Hello looking for help or to be pointed in the right direction.
So I'm looking at transmitting tubes and trying to figure out plate resistances for single tubes. If we look at the 805 tube datasheet, and see for example at 1250V and zero bias the Rp is listed at 1675ohm per tube, but if I check (delta V)/(delta I) it looks like it should be 10kohm, and when I double check Rp=mu*gm=50*4800uohm=225ohm (@100ma). Could someone please correct me or point me in the right direction, how do you calculate Rp on a zero bias / transmitting tube?
Here's a datasheet:
http://frank.pocnet.net/sheets/111/8/805.pdf
So I'm looking at transmitting tubes and trying to figure out plate resistances for single tubes. If we look at the 805 tube datasheet, and see for example at 1250V and zero bias the Rp is listed at 1675ohm per tube, but if I check (delta V)/(delta I) it looks like it should be 10kohm, and when I double check Rp=mu*gm=50*4800uohm=225ohm (@100ma). Could someone please correct me or point me in the right direction, how do you calculate Rp on a zero bias / transmitting tube?
Here's a datasheet:
http://frank.pocnet.net/sheets/111/8/805.pdf
So it's plate resistance is that high, oh that's too bad about that. I thought maybe the load resistance was more in line with plate resistance with the effective load resistance being 6.7k.
I guess the Shishido design must use a tonne of feedback to get away with a 3.5K transformer, I'd like to try to avoid that.
Thanks for the help, I see my mistakes now.
Also thanks for showing me the math on the gm=mu/rp, I see my mistakes there too.
I guess the Shishido design must use a tonne of feedback to get away with a 3.5K transformer, I'd like to try to avoid that.
Thanks for the help, I see my mistakes now.
Also thanks for showing me the math on the gm=mu/rp, I see my mistakes there too.
> Rp is listed at 1675ohm per tube
That is LOAD per tube.
Essentially 1/4 of Load plate to plate.
It is not the tube internal plate resistance.
> Rp=mu*gm=50*4800uohm
You may be mixing Ohms and Mhos?
{EDIT} I see S-A helped you while I was typing.
They are 1/x of each other.
Yes, the zeros are hard to track.
5,000 micro Mho is 200 Ohms.
4,800 micro Mho is about 200 Ohms (208.33 Ohms).
Mu times 208 is 10,400 Ohms.
> (delta V)/(delta I) it looks like it should be 10kohm
Which agrees with Mu*(1/4800uMho).
However the tube is bent all over. A dartboard study shows 4K to 30K.
Are you really proposing a single 805 in class A?
The 805 (and many others) has abandoned ALL attempt at SE linearity, in favor of huge output. The even-order distortion may be large (even huge). For RF this can be handled with tuned circuits. For audio we can run push-pull and cancel the even-order.
> zero bias
If you idle at zero, the signal must swing positive. Positive grid sucks.
Yes, there are A2 triode amps. Mostly these are "small 2" amps, just a couple mA of grid current. This is another world.
One tube, zero bias, has infinite grid impedance for negative swings and very-low grid impedance for positive swings. "Very low" may be 2K down to 600r. Very few drivers can handle this abrupt change of load at zero crossing. In push-pull (transformer) the impedance is "always" low, both ways.
Taking the yellow loadline we need about +50V on grid, which looks like 60mA grid current. The grid impedance is going below 1K. We want a driver which will supply 60mA with like 100 Ohms source impedance. A EL84 triode might get there. Actually the 805 has more gain above zero than below, and with the low positive grid impedance, for any specific load and drive, there is a non-zero driver impedance which will minimize total distortion. THD will not stay low for other drive levels. A dynamic driver impedance could help, and a tube driver naturally has the right curve. But getting it "optimum" will be a lot of hard hot work.
There's a name for cathode follower driving heavy A2 grid. There was even a tube with both units preconnected. It attempted to compete with early power pentodes. It didn't.
Plate resistance and speaker damping-- the yellow loadline is 12K, so damping for *small* signals is about unity. However at the extremes, where you may especially need damping, it goes to 3 and 0.4. The asymmetry may encourage cone-shift on loud passages; I'm really not sure of that.
This is a "skinny" power triode. It needs BIG plate and grid voltages to pass any useful current. For single-ended use it is much easier to work with a "fat" triode, high conductance at moderate voltages. HK257 was an excellent fat triode, and there are Class A (1, not 2!) conditions on the datasheet. While HK quit business long ago, I understand that 4E27A/5-125B is nearly the same, and I have seen a 4E27A SE amp.
That is LOAD per tube.
