cathode switching for power tube selection?
Moderators: pompeiisneaks, Colossal
Re: cathode switching for power tube selection?
do you really expect to need 5W resistor there? I built this amp with 10K 1/2W resistor there and it hasn't popped yet. I didn't measure the cathode V with the 10K to calculate the power dissipation in that resistor.
Re: cathode switching for power tube selection?
Hi Josh,krash wrote:do you really expect to need 5W resistor there? I built this amp with 10K 1/2W resistor there and it hasn't popped yet. I didn't measure the cathode V with the 10K to calculate the power dissipation in that resistor.
I used 5w in mine and I'll try to explain how I got there. My cathode switching is based on a circuit from TUT1 (fig.6-60) which is "Combined Electronic Stand-By & Output Tube Selection". It uses MOSFET's as the switching element. To further complicate matters I automated the standby so that after the amps been powered up for 30 seconds it comes off standby. Did i say i move from simple to complicated at a breakneck speed.
RG Keen has a standby switching circuit using MOSFET's very similar to what I used (on the cathodes anyways) and in the accompanying article he states to use a 3-5watt resistor across the MOSFET. (Although the schematic shows a 2watt.) I used 5watt resistors on the cathodes so figured i'd just go with 5watt for the 10k as well. So all in all there was no expectation, no thought, just followed. (Nowhere in TUT could I find a rating for the 10k.)
The Keen article. Mosfet Follies, scroll down to A MOSFET Standby Switch.
http://www.geofex.com/Article_Folders/m ... tfolly.htm
I haven't done a measurement and calculation either. Will do it next time I open it up, but that could be quite some time down the road. It does continue to work fine, no issues with it so far.
dave
Re: cathode switching for power tube selection?
Dave,davent wrote: The schematic shows switching between two single output tubes but it would be set up the the same way for two pairs of output tubes.
dave
Thanks for sharing your thoughts on your 18W. That helps - I think I will add that to my next 18W build.
What schematic are you referring to in the above quote?
-John
Re: cathode switching for power tube selection?
thanks for the info, Dave.
One would think that if 5W is truly called for, and I put in a 1/2W, then I would have smoked my 1/2W resistor very quickly, just in a few minutes if running 10x the amount of current through it for which it is rated. But I don't plan to open this amp back up or do any extended testing. I am just going to have faith that it will work. Worst thing that could happen is the resistor pops open and then it just opens those cathodes when switched to 1/2 power mode. No biggie.
Anyway I thought for a few minutes about how much current to expect through this cathode and gave up. I came up with anywhere from 2 watts of power across this R on down. What's unknown is how the tube will conduct from the cathode with a 10K resistor connected. This question is not illuminated by the tube data sheet.
Since the screen grid in my amp is connected to a 2.2K resistor I figure most of the plate current will flow from the screen grid, and likely with a 10K resistor connected to the cathode, it will not be sufficiently negative with respect to the screen grid to source a significant amount of current. Also I figure that it will be far and away positive with respect to the control grid and the tube will be in cutoff the entire time anyway so I wouldn't expect it to conduct any electrons from the cathode. I freely admit that my understanding of how this tube is going to conduct in anything but typical fashion is completely inadequate for analysis of this problem.
I guess I probably SHOULD open it up and measure it before delivering the amp to the customer.
One would think that if 5W is truly called for, and I put in a 1/2W, then I would have smoked my 1/2W resistor very quickly, just in a few minutes if running 10x the amount of current through it for which it is rated. But I don't plan to open this amp back up or do any extended testing. I am just going to have faith that it will work. Worst thing that could happen is the resistor pops open and then it just opens those cathodes when switched to 1/2 power mode. No biggie.
Anyway I thought for a few minutes about how much current to expect through this cathode and gave up. I came up with anywhere from 2 watts of power across this R on down. What's unknown is how the tube will conduct from the cathode with a 10K resistor connected. This question is not illuminated by the tube data sheet.
Since the screen grid in my amp is connected to a 2.2K resistor I figure most of the plate current will flow from the screen grid, and likely with a 10K resistor connected to the cathode, it will not be sufficiently negative with respect to the screen grid to source a significant amount of current. Also I figure that it will be far and away positive with respect to the control grid and the tube will be in cutoff the entire time anyway so I wouldn't expect it to conduct any electrons from the cathode. I freely admit that my understanding of how this tube is going to conduct in anything but typical fashion is completely inadequate for analysis of this problem.
I guess I probably SHOULD open it up and measure it before delivering the amp to the customer.
Re: cathode switching for power tube selection?
Morning John,jhaas wrote:Dave,davent wrote: The schematic shows switching between two single output tubes but it would be set up the the same way for two pairs of output tubes.
dave
Thanks for sharing your thoughts on your 18W. That helps - I think I will add that to my next 18W build.
What schematic are you referring to in the above quote?
-John
There is an attached schematic to my post but to see attachments you have to be logged into the Ampgarage site. So... I'm guessing you may need to to register on Ampgarage, if i remember correctly painless and quick.
Here's a link to a better representation of what i put up. Again probably need to be registered on 18watt to view.
http://www.18watt.com/modules.php?set_a ... _photo.php
dave
Re: cathode switching for power tube selection?
Morning Josh,krash wrote:thanks for the info, Dave.
