Separate names with a comma.
Discussion in 'Amplifiers' started by 357, Jul 2, 2018.
Can I run 1 tube in this amp at a time to test them?
That has the 8875 tubes I think which are unobtanium now. That amp was actually in the category of one of the worst amps ever made because of poor design. Eimac even warned Dentron about it but they ignored it. The amp is very prone to VHF parasitics so watch it carefully. You can run just one tube but be aware that the tune and load controls will be waaay off from what you would expect due to the different load impedance and the drive level will be half that of normal.
"That has the 8875 tubes I think which are unobtanium now!"
Correct CK, last set of those I knew were sold for $1000+ for the Eimac NIB pair. Those were dropped by Eimac within couple yrs after those amps were produced.
This is why many of the 2500/2500B's were converted to GI7BT's or single 3cx800's also 4cx400 Svetlana's(also no longer in production)...if you find one with good set of 8875 tubes, be aware when driving those amps the grids are fragile plus certainly NOT for AM usage under any conditions IMHO.
All the Best
YES Gary. IIRC The grid dissapation is rated as ZERO watts!!
The MLA 2500 is probably the world's "most converted" amplifier, since the tubes were never made by more than one vendor. When Eimac dropped the 8875 from their catalog, that was it. I have a MLA converted to two Eimac 3CX800 tubes. Wanted a blower that would be quiet. Only way to get big air without noise is a big blower. Slung it under the cabinet. Had to build a plenum to force air pressure through the fins in the anode rings. Makes it look like it has a parasitic snail clinging to its underside. The 7-inch tall legs that allow air entry into the blower are pretty darned ugly, as well.
I have keyed these amplifiers with no drive using only one tube. This allows you to read the idle current (with no drive) for each tube, and judge how well they will balance the load. If one tube is stronger than the other, it will "hog" more than half of the load. It then overheats and fails prematurely. This is the curse of using two tubes in parallel. They really have to match, so the load gets shared evenly between them.
As for keying it with drive power this way, it's mostly a waste of time. You could load it up this way to compare one tube to another. Just reading the idle current for each tube separately is only half the story. This way you can compare the largely-reduced RF you can get this way, from one tube to the next. If one tube shows much more (or less) power than the other tube, they probably won't balance properly.
You won't see even one-third the normal power with just one tube. The input SWR will be high. As a test method, it can be useful. But putting it on the air that way probably is not.
Dennis just didn't want to build a tuned-input circuit into this one. Really didn't have room for it. That feature would have to be linked to the shaft of the band selector to select the correct tuned-matching circuit for each band. No way to make that fit.
The MLA has two tubes with an input impedance of about 250 ohms each. Those two cathodes in parallel get you an input impedance of roughly 125 ohms. There is a 100-ohm aluminum-body power resistor that has one side grounded, and the other side wired to the input pole of the antenna relay. This puts 100 ohms of resistance in parallel with the two tubes' cathode circuits. 100 ohms in parallel with 125 is close enough to 50 ohms for most radios. This also means that half your drive power never reaches the tubes. But this is about the right drive level to use with a 100-Watt PEP ham radio. This "quick and dirty" input circuit has risks, like a tendency to oscillate, as pointed out above.
One big thing to look for in any MLA is that big, shiny anodized-aluminum power resistor. You can't miss it, pop-riveted to the divider wall, with a big bare wire leading to the relay's input side. If you see that wire cut, or unsoldered, you can assume the tubes are toast. Taking that wire loose does two things. Increases the SWR the radio will be driving into when it's keyed, and increases how hard the tubes are driven.
As a rule, this gets done after the tubes have gone "SNAP!" a time or three, and the internal damage to the tubes has reduced power. Cutting that resistor gets you some of that power back, maybe enough to unload it before it dies completely.
These did get used on channel 6 back in the day. That resistor would always be found unhooked when the owner unloaded it.
It's not really an amplifier for AM. The heat those tubes can unload is reduced by the wide space between the fins. Not a lot of surface area. But it allows you to blow the air across them with a simple fan. No blower or sealed air-pressure plenum is needed like you do for closely-spaced fins. What makes it practical for SSB is that the average power you see will generally be under 500 Watts. And with a heat rating of 300 Watts per tube, you can get by this way.
But the continuous carrier power of your AM signal will overheat the tubes with a carrier power much over 150 Watts, or peak power much over 800 or so. Don't remember anyone using an amplifier that says "2500" on the front at 800 Watts PEP. That tube is capable of delivering a lot more power than the safe limit for its temperature.
But not for terribly long.
Oh, and yeah.. The "4CX400" russky tube? Made by Svetlana originally for the soviet military as type "GS36B".
