Tubedepot.com is now selling "B" grade 300B Western Electric tubes, these come in at a fraction of the cost. They have two flavors high current and low current. A 300B spec says the tube will settle at 60mA with 350V on the plate and -74V on the grid. The hi current "B" grade tubes settle 90-100 mA the low current I hear around 40mA. This is why WE sells them at drastically reduced cost. But if your Amp has adjustable bias these tubes can be biased back to a typical 70mA in use.
My question is what physically happened in manufacturing which caused either a hi or low current defect like this? Not really a defect just out of spec? Is it that the plate wound up the wrong distance from the filament or grid? Or that the grid wound up the wrong distance from the filament? I assume the vacuum is fine and the metals and structures are fine, right?
Anyway I bought a matched quad of the hi current B grade tubes for use in a PSE, direct bias, interstage Amp project I'm working on now. Mine came matched at 111mA all within .3mA of each other. I will simply bias them back to a normal 70mA at 380V is my intention. Couldn't pass up the $800 bargain for a quad of genuine WE's. Did I make a mistake? Is my other question.
https://www.tubedepot.com/products/western-electric-300b-b-stock
My question is what physically happened in manufacturing which caused either a hi or low current defect like this? Not really a defect just out of spec? Is it that the plate wound up the wrong distance from the filament or grid? Or that the grid wound up the wrong distance from the filament? I assume the vacuum is fine and the metals and structures are fine, right?
Anyway I bought a matched quad of the hi current B grade tubes for use in a PSE, direct bias, interstage Amp project I'm working on now. Mine came matched at 111mA all within .3mA of each other. I will simply bias them back to a normal 70mA at 380V is my intention. Couldn't pass up the $800 bargain for a quad of genuine WE's. Did I make a mistake? Is my other question.
https://www.tubedepot.com/products/western-electric-300b-b-stock
Last edited:
Seems like the high bias would put you in a less optimal portion of the load line for a given primary impedance on the transformer and may change the power you get at clipping. Without WE giving some grid bias curves for the hi or low bias tubes (or someone measuring them for us) you won't really know where to bias them. Just putting them at 70mA is a bit of a crap shoot for performance.
Low vacuum? Contaminants? Trying to save money on sourcing the materials for internal components? Made by the new hires that could care less after covid?
I am just tossing some ideas out there.
Mistakes are in the eye (and wallet) of the beholder. 😉
PS @nerdo. I checked the bias numbers on the WE site and they seem to be all over the place. Do they make sense to you?
I am just tossing some ideas out there.
Mistakes are in the eye (and wallet) of the beholder. 😉
PS @nerdo. I checked the bias numbers on the WE site and they seem to be all over the place. Do they make sense to you?
Last edited:
Apparently it's a coating they used on the plate:
https://hifihaven.org/index.php?thr...-not-the-same-as-originals.12961/#post-310215
https://hifihaven.org/index.php?thr...-not-the-same-as-originals.12961/#post-310215
I have an eTracer, when I get the time I'll (try to) figure out how to set up the etracer for a non standard test like this. I usually just pick the tube and do the trace from the library. This would involve making custom parameters for the test, a learning experience for me.
Wondering if long ago these would be in a failing spec bin and the employees would sneak them home. Considering they are WE rejects, they are worth selling at a lower price.
My project that I bought these for is a PSE, direct bias design. It also has dual separate power supply chassis (two), each of these supplies contains a small variac to vary the B+ from 0 to 500 volts in choke-input mode or 0 to 800 volts in capacitor input mode (the mode is set by a jumper that moves C1 to the front or back of the pi filter). The amp is stereo in a single chassis. I posted my schematic a few weeks ago. Each supply also has a gyrator-filtered bias supply of -100V which is adjusted in the separate stereo amp chassis with a pot. Each PSU can supply up to 300mA at any B+ voltage in the range with parallel 5AR4's. So I'm probably in a position to actually use these 100mA 300B's, as I can dial in any B+ and bias I want if found that running the tube at 90 or 100 mA is best. Or I'll chalk it up to a $800 education 🙂
I should contact them I suppose. Mostly to see if they have curves for any of these, having that would help.
