OK, my father, now nearing 70, an ex-avid HAM, gave me 7 matched GE 6JE6C (6LQ6), in their original box and all, that he had been saving to build a linear amp at some point in the 70's.
I thought, since he's not doing anything with HAM anymore, that I'd try to use these to build him a nice hi-fi amp that he could enjoy. Now, I know these could bring up to $70.00 on ebay or more, and that there are probably other tubes that could be used to produce a comparable or better result for cheaper, but he's dutifully saved them in pristine condition for the last 30 years ... they should be used for something he'll enjoy. Call me sentimental
I'm asking for your help, in recommending an existing project (hopefully with schem and maybe layout guide) that could be [slightly] modified to use these as the power output tubes. I have some experience building SE tube guitar amps and have recently built a K12 kit, but haven't really designed any tube circuits myself.
Anthony
I thought, since he's not doing anything with HAM anymore, that I'd try to use these to build him a nice hi-fi amp that he could enjoy. Now, I know these could bring up to $70.00 on ebay or more, and that there are probably other tubes that could be used to produce a comparable or better result for cheaper, but he's dutifully saved them in pristine condition for the last 30 years ... they should be used for something he'll enjoy. Call me sentimental
I'm asking for your help, in recommending an existing project (hopefully with schem and maybe layout guide) that could be [slightly] modified to use these as the power output tubes. I have some experience building SE tube guitar amps and have recently built a K12 kit, but haven't really designed any tube circuits myself.
Anthony
It's going to be tough to do anything with the 6JE6's screen grid voltage maximum of 220 (that's ABSOLUTE maximum). Unless someone has torture tested your specific model a la Tubelab and found that they can survive in an AF situation, my only advice is the following:
Sell the tubes in 3 matched pairs on eBay with pictures, preferably some proof that they've been matched, and use the money from these tubes to help you buy some EL34's or something, a MUCH more suitable AF power tube. Then you can choose from among dozens of proven schematics using the EL34.
I respect you wanting to use the tubes your father gave you but if you value sound quality, you might want to play it safe and use some real audio tubes.
Sell the tubes in 3 matched pairs on eBay with pictures, preferably some proof that they've been matched, and use the money from these tubes to help you buy some EL34's or something, a MUCH more suitable AF power tube. Then you can choose from among dozens of proven schematics using the EL34.
I respect you wanting to use the tubes your father gave you but if you value sound quality, you might want to play it safe and use some real audio tubes.
It's possible to make an excellent amp from these tubes (I use that family of tubes routinely), but it's one you'd have to design for yourself- the sort of topology and drive suitable for sweep tubes is NOT conventional and you won't find ready-made projects with full documentation like you will with more "normal" designs.
Soren's advice is good. If you want to keep the "look" of plate caps, use the "super" 6BG6 being sold by SND in any circuit which calls for 6L6GC, 5881, or 7027.
Soren's advice is good. If you want to keep the "look" of plate caps, use the "super" 6BG6 being sold by SND in any circuit which calls for 6L6GC, 5881, or 7027.
You could use a 100V zener diode with parallel elec. cap. from plate to screen (for voltage dropping from plate to screen) to make a triode somewhat similar to two 2A3 tubes in parallel (which does have audio schematics available for it). The 6JE6C will then have a triode Mu of 3, gm of 10,500 uMhos at 130 mA, and Rp of 200 Ohms at 130 mA and max. plate diss. of 30 Watt.
The 2A3 specs:
Mu 4.2
gm 5250 at 60 mA
rp 800 at 60 mA
15 Watt plate diss.
Doubled up (parallel) 2A3 spec:
Mu 4.2
gm 10,500 at 120 mA
Rp 400 at 120 mA
30 Watt plate diss.
So the 6JE6C could use a bit lower output xfmr. primary Z than the doubled up 2A3 due to the lower Rp and will need a bit more grid 1 drive voltage due to it's lower Mu than the 2A3.
The 100 V zener drop for the screen grid should eliminate any issues of screen dissipation. Check the data sheet for the 6JE6C and you will see that screen current draw is insignificant when its 100 V below the plate voltage.
A schematic for a 1+1 tube push-pull 2A3 could be adapted by adjusting the output xfmr. primary impedance to between 1/2 and 1/4 the schematic value (and doubled wattage rating), and increasing B+ voltage for all the tubes by 30% to provide enough grid drive. (check that other tubes are not over their voltage specs. then. ) Grid biasing voltages and autobias resistors will need changing for the new setup to get suitable idle currents in the tubes. You will have to experiment with the 6JE6C biasing voltage in triode mode, start with 50% extra (above 2A3 grid bias value given) and slowly reduce until wanted idling cathode current is reached.
