Hi Robert,
Truth is, custom stuff is nasty. There is always prototype work and each custom job needs additional brain work out of the comfort zone. Big manufacturers try to escape from custom work, optimize for custom work and/or increase cost.
In the first two years I made a living out of custom high-end work only, prototypes, and I had to learn how to build them during the order itself. It's risk taking and to be honest, I had a year on anti depressants due to learning how to cope with an avalanche of failures.
Now that I receive bulky OEM orders that give a comfort zone, life is calmer. Same goes to other folks and MM should be no exception.
So, yes, I understand MM. But I would also expect from them honest and direct communication. There's nothing wrong in refusal.
Cheers!
...
Hi 50AE,
Yep, I can understand (all of) that.
Truth is also that trying to figure a power supply, in PSUDII for example, is nasty (might have said something else here
) as well since transformer makers never give clues/data about the *regulation* of their (power) transformers...i.e., what happens to voltage and current from off-load to full load...so we start with 3% to 6%, so 4.5%. And where does that get us? Mostly to a confused destination, which is where we started. Otherwise O.K..Best I can do for the moment...
Sleep well,
Robert
Hi Rod, (et al),
Have you found that yet?
I have taken your lead today and went looking for EML 30A...lo and behold they were in stock and available at TubesUSA. Ordered a pair....I am just beginning to look over the data sheet, plate characteristics, operating points and beginning to figure how they might dovetail with my anticipated 300B operating point, which is in the region of lower output power...350 Va-k, 5K-7K OT primary, 5-7Watts.
I have much more to do to put it in a schematic, but I am lookin' ahead.
Thanks and Best,
Robert
Hello Robert - Yes, the plots were posted by @vega65; find them here: EML30A Spectrum
It can work with ca. 430-450V with an IT - but the relatively high ra of the EML30A demands a special quality item, probably at a high cost.
RC coupling might be worth looking at. The downside is a supply voltage of ca. 750V but that can be overcome by brute force parts, and great care with the insulation to ensure safety. High voltage coupling capacitors are not so difficult to buy nowadays. The Driver load resistor would be 18-24k, and should be determined by experiment & listening, while trying different values of anode current.
If you use fixed bias for the 300B endstage, a power-FET source follower is indicated, to drive the 50kΩ grid-leak resistor. For autobias 300Bs, the 250k grid leak R, and load capacitance might be manageable without a follower; and the high value driver pullup resistor limits the risk of blocking distortion - at least to some degree.
It can work with ca. 430-450V with an IT - but the relatively high ra of the EML30A demands a special quality item, probably at a high cost.
RC coupling might be worth looking at. The downside is a supply voltage of ca. 750V but that can be overcome by brute force parts, and great care with the insulation to ensure safety. High voltage coupling capacitors are not so difficult to buy nowadays. The Driver load resistor would be 18-24k, and should be determined by experiment & listening, while trying different values of anode current.
If you use fixed bias for the 300B endstage, a power-FET source follower is indicated, to drive the 50kΩ grid-leak resistor. For autobias 300Bs, the 250k grid leak R, and load capacitance might be manageable without a follower; and the high value driver pullup resistor limits the risk of blocking distortion - at least to some degree.
Last edited:
Hi Robert and Rod. I recently purchased and implemented the 30A as a driver for the GM70. At first, I did not like the 30A as it was sterile and thin sounding. After about 100-150 hours, the 30A is singing nicely. This tube needs 100+ hours break-in before it starts to have a sweet, pure tone. That said, it is very revealing (top to bottom). The bias methodology (and related parts) are very audible. I am using Filament bias and tried several different resistors before I finally settled on the Duelund Cast 5ohm. I have a pair of Koolohm resistors that I will audition next week in the filament bias position. The 30A will show everything about the resistor.
Operating points (from memory): 440Vdc, -5V bias, 15mA.
I am currently using a 200H IT (nano crystalline from Monolith Magnetics). Contemplating a new IT with higher current capabilities.
Operating points (from memory): 440Vdc, -5V bias, 15mA.
I am currently using a 200H IT (nano crystalline from Monolith Magnetics). Contemplating a new IT with higher current capabilities.
Glad to hear the EML 30A has turned sweeter, Pat!
My EML300Bs took even longer to reach their best.
Noise in the filament bias resistor refers to the input of the amplifier, so it really is more sensitive to it; the noise includes current noise intrisic to the part, which can be greater at 1.4A of filament current. Noisy elements (like the thick-film TO220 power resistors have) should be avoided, and cheap resistors may have worse quality in the lead-attach.
