to be close enough and S5 Electronics has a stash of 'em.
...the 6GV8 caught my attention. It has the same pinout as the 11MS8, similar (at least, they seem similar enough to me) characteristics, and a 7 watt plate dissipation like the 11BM8.
...I decided to go with the Great Britian and Italian pairs to start. Guess what - I'm glad to say the tube substitution experiment worked!Very cool mod! The more tubes this kit can be made to work with, the more interesting it gets! Will you install a switch to switch the filament between parallel and pairs in series? Is your power transformer noticably coolor (or warmer)? Napkin math suggests the filaments in the 6GV8 are drawing about 2w more (aggregate total) than the 11MS8s were.
Hmm. I'd be inclined to think it's one of many contributing factors. The B+ power supply might be worth watching on heavy bass, as is the room/speaker combo. To some extent, it's probably also a question of how much bass can a low-wattage signal really produce, since woofers can eat current for breakfast...
Shouldn't that be "Plate v = 195.0" ? Are you measuring plate voltage relative to ground or to the cathode? Also, I find it surprising that plates with such different currents would be at the same voltage-- are your measuring each plate's voltage relative to its cathode individually?
You're within spec on most of the pentodes, which is more than I can say for the moment. Dunno what's up with the sylvania pair there, though. But, yeah, this sounds great! Very cool initiative, to go and mod it for some totally different tubes!Okay, it's still not biased to 70%, but it's no longer roasting the tubes full tilt. The best part: Cheaper tubes and no more glowing plates!
Hopefully I can play with some more serious mods on my K-12M now that I've restored a set of monoblocks (which have no problem with low frequencies, TYMV). I've installed the pre-rectifier tranny snubber cap, I've upgraded all the signal caps to MKPs. Cathode resister upgrade with balance pot is definitely on the short list, now. Once that's in, I'll wish I had two meters so that I could balance the current easily in pair.
Of course, these restored 30w 7247/7868-based amps may be my new love for a while. I've spent more on a set of tubes and caps for it than my K-12M cost, so...
mono bloc with 11ms8
has anyone played with the little mono blocs..
KIT, MONOBLOCK TUBE AMPLIFIER
All tube design rivals sound of high end audio amplifiers costing 10 times as much! Designed by George Fathauer. Enjoy the thrill of high end audio without going broke and enjoy the satisfaction in building it yourself. Kit come complete will all parts and easy to follow instructions.
* Input Impedance: 100K ohms
* Output Impedance: 8 ohms
* Power Output: 8 watts per channel
* Frequency Response: 20 to 20,000 Hz
* Minimum Input for Full Output: 0.3 V
* Total Harmonic Distortion @ 1 W: <1%
Uses 11MS8 tubes included in the kit. Also included pine board for mounting PC board and transformers.
also from AE but these guys have the 11MS8
has anyone played with the little mono blocs..
KIT, MONOBLOCK TUBE AMPLIFIER
All tube design rivals sound of high end audio amplifiers costing 10 times as much! Designed by George Fathauer. Enjoy the thrill of high end audio without going broke and enjoy the satisfaction in building it yourself. Kit come complete will all parts and easy to follow instructions.
* Input Impedance: 100K ohms
* Output Impedance: 8 ohms
* Power Output: 8 watts per channel
* Frequency Response: 20 to 20,000 Hz
* Minimum Input for Full Output: 0.3 V
* Total Harmonic Distortion @ 1 W: <1%
Uses 11MS8 tubes included in the kit. Also included pine board for mounting PC board and transformers.
also from AE but these guys have the 11MS8
markgall said:
The impression i got from Voltseconds site was that the tube heaters are all run in series... that would mean a requirement for 44V -- now is there a 44 V winding or is there a big resistor somewhere dropping 66V from the 110V? A series resistor (or a bigger series resistor could burn off the extra 19V if you just swaped to 6BM8s.
Given that the existing power trafo is running hot, i'd cut the traces to all the heaters, wire tham in series using twisted wire, and add a separate 6.3V filament transformer.
dave
noisenyc said:
Apparently, it's now available for free download online as well-- the copyright of the 4th edition (1952) lapsed. Available in PDF, Radiotron Designers Handbook (4th edition)
Sweet.
noisenyc said:
Actually, I'm probably not going to bother trying to improve the bass response. I'm most likely going to end up using this amp with small bookshelf speakers after I finish tweaking it. I've got a small Sony scope and it might be interesting to see what's causing the bass to roll off - it's just not enough of a roll-off to warrant what might just end up being a hunt for expensive new iron.
