| Posted by grindfix | February 02, 2022 - 08:47 am |
| Thanks, Steve! |
| Posted by Steve L. | January 28, 2022 - 09:00 am |
| Hi Grindfix, Yes, this method is applicable to the 2A3 and 45 tubes. As you know, those tubes have directly heated cathodes, meaning the heater filament itself is used as the cathode. Hence, heater voltage can potentially interfere with bias voltages. The heater does not have to be AC powered. Yes, you could power the heater from a floating supply but that doesn't really make the problem of heater voltage interference go away. You still have to provide a way to keep the cathode at ground or a fixed voltage relative to ground. Perhaps the simplest way to handle this for test purposes is to connect the heater to ground and +2.5VDC. Then the effective cathode voltage is halfway between or +1.25V. Effective grid and plate voltages are thus relative to +1.25V. For example, to test a 2A3 tube with the databook values of grid at -62V and plate at +300V, you would actually set them to -60.75V and +301.25V. In the test procedure though, the grid voltage from the table is just used as an initial value and it's quickly adjusted to get the target quiescent plate current (Iq), which is 80mA in this case. If you want to measure the effective grid voltage which produced Iq, remember to offset the grid voltage by 1.25V. So if the grid voltage turned out to be (say) -58V, that was effectively -59.25V. I'll be happy to help if you have any other questions about this. |
| Posted by grindfix | January 26, 2022 - 03:06 pm |
| Hi Steve, Could Power Triodes of 2A3/45 family be tested using the above method? Does the heater have to be AC powered? Could the heater be powered by a (floating) independent isolated DC source? Thanks! |
| Posted by Steve L. | January 09, 2022 - 09:16 am |
| Hi Grindfix, Happy to help. I will gather-up what I have on the electronic RL and post that here either later today or tomorrow. Yes, the rheostat should be fine for the 6BQ5. Ed note: Grindfix contacted me by email using the address on the About page (shown when you click the "About..." button near the top of the home page.) You can read our discussion of a circuit for the electronic load and other topics here: http://www.tronola.com/misc/Thread about an electronic plate load for testing power tubes and more.pdf |
| Posted by grindfix | January 02, 2022 - 10:17 am |
| "I can elaborate on that if you wish..." Please, if you don't mind. I don't have any 6550 to work with but would like to make the tester versatile and include the most common tube types. I have a bunch of 6BQ5 types to go through and for the moment rheostat should work but if you have a Mosfet circuit to share, I'd like to consider it. Thanks for your help! |
| Posted by Steve L. | January 02, 2022 - 09:50 am |
| Hi Grindfix, Yes, the 10K grid resistance should be fine for all receiving tubes of which I'm aware. Even the big KT120 and KT150 tubes allow up to 51K. And yes, the 50W rheostat should be okay. At this frequency and low impedance, the inductance isn't a concern. The only thing is, heating and tempco are always something one can do better with. I would make an effort to keep it cool (say with a small fan). You could try to keep the test short and could measure RL right after the test. Also, I get worried if I'm using only a small part of the rheostat's range. One couldn't expect to dissipate 50W if the 2.5K rheostat were set to 875R, for example. Pessimistically, you might consider it as good for 17.5W in that case. 50W x (875/2500) = 17.5W What I really had in mind was building an electronic load, which would let me set RL accurately with low-level resistors and leave the power dissipation in a beefy MOSFET. I can elaborate on that if you wish... |
| Posted by grindfix | January 02, 2022 - 12:46 am |
| Hi Steve, Thanks for your answer. Do you recommend using the same 10k grid resistance for all power tube types? Also, are there any objections to using inductive wirewound ceramic base 2.5k 50w rheostat for RL? Thanks again. |
| Posted by Steve L. | December 25, 2021 - 06:24 pm |
| [Edit: After posting, I see that this and your "found it" followup crossed in the posting...] Hi Grindfix, Great question! To develop the parameters for another power tube type, we go to the manufacturer's datasheet and find an example which hopefully exercises the tube at or near its max capabilities. (They liked to show off what their tubes could do, so that's usually easy to find.) We also have to impose the limitations of our test equipment at this point. That has generally meant max plate supply voltage of 400V. Also, we usually want fixed bias operation. For example, looking at the (excellent) GE 6550A datasheet p.3, I chose 400V plate voltage. That gives RL=3500R. They usually specify the plate-plate impedance of the whole primary winding. Each individual plate sees a quarter of that or 875R in this case. So there is your RL. Then we need to determine the ideal power output for the tube and that is done with the Average Plate Characteristics curves on p.4, as described under our article subheading, "How is the ideal output power (Pouti) determined?" Note that the typical operation example had screen voltage at 270V but we must settle for the lower chart which has screen voltage Ec2=300V, since that's what we have data on. The Pouti comes from the chart, so it will be exactly right. I'll be happy to help if you have further questions. |
| Posted by grindfix | December 25, 2021 - 06:14 pm |
| Found it! |
| Posted by grindfix | December 25, 2021 - 05:39 pm |
| Hi, How is RL calculated for other output tube types? Thanks |