On 07/17/2010 04:44 PM, Ben wrote: Make sure neither valve has a heater to cathode leakage. Insure the DC bus is clean as a common old valve gear problem is dried up electrolytics. Try grounding the other side of the filament string or float the heater string (insure neither side of the heater winding on the power transformer is grounded. I've also put a 100ohm 2W pot across the heater winding and grounded the slider to be able to balance the heater voltage against ground. If all else fails run the heaters on DC. Allison

Tony Duell wrote: I found my hum ... Poor magnetic shielding between the power supply and the pre-amp output transformers. Not much I can do unless I can order custom shielded chokes and power transformers as well as shielded 10K to 600 ohm output transformers. I don't have the $$$ for that, but it does prove you can get rid of hum if you don't use cheap off the shelf parts, but parts made for low hum and noise. Ben.

Chuck Guzis wrote: I tried that, but not much change in the hum. The main problem is the power supply, I got all crappy open core inductors, that hum like mad. This will require a rebuild, but I don't have $$$ for that. Now I know why consumer products vs commercial products have a 10x difference in price. Now that I know where my hum is from, I can get back to computer design. No not a Z80, but 9/18 bit CPLD cpu design I plan to build. 6850 ACIA, Simple IDE interface 32KB ( 9 bits of ram ), 512 byte EEPROM for boot straps ( bits #3 is duplicated as bit #4 to give 9 bits),A cpu CPLD and a GLUE CPLD and a bit 74HCT for misc parts. Ben.

On 07/19/2010 12:12 AM, Ben wrote: Those that still remember the technology knew enough to loose mount transformers so they could be rotated for minimum hum pickup. Or they remote mounted the power transformer/power supply so the hum issue is less pronounced. Allison

I am curious as t owhy toy have output transformers in a pre-amp, but anyway. Have you treid moving them around. If they're of the normal clamp construciton, arranging the cores in a straight line is about the wrst thing you can do. Turning the mains transofrmer through 90 degress might well help. Make your own screens? -tony

On 17 Jul 2010, at 19:01, Tony Duell wrote: Me too.....

This _is_ an international list... Is it just me, or are the common US valve numbers pretty much meaningless. The common UK/European ones, used by Philips/Mullard [1] can be easily decoded to give the heater rating, electrode structure and base type. US ones don't seem to give the electrode structure in any meaningful way, even though that it probably the most important thing [2]. [1] There were other codes used by manufacturers in Europe. One farily common one was the Mazda code. It looks like a US valve number (digits, letters, digits), but it's decoded in a totally different way. The first digits give the heater voltage _unless they're 10, 20, or 30, which means 100mA, 200mA, 300mA for sereis-string sets), the letter gives the electrode structure (things like 'C' = frequency changer, 'L' = triode, 'P' = pentode, 'D' = diode) but you dont' double up letters for repeated structures -- a double triode is 'L' not 'LL' And the last digits simply give unique numbers to valves which have the first part identical. A 6D1 is a single diode (==EA50 IIRC), a 6D2 is a double diode (EB91/6AL5), a 30P4 is a TV line output pentode, and so on [2] Reminds me of a the large electronic component distributor over here that let you select 3 terminal fixed regulators by just aobut every possible characteristic (case type, number of pins, max temperature , etc) _other than the output voltage. Odd, that' the thing I think of first when specifying such a component. Ah, the type 80 I have come across... -tony

Most of the US tube numbering systems are today fairly meaningless, and yes, the best way to know what tubes you are looking at is to pretty much know what circuits you are looking at. The initial "standard" (an example would be UV-201) set by RCA was quickly corrupted by the independents, so the industry set up a new standard that we all (US) know and love. I suspect that the choice to standardize the filament voltage and active number of elements in the type number, yet leave out the basic function, was driven by the idea that many radio designs could easily morph, simply with a tube swap, in order to meet a new demand. An example might be changing a standard AC power radio into a farm set into a car radio - mostly with only a tube swap involving filament changes. In an early 1930s context, this starts to make sense. However, with the explosion of new types, I think this was all corrupted as well, and by the late 1930s it was pretty much as meaningless as it is today. The industry standard 55xx series, where most computer tubes reside, chucked all ideas of encoding stuff in the number, and in a way has remained very intact. The European system used this same idea, tried to encode a little more into the type numbers, but still managed to corrupt itself. An article in Tube Collector magazine outlined the system, and there are an astonishing amount of inconsistencies. In a way, it tried to be too clever. Of all these systems, the most simple ones tend to be the best - do not encode anything into the type numbers. If a tube user needs to figure out just what a specific tube type does, he probably has no business sticking his nose into the radio in the first place. The award for most retarded tube numbering must go to the British. Apart from the military's CV (common valve) system, which almost nobody but the military uses, the British system was a mess, with numbers and formats specific to the government agency or industry that wrote the databooks. Often these numbers would even foul each other. The best example is the VT90. If you ordered a VT90 each from the Army, Air Ministry, and Post Office, you would get three completely incompatible tubes. And during World War 2, if you ordered a VT90 in Britain, you might get a humble little US made 6H6 dual diode - a "fourth" VT90. -- Will