It also includes acquired noise: electromagnetic pickup - so magnetic materials could be unhelpful.
My EML300Bs took even longer to reach their best.
Noise in the filament bias resistor refers to the input of the amplifier, so it really is more sensitive to it; the noise includes current noise intrisic to the part, which can be greater at 1.4A of filament current. Noisy elements (like the thick-film TO220 power resistors have) should be avoided, and cheap resistors may have worse quality in the lead-attach.
It also includes acquired noise: electromagnetic pickup - so magnetic materials could be unhelpful.
Last edited:
Nice to hear that patience changed your experience. I was worried you had concluded to early about the EML 30a. Please follow up on bias resistor testing!
Hi,
I think I will stand well clear of 750v HT and using RC coupling.
I envision using EML30A to drive WE300B, so I am considering either choke loading the 30A or IT coupling to 300B.
For choke loading I have ISO RC-160-15W; for IT I have Hirata Tango NC20.
I mentioned design points for 300B to be low power (4-6 Watts)...I have OTs that can do any of the OPs listed in the table attached, and also two 5K:8 OTs. (Slagle 49 Ni and Tamura F7003).
I have done some thinking about where to run the 30A, and my first sketch with the ISO plate choke has a design point of:
Va-k = 400v, Ia = 18 mA and Vg = -4.5v, so Va = 404.5v and VB+ = 425.4v. Ra for the 30A should be ~ 6.5K.
Any thoughts on these approaches would be greatly appreciated.
Very Best, Robert
I think I will stand well clear of 750v HT and using RC coupling.
I envision using EML30A to drive WE300B, so I am considering either choke loading the 30A or IT coupling to 300B.
For choke loading I have ISO RC-160-15W; for IT I have Hirata Tango NC20.
I mentioned design points for 300B to be low power (4-6 Watts)...I have OTs that can do any of the OPs listed in the table attached, and also two 5K:8 OTs. (Slagle 49 Ni and Tamura F7003).
I have done some thinking about where to run the 30A, and my first sketch with the ISO plate choke has a design point of:
Va-k = 400v, Ia = 18 mA and Vg = -4.5v, so Va = 404.5v and VB+ = 425.4v. Ra for the 30A should be ~ 6.5K.
Any thoughts on these approaches would be greatly appreciated.
Very Best, Robert
Attachments
Hmm qouting from Hificollecives description:
"The special characteristic of this resistor is that it displays a negative temperature coefficient. For example, when a speaker's voice coil heats up, its impedance rises. The DUELUND Graphite Resistor counteracts this by utilizing a decreasing impedance."
So as the resitor gets warmer bias changes ?
recommendations for filament bias resistors anyone ?
"recommendations for filament bias resistors anyone ?"
Must to use non inductive resistors.
I usually use Mills, or Dale NH series.
Duelund graphite is useable, but I used it -once- (9R 10W silver) as 01a filament bias resistor (below 1W dissipation).
For me it's not "better" than Mills MRA 10.
Must to use non inductive resistors.
I usually use Mills, or Dale NH series.
Duelund graphite is useable, but I used it -once- (9R 10W silver) as 01a filament bias resistor (below 1W dissipation).
For me it's not "better" than Mills MRA 10.
I tried the Mills MRA 10, Mundorf MR10, Brown Devils, others. To my ears, the Duelund wins.
In a different circuit the best commercially available resistor was the Audio Note Silver Niobium Non-Magnetic Resistors. Unfortunately, these are only 2W. So, I have not used them in filament bias due to low wattage capacity. I need 10W minimum.
Now, that said, I also tried some DIY graphite resistors. This is the best sounding to my ears, yes better than the AN Niobium and Duelund. The challenge is they started to smoke (yes, smoke) when used in the 30A filament position due to power overheating. The graphite I used (from a HB pencil) could only handle about 2-5W (just a guess). But in a proper location, this is the way to go. Really the way to go.
In a different circuit the best commercially available resistor was the Audio Note Silver Niobium Non-Magnetic Resistors. Unfortunately, these are only 2W. So, I have not used them in filament bias due to low wattage capacity. I need 10W minimum.
Now, that said, I also tried some DIY graphite resistors. This is the best sounding to my ears, yes better than the AN Niobium and Duelund. The challenge is they started to smoke (yes, smoke) when used in the 30A filament position due to power overheating. The graphite I used (from a HB pencil) could only handle about 2-5W (just a guess). But in a proper location, this is the way to go. Really the way to go.
@andyjevans: This is an old thread, but did you ever try the 4P1L triode into LL1660 2:4.5 or 2.25:4 configuration?
