| Ty_Bower |
I've come across a couple old (1945 vintage) 6L6GA tubes. They're beautiful with their ST bottles and smoked glass sides. I'd like to build a stereo single ended amp out of them. In the same box came a few 6X5GT rectifiers. I'd like to use them in the power supply. I think I'm going to need a separate 6X5 to feed each 6L6GA?
Any hints, or pointers to a recommended schematic? |
|
|
| danlaudionut |
You could adapt Bas' SE EL34 schematic for 6L5 use.
Just change the cathode resistor to 720R.
Yes you will need one 6X5 per channel.
DanL |
|
|
| Tom Bavis |
| One easy choice - use TubeLab's Simple SE circuit. You COULD use the PCB if you use a single rectifier like 5V4G. With the 6X5, yes you'd need two. |
|
|
| Ty_Bower |
TubeLab's Simple SE is appealing. I've taken a good look at it. I like building kits where someone else has already done the design and proven the circuit is sound. It doesn't hurt that many others have already built it as well. I'm not completely married to the 6X5, it's just that I happen to have them.
What kind of power transformer would I need to run a pair of 6L6GA in the Simple SE board? George seems to like the Allied 6K7VG, but I'm thinking the B+ would end up too high for the 6L6GA. |
|
|
| tubelab.com |
| quote: | | George seems to like the Allied 6K7VG, but I'm thinking the B+ would end up too high for the 6L6GA. |
YES, big time! I plugged a pair of crusty old 6L6GA's into one of my amps, and they REALLY didn't like it. They were on the verge of runaway, and would slip into uncontrolled red glow mode whenever they wanted to. The old 6L6 (metal) and the 6L6GA's want about 380 to 390 volts with cathode bias to remain in spec. I have used 400 to 410 without issue, but the Allied 6K7VG can make 440 to 460 with a 5AR4. I have been using some surplus HP power transformers for amps like this, but I don't have any more left. They are 650 VCT.
I have a Hammond 272JX (600VCT) on my bench. Hammonds are known for running high. I will hook it up and try it over the weekend if I can find the time. |
|
|
| Eli Duttman |
Guys,
The AnTek AN-3T325.pdf toroidal power trafo should do nicely for this project. Wire the 6X5 heaters in series and connect them to the 12.6 VAC winding. |
|
|
| Ty_Bower |
I'm wondering if the AnTek AN-3T275 would be a better fit, and just use a 5AR4 rectifier. My problem with the AnTek transformers is that I never know where to buy them, other than wait for one to show up on eBay. Can I just call the guys and order one? Do they sell direct? I guess I should try sometime...
I'm also looking at the Hammond P-T270HX. AES has them on their website for $63. But, I've read that sometimes the Hammond 200 series have a mechanical hum and that kind of stuff really turns me off. :( |
|
|
| Eli Duttman |
Ty,
Use the 6X5s with the AnTek trafo, as a 5 VAC winding is not present.
You most certainly can order directly from AnTek. Like me, they're in NJ. ;) Look here. |
|
|
| Ty_Bower |
Does this look anything close to correct?
 |
|
|
| JoshK |
| The filaments should not be tied together to the cathodes. Seperate them in the schematic and things will start to look better. |
|
|
| audiowize |
I might offer a word of advice on this idea, since I've had something similar brewing in my head for a long time. 6L6's are great tubes, but what's also nice is that there are the following equivalents:
12L6
25L6
35L6
50L6
etc.
With a simple 2 pole rotary switch and a couple of cheap transformers (a 50L6 needs very little heater current), you could have essentially a limitless supply of NOS power tubes. For example NOS 50L6's at Antique are $6!
-Paul |
|
|
| Ty_Bower |
Ah, I see. I was thinking of the 5AR4. I guess I should always check the datasheets first. Is this better? It seems it may be more customary to connect the plates of the twin diodes together as shown, rather than the way I did in my first attempt. I'm not sure it makes any difference either way.
I believe I need to think harder about the expected B+ voltage and my intended bias point. The 6L6GA is only rated for 19 watts, and my PSU schematic is showing 21.3. Of course, the 355 volt number could be way off. I'm not exactly sure how to calculate the B+ using multiple rectifier tubes. |
|
|
| Tom Bavis |
| quote: | Originally posted by audiowize
I might offer a word of advice on this idea, since I've had something similar brewing in my head for a long time. 6L6's are great tubes, but what's also nice is that there are the following equivalents:
12L6
25L6
35L6
50L6
etc.
