-
You can now help support WorldwideDX when you shop on Amazon at no additional cost to you! Simply follow this Shop on Amazon link first and a portion of any purchase is sent to WorldwideDX to help with site costs.
D&A Warrior 1000
- Thread starter LoneWolf TN
- Start date
I've looked all over and havn't found one... But if I run across it I will definately post a link.
In 30 years, I've never once seen or heard of a diagram for the Warrior. If you took the diagram for the original Maverick, and the diagram for the original 12-tube Phantom you'd be close.
That one's a "trace it down as you go" linear. Not a problem if you're familiar with how D&A did things, and nothing's been hacked or removed. If you're qualified to build one, you're probably qualified to maintain it.
The real problem is,,,, well,,,, that it's a Warrior.
And if someone were to somehow track down a diagram, now THAT would be news. The owner's instruction sheet that came with it did NOT include a diagram. Just tuneup instructions.
73
That one's a "trace it down as you go" linear. Not a problem if you're familiar with how D&A did things, and nothing's been hacked or removed. If you're qualified to build one, you're probably qualified to maintain it.
The real problem is,,,, well,,,, that it's a Warrior.
And if someone were to somehow track down a diagram, now THAT would be news. The owner's instruction sheet that came with it did NOT include a diagram. Just tuneup instructions.
73
f
Shucks!!
Oh well, looks like I'm going to have to take the Warrior in to a guy in Alabama who has a great local reputation with tube amps. I have been going nuts ever since I purchased this amp off of mr. BIGHAWAIIAN and nothing more than about 700-900 watts of output. I think that two of the four transformers have been disabled. I did some real close snooping and there is a black wire missing from one of the ac plug prongs that plug into the back of the power supply from the linear. Two yellow wires coming from each of two of the transformers have been cut also. I recently bought some good tubes from FNKYCOLDMEDINA and I now have tubes in the Warrior that go from over 80% to over 100%. This thing should be doing close to 2000 watts. I'd love to send this beast to Nomad or Justin but the shipping would be outrageous.
LoneWolf TN
Shucks!!
Oh well, looks like I'm going to have to take the Warrior in to a guy in Alabama who has a great local reputation with tube amps. I have been going nuts ever since I purchased this amp off of mr. BIGHAWAIIAN and nothing more than about 700-900 watts of output. I think that two of the four transformers have been disabled. I did some real close snooping and there is a black wire missing from one of the ac plug prongs that plug into the back of the power supply from the linear. Two yellow wires coming from each of two of the transformers have been cut also. I recently bought some good tubes from FNKYCOLDMEDINA and I now have tubes in the Warrior that go from over 80% to over 100%. This thing should be doing close to 2000 watts. I'd love to send this beast to Nomad or Justin but the shipping would be outrageous.
LoneWolf TN
Hey Lonewolf,
To start with, those yellow wires are not used. They lead to a separate winding that will deliver 5 Volts to power the filament in a tube-type rectifier. When that transformer was designed, and sold from a catalog, this was the standard way of rectifying high voltage for tubes. Using tubes as rectifiers.
ALL D&A amplifiers use solid-state rectifier diodes, so those yellow wires were cut off short. D&A started out small, and bought "stock" universal-type transformers from a stock catalog. Once Ed Dulaney was selling enough of his linears to buy transformers in bulk, they began using a transformer that was built without this 5-Volt winding on it. If the transformers have those yellow wires you know that they are old.
One really, really BIG deal with any "Glass-Factory" linear has to do with the Plate Tune controls, and the position at which each one shows a "peak" on the wattmeter. D&A used capacitors that have NO end stop for all three 'Tune' controls on that amplifier. They spin freely 360 degrees, unlike the Final Load control. It will turn only halfway around before it runs into an end stop with a "clunk", and won't turn any farther.
These "360" capacitors can deceive an unwary eye. They can APPEAR to show a "peak" while you tune it, when in fact that circuit never really achieves the resonant peak that you need to get full power.
Here's the rule: As you crank one of these '360'-type Tune controls, you should observe TWO peaks in one whole turn of the knob. Both of these peaks may be right next to each other, or they may be 180 degrees apart from each other.
What matters is that you see TWO peaks in that one, full turn of the knob. I'm gonna skip the long explanation of how this "false peak" phenomenon occurs, or why it matters. I've only been threatening to put it up on a web page for several years, now. Sure would be easier just to post a link to it, once I got a "Round Tuit".
In a nutshell, if you observe ONLY ONE peak in that full turn of the knob, you would observe one of two things if you peeked behind the panel, at the plates on that control.
Either the plates would be as far apart as they go, or they will be meshed as close together as they will go. What happened was that the control ran out of range, before reaching a properly-tuned peak. Since there is no end stop, cranking the knob to either side of this "end" of its range LOOKS like a peak on a wattmeter, even though it isn't. If, in fact, the circuit DID reach a true resonant peak, there would be TWO physical positions of the shaft at which the plates will mesh that close together. One to the left of the 'end' of its travel, one on the right side of that end-of-travel. If your wattmeter peak occurs with the plates meshed exactly halfway, those two peaks will be exactly one-half of a turn apart, with the knob's pointer at 180 degrees one peak from the other peak. If the peak occurs near one "end", those two peaks will be closer together.
The real bottom line here is that there's not a lot you can do about it, if any of the three "Plate Tune" controls fails to achieve a proper "Twin" peak. The control hooked to the knob is only the "fine" tune for that set of tubes. The coil attached to it is the real "Coarse" tune. Changing tubes, from one set to another, will change where the Tune control hooked to them will peak. If someone changed tubes, and threw that peak one-half of a turn BEYOND one end-stop, you'll never see full power from the new tubes. Never mind why changing tubes can throw those controls off of a true peak. They do. And if the coils in that unit were all set with a different set of tubes, they need to be readjusted to suit the tubes that you now have installed in it.
Running a tube-type linear with a Plate Tune control turned away from the peak will stress the tubes. Some folks will attempt to "back down" a linear by deliberately cranking a Tune knob one side of the peak. This reduces the efficiency of the tubes, causing them to overheat inside.
On the other hand, the tube doesn't care whether the Tune control is turned "off" of the peak on purpose, or because it just won't turn "that far".
Here's a pic of the "weak link" that will pop in a 8950 when this happens. The tube's internal connection to the "ground side" pin is a thin strip of metal that comes from the tube's center element leading to pin #2. The excess heat from mistuning will cause that link to melt, like the element in a fuse.
I recommend being cautious with running this box, until it's clear that all three Plate Tune controls are achieving a "True" peak on the meter. Folks who keep tuning, and tuning, and complain that the power just keeps dropping off, will find that they have been blowing the "Fuse" at the bottom of their tubes, one by one. Once the fuse is blown on a tube, there's no way to stitch it back together.
If this fella in Alabama knows his stuff, he's already hip to everything I just told you. But if you keep hammering on this box in its current condition, you could make things worse.
73
To start with, those yellow wires are not used. They lead to a separate winding that will deliver 5 Volts to power the filament in a tube-type rectifier. When that transformer was designed, and sold from a catalog, this was the standard way of rectifying high voltage for tubes. Using tubes as rectifiers.
ALL D&A amplifiers use solid-state rectifier diodes, so those yellow wires were cut off short. D&A started out small, and bought "stock" universal-type transformers from a stock catalog. Once Ed Dulaney was selling enough of his linears to buy transformers in bulk, they began using a transformer that was built without this 5-Volt winding on it. If the transformers have those yellow wires you know that they are old.
One really, really BIG deal with any "Glass-Factory" linear has to do with the Plate Tune controls, and the position at which each one shows a "peak" on the wattmeter. D&A used capacitors that have NO end stop for all three 'Tune' controls on that amplifier. They spin freely 360 degrees, unlike the Final Load control. It will turn only halfway around before it runs into an end stop with a "clunk", and won't turn any farther.
These "360" capacitors can deceive an unwary eye. They can APPEAR to show a "peak" while you tune it, when in fact that circuit never really achieves the resonant peak that you need to get full power.
Here's the rule: As you crank one of these '360'-type Tune controls, you should observe TWO peaks in one whole turn of the knob. Both of these peaks may be right next to each other, or they may be 180 degrees apart from each other.
What matters is that you see TWO peaks in that one, full turn of the knob. I'm gonna skip the long explanation of how this "false peak" phenomenon occurs, or why it matters. I've only been threatening to put it up on a web page for several years, now. Sure would be easier just to post a link to it, once I got a "Round Tuit".
In a nutshell, if you observe ONLY ONE peak in that full turn of the knob, you would observe one of two things if you peeked behind the panel, at the plates on that control.
