Hello friends,
Considering this 6E5P for example, which has dual cathode+shield pins:
When the cathode is to be connected to a single other node, is using only one of the pins okay or is it better to connect both pins.
I always assumed the dual pin arrangement on each side of the socket is just for wiring convenience but I've often been proven wrong 😎
Thanks for any insights.
Considering this 6E5P for example, which has dual cathode+shield pins:
When the cathode is to be connected to a single other node, is using only one of the pins okay or is it better to connect both pins.
I always assumed the dual pin arrangement on each side of the socket is just for wiring convenience but I've often been proven wrong 😎
Thanks for any insights.
If they're truly just attached to the same point internally? I'd go for one connection.
Two might be better if it's a shield, for ultimate low impedance pathing. Yadaya.
But realistically? It probably is for either wiring convenience of because the pins were available and they might as well hook it up to the shield when it's there yeah?
Two might be better if it's a shield, for ultimate low impedance pathing. Yadaya.
But realistically? It probably is for either wiring convenience of because the pins were available and they might as well hook it up to the shield when it's there yeah?
Very possible, the tube was specified as HF amplifier, which would benefit from good shielding.
Maybe but then there's NC pins 3 and 7 unconnected to the shield...
As this was a soviet military part I suspect the pin arrangement was very application-specific, to allow close-by circuit connections be made on the NC pins instead of an external tag board. That's what I do with NC pins when possible.
I always connecting such pins (4 and 9) together with short wire under the socket.
In Jurassic times, tubes often had multiple pins for important electrodes to reduce inductance internally for high frequencies.
The two cathode connections indicate this is an RF type. The idea is that one cathode connection is for the cathode bias resistor/bypass capacitor while the other one received the ground end of screen and plate bypass capacitors. This keeps the voltage developed across the cathode return from appearing in series with the grid return. This keeps the effective AC resistance of the grid high to prevent loss of gain/selectivity.
Thanks a lot Miles for your reply. As a noob I had to read a bit more about screen and plate bypass.
I thought sceen bypass caps always needed to go to ground. If I understand you correctly this would be such an implementation with screen bypass Cg2 connected to the cathode correct?
I'm not sure what exactly is a plate bypass though. Do you mean a cap from B+/plate node?
Forgive me for this basic question...
I thought sceen bypass caps always needed to go to ground. If I understand you correctly this would be such an implementation with screen bypass Cg2 connected to the cathode correct?
I'm not sure what exactly is a plate bypass though. Do you mean a cap from B+/plate node?
Forgive me for this basic question...
In the case of this schemo, it won't make any difference if you connect the screen bypass to the cathode or ground since the cathode resistor is bypassed. For audio frequencies this won't make any difference.I thought sceen bypass caps always needed to go to ground. If I understand you correctly this would be such an implementation with screen bypass Cg2 connected to the cathode correct?
When it does count is when the cathode isn't either grounded or bypassed. If there is AC on the cathode, the same AC needs to appear on the screen if v2K is to remain constant, and you connect the screen bypass to the cathode. If it isn't, you get degeneration and increased distortion. In RF designs operating at the upper ham bands and VHF you just have to live with it, and bypass the screen to ground to keep those currents out of stray cathode inductance. Those few nanohenries of unavoidable cathode inductance don't matter at audio frequencies, but are very detrimental at high enough frequencies. This is one of the causes of low effective AC grid resistance that can drop as low as a few K. That dampens LC tuned circuits and/or kills the gain of a preceeding stage and ruins selectivity of LC tuners.
This is a resistor of 0.1Ra connected between the DC rail and Ra. A capacitor to ground is connected between the junction of the resistors and ground. The traditional way is to select XBypass= 0.1RBypass at the lowest frequency of interest. This isolates the gain stage from the DC rail and is a preventative of low frequency oscillation. This is especially important when using high gain devices such as pentodes or cascodes.
Interesting technique. Can it be assimilated to a final power supply RC cell specifically aimed at shunting any possible oscillation to ground?
I have an tubed MM phono preamp project in the works that uses a high gain cascode input stage, I'll keep this trick in mind.
For any design, it depends. For this design (attached) plate decoupling wasn't necessary. Since the cascode LTP is balanced, there is no current change on the DC rail since as one half is increasing, the other is decreasing by the same amount (enforced by active tail loading) and the LTP is the only gain stage there. Besides, I needed all the available voltage.
Depending on your MM preamp design, then plate decoupling will probably be necessary as the cascode is a high gain design, comparable to a small signal pentode (where the name comes from: "cascade" + "tetrode", minus the screen current "kinks") in regards to voltage gain, CMiller and high frequency performance. Unbalanced, multi-stage designs are more susceptible to instability.