-Paul |
No, these are NOT equivalents! It IS true in most cases that equivalent tubes have similar numbers... but in this case the 6V equivalent is 6W6. The 35L6 is rated a little lower than the others - 6W6, 12W6, 12L6, 25L6, 50L6 are very close except for heater voltage. A power pentode with maybe 10W plate dissipation, 150V maximum screen voltage. |
|
|
| audiowize |
| Ah, tis true. Perhaps a switchable cathode resistor would be a good idea also. The 6W6 is a fine sounding tube, but power becomes an immediate issue. |
|
|
| Eli Duttman |
| quote: | Originally posted by Ty_Bower
Ah, I see. I was thinking of the 5AR4. I guess I should always check the datasheets first. Is this better? It seems it may be more customary to connect the plates of the twin diodes together as shown, rather than the way I did in my first attempt. I'm not sure it makes any difference either way.
I believe I need to think harder about the expected B+ voltage and my intended bias point. The 6L6GA is only rated for 19 watts, and my PSU schematic is showing 21.3. Of course, the 355 volt number could be way off. I'm not exactly sure how to calculate the B+ using multiple rectifier tubes. |
That PSU schematic looks OK. You can bring the B+ rail voltage down by making the 1st filter cap. 4.7 μF. Keep the ripple level low by using a 6 H./200 mA. Triad C-14X choke from Allied.
John Atwood's experience suggests that using both plates inside a single envelope in parallel, as your drawing indicates, is best.
Look at the 6L6G data sheet. The 19 W. plate dissipation limit is design center. You can safely go as high a 20 W., without trouble. Don't simultaneously "push the envelope" of both max. dissipation and anode voltage.
Also, keep the fact that g2 makes a contribution to power O/P, when UL mode is employed, in mind. |
|
|
| Ty_Bower |
Questions about the Simple SE board: What's the biggest (diameter) capacitor that can fit at C1 and C2? Mouser obsoleted the part suggested on the TubeLab parts list (C1) and there is no Mouser suggested part for C2.
Could I use these Mouser parts? They seem kinda expensive, though.
647-UPW2W470MRD
647-UPW2W101MRD
|
|
|
| danlaudionut |
BTW I just love my SE 6L6G amp.
I use the Tungsol Re-Issue 6L6G
which is a 6L6GC in a coke bottle glass.
I also use the Genelex Re-Issue KT66
both are great but different sounding.
My voltages are too high for your use.
I drive mine with Mullard 6P5GTs with
an unbypassed cathode resistor.
Choke loaded and grid choke for 6L6G.
LCLC filter and 5V4GT rectifier.
Silk 5K OPT and TVC input.
DanL |
|
|
| Eli Duttman |
Ty,
I've got an assembly idea. :D You have discrete rectifiers on each side of the full wave CT PSU. IMO, you will have fewer headaches mounting the pair of 6X5s close to, but off of, the PCB. Ground all 4 6X5 anodes. Connect a 6X5 cathode to each end of the 600 VCT. Take the "raw" B+ from the CT of the rectifier winding. Now, you have a single line to the PSU filter and the space the 5AR4 socket usually goes into can be occupied by filter capacitance. |
|
|
| Ty_Bower |
Like this?
I like it. I had already figured the pair of 6X5 had to be mounted off the board anyway. I'm trying to figure out how to make the chassis mount rectifiers and the PC mount power tubes "look" right in the finished amp, but then again I've never been much of one to worry about the final cosmetics.
Do you think the 600VCT transformer is the right size to get about 350 VDC, 55 mA per tube (110 mA total)? |
|
|
| Eli Duttman |
The topology looks correct, but the 1st filter cap. is too big. Use a 4.7 μF. part. An even smaller value initially may be a good idea. If the rail is too "short" under load, add a little capacitance to the 1st position.
| quote: | | Do you think the 600VCT transformer is the right size to get about 350 VDC, 55 mA per tube (110 mA total)? |
By starting with a smallish 1st filter cap., you should be able to tweak the rail voltage exactly. |
|
|
| Ty_Bower |
Hmm... I've unleashed my Google-fu and turned up this article:
http://archive.ampage.org/threads/1...ier_tube-1.html
In a nutshell, he's got an amp with a 325-0-325 PT, a single 6X5, and a 51 mA load. He's seeing about 370 volts at the plate, and running about 18 watts dissipation. I think that might be where I'd like to go, and I have to wonder if I'd be happier with a 650VCT PT instead of the 600VCT. I'd hate to build the thing and find out I made the wrong choice.