Either the plates would be as far apart as they go, or they will be meshed as close together as they will go. What happened was that the control ran out of range, before reaching a properly-tuned peak. Since there is no end stop, cranking the knob to either side of this "end" of its range LOOKS like a peak on a wattmeter, even though it isn't. If, in fact, the circuit DID reach a true resonant peak, there would be TWO physical positions of the shaft at which the plates will mesh that close together. One to the left of the 'end' of its travel, one on the right side of that end-of-travel. If your wattmeter peak occurs with the plates meshed exactly halfway, those two peaks will be exactly one-half of a turn apart, with the knob's pointer at 180 degrees one peak from the other peak. If the peak occurs near one "end", those two peaks will be closer together.
The real bottom line here is that there's not a lot you can do about it, if any of the three "Plate Tune" controls fails to achieve a proper "Twin" peak. The control hooked to the knob is only the "fine" tune for that set of tubes. The coil attached to it is the real "Coarse" tune. Changing tubes, from one set to another, will change where the Tune control hooked to them will peak. If someone changed tubes, and threw that peak one-half of a turn BEYOND one end-stop, you'll never see full power from the new tubes. Never mind why changing tubes can throw those controls off of a true peak. They do. And if the coils in that unit were all set with a different set of tubes, they need to be readjusted to suit the tubes that you now have installed in it.
Running a tube-type linear with a Plate Tune control turned away from the peak will stress the tubes. Some folks will attempt to "back down" a linear by deliberately cranking a Tune knob one side of the peak. This reduces the efficiency of the tubes, causing them to overheat inside.
On the other hand, the tube doesn't care whether the Tune control is turned "off" of the peak on purpose, or because it just won't turn "that far".
Here's a pic of the "weak link" that will pop in a 8950 when this happens. The tube's internal connection to the "ground side" pin is a thin strip of metal that comes from the tube's center element leading to pin #2. The excess heat from mistuning will cause that link to melt, like the element in a fuse.
I recommend being cautious with running this box, until it's clear that all three Plate Tune controls are achieving a "True" peak on the meter. Folks who keep tuning, and tuning, and complain that the power just keeps dropping off, will find that they have been blowing the "Fuse" at the bottom of their tubes, one by one. Once the fuse is blown on a tube, there's no way to stitch it back together.
If this fella in Alabama knows his stuff, he's already hip to everything I just told you. But if you keep hammering on this box in its current condition, you could make things worse.
73
Hello Nomad Radio:
WOW! Truly a impressive posting!
Good going, the close up view of the 8950 showing the Cathode wire burning open sure brings back not so found memeries.
The Tuning Capacator info surly comes from years of experence.
I thought that one production of GE tubes used the Base Contact Wire that was internally bent and and welded to the Cathode, but I could be wrong.
I am thinking that Johnson Made a Warrior Amp, or was that Heathkit? That used a pair of 3-500 Triode Tubes in Grounded Grid. Now that would haul the mail at 2KW, if it had the transformer and drive. Or was the EF Johnson I am thinking about a Thunderbolt?!?!?!?!.....................
Ok I just looked up a Warrior Heathkit Amp, it uses 4 each 811 tubes, so it will not do 2KW. See:
http://www.heathkit-museum.com/ham/ha-10.shtml
Jay in the Mojave
WOW! Truly a impressive posting!
Good going, the close up view of the 8950 showing the Cathode wire burning open sure brings back not so found memeries.
The Tuning Capacator info surly comes from years of experence.
I thought that one production of GE tubes used the Base Contact Wire that was internally bent and and welded to the Cathode, but I could be wrong.
I am thinking that Johnson Made a Warrior Amp, or was that Heathkit? That used a pair of 3-500 Triode Tubes in Grounded Grid. Now that would haul the mail at 2KW, if it had the transformer and drive. Or was the EF Johnson I am thinking about a Thunderbolt?!?!?!?!.....................
Ok I just looked up a Warrior Heathkit Amp, it uses 4 each 811 tubes, so it will not do 2KW. See:
http://www.heathkit-museum.com/ham/ha-10.shtml
Jay in the Mojave
Hi Jay,
Thank you.
The D&A Warrior/Warrior II was proof that if you take a good concept and keep piling it on, eventually it gets top-heavy and falls over.
The D&A 'concept' of using one stock transformer for every four tubes worked great for the 4-, 8-, and 12-(10)-tube designs. The Raider, Maverick and Phantom models would deliver better-than-average service if you treated them right.
