Hi!
I have so far bought five tube testers, Hickok 800A, AVO mk3, AVO mk4, Longhin 460, and last a L3-3 from 1972 in mint condition.
I had some hope before each purchase to get the possibility to make test and matching the big 6C33C tube. But i now understand any tester are not powerfull enough.
So my question before I try to start to design and build a tester who have the specification to make the job, is it someone who know where to find a working solution for this problem?
All tube shops say they make tests and matching before delivery, but my experience is the tubes are complete hopeless to match as they are made for voltage regulation and have great differences inside own triod pair and between each other.
But some way to check mA/V must be possible.
Thanks
I have so far bought five tube testers, Hickok 800A, AVO mk3, AVO mk4, Longhin 460, and last a L3-3 from 1972 in mint condition.
I had some hope before each purchase to get the possibility to make test and matching the big 6C33C tube. But i now understand any tester are not powerfull enough.
So my question before I try to start to design and build a tester who have the specification to make the job, is it someone who know where to find a working solution for this problem?
All tube shops say they make tests and matching before delivery, but my experience is the tubes are complete hopeless to match as they are made for voltage regulation and have great differences inside own triod pair and between each other.
But some way to check mA/V must be possible.
Thanks
Yes it is..
You will need a big supply (0 - 300V, 300mA typical) to furnish plate current, and a meter to measure the plate voltage and current accurately.
In addition you will need a bias supply (0 - 200V, > 10mA) and a meter to measure the grid voltage.
Next you will need to wire up some sockets in such a way that you can test each section independently.
You will need a very large filament transformer to test and match more than one tube at a time. I'd wire a pair fully in series for 25.2V operation using a 25.2V/4A transformer, and if center tapped I would wire the center tap to the connection between the two sockets.
You will measure the plate current at a fixed plate voltage for a variety of grid voltages, you will repeat this at several different plate voltages.
You can record these values in tabular form and using them you can calculate the transconductance for the various operating points you have chosen.
Start with the maximum value of grid bias when you apply the plate voltage. Decrease in 5V or 10V increments and record the plate current at each grid voltage value. Do this until you reach something approaching the rated dissipation, but do not exceed it. Make sure that the grid voltages in all sets of measurements are the same each time..
This is a very laborious way of doing it, and you must do it for each and every tube so I would set it up to allow you to test two tubes at a time. You can wire all of the grids in parallel and just switch the plate lines, but make sure that the grid supply can supply any grid current that may occur - or verify as you debug that no significant grid current flows under any circumstances.
Do the [delta mA]/[delta V] for a number of different plate voltages and current deltas and make note of the specific conditions under which these transconductance calculations apply. Do this for every section of every tube and try to select ones that match at several different points.
When I was working on my OTL amps I purchased a whole pile of 6C33 - roughly about 32 of them, and was barely able to get two reasonably well matched non-identical sextets out of that quantity.
For the measurements I ran mine at 100 - 250V and currents of 50 - 300mA as memory serves. It's a really good idea to run them for a couple of hours with B+ applied and then about 24hrs at about half their rated dissipation before attempting to match them. Make sure this set up is safe and fuse every plate individually. Monitor very closely during the first few hours as emission stabilizes and the tubes burn in.
Take this as just a general suggestion, and be careful! Not responsible for fires, potential electrocution or other mis-haps. YMMV..
You will need a big supply (0 - 300V, 300mA typical) to furnish plate current, and a meter to measure the plate voltage and current accurately.
In addition you will need a bias supply (0 - 200V, > 10mA) and a meter to measure the grid voltage.
Next you will need to wire up some sockets in such a way that you can test each section independently.
You will need a very large filament transformer to test and match more than one tube at a time. I'd wire a pair fully in series for 25.2V operation using a 25.2V/4A transformer, and if center tapped I would wire the center tap to the connection between the two sockets.
You will measure the plate current at a fixed plate voltage for a variety of grid voltages, you will repeat this at several different plate voltages.
You can record these values in tabular form and using them you can calculate the transconductance for the various operating points you have chosen.
Start with the maximum value of grid bias when you apply the plate voltage. Decrease in 5V or 10V increments and record the plate current at each grid voltage value. Do this until you reach something approaching the rated dissipation, but do not exceed it. Make sure that the grid voltages in all sets of measurements are the same each time..
This is a very laborious way of doing it, and you must do it for each and every tube so I would set it up to allow you to test two tubes at a time. You can wire all of the grids in parallel and just switch the plate lines, but make sure that the grid supply can supply any grid current that may occur - or verify as you debug that no significant grid current flows under any circumstances.
Do the [delta mA]/[delta V] for a number of different plate voltages and current deltas and make note of the specific conditions under which these transconductance calculations apply. Do this for every section of every tube and try to select ones that match at several different points.
When I was working on my OTL amps I purchased a whole pile of 6C33 - roughly about 32 of them, and was barely able to get two reasonably well matched non-identical sextets out of that quantity.
For the measurements I ran mine at 100 - 250V and currents of 50 - 300mA as memory serves. It's a really good idea to run them for a couple of hours with B+ applied and then about 24hrs at about half their rated dissipation before attempting to match them. Make sure this set up is safe and fuse every plate individually. Monitor very closely during the first few hours as emission stabilizes and the tubes burn in.
Take this as just a general suggestion, and be careful! Not responsible for fires, potential electrocution or other mis-haps. YMMV..
Last edited:
Thank you for a good proposal. Connect four filaments in series was a brilliant solution, 24V heavy transformers are stock parts in my collection. And that in two (or more) groups give a capacity to burn in and test 4 tubes at same time. I think I should build an regulated/stabilized 25,2VDC 10A power supply to get best measurements.
And then i can use a variac and rectifier for anode voltage, up to +300VDC.
Grid voltage, -3 to -90V should I use signalgenerator and use a 1600 Hz frequence to get audio conditions?
Thanks
And then i can use a variac and rectifier for anode voltage, up to +300VDC.
Grid voltage, -3 to -90V should I use signalgenerator and use a 1600 Hz frequence to get audio conditions?
Thanks
Have you tried this:
Amplitrex Audio Products
It seems fit your requirements and also has the socket adapters for 6C33.
It provide up to 500VDC plate voltage but only 160 mA max, so probably not enough if you need 300mA as suggested by Kevinkr.
Amplitrex Audio Products
It seems fit your requirements and also has the socket adapters for 6C33.
It provide up to 500VDC plate voltage but only 160 mA max, so probably not enough if you need 300mA as suggested by Kevinkr.
We built our own tester for this about 20 years ago. It had a large filament circuit- I recommend 6.3V so no tube can hog the voltage in a series embodiment. This tube is quite sensitive!
We used a fixed 150V supply and used an voltmeter to monitor current across a cathode resistance, given a range of grid voltages applied. It works fairly well- we can match tubes and also weed bad ones out. We *did* have to sort out some idea of how a good tube behaved initially though...
We used a fixed 150V supply and used an voltmeter to monitor current across a cathode resistance, given a range of grid voltages applied. It works fairly well- we can match tubes and also weed bad ones out. We *did* have to sort out some idea of how a good tube behaved initially though...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.