More than you ever wanted to know about heat-dissipating tube shield mil specs...but just the item for those *HOT* 6BF5s in Collins equipment.

(You may need to change your font type to a constant-spacing one like Courier for the tables to line up properly.)

by Pete Wokoun Sr., KH6GRT (6/2004)

heat dissipating shields

We all have heard the benefits of using International Electronic Research Corp (IERC) type heat-dissipating shields in the R390A and other equipments to reduce tube operating temperatures. However, I haven't seen any information on just how how much they actually reduce the temperatures. Collins did some temperature studies but I haven't been able to find a copy of their study, possibly called service bulletin 303. I don't know if that study included heat dissipating shields. Searching thru the mil specs that these shields were made to I finally found some definitive temperature reduction figures. The specs are all in degrees C; they have been converted to degrees F in this presentation.

The mil spec heat-dissipating shields designated for retrofitting to existing equipment come from three mil specs: MIL-S-9372(USAF), MIL-S-19786(NAVY), and MIL-S-24251. These shields are designed to replace the shiny, nickel plated JAN types. Mil-S-9372 was an Air Force spec and MS24233, its mil standard for retrofit shields, was implemented January, 1958. MIL-S-19786 was a Navy spec and its amendment for retrofit shields was implemented May, 1964. Both these specs were cancelled in 1968 and replaced by mil spec MIL-S-24251 which covered all branches of the service and was implemented March, 1967. Shields made to any of these specs will have the mil spec part number on them. Here are those mil spec part numbers cross referenced to the well-known IERC numbers:

     SIZE        IERC #   MIL-S-9372   MIL-S-19786   MIL-S-24251
  ------------   ------   ----------   -----------   -----------
  Short 7 pin    5015B    MS24233-1    S0761*V00     M24251/6-1
  Med 7 pin      5020B    MS24233-2    S0762*V00     M24251/6-2
  Tall 7 pin     5025B    MS24233-3    S0765*V00     M24251/6-3
  Short 9 pin    6015B    MS24233-4    S0966*V00     M24251/6-4
  Med 9 pin      6020B    MS24233-5    S0967*V00     M24251/6-5
  Tall 9 pin     6025B    MS24233-6    S0968*V00     M24251/6-6
  Ex-Tall 9 pin  6027B    MS24233-7       ---        M24251/6-7
                                       *(X or C)

All the above sizes except the short and ex-tall 9 pin ones are used in the R390A. You can get information on how many of which ones on many web sites. The IERC numbers are normally used when searching for these shields. If someone other than IERC made them, they may only have the mil spec number and some other model number. I have some made by Waterbury Pressed Metal Company (WPM in the table below) that are this way. One I have made by Cinch Connector Company does carry the IERC number. I found documentation that the Atlee Corp also may have produced some of these shields. Their different model numbers are noted in the table below and cross referenced to the IERC numbers:

      SIZE         IERC #     WPM #      ATLEE #
    ---------      ------    --------    --------
    Short 7 pin    5015B     RS-215-1    A10041-1
    Med 7 pin      5020B     RS-215-2    A10041-2
    Tall 7 pin     5025B     RS-215-3    A10041-3
    Short 9 pin    6015B     RS-216-1    A10042-1
    Med 9 pin      6020B     RS-216-2    A10042-2
    Tall 9 pin     6025B     RS-216-3    A10042-3
    Ex-Tall 9 pin  6027B        --         ---

BTW, I noticed the last two digits in the IERC number correspond to their height in decimal inches. For example, the 5015 is 1.5 inches high, 5025 is 2.5 inches high, etc. Anyone know if the 50 and 60 designate anything?

Physically, from ones I have seen, the shield inserts (the part that contacts the tube) are of two types: a multi-sided cylinder (5-sided for 7 pin tubes and 6-sided for 9 pin tubes) or a round insert with a multitude of 1/16 inch fingers. I found both types on shields from both the -9372 and -24251 mil specs. The multi-sided inserts have an open top between the insert and outer shell whereas the mini-fingered insert has a top closed. I personally have not seen or heard about any shields that have the MIL-S-19786 markings.

multi-sided internal insert multi-fingered internal insert

Shields made to MIL-S-9372(USAF) (MS24233) were qualified to reduce the surface temperature of a test 'slug' by 36 degrees F, minimum (a 10-11% reduction). The test 'slug' was an alumimum piece shaped like a tube with an internal heater and 3 imbedded thermocouples. This 'slug' was heated up to 338 to 356 degrees F when the shield was applied. The average reading for all thermocouples had to be at least 36 degrees F less than the starting temperature. How well this test 'slug' with its greater thermal mass related to actual tubes I don't know.