THEIR "Warrior" model used sixteen tubes. four baldies driving four 6LQ6 driving eight 6LQ6. At a nominal power gain of times ten per stage, you'd figure 1 Watt of drive would give you 1000 Watts of output.
And there's the rub. That is exactly what it WANTS to do. Makes it one bucking bronco of an amplifier. Trickiest part of getting any real use from it is to persuade it to 'slack off' from that times-1000 power-gain tendency.
The Warrior was proof that too much of a good thing stops being good. Best solution I ever saw was to lose the four-tube predriver stage altogether, and drive the four-6LQ6 "driver" stage directly. Made it a 12-tube Phantom with a beefier-than-stock power supply. Much more stable and less likely to throw the rider, so to speak.
The other tongue-in-cheek recommendation we'd make for it was to use a 1- or 2-Watt walkie-talkie to drive it.
Since it was their most expensive, top-of-the-line model, they sold fewer of the Warrior than the other models. As a result, very few techs have any experience with it.
To add insult to injury, the "high-performance" aspect of the Warrior made it necessary to use the "black-cap" 6MJ6 version of the 6LQ6 tube in the final stage. The chrome-cap versions of the tube, with the skinny wire leading to the plate cap, and the thinner glass would have to be 'babied' to keep them from blowing out prematurely.
Trouble is, the "black-cap" 9-pin tube is the rarest and most expensive version. I've seen them listed at 90 bucks each, when you can find them. The price of eight good 6MJ6 tubes would buy a decent new amplifier, all by themeselves.
Every year that goes by, the other available methods for getting that kind of RF power get cheaper and cheaper in comparison.
Oh, and BTW there was a little-known 'cheat' for a 8950 that had the cathode link blown. Pretty sure all of the GE 8950 had that right-angle bend in the cathode link to pin 2. That was the weak spot, every time. But there was a SECOND pin connected to the cathode, pin 6. Probably put there to reduce the inductance of the cathode connection. You could put a jumper wire across the socket from pin 6 to pin 2, and if the rest of the tube had survived, VOILA! Returned from the dead, as if by magic.
Did that to our tube tester, but found that most 8950s with a "blown fuse" also had OTHER problems, by the time they had been stressed that far. Filed that trick away under the "nice try" category.
Someday I gotta put up a web page with pics explaining the "two peaks" rule. Gotta figure it wouldn't take quite so many words that way.
Maybe.
73
Thank you.
The D&A Warrior/Warrior II was proof that if you take a good concept and keep piling it on, eventually it gets top-heavy and falls over.
The D&A 'concept' of using one stock transformer for every four tubes worked great for the 4-, 8-, and 12-(10)-tube designs. The Raider, Maverick and Phantom models would deliver better-than-average service if you treated them right.
THEIR "Warrior" model used sixteen tubes. four baldies driving four 6LQ6 driving eight 6LQ6. At a nominal power gain of times ten per stage, you'd figure 1 Watt of drive would give you 1000 Watts of output.
And there's the rub. That is exactly what it WANTS to do. Makes it one bucking bronco of an amplifier. Trickiest part of getting any real use from it is to persuade it to 'slack off' from that times-1000 power-gain tendency.
The Warrior was proof that too much of a good thing stops being good. Best solution I ever saw was to lose the four-tube predriver stage altogether, and drive the four-6LQ6 "driver" stage directly. Made it a 12-tube Phantom with a beefier-than-stock power supply. Much more stable and less likely to throw the rider, so to speak.
The other tongue-in-cheek recommendation we'd make for it was to use a 1- or 2-Watt walkie-talkie to drive it.
Since it was their most expensive, top-of-the-line model, they sold fewer of the Warrior than the other models. As a result, very few techs have any experience with it.
To add insult to injury, the "high-performance" aspect of the Warrior made it necessary to use the "black-cap" 6MJ6 version of the 6LQ6 tube in the final stage. The chrome-cap versions of the tube, with the skinny wire leading to the plate cap, and the thinner glass would have to be 'babied' to keep them from blowing out prematurely.
Trouble is, the "black-cap" 9-pin tube is the rarest and most expensive version. I've seen them listed at 90 bucks each, when you can find them. The price of eight good 6MJ6 tubes would buy a decent new amplifier, all by themeselves.
Every year that goes by, the other available methods for getting that kind of RF power get cheaper and cheaper in comparison.
