Dear all, as far as I can remember this was a Andy Grove design (AN-UK)
but I cant figure out how much current this power supply can deliver.
I am planning on feeding 2 5687 tubes with it running on 255 volt.
What I could find on the ecl 82 was 40 mA is this correct?
Please help, thanks
but I cant figure out how much current this power supply can deliver.
I am planning on feeding 2 5687 tubes with it running on 255 volt.
What I could find on the ecl 82 was 40 mA is this correct?
Please help, thanks
Attachments
Well the 6x5 is only good for 70ma. So it will supply 70ma minus whatever the shunt regulator needs to supply a steady voltage. Remember that the regulator will vary its current flow in conjunction with the current through the load to achieve a close to constant current load on the power supply.
In this particular case, yes, the current is limited by the 6x5. If you draw any more than 70ma of continuous current, you could damage the rectifier tube.
In a shunt regulator configuration, what you would do is figure out how much current your 5687s can swing. Let's say each 5687 can swing 5ma up and 5ma down from your bias point. I doubt those numbers are accurate, but lets pretend.
That means your total load would swing up 20ma and down 20ma from your continuous load (bias). What the ECl82 will try to do is swing its current in equal and opposite magnitude to your load. That means when your 5687s try to draw 20ma more than bias, the ECL82 will draw 20ma less than its bias point which makes the total draw on the power supply constant.
The ECL82 will always be drawing more current than the maximum offset current of your load. So if your load could draw 40ma more than your bias current, the shunt tube will need to be biased to more than 40ma so that it has a chance to swing to near 0 in order to do its job.
Personally, I think the whole exercise is more trouble than it's worth and I would just design a basic LCLC power supply and call it good.
In a shunt regulator configuration, what you would do is figure out how much current your 5687s can swing. Let's say each 5687 can swing 5ma up and 5ma down from your bias point. I doubt those numbers are accurate, but lets pretend.
That means your total load would swing up 20ma and down 20ma from your continuous load (bias). What the ECl82 will try to do is swing its current in equal and opposite magnitude to your load. That means when your 5687s try to draw 20ma more than bias, the ECL82 will draw 20ma less than its bias point which makes the total draw on the power supply constant.
The ECL82 will always be drawing more current than the maximum offset current of your load. So if your load could draw 40ma more than your bias current, the shunt tube will need to be biased to more than 40ma so that it has a chance to swing to near 0 in order to do its job.
Personally, I think the whole exercise is more trouble than it's worth and I would just design a basic LCLC power supply and call it good.
Ah thank you very much for this answer!
I do like to try the simpler approach also but don't want to end up with a buzzing choke , the first choke
in the LCLC as that is my experience so far with an Audio Note pre amp kit that does not only has this noise but
also a large radiated magnetic field to deal with.
My question therefore is what to do this case.
Thanks again!
I do like to try the simpler approach also but don't want to end up with a buzzing choke , the first choke
in the LCLC as that is my experience so far with an Audio Note pre amp kit that does not only has this noise but
also a large radiated magnetic field to deal with.
My question therefore is what to do this case.
Thanks again!
Ahhhhhh. I see.
AN kits are junky. If you bought it from a reputable source, I wouldn't even put up with it. I would call the dealer and tell him to replace the choke.
But if you want the shunt regulator, you are more than likely going to have to design your own. I think what andy grove is using is called a feed forward shunt regulator. Here is a nice post from Tube Cad explaining the concept.
https://www.tubecad.com/2007/01/blo...l shunt regulator,side of the series resistor.
AN kits are junky. If you bought it from a reputable source, I wouldn't even put up with it. I would call the dealer and tell him to replace the choke.
But if you want the shunt regulator, you are more than likely going to have to design your own. I think what andy grove is using is called a feed forward shunt regulator. Here is a nice post from Tube Cad explaining the concept.
https://www.tubecad.com/2007/01/blo...l shunt regulator,side of the series resistor.
Looking at the circuit diagram this appears to me as a series regulator circuit with the series element being the pentode section of V1 configured as a triode.
350V is applied to the plate of V1 and the output current is supplied from V1's cathode.
The pentode section of V1 is in "series" with the load.
In a shunt regulator the regulation element passes current to ground from the output terminal.
In the circuit shown there is no path to ground from regulation element, V1.
The 6X4 is rates for 70mA.
The 6x5 is rated for 75mA. To further confirm what is available in output current the pass element, a ECL82 pentode is rated for 7W plate dissipation.
Given the circuit shows the pentode section of V1 with 350V on the plate and 260V on the cathode if we assume a 75mA output current goal this becomes
7Watts / (350V-260V) = 77.8mA maximum. So 75mA looks just Ok.
Re:"What I could find on the ecl 82 was 40 mA is this correct?"
The PCL82 data sheet (same pentode ratings as a ECL82) shows the the pentode section will deliver about 82mA with a screen voltage of 100V.
As this circuit runs the pentode in triode mode there will be about (350V volts on screen - 260 on cathode) or 90V on the screen so it is likely you will be able to deliver 70mA of current to the load with 90V across the the pentode section of V1.
As first I thought 70mA will likely be very close to all the tube can deliver so with aging perhaps a bit less with time.
On second look the grid of the pentode section can be supplied a small amount of grid current from R3. So there is liklly more margin than I assumed above with zero grid current. The more I look the more the design become slick for it's simplicity and function.
Then there are two sources of lost current.
R3 at 55V across 68k will draw 55v / 68k = .81mA. This is taken from the pentode output.
R1, R7 and the 50K pot that sets the output voltage draw 260V / (100k + 150k + 50K) = .87mA. This is taken from the pentode output.
So almost 2mA lost in the supply circuit.
So really the most I would expect will be about 70mA.
I personally would consider to de-rate this supply design to a bit lower value for continuous current so there is more margin for aging of the tubes but perhaps that is not required.
If there is enough current to supply a pair of 5687 looks depent on what current level you plan to run each section of 5687 at.
Hope this helps
Last edited:
The output triode looks to be biased at 60V on the cathode across a 3.3K resistor.
So 60V / 3.3K ohms = 18.2mA. Pretty low current level.
The input triode looks to have 1V on the cathode across a 470 ohm resistor.
So 1V / 470 ohms =2.1mA
Total = 18.2mA + 2.1mA = 20.3mA per channel or 40.6mA for two channels.
Should be no issue for the supply.
By the way this amplifier is going to be a pretty low audio power output. Just for my interest what are you driving?
So 60V / 3.3K ohms = 18.2mA. Pretty low current level.
The input triode looks to have 1V on the cathode across a 470 ohm resistor.
So 1V / 470 ohms =2.1mA
Total = 18.2mA + 2.1mA = 20.3mA per channel or 40.6mA for two channels.
Should be no issue for the supply.
By the way this amplifier is going to be a pretty low audio power output. Just for my interest what are you driving?
I see how he drew it now, and yes, this is correct.
And no, he did not use a 8 or 16 ohm tap as a joke. This circuit has quite a lot of gain that needs to be attenuated so that you don't send your power amp into clipping when you touch the volume control.
Thanks Tjj226 and Michael C,
I dont have a problem with too much gain from the 5687 stages as the transformer reduces the gain anyway and the
input transformer of my poweramp does aswell.
I dont have any experience with hashimoto chokes other than remembering high prices but thanks for the information.
I dont have a problem with too much gain from the 5687 stages as the transformer reduces the gain anyway and the
input transformer of my poweramp does aswell.
I dont have any experience with hashimoto chokes other than remembering high prices but thanks for the information.