4 x used 5867A (TB3/750) triodes

[multi-volti]

Anyone interested in 4 used 5867A (a.k.a. TB3/750)? No, I didn't mean 5687A.

350 W plate dissipation directly heated triode, 4 kv max, I think, transconductance 5 mA/V, mu 25, 5V, 14.1 A filament.

Philips Datasheet:
http://tubedata.itchurch.org/sheets/030/t/TB3-750.pdf

Typical application is RF heating. If you're into RF amps, a guy in the Netherlands configured a 'universal amp', with this tube (maybe a pair of them). http://www.xs4all.nl/~pa0fri/Lineair...0/FRI750uk.htm
The schematic shows the bias needed. They are NOT zero-bias like 3-500Z's.

1 is marked Siemens 5867A/RS1026. The others are unmarked because the guy I got them from thought he could fool people into thinking they were 3-400's or 3-500Z's. (same socket - Giant 5 pin E.F. Johnson). They are all in Amperex boxes.

I took them to a broadcast/industrial tube place in Northern NJ around 1980 and he had a Franken-tester he built for large tubes. He did what he called an 'emission test' (?), and rated the 3 Amperex tubes at "95% emission", and the Siemens at "80%".

They have been in a box in my (dry) basement ever since.

People have asked about DHT SE amps with these, but good luck with the OPT!

I need to get $100 for the 4 + shipping (8-10#, roughly $15 in US).

Thanks

Jeez, no wonder no interest on eBay (listing ended), I forgot to include the pictures!

http://murray.uptown-gallery.com/eba...750/index.html

Murray

[HollowState]

Originally posted by multi-volti

I took them to a broadcast/industrial tube place in Northern NJ around 1980 and he had a Franken-tester he built for large tubes. He did what he called an 'emission test' (?), and rated the 3 Amperex tubes at "95% emission", and the Siemens at "80%".

Hi Murray,
That would be JFF Research in Nutley which was run by the late John Fikus who was a good friend of mine. Two of the custom rack testers now reside in my basement after being rebuilt by myself. The one shown is a universal class A tester capable of 1500 watts dissipation. All elements are monitored and transconductance is measured at 1000 hertz. Sockets are interchangable plug-ins. Shown is the 5868, big brother of your 5867s

Victor

[kevinkr]

Pretty cool tester Victor.. Now I need to whip up something similar to test some 211 with..

[multi-volti]

HS:

Sorry to hear about John. I half wondered if someone would recognize my story.

I initially was going to do a RF project, then thought about audio, but it just doesn't seem practical to me...but I don't always try to be practical.

Something, someday...

[tubelab.com]

I initially was going to do a RF project, then thought about audio, but it just doesn't seem practical to me...but I don't always try to be practical.

I also have 4 of these tubes with two sockets. I have been saving them for a project that will never happen. I have several 833A's also. I found out when I decided to try an 833A SE amp that the OPT is the hard part not the tubes. There is also a little matter of about 1KW of heat energy that must be disposed of. This is a real problem in south Florida. I could be open to interesting trades for either tubes.

Now I need to whip up something similar to test some 211 with..

I thought that this was easy, just use my amp. Transconductance could be measured by varying the bias voltage and watching the current meter. Everything worked fine until I found an extremely gassy 211. The purple lightning bolt shot through the tube slapping the current meter needle all the way over (300 mA) and zapping the life out of the mosfet driver.

[kevinkr]

Originally posted by tubelab.com

<snip>

I thought that this was easy, just use my amp. Transconductance could be measured by varying the bias voltage and watching the current meter. Everything worked fine until I found an extremely gassy 211. The purple lightning bolt shot through the tube slapping the current meter needle all the way over (300 mA) and zapping the life out of the mosfet driver.

Hi George,
Yeah, I haven't built the amplifier yet - still scrounging around for the parts required. I'm going to use my isco 494 to provide plate voltage and either the lambda or a heath supply to provide the bias. I have enough meters to measure plate voltage and current along with the bias voltage. (May also measure grid current.) Hopefully I will not have the lightening bolt experience... I have a couple of VT4C and will be getting more in return for testing a whole bunch of them.. I just need to find a 10V -3A filament transformer and I'm good to go.. I'll take a couple of measurements and calculate the transconductance.