Essentially 1/4 of Load plate to plate.
It is not the tube internal plate resistance.
> Rp=mu*gm=50*4800uohm
You may be mixing Ohms and Mhos?
{EDIT} I see S-A helped you while I was typing.
They are 1/x of each other.
Yes, the zeros are hard to track.
5,000 micro Mho is 200 Ohms.
4,800 micro Mho is about 200 Ohms (208.33 Ohms).
Mu times 208 is 10,400 Ohms.
> (delta V)/(delta I) it looks like it should be 10kohm
Which agrees with Mu*(1/4800uMho).
However the tube is bent all over. A dartboard study shows 4K to 30K.
Are you really proposing a single 805 in class A?
The 805 (and many others) has abandoned ALL attempt at SE linearity, in favor of huge output. The even-order distortion may be large (even huge). For RF this can be handled with tuned circuits. For audio we can run push-pull and cancel the even-order.
> zero bias
If you idle at zero, the signal must swing positive. Positive grid sucks.
Yes, there are A2 triode amps. Mostly these are "small 2" amps, just a couple mA of grid current. This is another world.
One tube, zero bias, has infinite grid impedance for negative swings and very-low grid impedance for positive swings. "Very low" may be 2K down to 600r. Very few drivers can handle this abrupt change of load at zero crossing. In push-pull (transformer) the impedance is "always" low, both ways.
Taking the yellow loadline we need about +50V on grid, which looks like 60mA grid current. The grid impedance is going below 1K. We want a driver which will supply 60mA with like 100 Ohms source impedance. A EL84 triode might get there. Actually the 805 has more gain above zero than below, and with the low positive grid impedance, for any specific load and drive, there is a non-zero driver impedance which will minimize total distortion. THD will not stay low for other drive levels. A dynamic driver impedance could help, and a tube driver naturally has the right curve. But getting it "optimum" will be a lot of hard hot work.
There's a name for cathode follower driving heavy A2 grid. There was even a tube with both units preconnected. It attempted to compete with early power pentodes. It didn't.
Plate resistance and speaker damping-- the yellow loadline is 12K, so damping for *small* signals is about unity. However at the extremes, where you may especially need damping, it goes to 3 and 0.4. The asymmetry may encourage cone-shift on loud passages; I'm really not sure of that.
This is a "skinny" power triode. It needs BIG plate and grid voltages to pass any useful current. For single-ended use it is much easier to work with a "fat" triode, high conductance at moderate voltages. HK257 was an excellent fat triode, and there are Class A (1, not 2!) conditions on the datasheet. While HK quit business long ago, I understand that 4E27A/5-125B is nearly the same, and I have seen a 4E27A SE amp.
Attachments
What do you know about cathode driven grounded grid triode, the 203A might be something I would consider, Rp should be about 5k, if plate voltage is 800V and grid is grounded that's approx 100ma Ip, about what I'm looking for with decent gain=25. I'm considering that if I transformer couple, the plate current on the secondary can be used to cancel primary like in an inverted interstage without changing the phase of the windings (I hope I said that correctly), also with a grounded grid there is no need to supply current for A2. It seems like there's a few benefits, but it's not common, why? I have lots of questions about that sort of thing too.
First of all does cathode driven grounded grid sound good?
What kind of input impedance would you see on the cathode as input?
Would the impedance swing like the grid in class A2?
Thanks
First of all does cathode driven grounded grid sound good?
What kind of input impedance would you see on the cathode as input?
Would the impedance swing like the grid in class A2?
Thanks
"the 203A might be something I would consider"
Looks like the same tube, but 1/2 the Rp. Driving the cathode requires supplying the cathode current, albeit with an interstage xfmr that could be reduced (and class A op would avoid sudden current transitions compared to A2 grid drive).
To get the driver and output currents to cancel in the interstage xfmr may be tricky, that might compromise the linearity of the driver stage if the operating point is forced off oddly. The driver stage V to I transfer will then determine the final output current, so one will be listening to the driver essentially (the setup is effectively a cascode with driver at the bottom, although inverted phase in the interstage case), hopefully still operating in a linear range.
If the driver is essentially a V to V gain device, then the 203A will be adding its own distortion characteristics, which look like heavy 2nd or 3rd harmonic (depending on op point).
5K Rp is helpful over 10K, but this still looks like either parallel tubes for an off the shelf SET OT, or a custom $$$ OT. Borderline still.
https://frank.pocnet.net/sheets/111/2/203A.pdf
....