Oh, and BTW there was a little-known 'cheat' for a 8950 that had the cathode link blown. Pretty sure all of the GE 8950 had that right-angle bend in the cathode link to pin 2. That was the weak spot, every time. But there was a SECOND pin connected to the cathode, pin 6. Probably put there to reduce the inductance of the cathode connection. You could put a jumper wire across the socket from pin 6 to pin 2, and if the rest of the tube had survived, VOILA! Returned from the dead, as if by magic.
Did that to our tube tester, but found that most 8950s with a "blown fuse" also had OTHER problems, by the time they had been stressed that far. Filed that trick away under the "nice try" category.
Someday I gotta put up a web page with pics explaining the "two peaks" rule. Gotta figure it wouldn't take quite so many words that way.
Maybe.
73
d
My D&A Warrior 1000 has four 6jt6 tubes driving 12 6lq6 tubes, so it should do right around 2kw.
Nomad, yes, that was impressive.
Thanks, LoneWolf TN
My D&A Warrior 1000 has four 6jt6 tubes driving 12 6lq6 tubes, so it should do right around 2kw.
Nomad, yes, that was impressive.
Thanks, LoneWolf TN
Hello Nomad Radio:
More great info.
Does someone have a few photos of the D&A Warrior, that sounds like a rare one. Is the Warrior in the same size box as the Phantom?
I am still looking for a Phantom, Pride 300/1KW or a big box or whatever to gut, to put in a couple of 4CX800 tubes.
Jay in the Mojave
More great info.
Does someone have a few photos of the D&A Warrior, that sounds like a rare one. Is the Warrior in the same size box as the Phantom?
I am still looking for a Phantom, Pride 300/1KW or a big box or whatever to gut, to put in a couple of 4CX800 tubes.
Jay in the Mojave
Jay in the Mojave said:
ive got a phantom (10 tube)AND a maco 300 up for sale
both work tho so im not sure youd want to gut them
scott
I have almost no tube experience, but from what I read in your posts, Nomadradio, the 8950 "cheat" could be done right from the get-go to prevent the cathode lead from melting in the first place, or am I wrong?
Hi DTB,
The excessive current that melts the cathode-pin link always seems to occur right at the bend. Makes me think that it's RF current that's making it hot, along with the tube's excessive DC current when the Tune knob is not peaked.
Splitting the overload two ways would probably make it take longer to blow, but not by much.
So long as pin 6 is not connected at the socket, you still have a second chance. Hook up both, and it's likely you won't.
And HI Jay,
I've got the rating of one D&A transformer's HV secondary figured at around 350 Watts. Considering that the "500" number on the Phantom was meant to describe, more or less, the average power you'd see on a lazy meter, at most,
this suggests an input power of roughly 1000 Watts, spread across the three trasformers.
In rough, back-of-the-envelope terms, the ten tubes in Phantom are being run at a nominal 50 Watts on the plate of each. Not enough to make them cherry in less than a minute or two, but more than the "35" Watt rating that applies to continuous use. Ten times 50 is, it turns out, 500.
The overal efficiency of the two stages they used is roughly 50%. For an average output power of 500, this implies a nominal 1000 Watts of DC input to the tubes.
On the other hand, the 800-Watt rating of the GU-74b's anode is over half-again as large as the original glass-factory setup. If it wants to pull double that (1600 W.) from the HV supply, those 3 D&A transformers won't like it for long.
Putting all three HV secondaries in parallel, and feeding them into a full-wave doubler should get you around 1900 Volts DC. This isn't quite full-bore plate voltage for a GU-74b/4CX800, so maybe a single one of the would pull less than the nominal 1600 Watts input power that it should be capable.
I had the three stock D&A transformers figured as a good bet for just one of those tubes, so long as it doesn't get pushed too hard. The recomendations I've seen to put some resistance in the DC cathode return of the tube will tend to limit the size of accidental surge currents. Seems to me Svetlana recommended 10 or 25 ohms for a smaller tetrode. The correct choice for that part should keep the tube from hammering the power supply. You'd expect at least as much as the stock tubes would muster, more or less. Don't know about more, haven't tried this yet.
I figured that taking a small dual(120)-primary filament transformer, and feeding 12.6 from one of the D&A transformers, would get you 240 Volts AC from the two "primaries" in series. Into a bridge, for the screen supply.