Ideally one would want a variable B+ and grid bias power supply to set this up using the 6JE6Cs, so as to find the best operating point.
Don
The 2A3 specs:
Mu 4.2
gm 5250 at 60 mA
rp 800 at 60 mA
15 Watt plate diss.
Doubled up (parallel) 2A3 spec:
Mu 4.2
gm 10,500 at 120 mA
Rp 400 at 120 mA
30 Watt plate diss.
So the 6JE6C could use a bit lower output xfmr. primary Z than the doubled up 2A3 due to the lower Rp and will need a bit more grid 1 drive voltage due to it's lower Mu than the 2A3.
The 100 V zener drop for the screen grid should eliminate any issues of screen dissipation. Check the data sheet for the 6JE6C and you will see that screen current draw is insignificant when its 100 V below the plate voltage.
A schematic for a 1+1 tube push-pull 2A3 could be adapted by adjusting the output xfmr. primary impedance to between 1/2 and 1/4 the schematic value (and doubled wattage rating), and increasing B+ voltage for all the tubes by 30% to provide enough grid drive. (check that other tubes are not over their voltage specs. then. ) Grid biasing voltages and autobias resistors will need changing for the new setup to get suitable idle currents in the tubes. You will have to experiment with the 6JE6C biasing voltage in triode mode, start with 50% extra (above 2A3 grid bias value given) and slowly reduce until wanted idling cathode current is reached.
Ideally one would want a variable B+ and grid bias power supply to set this up using the 6JE6Cs, so as to find the best operating point.
Don
If your Dad is like me, then after years of listening to SSB, CW, etc., he is probably not a particularly critical listener and it's not going to take a super amp to make him happy.
You could triode strap his tubes in the conventional manner and just use a low plate voltage consistent with the screen limitations. I think the screens would probably take more than the 220 volts listed as the maximum rating. TV sweep tubes were intended for low plate voltage, high peak cathode current applications, anyway. There is a published 6T9 amp that ran in Nuts and Volts a few years ago (the schemtic is on the web at sparetimegizmos.com, IIRC) that might be useful to you. A 6T9 is just a small pentode sweep tube and a triode in a compactron bottle.
Or run them in pentode mode with a true screen supply in the power supply. Take another look around your dad's stuff. If he was planning on using these tubes in something other than a grounded grid amp, he may have a suitable power transformer with appropriate plate and screen voltage secondaries already in his shack. If not, power transformers with an extra secondariy just for screen voltage are still available at hamfests. I bought about a ten pounder a couple of weeks ago for $1.
Win W5JAG
You could triode strap his tubes in the conventional manner and just use a low plate voltage consistent with the screen limitations. I think the screens would probably take more than the 220 volts listed as the maximum rating. TV sweep tubes were intended for low plate voltage, high peak cathode current applications, anyway. There is a published 6T9 amp that ran in Nuts and Volts a few years ago (the schemtic is on the web at sparetimegizmos.com, IIRC) that might be useful to you. A 6T9 is just a small pentode sweep tube and a triode in a compactron bottle.
Or run them in pentode mode with a true screen supply in the power supply. Take another look around your dad's stuff. If he was planning on using these tubes in something other than a grounded grid amp, he may have a suitable power transformer with appropriate plate and screen voltage secondaries already in his shack. If not, power transformers with an extra secondariy just for screen voltage are still available at hamfests. I bought about a ten pounder a couple of weeks ago for $1.
Win W5JAG
Pentode mode works best when g2 B+ is well regulated. A stack of 2X VR105s neatly takes care of the sweep tube screen grids, while running off the plate B+ supply. No extra PSU "iron" or windings is needed.
NFB of some kind is mandatory in pentode mode, in order for the amp to have an adequate damping factor. Loop NFB or Bandersnatch's "exolinear" topology are options. If small signal circuitry gain is adequate, the 2 methods can be combined to excellent effect.
NFB of some kind is mandatory in pentode mode, in order for the amp to have an adequate damping factor. Loop NFB or Bandersnatch's "exolinear" topology are options. If small signal circuitry gain is adequate, the 2 methods can be combined to excellent effect.
hey-Hey!!!,
The 6JE6C is a *FINE* audio final. Brilliant actually. Stuff like EL34's are able to be used with the usual short-cuts audio folks are wont to use...this doesn't mean that they're better.
Rig it as a pentode, with a regulated screen supply. Perhaps a shunt regulated circuit with a 0D3/VR150. Put B+ in the neighborhood of 400V for an AB1 amp with 4k a-a output TX, and 300V with a 6k OPT for class A.