It's actually quieter now than when it was mounted to the block of wood the kit came with. I figured if it was a problem, there's enough room to run some sheet metal between the power transformer and the PCB to cut down on noise. As it stands though, it's not picking up any noise that it seems to want to amplify - there's just the same mild persistant 60hz hum (at zero volume and doesn't get louder when you crank it up) that everyone else seems to be having with this kit. I haven't tried installing any of the mods that voltsecond did, so perhaps they'll cut down on the hum.
chromal said:
I've played a few test tracks designed for car audio systems through the amp to get an idea where the low frequency response rolls off. The amp's accuracy seems to plummet with the pitch, with the lowest notes before roll off, sounding very loose and flabby. I should mention I'm using the Jolida for comparison on the same speakers, so it's not exactly a fair fight. I obviously don't expect miracles out of a $140 kit amp. (at least, not until I'm done modding it
)Each PP pair shares a common 3 watt cathode resistor. Besides the fact it's too small and it's underrated, a single resistor means both tubes have the same cathode voltage regardless of one being more of a current hog. I took my voltage measurements from B+ right at the main capacitor and the cathode-connected side of the resistor. The current draw ratings aren't derived from a formula, they're the actual readings I took from my multitester between B+ and the plate.
They're just not matched. I got lucky with the RCA tubes - pairs of same manufacture date at the same plant that have really close characteristics, I'd imagine. I bought the tubes as "unmatched, untested" to save a few bucks. Those are probably the most closely matched "unmatched" tubes I've ever seen.
markgall said:
I cut the parallel trace feeding the second pair of filaments, and rewired it to switch between all parallel and parallel pairs in series. I'll include a schematic drawing below.
Okay, time for pictures:
Filament voltage mod schematic
6GV8 tubes installed (flash picture)
6GV8 tubes installed (ambient light only)
6GV8 tubes installed (complete darkness, pentode side)
K-12m initial test set up (Worthless computer speakers, wasn't taking any chances!)
I strongly recommend the snubber cap between the power transformer's secondary winding leads. I haven't gotten the parts for the UF diode or filament snubbers yet, but those would also be on the short .Powercntrl said:
I think maybe I could have said this more clearly-- my point is, 195v seems a bit high for your cathode voltage. That would mean 650mA is passing over that 2W 300ohm cathode resister (and that the two plates were each passing somewhere around half that). 195v would make more sense at the plate. As for your plate voltage measurements, wouldn't it be more accurate to get the voltage between the plate and the cathode, rather than B+ and cathode? Your output transformer probably has between 75-85 ohms of resistance (depending upon which side of the primary coil you're measuring across), so there is a voltage drop between B+ and the plate of any given tube, given the current flow. Another thing to keep in mind is that some current is still flowing through the OT when you do the 'shunt' current reading, unless you cut a trace or something...
Cool photos!
chromal said:
The lower current draw of the 6GV8s seemed to bump the cathode voltage up a bit (check my earlier post). B+ is around 210v. The mains voltage here is usually around 126VAC so it's highly likely you're seeing lower voltages if your mains aren't quite as high.
I don't really know what I'd do with that measurement. It seems like it would be inaccurate due to the cathode being only seperated from ground by the resistor and the plate from B+ by the output transformer. Since the output transformer is connected directly to B+ on the center tap, B+ to the cathode is accurate measurement of the voltage the tube is seeing.
I guess I could use another multitester to test the resistance of my multitester in miliamp mode, but if current is flowing through the output transformer instead of the meter - it would give lower test results. Between how hot the plates got with the stock tubes, I believe the dissipation was being exceeded and the measurements I made agreed with it. The 6GV8 tubes looked like they were running well within spec and even with a long exposure, I couldn't see any glow on the plates (which my camera actually does a better job seeing).
If someone with the 11BM8 version could do the same tests (please follow all appropriate HV safety precautions!), I'd be interested in seeing how hard that kit runs its tubes.
hey fellers.
the shunt method is close enough for rock and roll.
the plate voltage you use to calculate is measured at the plate connection of the tube. in cathode bias arrangement subtract the cathode voltage from the plate voltage figure to get the actual plate voltage the tube is seeing, and beyond that...don't sweat it!
the shunt method is close enough for rock and roll.
the plate voltage you use to calculate is measured at the plate connection of the tube. in cathode bias arrangement subtract the cathode voltage from the plate voltage figure to get the actual plate voltage the tube is seeing, and beyond that...don't sweat it!