I guess a lot depends on the power transformer's own source resistance. I don't really know what it is for the Antek toroids, but I've got to expect it's pretty low. |
|
|
| Tom Bavis |
I wonder about the heater cathode limit in the circuit above... if the voltage divides equally (not a given...), there will be 300V AC from each 6X56 cathode to heater. Connecting the cathodes together in the "conventional" connection (plates to ends of HV, CT as negative) will put both cathodes at the same voltage, and there won't be any voltage to the floating heater winding. Would probably be OK with damper diodes, but the insulation in 6X5s is not that good.
A quick sim in PSUD says that the 300V transformer will give about 330VDC - close enough, I think. But 325V might be closer to what you wnat... |
|
|
| Ty_Bower |
| quote: | Originally posted by Tom Bavis
I wonder about the heater cathode limit in the circuit above... |
Good point. I'll look more closely.
| quote: | | A quick sim in PSUD says that the 300V transformer will give about 330VDC - close enough, I think. But 325V might be closer to what you wnat... |
How do you get PSUD to properly simulate the voltage drop when using a pair of 6X5? |
|
|
| Ty_Bower |
| quote: | Originally posted by Tom Bavis
...in the circuit above... there will be 300V AC from each 6X56 cathode to heater. Connecting the cathodes together in the "conventional" connection (plates to ends of HV, CT as negative) will put both cathodes at the same voltage, and there won't be any voltage to the floating heater winding. |
In the suggested circuit, you'll have 300 VAC H-K, because you're applying 300 VAC to the cathode? And with the "normal" method the H-K eventually ends up at zero volts because the cathodes are at DC and the heater winding floats up to meet it? |
|
|
| Eli Duttman |
Tom,
I don't think there will be a problem with 6X5 heater to cathode insulation breakdown. Look at the data sheet. The design center limit is 450 VDC. :) |
|
|
| tubelab.com |
| quote: | | Questions about the Simple SE board: What's the biggest (diameter) capacitor that can fit at C1 and C2? |
C1 can go as big as 22.5 mm the hole spacing is 10 mm. C2 can go to 35 mm, with a 10 mm hole spacing.
| quote: | | Could I use these Mouser parts? They seem kinda expensive, though. |
They claim a low impedance, which is a good thing, maybe that is why they are expensive. DigiKey lists a 150 uF 450 volt Panasonic for $4.32 each. P13913-ND
They will fit the board, but the 6X5 doesn't like much over 5 uF for the input cap. I have experimented with the 6X5 since I have a few hundred of them. A 47uF input cap, and a transformer with a low winding resistance (like the Antek) WILL lead to a fireworks display inside the tube. The 6X5 has a higher voltage drop than the typical 5AR4. If you want to use the circuit that you show use a small (4.7 uF 450 volt) leaded capacitor, for C1 and a big (150 uF 450 volt) cap for C2.
450 volt capacitors often show up on the surplus market cheap. Now that I have acquired 25 of these I will let everybody in on it. They work in a Simple SE like you are planning, don't try to use them with the big transformers, they will spill their guts. Keep them under 400 to 420 volts. It is the same as the DigiKey part.
http://www.allelectronics.com/make-...ICITOR/-/1.html
| quote: | | Do you think the 600VCT transformer is the right size to get about 350 VDC, 55 mA per tube (110 mA total)? |
I will hook my Hammond 272JX up tomorrow and see what I get. It is a 600VCT transformer, but these are Hammond volts, which are about 650 to 700 real volts. I get 390 volts of B+ in a Tubelab SE using thie transformer. All of the Anteks that I have here are much bigger 4T360 and 4T400, or smaller 1T200 than what you need. |
|
|
| Ty_Bower |
| quote: | Originally posted by tubelab.com
...but the 6X5 doesn't like much over 5 uF for the input cap. I have experimented with the 6X5 since I have a few hundred of them. A 47uF input cap, and a transformer with a low winding resistance (like the Antek) WILL lead to a fireworks display inside the tube. |
Thanks for the informative and helpful reply. It's got me wondering if I'd be happier in the long run if I just give up on the 6X5 idea and go with the 5AR4. I've already got the 6X5 for "free", but I'm sure I've got an extra 5AR4 laying around somewhere too.
Of course, the 5AR4 is going to need a 5 volt winding from somewhere... |
|
|
| Tom Bavis |
| quote: | Originally posted by Eli Duttman
Tom,
I don't think there will be a problem with 6X5 heater to cathode insulation breakdown. Look at the data sheet. The design center limit is 450 VDC. :) |
Agreed... it's rated +/-450V ... but which is better... 300 or more VAC heater - to - cathode... or essentially zero? LOTS of '30s-'40s Zenith power transformers died from H-K shorted 6X5s - ask any old-time radio repairman!