The drawback to that trick is that if the plate voltage were to quit, the screen would now draw excessive current, maybe damaging the screen grid.
The other way to do it would be to take 900-plus volts DC from halfway-up the HV doubler, and use a (large) dropping resistor to feed the screen. That way, at least a loss of HV to the anode is less likely to create trouble. Ya just have to find someplace to put some hot, large resistors.
It will take the 12.6 winding from one xfmr to run the tube's heater. The rating should be about right.
The remaining two can be put in series, and into a half-wave doubler for negative grid bias. If one end of those two windings in series is grounded, you can put a simple half-wave rectifier on it to run 12-Volt DC relays and lights, preamp, etc.
I've even got the HV doubler angle covered. My Pride HV board is configured to assemble as a full-wave doubler. At a nominal 1900 Volts it will be running the parts at only 80% of the rated voltage.
So many cool projects, and only one life to try them.
Oughta put it up in the "Projects" section if I ever got serious. Sure wouldn't go to all that trouble without taking pics.
The excessive current that melts the cathode-pin link always seems to occur right at the bend. Makes me think that it's RF current that's making it hot, along with the tube's excessive DC current when the Tune knob is not peaked.
Splitting the overload two ways would probably make it take longer to blow, but not by much.
So long as pin 6 is not connected at the socket, you still have a second chance. Hook up both, and it's likely you won't.
And HI Jay,
I've got the rating of one D&A transformer's HV secondary figured at around 350 Watts. Considering that the "500" number on the Phantom was meant to describe, more or less, the average power you'd see on a lazy meter, at most,
this suggests an input power of roughly 1000 Watts, spread across the three trasformers.
In rough, back-of-the-envelope terms, the ten tubes in Phantom are being run at a nominal 50 Watts on the plate of each. Not enough to make them cherry in less than a minute or two, but more than the "35" Watt rating that applies to continuous use. Ten times 50 is, it turns out, 500.
The overal efficiency of the two stages they used is roughly 50%. For an average output power of 500, this implies a nominal 1000 Watts of DC input to the tubes.
On the other hand, the 800-Watt rating of the GU-74b's anode is over half-again as large as the original glass-factory setup. If it wants to pull double that (1600 W.) from the HV supply, those 3 D&A transformers won't like it for long.
Putting all three HV secondaries in parallel, and feeding them into a full-wave doubler should get you around 1900 Volts DC. This isn't quite full-bore plate voltage for a GU-74b/4CX800, so maybe a single one of the would pull less than the nominal 1600 Watts input power that it should be capable.
I had the three stock D&A transformers figured as a good bet for just one of those tubes, so long as it doesn't get pushed too hard. The recomendations I've seen to put some resistance in the DC cathode return of the tube will tend to limit the size of accidental surge currents. Seems to me Svetlana recommended 10 or 25 ohms for a smaller tetrode. The correct choice for that part should keep the tube from hammering the power supply. You'd expect at least as much as the stock tubes would muster, more or less. Don't know about more, haven't tried this yet.
I figured that taking a small dual(120)-primary filament transformer, and feeding 12.6 from one of the D&A transformers, would get you 240 Volts AC from the two "primaries" in series. Into a bridge, for the screen supply.
The drawback to that trick is that if the plate voltage were to quit, the screen would now draw excessive current, maybe damaging the screen grid.
The other way to do it would be to take 900-plus volts DC from halfway-up the HV doubler, and use a (large) dropping resistor to feed the screen. That way, at least a loss of HV to the anode is less likely to create trouble. Ya just have to find someplace to put some hot, large resistors.
It will take the 12.6 winding from one xfmr to run the tube's heater. The rating should be about right.
The remaining two can be put in series, and into a half-wave doubler for negative grid bias. If one end of those two windings in series is grounded, you can put a simple half-wave rectifier on it to run 12-Volt DC relays and lights, preamp, etc.
I've even got the HV doubler angle covered. My Pride HV board is configured to assemble as a full-wave doubler. At a nominal 1900 Volts it will be running the parts at only 80% of the rated voltage.
So many cool projects, and only one life to try them.
Oughta put it up in the "Projects" section if I ever got serious. Sure wouldn't go to all that trouble without taking pics.
Hey Lonewolf,
To start with, those yellow wires are not used. They lead to a separate winding that will deliver 5 Volts to power the filament in a tube-type rectifier. When that transformer was designed, and sold from a catalog, this was the standard way of rectifying high voltage for tubes. Using tubes as rectifiers.
