[Amps] looking for assistance with MLA-2500 tank circuit rework

Mon Dec 31 22:57:58 EST 2012

The amp came already modified with the 3cx800s.  It was done up by W4EMF.  One of my EBay finds that has found my bench.

From: Mark - K4SO 
Sent: Monday, December 31, 2012 10:09 PM
To: Chris Miller 
Subject: Re: [Amps] looking for assistance with MLA-2500 tank circuit rework

On 12/31/2012 8:54 PM, Chris Miller wrote:

  Yes that is the technique that I was using.  I was using a miniVNA box, but that should be ok.

  I pulled the value of the resistor off of a spec sheet for the tubes that I downloaded, however the sheet is not clear as to whether or not they were showing specs for one or two tubes.  Although, as I look at the specs here again, it has to be for two tubes as they show 1100W PEP input while in the max ratings it shows 800W max plate dissipation.  However, I have very limited experience with playing with tube spec sheets.  My schooling was during the transition period between discreet components and large scale IC so there was not a lot of tube training.

   I thought I had looked at the spec for the 8875 to compare, but I can not remember now.  I may have just given it a quick glance to check the capacitance values.  Likely explains the harshness that I seemed to note on its signal on the monitor receiver I had running during the dummy load test.  It probably was beating the hell out of the filtering.  I will have to figure a way of preventing that, that is, something that is a hard limit rather than leaning on the ALC hard.  As I look at the manual specs it does look like they were trashing the heck out of it on sideband.  they show 2000W PEP input while showing 1000 W DC input CW

  I am going to find that 8875 spec sheet again.

  From: Mark - K4SO 
  Sent: Sunday, December 30, 2012 4:23 PM
  To: Chris Miller 
  Subject: Re: [Amps] looking for assistance with MLA-2500 tank circuit rework

  On 12/28/2012 10:49 PM, Chris Miller wrote:

    Hi to all, 

    I have taken on a project that has been bugging me for a bit.  I have a MLA2500 that I picked up that has been retrofitted with a pair of 3cx800 intending to have a transportable amp.  From what I have read the tubes are supposed to be a close match for the original tubes, as far as tube specs go.  However, I have found that the amp on 160 through 40 has to have the load control set to 1 to get close to load on power.  20 through 10 appeared to work o.k. 

    I decided to sweep the tank through the antenna, with a 2.7k ohm resistor in place on the anode to simulate tube load and found that resonances appeared off on lower bands.  Figuring I had a good handle on the process, I prototyped up a new tank coil to make mistakes on.  I then playing with tap points and turns until I achieved what looked like a good set of patterns for each band, for the range of tuning. 

    When I attempted to test my results, on dummy load conditions, all bands are now showing insufficient capacitance for load on all bands with a little bit of range at near max capacity on 10 near the upper end of the band. 

    Since that seemed to be an epic failure, I must be missing steps in the process.  Is there something I should be watching for or doing so that what I see as tank resonance actually lands where it should on real world testing?  Am I now at the point where I have to do some trial additions of capacitance to the load control?  Is this amp worth the effort to make it work properly (realizing it is a bit of a sardine can in there)?  Has anyone else done the retrofit on these amps to the 3cx800 tubes, like this amp has, and did the amp still resonate as the original tubes did? 

    Thanks for any help 

    Chris VE3CEA 

    _______________________________________________ 
    Amps mailing list 
    Amps at contesting.com 
    http://lists.contesting.com/mailman/listinfo/amps 

  Chris,

  I used this technique for a 6M amp, with a pair of 4CX250Bs, and it was amazingly accurate. Things are a lot touchier at those frequencies, and a couple stray pfs can make a difference.

  I used an MFJ-259B, looking for 1:1 SWR, while holding the TR relay closed., but I must admit that the value of resistor used for tube simulation is now a distant memory (in other words, I have no idea, hi).

  I'm just encouraging you that the technique will work, at least in my particular circumstances.

  I do note that the MLA2500, while often used to create some pretty large output wattage figures, used a pair of tubes rated at less than the dissipation of just one 3CX800. The 8875 was rates at 300W plate dissipation. If you get this thing going, I think you'll cook the power supply in pretty short order, unless you keep the throttle back. I think a lot of the 8875 tube failures were because they would produce a fair amount of power without much overhead (or really well beyond their ratings), until they popped.

  Maybe someone can verify the value of the simulator resistor. Oh, I assume you put the cabinet back on as well...

  I do see the Cout is almost the same for the two tubes.

  HNY es GL,

  73, Mark - K4SO

Chris,

Well, I'm a hobbyist only, so my tube knowledge is not formal. OTOH, the spec. sheet for the 8875 is for the 8873 (conduction cooled), 8874, traditional convection cooled, as in up from the bottom, and out from the top, and the 8875, which is a cross flow scheme. The tube is basically the same, but the cooling does affect the dissipation. The 8875 is only 300W per tube. They were thrashed, but that's Dentron in those days, and when the tubes were plentiful and not prohibitively expensive, I guess it was more reasonable. I remember the heyday of this amp thougth, and guys were pretty satisfied with themselves about the power output.  Still, a pair of 8874s in an Alpha in that era, was not expected to put out as much power, but could. They always left more headroom. I had 3 of the 3-tube Alphas running 8874s, and they would run 150W out, esp. on sideband, almost indefinitely. but that would have been twice the plate dissipation...

Here are the spec. sheets I was using:

http://www.g8wrb.org/data/Eimac/8873.pdf

I was assuming you had the 3CX800s available, and they fit the sockets, so that is why you were using them, wth a filament transformer swap. I wished I'd put a pair in my Alpha 78, instead of jumping to the 89, and they are nice, low-drive, tubes that make good power. I have a pair of pulls for spares. so I'm going with them for as long as they'll carry me.

Your technique is good, and will work, so double check the specs, and stay with it.

GL agn!

73, Mark - K4SO