I just simulated it with 55 mA - half the load. Transformer drop will be double with 110 mA, but it will be small in any case. Make sure your input cap is good for 150 mA or so ripple current - not all small caps will be. You can get the ripple current from PSUD, but a first approximation is same as DC load current. |
|
|
| chrish |
Don't have much time, just skimming this thread while taking a dinner break from a university assignment...
I have ordered some custom made power transformers from Edcor 2x120 volt primary, 330-0-330v 200mA, 6.3v 5A, 5v 3A secondary for use with 6L6 etc. As the design work has been paid for by me, these transformers should now be on their database and no design fee for custom order. I have not received them yet, so cannot comment on their build quality. Have been happy with the quality of their output transformers though. Oh, price quoted was $63.51
Hope this helps.
Cheers,
Chris |
|
|
| Ty_Bower |
| quote: | Originally posted by tubelab.com
I will hook my Hammond 272JX up tomorrow and see what I get. It is a 600VCT transformer, but these are Hammond volts, which are about 650 to 700 real volts. I get 390 volts of B+ in a Tubelab SE using thie transformer. |
I've got a mystery transformer running in a PP EL84 amp. Total load is about 140 mA, and the transformer measures about 330-0-330 with no load. B+ on this amp ended up around 400~410 VDC, depending on line voltage (which tends to run high at my house). The power supply is solid state rectified, with a CRC filter (110 uF, 50 ohm, 110 uF0.
I'm thinking even the 272JX is going to be too hot for 6L6GA. I'll probably need to go with something like the 270HX. |
|
|
| Mr. Triatic |
Here's mine 6L6G amplifier..
(6P3S = 6L6G) |
|
|
| tubelab.com |
| quote: | | It's got me wondering if I'd be happier in the long run if I just give up on the 6X5 idea and go with the 5AR4. |
That would be my choice since I have toasted a few 6X5's and I am still using the same 5AR4 that I started with about 6 years ago. I decided that the 6X5 just isn't really suited for power amp use except for maybe a 6V6 amp. It does impart a unique sound to a 6V6 guitar amp.
| quote: | | I'm thinking even the 272JX is going to be too hot for 6L6GA. I'll probably need to go with something like the 270HX. |
I dug up a pair of golden oldie 6L6G's. Not matched, not even the same brand, but they work. I tried 3 different power transformers. All tests were done in triode mode with a 5K ohm OPT since thats what I had hooked up to the amp. You could get about 20% more power with a 3K ohm OPT, and about 50 to 70% more in UL mode. The choke is a Triad C-14X with a DC resistance of 150 ohms. I stuffed it inside the shell of an old power transformer. The Simple SE board is assembled as per the parts list including 560 ohm cathode resistors.
First the 272JX. The 300 - 0 - 300 volt transformer makes 325 volts AC here. The B+ before the choke is 385 volts, and 365 volts after the choke with a total current of 133 mA. The plate voltage is 360 volts. The cathode voltages are 26.7 volts and 27.0 volts, so the tube itself sees 333 volts. Tube current is 48 mA which works out to 16 watts plate and screen dissipation. Power output at 5% distortion is 3.1 watts. The sound is clean, and sounds similar to other 6L6 type amps that I have built, but not quite as loud or dynamic.
I don't have a 270HX, but I do have a 270FX, so I hooked it up. The 275 - 0 - 275 volt transformer makes 301 volts AC here. The B+ before the choke is 345 volts and 326 volts after the choke with a total current of 124 mA. The plate voltage is 322 volts. The cathode voltages are 24.4 volts and 24.2 volts, so the tube itself sees 298 volts. Tube current is 43 mA which works out to 12.7 watts plate and screen dissipation. Power output at 5% distortion is 2.6 watts. The sound is clean, but the bass is a bit weaker than before.
I have some Allied 6K56VG's which are made by Hammond, so I hooked one up. The 275 - 0 - 275 volt transformer makes 288 volts AC here. The B+ before the choke is 334 volts and 317 volts after the choke with a total current of 118 mA. The plate voltage is 312 volts. The cathode voltages are 23.3 volts and 22.8 volts, so the tube itself sees 289 volts. Tube current is 41 mA which works out to 11.6 watts plate and screen dissipation. Power output at 5% distortion is 2.3 watts. The sound is kind of wimpy. If I was going to run these tubes on this low of a voltage I would bump up the current and use a 3K ohm OPT.