Okay, here's my absolutely crazy idea to convert this amp to fixed bias:
Delete the 470k (R7, R6, R15, R16) resistors in series between the pentode grids and ground.
Add 1k resistors in series between the .22uf capacitors (C2, C3, C6, C7) and the respective pentode grids.
Replace the 300 ohm 3w cathode resistors (R8, R17) with 1 ohm resistors of sufficient wattage.
Apply pot adjusted negative voltage between the coupling capacitors and the newly added 1k resistors.
First on the agenda was finding a source of bias voltage. Since I don't have much room left inside my case as it is (and I think I'm going to need to move the fuse holder for more space), I decided on seeing what would happen if I used the transformer out of a 24 volt wall wart (after removing the non-transformer parts, of course!) backwards, attached to the 12v filament supply. It worked, I got roughly 57 volts AC out of the primary windings (time to rename them, I guess!). I'm a little worried about overtaxing the stock transformer, but I figure just getting rid of those hot cathode resistors will more than make up for the small toll the bias transformer will take.
Now, here comes the part where I'm not entirely sure. I decided to copy the biasing scheme Jolida uses (specifically, the one I found of their EL34 amp) - if the pot fails open, the tubes become biased very cold. IMHO, much better than losing bias voltage and watching glass melt all over the place. Since I don't have a lot of parts lying around, I simulated the circuit on the computer to see what it'll do. Turns out as I copied it, the Jolida circuit produces far too much negative voltage even with the pots turned all the way up. No good. I also realized I haven't a clue what 11MS8 or even the 6GV8 like to be biased at. Between reading google groups for posts about the 6BM8 and looking at the tube database specs for the 11BM8, it seems for this family of tube at around 200v B+, a range of around -10v to -18v should be right in the ballpark. I hope.
In the sim, I decided to tweak the resistors a bit and changed out the 30k pots for 10k ones which can actually be purchased at RadioShack for some instant gratification. I sorta picked a rectifier diode at random, it had a very low voltage drop so I see no reason it can't be substituted for something easier to get. End result: -9.211v to -22.18v according to the simulator. Next step is going to be actually building it (at least for one channel) and seeing what happens.
Delete the 470k (R7, R6, R15, R16) resistors in series between the pentode grids and ground.
Add 1k resistors in series between the .22uf capacitors (C2, C3, C6, C7) and the respective pentode grids.
Replace the 300 ohm 3w cathode resistors (R8, R17) with 1 ohm resistors of sufficient wattage.
Apply pot adjusted negative voltage between the coupling capacitors and the newly added 1k resistors.
First on the agenda was finding a source of bias voltage. Since I don't have much room left inside my case as it is (and I think I'm going to need to move the fuse holder for more space), I decided on seeing what would happen if I used the transformer out of a 24 volt wall wart (after removing the non-transformer parts, of course!) backwards, attached to the 12v filament supply. It worked, I got roughly 57 volts AC out of the primary windings (time to rename them, I guess!). I'm a little worried about overtaxing the stock transformer, but I figure just getting rid of those hot cathode resistors will more than make up for the small toll the bias transformer will take.
Now, here comes the part where I'm not entirely sure. I decided to copy the biasing scheme Jolida uses (specifically, the one I found of their EL34 amp) - if the pot fails open, the tubes become biased very cold. IMHO, much better than losing bias voltage and watching glass melt all over the place. Since I don't have a lot of parts lying around, I simulated the circuit on the computer to see what it'll do. Turns out as I copied it, the Jolida circuit produces far too much negative voltage even with the pots turned all the way up. No good. I also realized I haven't a clue what 11MS8 or even the 6GV8 like to be biased at. Between reading google groups for posts about the 6BM8 and looking at the tube database specs for the 11BM8, it seems for this family of tube at around 200v B+, a range of around -10v to -18v should be right in the ballpark. I hope.
In the sim, I decided to tweak the resistors a bit and changed out the 30k pots for 10k ones which can actually be purchased at RadioShack for some instant gratification. I sorta picked a rectifier diode at random, it had a very low voltage drop so I see no reason it can't be substituted for something easier to get. End result: -9.211v to -22.18v according to the simulator. Next step is going to be actually building it (at least for one channel) and seeing what happens.
An externally hosted image should be here but it was not working when we last tested it.