Looks like the same tube, but 1/2 the Rp. Driving the cathode requires supplying the cathode current, albeit with an interstage xfmr that could be reduced (and class A op would avoid sudden current transitions compared to A2 grid drive).
To get the driver and output currents to cancel in the interstage xfmr may be tricky, that might compromise the linearity of the driver stage if the operating point is forced off oddly. The driver stage V to I transfer will then determine the final output current, so one will be listening to the driver essentially (the setup is effectively a cascode with driver at the bottom, although inverted phase in the interstage case), hopefully still operating in a linear range.
If the driver is essentially a V to V gain device, then the 203A will be adding its own distortion characteristics, which look like heavy 2nd or 3rd harmonic (depending on op point).
5K Rp is helpful over 10K, but this still looks like either parallel tubes for an off the shelf SET OT, or a custom $$$ OT. Borderline still.
https://frank.pocnet.net/sheets/111/2/203A.pdf
....
Last edited:
That 4E27A would get you down around 1200 Ohms Rp at 150 mA in triode mode.
Mu 5
(gm 2150 at 50 mA -> approx. gm 4156 at 150 mA)
gm2 = 4156/5 = 831 Triode Rp = 1/gm2 = 1200 Ohms
https://frank.pocnet.net/sheets/049/4/4E27A.pdf
-----------------
Late correction to above post 7: delete: "although inverted phase in the interstage case"
the phase is the same
Mu 5
(gm 2150 at 50 mA -> approx. gm 4156 at 150 mA)
gm2 = 4156/5 = 831 Triode Rp = 1/gm2 = 1200 Ohms
https://frank.pocnet.net/sheets/049/4/4E27A.pdf
-----------------
Late correction to above post 7: delete: "although inverted phase in the interstage case"
the phase is the same
Last edited:
With 20% plate-to-grid feedback, rp of the 805 could be almost as low as a 300B. You would have to have a power grid driver inside the feedback loop but that could be done. At one point, I entertained the possibility of doing an SE amp like that with an 811A.
The biggest problem would be getting output transformers. But if you weren't greedy you could get almost 20W out of some fairly inexpensive 5k transformers by operating at 500V and 100mA, and you would be operating the power tubes at less than 1/2 of rated power. More power output is going to require a large ($) investment.
All of this could be done, but something similar could also be done a lot cheaper and with much less distortion by a much smaller beam tube.
The biggest problem would be getting output transformers. But if you weren't greedy you could get almost 20W out of some fairly inexpensive 5k transformers by operating at 500V and 100mA, and you would be operating the power tubes at less than 1/2 of rated power. More power output is going to require a large ($) investment.
All of this could be done, but something similar could also be done a lot cheaper and with much less distortion by a much smaller beam tube.
> with a grounded grid there is no need to supply current for A2
Maybe a good time to brush up on basic tube operation.
In GG the driver must supply ALL the cathode current over the whole cycle. Brutal.
> What kind of input impedance would you see on the cathode as input?
Essentially 1/Gm. Rp/Mu. Under 1K for most power tubes. Peak voltage about the same as peak grid voltage.
Grid offers a LOT of gain. Drive that if you possibly can.
Tuned-circuit radio work allows storing energy to cover peak current, and ignoring linearity. If the sheet only shows radio C, it will be an unhappy audio amp. Radio B is a little better but will still need heroic drive. B Modulator service is getting closer, and these will be fine for big push-pull amps. But only a few of these make good SE amps. They either have very low current, low current and high voltage (bad for OT design), or need nearly as much peak power at the input as you get from the output.
BTW: look on the HK257 datasheet for SE Class A conditions (omitted on the 4E27A sheet). They call it video or television for a historical reason, but it is valid for audio.
Maybe a good time to brush up on basic tube operation.
In GG the driver must supply ALL the cathode current over the whole cycle. Brutal.
> What kind of input impedance would you see on the cathode as input?
Essentially 1/Gm. Rp/Mu. Under 1K for most power tubes. Peak voltage about the same as peak grid voltage.
Grid offers a LOT of gain. Drive that if you possibly can.
Tuned-circuit radio work allows storing energy to cover peak current, and ignoring linearity. If the sheet only shows radio C, it will be an unhappy audio amp. Radio B is a little better but will still need heroic drive. B Modulator service is getting closer, and these will be fine for big push-pull amps. But only a few of these make good SE amps. They either have very low current, low current and high voltage (bad for OT design), or need nearly as much peak power at the input as you get from the output.
BTW: look on the HK257 datasheet for SE Class A conditions (omitted on the 4E27A sheet). They call it video or television for a historical reason, but it is valid for audio.
Attachments
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.