Hi.
I have some 6336s that I want to use in an OTL amp.
Can anyone suggest a DIY circuit I could build to test and compare them?
Andy
I have some 6336s that I want to use in an OTL amp.
Can anyone suggest a DIY circuit I could build to test and compare them?
Andy
All you really need is a couple of good power supplies and meters.
Set up one supply to provide grid bias, set up the other to provide plate current. Measure both plate current and plate voltage.
Measure grid voltage.
To test allow 6336 to warm up, then apply grid bias about -200V to start, apply the B+ (up to 400V)
Reduce bias voltage in say 5 or 10V increments and record the plate current at each value. Don't exceed current or dissipation rating of the tube. (Some math required to determine limits)
Set bias back to -200V, decrease plate voltage by say 50V and do this again. Continue to reduce plate voltage down to say 100V or even 50V.
Reduce plate voltage in increments following above procedure until you have enough points.
You can graph the data in excel, on paper, or make a tabular form.
(Characteristic curve if you collect enough data)
Do the same for all tubes you are testing, don't forget to test each section and be particularly careful to make sure plate voltage is not applied when bias voltage is not present. (Muy importante..)
You can compare selected points or you can compare graphs.
You can also calculate the transconductance since you know change in grid voltage vs change in plate current.
This is a lot of work, but for OTL use in particular may allow you to match tubes quite closely over some part of their dynamic operating range, depending on how many samples you have. I used this technique quite successfully with 6C33 for my ill fated OTL project.
Bear in mind that curve tracers are dynamic and quick, and hence can operate outside of plate currents and voltages that are safe with the manual technique above. I.E. Don't exceed the static plate dissipation rating for an extended period of time.
Be sure to mark each tube so that you can identify it and its associated data, somewhere on the base shell should be fine.
A fuse in the plate circuit rated at 125% of the max current rating of the 6336 will prevent serious damage in the event of a mis-hap during testing.
Don't change anything in the set up until you have completed testing as differences in meter accuracy may invalidate the results unless your meters are accurate.
Sounds tedious, but it isn't all that bad if you do just a few at a time.
Set up one supply to provide grid bias, set up the other to provide plate current. Measure both plate current and plate voltage.
Measure grid voltage.
To test allow 6336 to warm up, then apply grid bias about -200V to start, apply the B+ (up to 400V)
Reduce bias voltage in say 5 or 10V increments and record the plate current at each value. Don't exceed current or dissipation rating of the tube. (Some math required to determine limits)
Set bias back to -200V, decrease plate voltage by say 50V and do this again. Continue to reduce plate voltage down to say 100V or even 50V.
Reduce plate voltage in increments following above procedure until you have enough points.
You can graph the data in excel, on paper, or make a tabular form.
(Characteristic curve if you collect enough data)
Do the same for all tubes you are testing, don't forget to test each section and be particularly careful to make sure plate voltage is not applied when bias voltage is not present. (Muy importante..)
You can compare selected points or you can compare graphs.
You can also calculate the transconductance since you know change in grid voltage vs change in plate current.
This is a lot of work, but for OTL use in particular may allow you to match tubes quite closely over some part of their dynamic operating range, depending on how many samples you have. I used this technique quite successfully with 6C33 for my ill fated OTL project.
Bear in mind that curve tracers are dynamic and quick, and hence can operate outside of plate currents and voltages that are safe with the manual technique above. I.E. Don't exceed the static plate dissipation rating for an extended period of time.
Be sure to mark each tube so that you can identify it and its associated data, somewhere on the base shell should be fine.
A fuse in the plate circuit rated at 125% of the max current rating of the 6336 will prevent serious damage in the event of a mis-hap during testing.
Don't change anything in the set up until you have completed testing as differences in meter accuracy may invalidate the results unless your meters are accurate.
Sounds tedious, but it isn't all that bad if you do just a few at a time.
- Status
- Not open for further replies.