The E-Linear circuit will need an output with Ultra-Linear taps, preferably more than one pair so you can tweak and adjust. I have a similar amp in the build que using the 26LW6( another big TV sweep tube ).
You can also get Ultra-Linear operation of the 6JE6C with a tertiary winding OPT like the Acrosound TO-350. Remember it is the g2-cathode voltge varying at some fraction of plate voltage that defines U-L; you don't need B+ level taps in the plate winding. B+ to the CT, cathodes to the third winding and g2 to an appropriate fixed voltage. There's more of these outputs, Chicago B0-14 for one. The cathode winding will also give some needed output Z reduction.
cheers,
Douglas
The 6JE6C is a *FINE* audio final. Brilliant actually. Stuff like EL34's are able to be used with the usual short-cuts audio folks are wont to use...this doesn't mean that they're better.
Rig it as a pentode, with a regulated screen supply. Perhaps a shunt regulated circuit with a 0D3/VR150. Put B+ in the neighborhood of 400V for an AB1 amp with 4k a-a output TX, and 300V with a 6k OPT for class A.
The E-Linear circuit will need an output with Ultra-Linear taps, preferably more than one pair so you can tweak and adjust. I have a similar amp in the build que using the 26LW6( another big TV sweep tube ).
You can also get Ultra-Linear operation of the 6JE6C with a tertiary winding OPT like the Acrosound TO-350. Remember it is the g2-cathode voltge varying at some fraction of plate voltage that defines U-L; you don't need B+ level taps in the plate winding. B+ to the CT, cathodes to the third winding and g2 to an appropriate fixed voltage. There's more of these outputs, Chicago B0-14 for one. The cathode winding will also give some needed output Z reduction.
cheers,
Douglas
Following along the trail blazed by Bandersnatch leads us to the Plitron PAT-4070-SSCR O/P trafo. The "4070" has a separate 40% screen grid winding and the 4 KOhm (end to end) impedance Doug suggested. Plitron O/P trafos are expensive, but look at the bandwidth you get.
A quick look at the 6JE6-C data sheet suggests that the splitter/driver used in "El Cheapo" might be adequate for this job.
Let's see, Plitron O/P trafos, Mullard made 'T7s, and those nice GE made O/P tubes yield a "killer" amp. Better hoard some more 6JE6-Cs. This "puppy" will do nicely for decades.
A quick look at the 6JE6-C data sheet suggests that the splitter/driver used in "El Cheapo" might be adequate for this job.
Let's see, Plitron O/P trafos, Mullard made 'T7s, and those nice GE made O/P tubes yield a "killer" amp.
Better hoard some more 6JE6-Cs. This "puppy" will do nicely for decades.hey-Hey!!!,
A quick look at the GE 6JE6C data sheet shows a pentode plate Z of 5k5 Ohms. Hardly more than a 211. Certainly less than the 211 running low current and high voltage... Still, for this particular design, running a high-slope diagonal across the plate characteristics at half the a-a load is the way to go. Keep it Class A bias; drawing an AB load line is a lot more difficult.
Also, see Pete Millett's write up on his SE E-Linear amp, the driver stage should be a high-Z element. Cascodes or pentodes; it doesn't do well with straight triodes.
cheers,
Douglas
A quick look at the GE 6JE6C data sheet shows a pentode plate Z of 5k5 Ohms. Hardly more than a 211. Certainly less than the 211 running low current and high voltage...
Still, for this particular design, running a high-slope diagonal across the plate characteristics at half the a-a load is the way to go. Keep it Class A bias; drawing an AB load line is a lot more difficult. Also, see Pete Millett's write up on his SE E-Linear amp, the driver stage should be a high-Z element. Cascodes or pentodes; it doesn't do well with straight triodes.
cheers,
Douglas
Bandersnatch said:hey-Hey!!!,
A quick look at the GE 6JE6C data sheet shows a pentode plate Z of 5k5 Ohms. Hardly more than a 211. Certainly less than the 211 running low current and high voltage...
Also, see Pete Millett's write up on his SE E-Linear amp, the driver stage should be a high-Z element. Cascodes or pentodes; it doesn't do well with straight triodes.
cheers,
Douglas
Doug,
Using the Plitron "4070" trafo and "mundane" UL topology, with 400 V. on the plates, 150 V. on the screen grids, and -20 V. on the control grids of the 6JE6Cs seems to yield a reasonable Class "A" operating point. FWIW, that's well within the capability of the "El Cheapo" splitter/driver. No line stage needed, when a CDP is the source.