I'd be really curious to see curves for any of these, just as a curiosity for what has 'gone wrong' with them. I wonder if the high-current tubes would have problems with the curves bunching up at high voltages. That is, you could add more negative bias to bring the current down, but then when swinging negative into cutoff, what happens. Conversely the low-current tubes, if you had a circuit that could drive DC into them for A2 operation, maybe you could compensate for the low bias voltage.
All that said, I can't imagine spending my money on these with absolutely no warranty when I could get some perfectly nice JJ 300B's for less.
All that said, I can't imagine spending my money on these with absolutely no warranty when I could get some perfectly nice JJ 300B's for less.
I knew it, and I am not surprised at all. Money saving operation backfired. Great job WE.
Buy Takatsuki or the current Elrog Mo and be done with it. If you're spending the money, spend the money for real thing.
Caveat: I don't have any connection to WE, so I don't know the real story.
But that won't stop me from guessing! Here's how I see it:
First off, the current being widely quoted is the one on the box - every tube is tested, and those that fall outside the spec are rejected. (Rumor says that in early days, they smashed as much as half the production.) The current test gives the current at 300v plate, -61v grid. It appears that the acceptance band is 40-90mA, with the center being at 60mA. The closest tolerance is probably the grid to cathode spacing, which dominates the transconductance (gm). That is very small relative to the grid to plate spacing, which mostly controls the gain (mu). I have admittedly over-simplified there but you get the picture.
There are hints that the low-current tubes are from before the graphene, was adopted, and most of the high-current ones are since the graphene. The old plates were already carbonized to reduce secondary emission; graphene is just a somewhat more efficient carbon-based reducer of secondary emission. This offers significant improvement in life span, with the added effect of getting more electrons to stick to the plate when they land instead of scittering around the glass envelope. This increased current is (sort of) equivalent to increased transconductance (gm).
Now the kicker is that the current test is very sensitive - gm varies with the cube root of current, so that 50% current rise from 60 to 90mA means only a 14% increase in gm. A similar percentage of bias adjustment will restore the the 60mA, and with cathode bias you might not need any adjustment at all. Incidentally the ratio 40mA/60mA is identical the the ratio 60/90mA, and 60mA is thus the geometric mean. The acceptance criterion 14% of bogey gm is a tighter spec than typical mil-spec I have seen - as much as twice as strict. Don't get me wrong, matching those current numbers is still important for amps with shared bias, such as some push-pull or PSE designs.
I think WE is keeping the test criterion of 40mA-90mA so they can say these tubes meet the traditional spec; perhaps the percentage that exceed 90mA has increased so they want to recoup some of their losses. The real reason for the low price is probably that beautiful wood box, the 5-year guarantee, and lack of any other support. The real danger for us is when the B-stock is being sold as prime. Caveat Emptor!
But that won't stop me from guessing! Here's how I see it:
First off, the current being widely quoted is the one on the box - every tube is tested, and those that fall outside the spec are rejected. (Rumor says that in early days, they smashed as much as half the production.) The current test gives the current at 300v plate, -61v grid. It appears that the acceptance band is 40-90mA, with the center being at 60mA. The closest tolerance is probably the grid to cathode spacing, which dominates the transconductance (gm). That is very small relative to the grid to plate spacing, which mostly controls the gain (mu). I have admittedly over-simplified there but you get the picture.
There are hints that the low-current tubes are from before the graphene, was adopted, and most of the high-current ones are since the graphene. The old plates were already carbonized to reduce secondary emission; graphene is just a somewhat more efficient carbon-based reducer of secondary emission. This offers significant improvement in life span, with the added effect of getting more electrons to stick to the plate when they land instead of scittering around the glass envelope. This increased current is (sort of) equivalent to increased transconductance (gm).
Now the kicker is that the current test is very sensitive - gm varies with the cube root of current, so that 50% current rise from 60 to 90mA means only a 14% increase in gm. A similar percentage of bias adjustment will restore the the 60mA, and with cathode bias you might not need any adjustment at all. Incidentally the ratio 40mA/60mA is identical the the ratio 60/90mA, and 60mA is thus the geometric mean. The acceptance criterion 14% of bogey gm is a tighter spec than typical mil-spec I have seen - as much as twice as strict. Don't get me wrong, matching those current numbers is still important for amps with shared bias, such as some push-pull or PSE designs.