One would think that if 5W is truly called for, and I put in a 1/2W, then I would have smoked my 1/2W resistor very quickly, just in a few minutes if running 10x the amount of current through it for which it is rated. But I don't plan to open this amp back up or do any extended testing. I am just going to have faith that it will work. Worst thing that could happen is the resistor pops open and then it just opens those cathodes when switched to 1/2 power mode. No biggie.
Anyway I thought for a few minutes about how much current to expect through this cathode and gave up. I came up with anywhere from 2 watts of power across this R on down. What's unknown is how the tube will conduct from the cathode with a 10K resistor connected. This question is not illuminated by the tube data sheet.
Since the screen grid in my amp is connected to a 2.2K resistor I figure most of the plate current will flow from the screen grid, and likely with a 10K resistor connected to the cathode, it will not be sufficiently negative with respect to the screen grid to source a significant amount of current. Also I figure that it will be far and away positive with respect to the control grid and the tube will be in cutoff the entire time anyway so I wouldn't expect it to conduct any electrons from the cathode. I freely admit that my understanding of how this tube is going to conduct in anything but typical fashion is completely inadequate for analysis of this problem.
I guess I probably SHOULD open it up and measure it before delivering the amp to the customer.
You've got me in way over my head here!
If this resistor were to die, does it become a short or does it show infinite resistance, so no connection? If it became a short it would mean that pair of tubes were always on and that I would imagine could be a problem for the OT, it would always see the two pairs of output tubes on instead of one pair at a time. If in dying the resistor became infinite, when the corresponding pair of tubes were switched out, the absence of the 10k would mean that the pair of tubes were totally off, no cathode current. Does this make any sense?In the original TUT schematic using a switch to select tube pairs, there were no resistors shown across the switch but it was suggested as a way to minimize popping when the switch was changed. A discussion on 18watt talked about the possibility of cathode stripping if a pair of tubes were left hanging with no connection of the cathodes. There, it was suggested to use the 10k's as a means to keep both pairs of output tubes on, one pair ever so slightly and thus prevent cathode stripping. I think it was in that thread where I came across the Keen link.
Scratching my head,
dave
Re: cathode switching for power tube selection?
the resistor would likely fail open. Normally resistors don't fail short.
In this case, the amp is a 4xEL84 amp, and the switch switches between a 68 ohm cathode R connected to all 4 tubes, or a 150 ohm connected only to two tubes, and 10K to the other two. It's basically a "half power" switch. AKA a "barely audible difference" switch.
If the 10K resistor failed short, then it would result in the 150 ohm resistor connected all the time and you would draw all of the bias current through the failed resistor. It would not remain in this fail-short mode very long. The result would be a combined cathode resistance of about 47 ohms which would be pretty hot bias for the amp, and it would go through power tubes real quick but mostly the owner of the amp would notice that the half power switch stopped working (stopped having any effect) and he'd send the amp back for a warranty repair
But if you ran it this way very long, then drawing that much current through the failed resistor would certainly make it eventually fail open.
So on that note, I think I'll open it up and measure, just in case. I don't want it coming back. If it turns out that it's under-rated, then I'll just take out the 10K resistor and take my chances with cathode stripping.
In this case, the amp is a 4xEL84 amp, and the switch switches between a 68 ohm cathode R connected to all 4 tubes, or a 150 ohm connected only to two tubes, and 10K to the other two. It's basically a "half power" switch. AKA a "barely audible difference" switch.
If the 10K resistor failed short, then it would result in the 150 ohm resistor connected all the time and you would draw all of the bias current through the failed resistor. It would not remain in this fail-short mode very long. The result would be a combined cathode resistance of about 47 ohms which would be pretty hot bias for the amp, and it would go through power tubes real quick but mostly the owner of the amp would notice that the half power switch stopped working (stopped having any effect) and he'd send the amp back for a warranty repair
But if you ran it this way very long, then drawing that much current through the failed resistor would certainly make it eventually fail open. So on that note, I think I'll open it up and measure, just in case. I don't want it coming back. If it turns out that it's under-rated, then I'll just take out the 10K resistor and take my chances with cathode stripping.
Re: cathode switching for power tube selection?
Don't mean to hijack this thread, but the site below has a lot of good info on amp building in general, and one pertinent discussion on the "myth" of cathode stripping.
Go to his MUSIC page and use the Search on his site for "How to Use a Standby Switch"
http://www.psg.com/~dlamkins/
Go to his MUSIC page and use the Search on his site for "How to Use a Standby Switch"
http://www.psg.com/~dlamkins/
Re: cathode switching for power tube selection?
I have a Blue Angel. It 'switches on the fly' without popping or banging.
Another cool cathode switching thing is seen in the 50/100 switch of the Dumble. The non-used tubes have the cathode connected to a resistor and a LED that lights up with a little residual cathode current, indicating '50 watt' mode.
Re: cathode switching for power tube selection?
the LED is a cool idea.
BTW I think in the amp I just built with the half power switch (using this method for cathode switching) has mostly a feel difference and not a volume difference going from full to half power. This crosses over into the "100 watts" thread. 4 tubes sound and feel different than two, that's the main difference.
BTW I think in the amp I just built with the half power switch (using this method for cathode switching) has mostly a feel difference and not a volume difference going from full to half power. This crosses over into the "100 watts" thread. 4 tubes sound and feel different than two, that's the main difference.