Hi all,
I am working up an idea for a direct coupled Plate to grid feedback amp using Toroidals as outputs. It will be a PP, amp which requires exact current matching between halves. i have always achieved this before by using cathode CCS and bypass caps - but I would like to attempt to do it without the caps this time. I have designed a conceptual building block using a CCS above the transformer winding and wonder why this has never been tried and what I am missing if it hasn't. I suspect the inductive kick at the top of the transformer is probably enough to ensure the CCS has a short and brutal life - but I could be wrong.
Any thoughts ??
Shoog
I am working up an idea for a direct coupled Plate to grid feedback amp using Toroidals as outputs. It will be a PP, amp which requires exact current matching between halves. i have always achieved this before by using cathode CCS and bypass caps - but I would like to attempt to do it without the caps this time. I have designed a conceptual building block using a CCS above the transformer winding and wonder why this has never been tried and what I am missing if it hasn't. I suspect the inductive kick at the top of the transformer is probably enough to ensure the CCS has a short and brutal life - but I could be wrong.
Any thoughts ??
Shoog
Attachments
A CCS works by adjusting the voltage across itself until the rest of the circuit provides exactly the right current. A cathode CCS only has to change by a few volts to achieve this. A CCS in the anode circuit of a pentode might want to change its voltage by several kV, but it can't so it will either drop almost no voltage (but current will still be too low) or almost all the supply voltage (so the valve is pushed into bottoming).
To get the best results from a CCS it needs to work against a low impedance, not a pentode anode!
To get the best results from a CCS it needs to work against a low impedance, not a pentode anode!
But the transformer is acting as a shock absorber taking up all the voltage variation developed across the pentode.
Alternatively assuming that the Pentode was replaced with a triode - then theoretically it would work ?? Gary Pimm suggests such an arrangement for a PP output stage, but he only uses one CCS for both sides. My idea would be to split the primary and use two CCS's to enforce current balance within the transformer.
Shoog
Alternatively assuming that the Pentode was replaced with a triode - then theoretically it would work ?? Gary Pimm suggests such an arrangement for a PP output stage, but he only uses one CCS for both sides. My idea would be to split the primary and use two CCS's to enforce current balance within the transformer.
Shoog
The transformer doesn't absorb DC.
A triode would work better (i.e. less badly) for DC, but there is still the problem that for signals the CCS is in series with the transformer so the CCS needs to be bypassed by a capacitor. I can't think of any reason to put the CCS in the anode circuit. I don't even like CCS in the cathode circuit of an output valve, as a CCS sets the wrong parameter: average current instead of quiescent current (but that is a separate issue).
A triode would work better (i.e. less badly) for DC, but there is still the problem that for signals the CCS is in series with the transformer so the CCS needs to be bypassed by a capacitor. I can't think of any reason to put the CCS in the anode circuit. I don't even like CCS in the cathode circuit of an output valve, as a CCS sets the wrong parameter: average current instead of quiescent current (but that is a separate issue).
I can see that placing a Cap to earth would make this work.
However I do not see that the transformer is incapable of developing an AC signal if it is shunt fed a constant current - again I return to the analogy of a spring with one fixed end and one free end.
Fundamentally I am more than willing to admit that this is no a runner - but the idea that the +B of a normal amplifier with a large cap is behaving any differently to a CCS, with respect to AC, just doesn't seem logical. There is no AC signal at the last cap of a valve power supply. The fundamental difference between the two is the relative impedance.
Shoog
However I do not see that the transformer is incapable of developing an AC signal if it is shunt fed a constant current - again I return to the analogy of a spring with one fixed end and one free end.
Fundamentally I am more than willing to admit that this is no a runner - but the idea that the +B of a normal amplifier with a large cap is behaving any differently to a CCS, with respect to AC, just doesn't seem logical. There is no AC signal at the last cap of a valve power supply. The fundamental difference between the two is the relative impedance.
Shoog
Last edited:
Constant current source by definition provides constant current. That means, it's dynamic resistance according to Ohm's law is dV/dI, so if dI =0 (constant current, zero current change with changing voltage) that means infinitely high dynamic resistance. According to Ohm's law, voltage variations on transformer will be equal to dV = dI * Z, i.e. dV = 0 * Z. Voltage drop no transformer in series with infinite resistance will be zero.
Using your analogy: with a CCS at the anode you have a spring with both ends free.
I have always found CCS loading to be a great solution so I cannot share the opinion that it intrinsically creates a DC problem. You may have different mileage.
Fundamentally this allows me to eliminate the cathode bypass cap needed to make it work when the cathode is loaded with a CCS.
Shoog
Fundamentally this allows me to eliminate the cathode bypass cap needed to make it work when the cathode is loaded with a CCS.
Shoog
You have stricter requirements to stability of current of your tube stage. It is more tricky with pentode than with triode.
No. Bypass cap works on AC. On AC your stage is still loaded on transformer. No difference, do you supply the power directly, or through CCS/capacitor filter.
CCS loading for small-signal circuits is very different, as there you are not trying to use a wide voltage range and you don't mind losing some voltage. Power outputs are different, which is why they use transformers or chokes so full voltage is applied to the anode under quiescent conditions. An anode CCS stops this, so immediately ruins efficiency but with no gain in any other respect. If you want to use a CCS in an output stage then it should go in the cathode and be bypassed.
If you want to avoid cathode bypass caps then use fixed voltage bias or LEDs.
If you want to avoid cathode bypass caps then use fixed voltage bias or LEDs.
You misunderstand the requirement - it is to perfectly control the DC standing current and an LED bias will not achieve this.
The other question then is what compliance voltage is this Cap bypassed CCS power supply going to take up ?
Shoog
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.