The 6C33CB has an operating environmental temp limit of 260ºC, so 180ºC should not be a problem.
These two just had more gas than the getter could absorb.
WOW, I missing completly this detail.
Seems this drastically closes the case of baking tubes.
Some old 6C33 versions had a max ambient temp of +70ºC.
So a safe temp to baking tubes could be +50ºC for one hour seems to me.
The maximum environmental temperature that I found in an English datasheet could be a error in translation from Russian to English.
Normally datasheets list the maximum hottest point on the glass during operation, typically around 250ºC. Could be that this was meant instead of the temperature around the valve.
An ambient temperature of 50-70ºC should be read as the temperature around the valve when it is in operation, and producing heat by itself as well. If the difference in temperature between the valve and the air around it is relatively small, the valve cannot get rid of its heat and will overheat. During baking the valve is not producing heat of its own, so this does not apply.
BTW, I have found conflicting information about the temperature at which barium will react with the different gasses: with oxigen and carbondioxide it will react at room temperature. For the reaction with nitrogen I found comments that the reaction will take place at room temperature, and another site lists a temperature range of 200-500ºC. The last is the only explaination that I can come up with for the improvement of the vacuum with baking.
Anything that buffers airflow over the glass in the region of the getter will raise the local thermal resistance. Perhaps a layer of something temperature resistant for say 120C (eg. nomex paper, or semiconductor thermal pad for a T03P), followed by a layer of foam of any kind, would easily achieve the desirec effect. Keep the pad in place with a loop of wire or something easy to slip-on/remove.
Thanks for this information.
When I see those getters disappearing after two baking in two tubes I though it was not good.
I feel 180ºC is too much for the 6C33 getter.
Yes, raising the getter temp is aiming for a middle ground, which is effectively the tube dissipating a nominal power level from the anode - and hence why nearly all of us don't worry about applying some form of baking. If the heater is the only dissipator in the valve, then adding thermal insulation around the getter should move the getter up in temp towards that middle-ground.
i understand that any baking is dangerous, it can release "old" gases (?)
I have not conducted any experiments with "baking" tubes, but this thread did prompt me to make a crude burn-in jig for some old 6SN7s I wanted to test - I wired them as diodes (grid tied to plate), a 1k resistor in series with each plate for current balancing, dialled in 20mA per section with bench supply and let one sit there for a while (~1 hour). Initial test showed some grid leakage and imbalance between sections on AVO MkIV. Set up same tube in jig as described, then retested, leakage gone, and readings showing similiar Gm and plate current between sections. IMHO it needed to be heated for a time with plate voltage applied. Interesting idea to keep perhaps, for dealing with used tubes that have been stored for a while.
I've gotten similar results (albeit w/different valves) using nothing but filament and, although I do "burn-in" afterwards - (≈½ max. W for 24-48 hrs) - the difference in grid leak is measured before that.
I built a little rig that lets me "heat" 4 at a time. I put a set of valves in before I go to bed, then apply B+ in the morning, measure grid leak, transconductance, and µ a couple hours later (for matching) and leave 'em there till the next day (or the day after if I don't have too many to do).
I buy a lot of NOS Russkies - (love 'em!) - and while I used to have trouble with maybe 10% of 'em, I haven't had a problem with any since starting this regimen.
Last edited:
Hey radiotron, I just re-read your post and I'm intrigued by your results.
My little heater rig takes care of grid leak and any other issues that 30+ years of "cold storage" may have caused, but I'm not getting balanced sections out of the deal.
So, by shorting your supply with diode-strapped tubes, you're essentially burning them in A2 (positive grid), albeit with a low (I assume your bench supply is 25-50V max) current-limited B+.
I've been running 'em high E/low I - (e.g. I'm running 10K per anode @ 300V on my 6Н6П's which is about 11-12mA) - for burn-in. Do you suppose that accounts for the difference? (i.e. with the grid passing current as well as the plate, are the 1K resistors re-distributing the balance between the two permanently?)
So, how much does the balance improve? Can you quote some "before and after" figures, please?
Also, if paralleling sections can balance them, could you not parallel a couple valves as well, and end up with a matched set this way?
My little heater rig takes care of grid leak and any other issues that 30+ years of "cold storage" may have caused, but I'm not getting balanced sections out of the deal.
So, by shorting your supply with diode-strapped tubes, you're essentially burning them in A2 (positive grid), albeit with a low (I assume your bench supply is 25-50V max) current-limited B+.
I've been running 'em high E/low I - (e.g. I'm running 10K per anode @ 300V on my 6Н6П's which is about 11-12mA) - for burn-in. Do you suppose that accounts for the difference? (i.e. with the grid passing current as well as the plate, are the 1K resistors re-distributing the balance between the two permanently?)
So, how much does the balance improve? Can you quote some "before and after" figures, please?
Also, if paralleling sections can balance them, could you not parallel a couple valves as well, and end up with a matched set this way?
Last edited:
Hi digitrax, you are correct about the power supply.