Capacitor Upgrades
I want to do some more mods on this project and then put it away for a while and do some other projects.
As suggested by Blackie, I will try removing the DC blocking capacitors C1 and C4.
I would also like to upgrade the coupling capacitors C2, C3, C6 and C7 from polyester film to film and foil capacitors. The only problem I forsee is fitting these rather large film and foil capacitors on to the already crowded PCB.
I would appreciate some advice on these two items:
1. Capacitors C10 and C11 are 33pf 100v ceramics. Is it worthwhile to upgrade these?
2. I am thinking of installing snubbers on the electrolytic capacitors? Is this worthwhile pursuing?
I have the following potential capacitors to complete items 1 and 2 above.
Several 820 pf 5% 630V Polystyrene
Handfull of 0.1 uF 2% 250V Metalized Polypropylene Film
Two .002uF 10% 1000V orange drops
Two 0.003uF 10% 1000V Sprague Black Beauties
I prefer to use the polystyrene ones since I have about 100 of them kicking around
Can I use the 820 pf to replace the 33pf ceramics or is the difference too much?
For snubbing, will the polystyrene work well or should I use the 0.1 uF Metalized Polypropylene Film?
Thanks,
Gio.
I want to do some more mods on this project and then put it away for a while and do some other projects.
As suggested by Blackie, I will try removing the DC blocking capacitors C1 and C4.
I would also like to upgrade the coupling capacitors C2, C3, C6 and C7 from polyester film to film and foil capacitors. The only problem I forsee is fitting these rather large film and foil capacitors on to the already crowded PCB.
I would appreciate some advice on these two items:
1. Capacitors C10 and C11 are 33pf 100v ceramics. Is it worthwhile to upgrade these?
2. I am thinking of installing snubbers on the electrolytic capacitors? Is this worthwhile pursuing?
I have the following potential capacitors to complete items 1 and 2 above.
Several 820 pf 5% 630V Polystyrene
Handfull of 0.1 uF 2% 250V Metalized Polypropylene Film
Two .002uF 10% 1000V orange drops
Two 0.003uF 10% 1000V Sprague Black Beauties
I prefer to use the polystyrene ones since I have about 100 of them kicking around
Can I use the 820 pf to replace the 33pf ceramics or is the difference too much?
For snubbing, will the polystyrene work well or should I use the 0.1 uF Metalized Polypropylene Film?
Thanks,
Gio.
Powercntrl said:
looking good, build it and all will be revealed.
the amp will gain some voltage swing to the plates due to the cathodes now being at ground potential...meaning, more wattage.
keep your eye on those maximums. you can cheat plate voltage within reason with good tubes but keep the dissipation in line or...glowy glowy plates!
but do remember the message of the ancient ones...low and hot might sound better than high and cold.
Go for it. I put in some Soren MKP 0.22uF caps and didn't have any problems, though I had to do some creative lead bending...I should probably inject some silicone cement beneath each cap to cut on vibration... Do watch out with C6 and R14-- if the cap you use is physically larger and/or axial, it will have a tendancy to have one lead get uncomfortably close to R14, B+ voltage resistor. Maybe if you pre-fit with some heat-shrink tubing on the C6 lead near R14.. (I kind wish I had that foresight...)
Pending changes on my own K-12M... well, W.J. Ford Surplus / Test Equipment Canada, in Ontario, was kind enough to match some 11MS8s pentode sections for me. I've got a couple of wirewound 250-500 ohm resistors, so I'll probably play with cathode voltage. I haven't found a pot I'm quite comfortable with for cathode balance, though I'm increasingly tempted to give each cathode its own 5w pot.. Then again, a CCS or grid bias might be the way to go...
In truth, I've been mostly listening to my 7247/7868 Bogen monoblocks (so very sweet), but I'm looking for some small bookshelves to couple with the K-12M and stick in my bedroom, so...
I've very interested in a) running the tubes a little less hot, b) going for a triode strapped pentode configuration. c) Building a custom chassis for the little guy...
Well, a 100 Ohm 10w for balance, or alternately two 50 ohm 5w. But, on the other hand, maybe I don't care, now that I've got supposedly balanced tubes... I think I plan to experiment with some different sizes. I also want to build a chassis that makes soldering/desoldering less of a pain. (e.g.: access from bottom.)
Not necessarily. VoltSecond substituted CCSs for R8 and R17. Read about it near the bottom of this page. Very interesting stuff. I'll probably want to try all three approaches and see if I favor one!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.