The data sheet for that tube says the control grid is rated for zero point 2 (0.2) Watts. Not quite zero, but close enough.
But the Svetlana data sheet for the re-branded 4CX400 says the grid is good for two Watts.
My take is that the military rating is conservative. And if you hit the civilian-branded version or the original military-numbered tube with too much drive, you'll damage it.
I suspect that the 2-Watt rating in the Svetlana civilian data sheet assumes a low duty cycle. They were marketing these tubes to amateurs 20 years ago when they rebranded this one. Odds are that the military data sheet assumes a continuous-duty condition for those numbers.
If you treat the 4CX400 like a 4CX250, you probably won't hurt it.
I had one a few years ago. After I realized the tubes were not something I wanted to deal with down the road, I sent it down the road. Never turned it on.
Its this amp I re-did for my buddy.
When I worked on it it had cracked solder on the filament pins.
I loaded it up and didn't get the 1000w more like 400 and I noticed the power would drop off about 100w.
I think it was run with cold filaments due to the bad connection.
I guess I will not buy it.
See pin 5 before I soldered it up.
When he first talked on it, I could hear something kicking in and out, I think it was the filament.
He had new tubes in the box that came with sockets but he sold them for a couplr hundred bucks.
OK he brought this over and we ran it on 240V
Up on 10M it tuned 1KW and we were seeing 1200+ SSB.
I'd imagine it would do better on 40 or 80.
Grid current was minimal.
I think it was the 120V sag that it didn't like before
yes indeed chris ..was a very happy dude to see over the 1k on your lp100 pal ..you did a very good job on the harbrach soft start ..will be picking that amp on my return to beautiful bc ..found 3 more 8875s over here ..haven't tested them yet ..fingers crossed ..73 pal
If you ever have to convert one of these amps to run a different tube, it's worth the extra time and money to install the single Eimac 3CX800A7 as opposed to the twin GI7B. While the cooling and electrical specs are a closer match for the GI7B, those Russian surplus tubes are worth exactly what they are charging for them. Almost nothing.
After testing lots of the 7B, 7BT, 46B, and the 35, I've found two out of three show signs of a poor vacuum. One in five are so bad that the filaments won't light and look almost like a short. About one out of three perform as expected and that makes matching more than one in a single stage difficult. That's after the recommended conditioning process.
When it comes to power output, the most important part is the plate transformer. When it comes to reliability, the most important part is your choice of tube. Eimac is still the king when it comes to power grid tubes. They don't just last longer, they take more abuse too.
If you go with old Russian surplus, do not operate outside of the rated specs for grid, plate voltage and dissipation even at lower frequencies. When they were new with a good vacuum, they could handle that. Some still can but most will fail from internal arcing if you don't run them conservatively.
To answer the original question, yes you can test with one tube but I would only do this if I detected a short on the plate supply when B+ or bias is applied to the tubes. Operating the plate supply with no tubes and then one at a time can easily identify a tube that can no longer handle plate voltage. It's still no guarantee the tube can handle RF but that can be more difficult to test with the input and output impedances doubled on one tube.
Ameritron sells a grid protection kit that can and should also be added to this amplifier. This entire line of tubes from the 8873 to the 8877 are specifically designed to have extremely low IMD and good linear characteristics. That means fragile grids that need protection today when the tubes are this expensive.
Can anyone explain what D14 and D15 are doing in this circuit? I just noticed the schematic for this amplifier has the filament tied directly to RF ground. That is not so good on these tubes because the cathode to filament spacing is the closest gap inside the tube. It does save having to add a filament choke like you would on a 3-500Z. If you get filament to cathode shorts in this amp, that is why.
Can anyone explain what D14 and D15 are doing in this circuit?
I took a look at the schematic and I see these diodes. Kind of unusual to see them there and I can only guess what the designer had in mind with them. I am thinking that "maybe" his thought was if one of the tubes arced the back to back didoes would act to snub the transient to the other tube as protection, much like back to back didoes protect meters from transients. Just a guess.
Normally when we see the back to back diodes they are a form of protective shunt. Typically placed across the input of receivers, meters and other sensitive devices to prevent overload. In those cases the diodes are placed in parallel, across the signal to ground. Here we have the diodes in series with the bias and drive.
I see at least three problems with them in this location. It adds .6 volts of negative bias on V2 making it draw less idling current compared to V1. It places a .6 volt drop in the RF drive feeding V2. Who knows what type of switching transients the two diodes could be having on the signal feeding V2? One diode has forward bias current supplied through the 8.1 volt Zener while the other is switching on and off with every half cycle of RF drive...