Re balanced sections after the diode treatment, I'm not claiming to make something out of nothing, once a tube is low emission, nothing much to be done, but in the case of a stored tube, I reasoned that under initial test after storage, a cathode in a dual section tube could simply be a bit slower to reach maximum emission compared to other.
"Creeping" current and Gm readings on AVO seemed to confirm this, making accurate measurements difficult, so I figured I could speed up the process by heating the tube and pushing max rated current through the cathode (20mA for a 6SN7) for an hour or so, giving it some "exercise" if you like.
Results with one tube: 9mA and 7mA before burn-in jig, both sections at 9mA afterwards. similiar result with Gm, that settled as well, both now ~2.8 mA/V, as it should, since Gm changes with plate current.
Next job is to alter the jig to test the power pentode section of some tired old 6GW8s I found in a box, there are several, and I have already prelim test numbers on them, so we'll see if this idea is worthwhile with a different tube altogether.
Re balanced sections after the diode treatment, I'm not claiming to make something out of nothing, once a tube is low emission, nothing much to be done, but in the case of a stored tube, I reasoned that under initial test after storage, a cathode in a dual section tube could simply be a bit slower to reach maximum emission compared to other.
"Creeping" current and Gm readings on AVO seemed to confirm this, making accurate measurements difficult, so I figured I could speed up the process by heating the tube and pushing max rated current through the cathode (20mA for a 6SN7) for an hour or so, giving it some "exercise" if you like.
Results with one tube: 9mA and 7mA before burn-in jig, both sections at 9mA afterwards. similiar result with Gm, that settled as well, both now ~2.8 mA/V, as it should, since Gm changes with plate current.
Next job is to alter the jig to test the power pentode section of some tired old 6GW8s I found in a box, there are several, and I have already prelim test numbers on them, so we'll see if this idea is worthwhile with a different tube altogether.
Thanx, rt!
I'm a little reticent to give three-decade-old stored valves a "full current workout" straight outta the gate, but think that a 3-stage (filament only, ½ Wmax, full current/low volt) workout could be the best of all worlds.
It wouldn't be hard to add another switch to my jig to short grid to plate and throw, say 1.2K, in parallel w/the 10K Rp's and flip the "HT in" between my B+ supply and the bench supply (pre-set to Imax of both sections).
Also, I take your point about the balance: it would be foolish to assume you're doing any sort of actual redistribution of currents (on an ongoing basis), but if you have a valve whose sections have fallen out of their original state of balance (through age/disuse), then it only makes sense that breaking them in will bring them back to where they were/should be.
I was initially thrown by the fact that you got similar Gm & I readings after your "workout" - (but then, of course, you're dealing with NOS SN7's, as opposed to the beautiful, but wildly-varying "Russian wonders"
).
Anyway, thanks for sharing. Hopefully, between us and all the others who've contributed, we can save some innocent valves from being baked to death.*
*Note: Not intended as a criticism to anyone who "bakes"; as I said, I used to do it as well.
It's just that I've found it's not absolutely necessary. Other alternatives are equally effective.
I'm a little reticent to give three-decade-old stored valves a "full current workout" straight outta the gate, but think that a 3-stage (filament only, ½ Wmax, full current/low volt) workout could be the best of all worlds.
It wouldn't be hard to add another switch to my jig to short grid to plate and throw, say 1.2K, in parallel w/the 10K Rp's and flip the "HT in" between my B+ supply and the bench supply (pre-set to Imax of both sections).
Also, I take your point about the balance: it would be foolish to assume you're doing any sort of actual redistribution of currents (on an ongoing basis), but if you have a valve whose sections have fallen out of their original state of balance (through age/disuse), then it only makes sense that breaking them in will bring them back to where they were/should be.
I was initially thrown by the fact that you got similar Gm & I readings after your "workout" - (but then, of course, you're dealing with NOS SN7's, as opposed to the beautiful, but wildly-varying "Russian wonders"
).Anyway, thanks for sharing. Hopefully, between us and all the others who've contributed, we can save some innocent valves from being baked to death.*
*Note: Not intended as a criticism to anyone who "bakes"; as I said, I used to do it as well.
It's just that I've found it's not absolutely necessary. Other alternatives are equally effective.
Last edited:
As in regular use the tubes and getter also gets hot, I would say not all bakings are bad, only those in hi temperature as happen to me.
I think a low temp 50ºC to 100ºC would be OK.
Not to kick a dead horse, but I found this quite by accident:
http://www.cpii.com/docs/related/23/Tube Maintenance.pdf
It's from Econco, a company whose sole function is rebuilding/restoring tubes. AFAIK, Econco is the only company in the world who restores the mighty RS3500CJ - world's most powerful triode (12V heater @ 1690A, 300,000W plate dissipation).
Anyway, here is what they have to say about degassing:
http://www.cpii.com/docs/related/23/Tube Maintenance.pdf
It's from Econco, a company whose sole function is rebuilding/restoring tubes. AFAIK, Econco is the only company in the world who restores the mighty RS3500CJ - world's most powerful triode (12V heater @ 1690A, 300,000W plate dissipation).
Anyway, here is what they have to say about degassing:
So apparently, the "filament-only; then half-power" method is what the pros use as well. No baking needed.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.