[HollowState]

Originally posted by kevinkr

I have enough meters to measure plate voltage and current along with the bias voltage. (May also measure grid current.) Hopefully I will not have the lightening bolt experience

Definitely use a microammeter in the grid bias circuit to monitor current as this indicates any gas present. I use a 200uA movement with back to back rectifiers across it for protection.

The acid test for general condition is to lower the filament 10% under test and watch plate corrent. A good tube will not move very much at 10% down. Only a few percent. If plate current falls a lot and acts "spongy" as the filament drops, the tube is weak and is near (or at) end of life. There is no fine line regarding plate current fluctuation with this test. It is a judgment call that will become evident through continued testing. You develope a mental and visual curve.

I measure transconductance by amplifing 1000~. I feed the grid with 1 volt rms, and measure AC signal in millivolts across a 10 ohm resistor at the negative side of the plate supply just above ground with an HP 3400A. (1mV=Gm of 1000)

You will need a well filtered supply to avoid measuring hum. In my setup I wanted the plate supply to fall rapidly when I shut it off for volume testing. So I use minimal filtering and bleeders. Because of this, I measure the output across the 10 ohms through a narrow 1000~ bandpass filter. (A UTC BMI-1000) This filter has a very high Q which raises the the voltage at it's output, so I must lower it to the correct level with a 10K 10 turn pot.

Regarding gassy tubes, yes they're out there. I've learned to raise the plate voltage up slowly to avoid whacking everything. This is also why I use simple unregulated supplies.

As a aside, there's a seller on feebay that uses a small tester to help sell tubes. I dont know the particulars, but it looks cute, so food for thought.

Victor

[multi-volti]

Say, George,

to relax the requirements on OPT fidelity, what about a guitar amp for a 5867A (leaving some spares!)?

There are still some unpleasant insulation issues in such a transformer...I have no idea what the primary impedance would need to be...someone once suggested to me a 15k ohm primary...
assume 8 ohm primary...maybe 16 would be easier...

Going out on a limb...(15000/8)^0.5 = 43.3:1 turns ratio. Not much closer to a solution, but maybe spare HV xfmr?

Murray

[multi-volti]

I got interrupted...(analog interruption)

15k:8 ohms = 43.3:1 turns ratio: a 120 primary would be about 5200 V secondary.

15k:16 ohms = 30.6:1 turns ratio: 120 primary and 3675 V secondary.

I have a 5500:208 transformer I probably couldn't lift into my car to drive anywhere to ship, so it's probably staying at home...that would be 26:44:1 turns ratio, 699:1 Z ratio, 16 ohm secondary for 11k primary. No center tap...I'm sure it isn't going to like 200-300-400 mA DC for SE. and can't do p-p.

[tubelab.com]

to relax the requirements on OPT fidelity, what about a guitar amp for a 5867A (leaving some spares!)?

I had an OPT custom wound for the 833A. It was a 5K ohm primary gapped for 300 mA. I intended to use it in a hifi amp, but the frequency response wasn't good enough. I only had one made, so a stereo amp is out of the question. I also have the power transformer from a Harris transmitter (obtained from another forum member). I have dreamed of building a 200 watt SE guitar amp since I have all of the parts and have already breadboarded the thing. It sounded awesome and could be heard 4 blocks away. It is currently pretty far down on the "projects" list since, as you pointed out, it is rather impractical.

Not much closer to a solution, but maybe spare HV xfmr?

When I was a young teenager I made guitar amps from old TV sets (they were freely discarded). I used the vertical output transformer for an OPT in SE "Champ" style amps. They are gapped for SE operation. I used power transformers for OPT's in push pull amps. Connect the plates to the HV winding, CT goes to B+. All filament windings are wired in series and connected to the speakers. I really didn't understand impedance matching or frequency response then, but some of those amps ROCKED. With several speakers connected in parallel, the impedance was probably close.

Some power transformers may be OK for OPT use in a P-P guitar amp. Some are really bad. The worst is the transformer from a microwave oven. HF 3db point is about 200 Hz. The big industrial control transformers might be another possibility, but I haven't tried them yet. In any case some experimentation will be needed to determine the required impedance, and the suitability of a given transformer. All the usual "don't kill yourself" warnings apply.

Either way, if I decide to build the "big one", I will use the 833A. I have several known good ones, they look cool, and I have already seen 200 watts flow frome one in a SE amp.