Shields made to MIL-S-19786(NAVY) were qualified using an instrumented glass tube called a Thermion. Apparently these were tube-sized things containing a heater and thermocouples. It was heated to its test temperature when the shield was applied. The shields designated for retrofit service were only required to reduce the temperature of the thermion between 10 and 25% (symbol 'X' in the tables). However, the shields worked so well they were qualified to the next higher reduction of 25-38% (symbol 'C' in the tables). Specific temperatures for this spec are as follows:

                     Bare Bulb    Shield Temp Reduction (Minimum)
  MIL-S-19786 #      Test Temp       (X) 10-25%      (C) 25-38%
  ---------------  -------------   -------------    ------------
  S0761 (short 7)  293 degrees F    27- 65 deg F     65- 99 deg F
  S0762 (med 7)    437 degrees F    41-101 deg F    101-154 deg F
  S0765 (tall 7)   455 degrees F    43-106 deg F    106-161 deg F
  S0966 (short 9)  266 degrees F    23- 59 deg F     59- 89 deg F
  S0967 (med 9)    446 degrees F    41-104 deg F    104-157 deg F
  S0968 (tall 9)   347 degrees F    32- 79 deg F     79-120 deg F
  Note:  The V00 in the -19786 mil part number refers to a
         vertically mounted shield with no separate base provided.

Shields made to Mil-S-24251 were qualified using actual electron tubes. The temperatures were measured from a thermocouple imbedded into the test tube's glass at its hottest spot. The hot spot location was determined by temperature sensitive paints. Like in the previous specs, the test tube was heated to its test temperature when the shield was applied. The shield had to reduce the bulb temperature by at least the amount indicated in the following table:

                             Bare Tube       Shield Temperature
  MIL-S-24251 #           Test Temperature   Reduction (minimum)
  ----------------------  ----------------   -------------------
  M24251/6-1 (short 7)     239 degrees F     45 degrees F (19%)
  M24251/6-2 (med 7)       419 degrees F     72 degrees F (17%)
  M24251/6-3 (tall 7)      464 degrees F     81 degrees F (17%)
  M24251/6-4 (short 9)     266 degrees F     45 degrees F (17%)
  M24251/6-5 (med 9)       437 degrees F     99 degrees F (23%)
  M24251/6-6 (tall 9)      446 degrees F     81 degrees F (18%)
  M24251/6-7 (ex-tall 9)   455 degrees F     81 degrees F (18%)

Typical tube operating temperatures I expect are somewhat less than these test temperatures which maximized tube dissipation. This would lead to somewhat less than the above temperature reductions in actual situations. However, I think these tests were closer to actual conditions than the 'slugs' and Thermions used in previous testing.

The mil spec Mil-S-24251 remains in effect today. However, there are no products on its qualified products list. What that means is no one currently makes any of these shields because the military doesn't have a need for any. Personally, I think shields made to any of these mil spec are going to perform similiarly because they're not all that different from each other.

There are other types of mil spec heat-dissipating shields even of improved design but they are not designated for general backfitting into existing equipments. These shields and their sockets were designed from the start as an integral part of their equipment. As such, significant quantities to use in other equipments are probably not available.

So, what does all this mean? Here are my thoughts: These temperature reductions listed that the shields had to meet are all minimums so actual reductions cannot be determined. Physically these shields seem to remain pretty much unchanged throughout the years; it was the mil specs that were changing. And mil specs are sometimes written just to document what is normally used and available! From the mil spec 19786 qualified products list the manufacturers had test data that supported their products qualification of 25-38% reductions in bulb temperatures. This range also allowed them to meet the newer mil spec 24251 minimum reductions. So I would venture to say a typical bulb temperature reduction of 20-25% is realizable with the heat-dissipating shields. Having a temperature reduction figure only leads to a further question: By decreasing the operating temperature of a tube by some amount, how much improvement in tube life does this lead to? This becomes harder to answer than determining how much cooler the tube operates. But one can generalize by saying any increase in tube life by lowering bulb temperature is beneficial.

The most informative article I was able to find on-line which related tube bulb temperatures to tube life was pearl_tube_coolers.pdf on the www.pearl-hifi.com website. Although much of the website borders on the more esoteric nuances of high-end audio, this paper presents some of the earlier works done by GE and IERC on tube temperatures and life spans that are difficult to find these days. An example from an IERC study in that article: a 6AQ5(6005) tube operating near maximum plate dissipation has a bare bulb temperature almost 460 degrees F. Enclosed in a bright JAN shield its bulb temperature rises to 600 degrees F. With an IERC type B cooler installed the bulb temperature drops to 365 degrees F. This is a 20% drop from its bare bulb temperature and an 39% drop from its JAN shield temperature. This related to a tube survival rate after 500 operating hours of 35% using no shield, to less than 5% using the JAN shield, to over 95% still working using the IERC type B cooler. In another example from a GE study: From a batch of 200 6AQ5(6005) tubes running at 502 degrees F, 15% were still operational after 2500 hours. A second batch running at 428 degrees F, 74 degrees cooler or about a 15% reduction in bulb temperature, still had 90% operational after 5000 hours. It seems "small decreases in bulb temperatures often result in seemingly disproportionately large increases in tube life". The article is also interesting in that it touches on other factors like filament voltage, forced air cooling, and temperature gradients that also have an influence on tube life.