I know you are not fond of loop NFB around the O/P trafo. Frankly, I don't think there is anything to worry about at the 9% of the speaker drive voltage that the "El Cheapo" topology uses.
I sorta recall an ultralinear amplifier design in Glass Audio magazine some years back that used these type of horizontal output tubes. It was called the "Snubber" I think, since it used snubber circuits on a Hammond 1650T transformer to clean up some UL transients. I will try to find it, but my magazine collection is in extreme disorder after moving.
Don
Don
Eli Duttman said:
The 400V plate voltage and 2k load would need a lot of idle current to achieve Class A. Assume something like 325V of swing, and we're looking at needing 160 mA idle current per tube to keep it from cutting off.
We'd also need a lot less than 150V on g2. Likely something like 100V would yeild a peak g1=0V plate current of ~340 mA.
cheers,
Douglas
hey-Hey,
For a class A operating point, it looks like a higher load a-a should be used. Say ~7k a-a, or a 3k5 load for each plate. 330V of B+ will leave about 300V of swing towards the top horizontal part of the curves. You set this g1=0V line with the g2 voltage. Since we've got a 30W plate, and there will be little g2 current, we'll run the plate to 30W. That's 90 mA. Then we need the g2 voltage so that g1=0V line comes to ~180 mA, or twice the idle current. For the pentode, plate current= C*g2_voltage^1.5. At 200V plate voltage, and 125V on g2, C is ~.0005. So for our 200 mA peak leaves g2 to be ~55V. Try 60 or so for the first try....
So, 7k a-a, B+ of 330, g2 of 60, and 90 mA of plate current you'll have a class A operation. Does it look like there are very many real Class A amps out there now?
cheers,
Douglas
For a class A operating point, it looks like a higher load a-a should be used. Say ~7k a-a, or a 3k5 load for each plate. 330V of B+ will leave about 300V of swing towards the top horizontal part of the curves. You set this g1=0V line with the g2 voltage. Since we've got a 30W plate, and there will be little g2 current, we'll run the plate to 30W. That's 90 mA. Then we need the g2 voltage so that g1=0V line comes to ~180 mA, or twice the idle current. For the pentode, plate current= C*g2_voltage^1.5. At 200V plate voltage, and 125V on g2, C is ~.0005. So for our 200 mA peak leaves g2 to be ~55V. Try 60 or so for the first try....
So, 7k a-a, B+ of 330, g2 of 60, and 90 mA of plate current you'll have a class A operation. Does it look like there are very many real Class A amps out there now?
cheers,
Douglas
Just a quick question,....What kind of power output do you need ....I assume you need to make 2 amps for Stereo operation?? If so, would you put both channels in the same chassis with a common power transformer....??
Do you really need Class A P-P ???
This should be straight forward design.....simply make screen regulator to keep the screens happy, under 200V....then choose a plate load depending on the applied B+ .....
Chris
Do you really need Class A P-P ???
This should be straight forward design.....simply make screen regulator to keep the screens happy, under 200V....then choose a plate load depending on the applied B+ .....
Chris
Bandersnatch said:
Doug,
A VR105 would take good care of the screen grid supply.
The design max. plate dissipation limit is 30 W. With 400 V. on the plate, I make the Class "A" idle current to be no greater than 75 mA.
I wonder how much abuse the 6JE6C's control grid can take. Is a Class "_2" operating point on the horizon? L_RD knows the 6JE6C is not cathode current constrained.
FWIW, the full 2 VRMS I/P from a CDP and more will not drive the "El Cheapo" splitter/driver into clipping. Add Tubelab's "Power Drive" or circuitry along the lines suggested in MOSFET Follies and positive grid current is not problematic. IMO, the superior IRFBC20 MOSFET is appropriate for that kind of circuitry.
Eli Duttman said:
Hey Eli,
It works a bit differently. I'll assume the plate could swing to zero volts...
Define Class A; equal current increase to the current decrease as the grid swings up and down.1. DC conditions; 400V, 75 mA
2. So what's the load? V=i*R...5333 ohms. For PP double that for a-a or 10666 ohms. 400/.075=[a-a/2]
Higher load( lower numeric ) will have more current swing up than down...that's AB. Increasing idle current raises plate dissipation( usually past surviveable limits )
3. (a)Details to complicate things: Estimating lower limit of plate voltage swing to apply to (2.) in the voltage slot. We can get close with the curves, and a guess can be conservative.(b)The tubes aren't anything like linear; upward current swing isn't going to equal the lower.(c)U-L complicates the assessment of working g2 voltage in locating the g1=0volts line and thus the upper current limits of the anode swing.
cheers,
Douglas
seen many real Class A amps lately...heh-heh-heh!!!!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.