Only the 7xxxx is not the Loctal equivalent of the 6xxxx in genral. IIRC, the 6Q7 became the 7C6, the 6V6 became the 7C5. No logic at all. And IIRC for octal valves, the metal envelope (early octals being those RCA metal things, of course) counted as an element. In other words, if oyu know what the valve is, fine, but there's no way you'll work it out from the type number. But if somebody says to me 'EFM1', I know it's a combination signal pentode and 'magic eye' indicator (yes, such a device did exist!) with a 6.3V heater on a P side contact base. And a UCH81 is a triode-hexode freqeucny changer with a 100mA series-string heater on a B9A base. And so on -tony

To nitpick, not entirely true. In the original RCA system, U means unit, V mean vacuum tube, and the 201 means it is a type 201. The A indeed means the improved version. Capacitors were UC, resistors were UR, and so forth. I think the type numbers were unique - there are UV-201s, but I do not think there are any UR-201 or UC-201, for example. Yes, RCA themselves renamed the tube 01A, but this was after the independents hijacked the system. They did so very effectively, and even RCA fell in line by the early 1930s. Be careful, here. Specialized and transmitting type numbers were indeed in their own world, but almost all US manufacturers stuck to the system for receiving types. Hytron was no exception. A Hytron 6SQ7 will be an industry standard 6SQ7. Considering all of the subcontracting and tube swaps done by RCA and the independents, even back in the 1930s, it was extremely wise for everyone to stick to the rules. Once again, be careful. For receiving types, Raytheon used the same system everyone else did. Occasionally they would put their vanity prefix on a high quality premium tube (CK-36 is a standard 36). This is one of the very few times the industry really broke their own rules. There are only a few instances of this happening, and the mostly deal with battery tubes all around 1933 or so. Past these few screwups, receiving tube number duplication was pretty unknown in the US, save a few times a European or Japanese type was unofficially slipped into the system without industry approval. Those tubes are almost never seen today. You mean UV-866, correct? Hehe... Ironically, yesterday I was in Boston buying and selling tubes. And of course I got lost after straying from Mass Ave. So, yes, the US system has a lot of quirks and burps, but for the most part, the consistency from even the early days thru the end was very impressive. The US truly is (was?) the land of standardized parts. -- Will

On 20 Jul 2010 at 20:03, Tony Duell wrote: That's true, but only in those cases where the shell (or any other element) is brought out to a base pin. The 1934 standard says that an element has to be "useful" and have a connection to be counted. So, a 6L6 has heater, cathode, control grid, screen grid, plate and shell all connected to base pins. The beam-forming electrodes aren't counted, so, for instance, the 6F5 triode has exactly one fewer "useful" electrode, even though it has two fewer elements. The other requirement is that the tube must have been introduced initially in the metal shell form (I can't think of any that were glass, then metal, but there may be some.) An "S" as the first part of a middle two-letter pair signifies a single-ended tube, which was not universally followed. Rectifiers *usually* have a high-middle letter (e.g. 5U4, 5Z3), but sadly, there are many exceptions. The most reliable part of the number is the suffix. e.g. G = glass, GT = short glass, GA = improved glass version, GY = micanol base, etc. In 1942, the RMA introduced a scheme for special-purpose and transmitting tubes, that was called the "1A21" system. The first number represents the power rating, the second, the tube type (e.g. diode, triode, etc.), the second and third numbers are assigned in the order of introduction, starting with 21. So the gas thyratron 2D21 tells us nothing more than it's a tetrode rated for 10 watts or less. That standard lasted only until 1944. --Chuck

Right,.. It's getting to the point where all you can do is look in the valve data books.... I still much prefer the Philips coding scheme. it tells you a lot of useful information, and exceptions are not common (put it this way, I've never seen a valve which doesn't fit the scheme correctly). It appears that sometimes you count the tap on a heater (e.g. 35Z3 against 35Z4), sometimes you don't (12AX7 has 2 cathodes, 2 grids, 2 anodes and a tapped heater). My guess would be tuning indciators (magic eyes, or whatever you call them). A glss window is essential for that type of device, and I would assume many exist only in glass envelopes. There are many single-neded valves (no top cap) without the 'S'. Which is not normally the most important piece of information about a valve :-). [I thought 'GT' was 'Glass, Tubular', meaning a straight-sided envelope] Ah, so trhat's where that number comes from.... We call it an EN91 (6.3V heater, Thyratron, B7G base). -tony

Or, sure there were coding schemes for CRTs that made snese. But I don't think any manufacturer used the same code to cover both CRts and other valves. Which amkes sense. The important characteristics of a valve and of a CRT are rather different. -tony