ALL D&A amplifiers use solid-state rectifier diodes, so those yellow wires were cut off short. D&A started out small, and bought "stock" universal-type transformers from a stock catalog. Once Ed Dulaney was selling enough of his linears to buy transformers in bulk, they began using a transformer that was built without this 5-Volt winding on it. If the transformers have those yellow wires you know that they are old.
One really, really BIG deal with any "Glass-Factory" linear has to do with the Plate Tune controls, and the position at which each one shows a "peak" on the wattmeter. D&A used capacitors that have NO end stop for all three 'Tune' controls on that amplifier. They spin freely 360 degrees, unlike the Final Load control. It will turn only halfway around before it runs into an end stop with a "clunk", and won't turn any farther.
These "360" capacitors can deceive an unwary eye. They can APPEAR to show a "peak" while you tune it, when in fact that circuit never really achieves the resonant peak that you need to get full power.
Here's the rule: As you crank one of these '360'-type Tune controls, you should observe TWO peaks in one whole turn of the knob. Both of these peaks may be right next to each other, or they may be 180 degrees apart from each other.
What matters is that you see TWO peaks in that one, full turn of the knob. I'm gonna skip the long explanation of how this "false peak" phenomenon occurs, or why it matters. I've only been threatening to put it up on a web page for several years, now. Sure would be easier just to post a link to it, once I got a "Round Tuit".
In a nutshell, if you observe ONLY ONE peak in that full turn of the knob, you would observe one of two things if you peeked behind the panel, at the plates on that control.
Either the plates would be as far apart as they go, or they will be meshed as close together as they will go. What happened was that the control ran out of range, before reaching a properly-tuned peak. Since there is no end stop, cranking the knob to either side of this "end" of its range LOOKS like a peak on a wattmeter, even though it isn't. If, in fact, the circuit DID reach a true resonant peak, there would be TWO physical positions of the shaft at which the plates will mesh that close together. One to the left of the 'end' of its travel, one on the right side of that end-of-travel. If your wattmeter peak occurs with the plates meshed exactly halfway, those two peaks will be exactly one-half of a turn apart, with the knob's pointer at 180 degrees one peak from the other peak. If the peak occurs near one "end", those two peaks will be closer together.
The real bottom line here is that there's not a lot you can do about it, if any of the three "Plate Tune" controls fails to achieve a proper "Twin" peak. The control hooked to the knob is only the "fine" tune for that set of tubes. The coil attached to it is the real "Coarse" tune. Changing tubes, from one set to another, will change where the Tune control hooked to them will peak. If someone changed tubes, and threw that peak one-half of a turn BEYOND one end-stop, you'll never see full power from the new tubes. Never mind why changing tubes can throw those controls off of a true peak. They do. And if the coils in that unit were all set with a different set of tubes, they need to be readjusted to suit the tubes that you now have installed in it.
Running a tube-type linear with a Plate Tune control turned away from the peak will stress the tubes. Some folks will attempt to "back down" a linear by deliberately cranking a Tune knob one side of the peak. This reduces the efficiency of the tubes, causing them to overheat inside.
On the other hand, the tube doesn't care whether the Tune control is turned "off" of the peak on purpose, or because it just won't turn "that far".
Here's a pic of the "weak link" that will pop in a 8950 when this happens. The tube's internal connection to the "ground side" pin is a thin strip of metal that comes from the tube's center element leading to pin #2. The excess heat from mistuning will cause that link to melt, like the element in a fuse.
I recommend being cautious with running this box, until it's clear that all three Plate Tune controls are achieving a "True" peak on the meter. Folks who keep tuning, and tuning, and complain that the power just keeps dropping off, will find that they have been blowing the "Fuse" at the bottom of their tubes, one by one. Once the fuse is blown on a tube, there's no way to stitch it back together.
If this fella in Alabama knows his stuff, he's already hip to everything I just told you. But if you keep hammering on this box in its current condition, you could make things worse.
73
Please Nomad, tell me about the tune control, I just rebuilt my scout and put in new 6je6 tubes.
The tune is maxed out with single peak.
I squeezed the coarse coil but it's not enough.
Do I jump 1 more or 1 less wrap?
thanks
Last edited:
- No one is chatting at the moment.
-
-
dxBot:63Sprint has left the room.
-
dxBot:kennyjames 0151 has left the room.
-
-