OK, Which would I choose? I would go with the big guy, the 272JX, but that is just me, and I have one. You could use the 272HX or even the 272FX to lose a few more volts. You could use a 5U4G for the look, and lose a few volts over a 5AR4. Technically 333 volts is over the screen grid voltage rating, but this is usually OK in triode or UL mode. The 270HX would probably give you afew more volts than my 270FX, and this may be enough for your application especially if you are using a lower impedance OPT.
I recommend reducing the 10K ohm resistors in series with the CCS chips to 1K ohm on lower voltages. The CCS can run out of headroom on low B+ voltages with some 12AT7's.
While I had all this good stuff hooked up, I stuffed some other tubes in the amp. I tried some 6Y6's because they were in the same box as the 6L6G's. They really didn't like any of this, but they have a max spec of 200 volts. I usually use the Allied 6K7VG with 6V6's and 6K6's ant they work great, but I wanted to see how the tubes handled the big transformer.
The Allied 6K56VG makes 290 volts AC on the lighter load of a 6V6. The B+ before the choke is 335 volts and 320 volts after the choke with a total current of 100 mA. The plate voltage is 316 volts. The cathode voltages are 20 volts and 19.5 volts, so the tube itself sees 296 volts. Tube current is 36 mA which works out to 10.5 watts plate and screen dissipation. Power output at 5% distortion is 1.8 watts. The sound is excellent. These are ideal conditions for the 6V6.
The 272JX makes 326 volts AC on the lighter load of a 6V6. The B+ before the choke is 386 volts and 370 volts after the choke with a total current of 105 mA. The plate voltage is 366 volts. The cathode voltages are 23.3 volts and 22.4 volts, so the tube itself sees 343 volts. Tube current is 41 mA which works out to 14 watts plate and screen dissipation. Power output at 5% distortion is 2.5 watts. These are over the limit conditions for the 6V6. Fender ran the 6V6 this hard in the Champ and other amps, so good tubes should live. The sound is excellent. Bass and dynamics are better than the "normal" voltages.
Would I put my NOS pair of 6V6G's in here? No way! I don't mind toasting a Russian tube or two. Somewhere I have a NOS bulk pack of Sylvania waver base 6V6GTB's. I bought the box about 10 years ago, unopened. Upon opening the box, I found about 25 of the tubes had lost their vacuum. Since then about 20 more have turned white. I will toast some of these, since I can't trust them to live even in a nice cozy cardboard box.
The 6K6's bias up like the 6V6, but sound a bit different. |
|
|
| Ty_Bower |
| quote: | Originally posted by tubelab.com
All tests were done in triode mode with a 5K ohm OPT...
OK, Which would I choose? I would go with the big guy, the 272JX, but that is just me... |
Thanks for all your testing, and the detailed writeup. I plan on using a 5K OPT. You've convinced me I probably want the 272JX. What's your typical AC line voltage? Mine runs around 124 most days.
Is there a template for cutting out a chassis to fit the Simple SE board? Just the edges of the board, and the locations of the tube holes and mounting screws is all I really need. I've had good luck making Visio drawings if I've got the dimensions.
 |
|
|
| tubelab.com |
| quote: | | What's your typical AC line voltage? Mine runs around 124 most days. |
I forgot to post it. It was 122.2 and varying up and down about .5 volt during the testing. The outside temp was about 90F this afternoon. The voltage varies as the AC units cycle. 8 houses share a pole transformer, and mine is the furthest house from the transformer. During the winter my voltage can hit 124 or 125. On an August afternoon it can go down to 115.
| quote: | | Is there a template for cutting out a chassis to fit the Simple SE board? |
I have a PDF of the silkscreen layer of the PC board. This is what I use for a drilling template. Print it, tape it to the chassis, and drill (or center punch) right through it. It is too big to attach here, and the forum doesn't seem to have a way to add attachments to its email, so send me an email and I will reply with the file. |
|
|
| Ty_Bower |
| quote: | Originally posted by Ty_Bower
Could I use these Mouser parts?
647-UPW2W470MRD
647-UPW2W101MRD
|
I took a closer look. One of those caps has 12.5mm lead spacing, which isn't going to work well with the Simple SE board. I think maybe these would be a better choice:
647-UPW2W470MRD, 47uF, 450V, $3.67/ea.
647-UPT2W121MRD, 120uF, 450V, $4.40/ea.
This is also an option, but is currently backordered:
647-UPT2W151MRD, 150uF, 450V, $5.81/ea.
Keep in mind that all these Nichicon caps are only rated 450V, so don't try to use them with the "big" power transformer. |
|
|
|