I think WE is keeping the test criterion of 40mA-90mA so they can say these tubes meet the traditional spec; perhaps the percentage that exceed 90mA has increased so they want to recoup some of their losses. The real reason for the low price is probably that beautiful wood box, the 5-year guarantee, and lack of any other support. The real danger for us is when the B-stock is being sold as prime. Caveat Emptor!
I am so surprised to hear the Tube Depot is selling the B-stock of the WE300B. I believe Western Electric shall stop them from selling because it will cause serious damage of the WE brand name image. Keeping the tight specification and quality of the WE 300B tubes is what we pay for the premium price rather than those dirt cheap 300B's made in China and Russia. Once the B-stocks are out in the market, I won't pay even $500 for the WE300B for sure.
I was the importer in Italy of new ( at that time) 300B, year 2000 (around)
I know the story where a man from Turin, Mr. Doleatto, was involved in facts related at the first production.
It was funny.
Mr. Doleatto was a owner of a little farm in Turin where some tube were built, mainly for military. Three floor of equipments and tools and one room for vacuum. With a tube tester big as almost 2 square meter!!!!
He passed away years ago and was a big seller of instruments, tube and components with a big shop in the same building, closed years ago.
The brand Western Electric was taken from a new company and nothing have to do with Westrex
The firsts samples of tube was very good; I tested them with my Sofia.
Then some trouble happen until some years ago.
I don't know if the new production is very fine but the "B" version at low price maybe something tell us
Walter
I know the story where a man from Turin, Mr. Doleatto, was involved in facts related at the first production.
It was funny.
Mr. Doleatto was a owner of a little farm in Turin where some tube were built, mainly for military. Three floor of equipments and tools and one room for vacuum. With a tube tester big as almost 2 square meter!!!!
He passed away years ago and was a big seller of instruments, tube and components with a big shop in the same building, closed years ago.
The brand Western Electric was taken from a new company and nothing have to do with Westrex
The firsts samples of tube was very good; I tested them with my Sofia.
Then some trouble happen until some years ago.
I don't know if the new production is very fine but the "B" version at low price maybe something tell us
Walter
Back in the early 70's nearly every ham radio operator wanted a Motorola HT-220 walkie talkie. We had a worker whose full time job was to crush every scrap radio before it went into the trash because people were digging radios out of the trash at the landfill dump, cleaning and reselling them.
Selling their "B" stock will dilute the value of the entire product line unless they are permanently marked as "B" stock.
There are several reasons for a variation in current, but electrode spacing and vacuum purity are the two big players. If the elements in the tube are not properly aligned more current can flow in a small spot area which can lead to red spots on the plate even when operated within the specified ratings. Many 6AQ5 tubes make red spots since all of the elements need to be perfectly concentric which is hard to do. Every element is round in the 6AQ5 and they are all quite close to each other.
Last edited:
I believe they stamp a B into the base. That's according to the Blue Glow Electronics video where he shows some of them with the stamp.
Someone needs to do a curve trace of one of these "high" current tubes to see what you are getting. If the grid or filament are off center, then one would expect to see "roll-over" or flattening of the curves at the HV end, giving high 2nd Harmonic.
Carbonizing the anodes in odorless methane, benzine, etc can be a very dangerous process. I gather WE opted for an easier approach by using graphene.
I might give these B-stocks a try in my Elekit TU-8600R 300B SET, which runs 360V plate voltage.
Is there any reason to buy the Hi-IB over the cheaper Low-IB B-stock tubes for the Ekekit, which has the "new active automatic bias"?
"Unlike traditional cathode bias, where a resistor and capacitor are placed at the cathode to set the bias voltage, the TU-8600R employs a separate, dedicated bias power supply and an active circuit. This circuit senses the plate current (with the current detection moved from the cathode side to the B+ power supply side for improved sound quality) and actively controls the grid voltage to maintain optimal bias."