|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
63.246.183.22
This should be a simple question for someone. I just want to confirm my understanding of the changes that occur in an output stage during the A1 to A2 transition. Please correct me if any of the following are incorrect:
When an output stage transitions from A1 to A2...
1. The output tube draws grid current.
2. The driver delivers power to the grid.
3. Average cathode current increases.
4. Average anode current does not increase.
Is this all good? If so, I would also be interested to know whether there are anomalies in practice that do cause anode current to change when the tube enters A2. I assume that would be measurable as THD, but maybe I haven't fully thought through the process.
--------------------------
Buy Chinese. Bury freedom.
Follow Ups:
Going to have to disagree on perhaps a bit of hair splitting.
As grid voltage increases incrementally, the plate current increases. I see no mechanism that disconnects this transconductance around the transition to a positive grid.
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
nt
Not sure I understand your meaning. Grid voltage always swings symmetrically around the static DC bias voltage. That means average grid voltage never changes. My question is whether average plate current behaves the same, even when the grid is driven hard enough to draw current on positive peaks. Extracting from this, if an amplifier exhibits low distortion when operating in the A2 region, can it be assumed that average anode current is essentially the same as in A1?
--------------------------
Buy Chinese. Bury freedom.
The average plate current should be a constant, within the linearity capability of the tube. So yes, to your question. I thought it was clear that the plate current is going to be dependent on grid-cathode voltage.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
"I thought it was clear that the plate current is going to be dependent on grid-cathode voltage. "
But that's the rub. No?
The driver stage has to handle the grid current without changing the grid-cathode voltage beyond that of the music signal.
I hope I said that right.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
The usual way of hooking up an A2 tube is by grid drive. If it is not in the 'always positive' region, the driver now needs to deal with a step-change in load when it goes positive. Driving the cathode offers a much more consistent load, but it is usually an order of magnitude lower than the grid when positive...LOL
The HY51 is a power tube just dying for cathode drive. I have enough for an amp project, and some 3k:500 output/interstage iron that will be brought to bear when I have the time.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
> > Grid voltage always swings symmetrically around the static DC bias voltage. < <
Your answer lies in this sentence. It's evident from the plate curves of the 211 tube, that the grid lines on either side of the 0V point are fairly symmetrical, so that along a 10K load line going from say -25V --> 0V --> +25V the increase in plate current appears to be fairy linear. The attached graph shows an OP at 60mA and 900V. Of course, at this operating point there will be no swinging past about +35V on the grid - you're already approaching zero volts at the plate.
Thanks, I do understand load lines and such. I'm just not sure the extent to which I can trust curves for this particular purpose. Your comments got me thinking about all this from a different angle; I think there's no reason I can't use grid voltage as a monitoring point rather than anode current. If distortion is low, they should follow each other. The only difference is that one lets me stabilize bias voltage, the other stabilizes bias current. Neither is necessarily better, as long as there's an adjustment. The OPA445 is a high voltage opamp that can run directly from my -80V bias supply. So, monitoring the grid itself is probably the simplest approach.
--------------------------
Buy Chinese. Bury freedom.
I know you understand loadlines, but I took what you said at face value and had a look at the plate curves. Think about it, it's the simplest and most obvious place to look - and admittedly it didn't come to me right away. I can understand what you're saying about trusting the curves, but then again why not? While tubes might've been hand picked for the purpose of obtaining and publishing test results, I would think that deviation from linearity within any specimen's performance envelope should be pretty much constant along a large sample of the type.
So, having said that and while I don't know what your proposed design looks like, a suggestion I have (haven't tried it so can't vouch for its results) would be to place a grid leak resistor between the power tube's grid and its cathode. There will be verry little current trickling at negative grid voltages and a bit more when the grid starts conducting, but it will effectively tie the grid voltage to that of the cathode so that when the cathode gets disproportionately elevated from the additional current when it enters A2, it will also pull up the grid thus maintaining the same differential.
I forgot to include an image, so here's the scheme as implemented on a SRPP Loftin-White circuit.
Edits: 10/10/16
My circuit is quite a bit different, so that won't work. The goal of this project is to eliminate a warmup drift that takes place during the first 20-30 minutes of operation. The drift is being caused by the driver tube itself. At this point, I've sketched a few preliminary controller designs, but the fundamental concept is the same. Monitor the output stage with an opamp and apply the correction at the driver grid. I guess the only decision to make now is whether to sense the output's grid voltage or anode current. Here's the basic amplifier circuit:
--------------------------
Buy Chinese. Bury freedom.
Okay, I understand now what you're after. Nice design you have come up with. I had a similar DC coupled scheme a couple of years back that was never built. It had an el34 as the driver with a 5K resistor at the cathode instead of a choke and used a bipolar supply at (+) / (-) 330V.I have one question; is your current 211 amp running at the same OP as this schematic indicates? If so, it appears that it's biased at around 95mA give or take, witch has the tube running at close to max dissipation. I understand you're using NOS tubes but how is their in-circuit longevity so far?
Edits: 10/11/16
In practice, -52V at the grid biases the 211 to approximately 70-75mA. I'm using the common GE VT-4C. Longevity is yet to be fully explored, but this is within the manufacturer's ratings for Class A service.
--------------------------
Buy Chinese. Bury freedom.
Thank you for the info. 70~75% of max plate dissipation is certainly within the limits for long tube life. These being NOS US made tubes, they should provide many hours of good service under these conditions. I'm sure your new project will turn out great.
Your understanding is the same as mine, so I'll conclude were are both right. :^) The cathode current includes the grid current plus the plate current.
The big effect that will shift the operating point occurs if the driver is cap-coupled to the power tube grid. Grid current will instantly charge the cap, and the power tube grid resistor to ground will slowly discharge it.
Thanks Paul. The exact operation of the output stage has become an issue because I need to design autobias for an A2 SET. The basic concept is to monitor the current into the output stage with an opamp, then adjust bias voltage to the driver, which is direct-coupled. The problem is that cathode current isn't constant in an A2 amplifier. Extended operation at higher power levels in A2 would be misinterpreted by the control circuit. So, I've been thinking about monitoring anode current instead. In my case, the HV supply for the output tube doesn't power anything else. A small resistor in the supply's ground return would create a voltage drop that corresponds to anode current. If anode current doesn't change as a result of operation in A2, I can use the resistor drop for bias monitoring and adjustment.
--------------------------
Buy Chinese. Bury freedom.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: