I realized that the yellow plot is for the H2, so the 50Hz point does not exists, I saw it too late, out of edit time, sorry.
As I said before, the magnetic field B goes as 1/f, so at 1KHz, B should be 20 times lower than at 50Hz, the same should be for the induced AC voltage, the only plot consistent with this is the yellow, so I smell an error somewhere.
Last edited:
Taking advantage of the fact that magnetic field B goes as 1/f and also does the induced AC voltage, here is an attempt to powering heaters with a "more-or-less sinusoidal" 100KHz voltage
HF filament supply
There are no measurements though, but I think that is way easier a DC heater supply, as we discussed before, for DC, ∂B/∂t=0 and no induction...
HF filament supply
There are no measurements though, but I think that is way easier a DC heater supply, as we discussed before, for DC, ∂B/∂t=0 and no induction...
I think your error is to associate B with a mechanism that causes noticeable hum related output voltage levels from a typical 12AX7 gain stage.
If you have in mind a mechanism where the physical shape of the heater wire generates an external field that modulates the nearby cathode-anode current flow, then for indirectly heated valves such as 12AX7, the hum generated by that mechanism was suppressed by valve manufacturers to such a low level as to not be noticeable.
Depends on components that you use, including power transformer. And on construction, like how much can MOSFET dissipate.
I will not bother to explain all fundamentals again, but regardless of cancellation mechanisms, the magnetic field B goes as 1/f, and also does the induced AC voltage, your yellow plot shows this quite nicely, but is the only one which does it.
I think that you have an experimental error somewhere, and your experimental results are not consistent, if you want to deny it, is fine, but in such conditions you cannot argue that induced hum is under the noise floor for 50Hz AC heaters, the yellow plot contradicts your claim.
I never use that capacitor, nor I will because I use DC on heaters, but that is not the point.
The point is that you designed an experiment to show AC on heaters does not contribute to hum in high gain sensitive stages and your measurements are not consistent.
If powering heaters with AC needs an electrolytic capacitor as a prerequisite, seems to me that AC on heaters is not *that* cool.
Last edited:
I never bypass the cathodes on my input stages, and have not hum. The bypass isn't a prerequisite for quiet operation, just one of several ways of dealing with it. I often lift the DC reference to 25% of my supply voltage for concertina splitter circuits, and that works fantastic for eliminating hum anyways. Even when I don't lift, properly grounded filaments don't cause audible hum on anything other than super high gain setups.
Let's clear up your misconception on that for starters. The testing was done to benchmark the level of hum contribution from an AC powered heater, as best I could with the instrumentation at hand, and with an awareness that the presented results are for just one sample valve and knowing that noise floor levels of different valves (including "12AX7") can vary quite a bit (as shown by Scott Reynolds from Tavish Design in some DIY threads). I try not to push any wheelbarrows, and have not voted in the thread's poll, and aim to make any comments as objective views on the topic.
Everyone can make up their own mind as to what circuit configuration they want to use, and why. For some, a bypassed cathode resistor on the first stage is seen as mandatory to gain the greatest SNR level from the total amp, and hum reduction is just a nice byproduct and not the raison d'etre. For some others that don't bypass their first stage cathode resistor, as indicated by at least one thread comment, their use of AC heaters has not shown up a hum concern.I never use that capacitor, nor I will because I use DC on heaters, but that is not the point.
.....
If powering heaters with AC needs an electrolytic capacitor as a prerequisite, seems to me that AC on heaters is not *that* cool.
Ciao, Tim
Last edited:
That's funny, your measurements are flawed and the misconception is mine
Again, was you who introduced the bypass capacitor, not me.
La raison d'être du condensateur est de cacher vos erreurs de mesure, n'est ce pas, mon ami?
The most curious thing is that we end discussing about bypass capacitors, the AC crowd is funny, they need horrendously complicated twisted wiring, tuned humdingers, cable orientations, etc. but they claim that it is for simplicity and better sound...
Huh, simplicity...
Heater Wiring - the Good the Bad and the Ugly
Nor to speak about intermodulation... better sound...
Last edited:
popilin, I'm happy to work through any flaws in the test jig and method used for measurements. Can you point to any circuit configuration, or power supply, or measurement instrument concerns you have? Can you point to any other initial setup conditions that can be applied as a confidence benchmark test?
The output spectrum with battery powered DC heater shows no noticeable difference to the spectrum when using AC heater, except for the heater C related hum signals in spectrum.
I've compared results with two 12AX7 now - both show nominal bias and anode voltage levels, and no noticeable difference in spectrum noise floor. Both show similar heater related hum frequency levels, within a few dB, which is likely related to the spread in Rhk characteristic that I had measured on ten 12AX7 samples years ago (ref at end). I could do the same check on another 6 or so to get the statistical aspect up to a high confidence level.
https://www.dalmura.com.au/static/Hum%20article.pdf
The output spectrum with battery powered DC heater shows no noticeable difference to the spectrum when using AC heater, except for the heater C related hum signals in spectrum.
I've compared results with two 12AX7 now - both show nominal bias and anode voltage levels, and no noticeable difference in spectrum noise floor. Both show similar heater related hum frequency levels, within a few dB, which is likely related to the spread in Rhk characteristic that I had measured on ten 12AX7 samples years ago (ref at end). I could do the same check on another 6 or so to get the statistical aspect up to a high confidence level.
https://www.dalmura.com.au/static/Hum%20article.pdf
I think trobbins experiment shows that small triodes have no hum induced by AC. There is absolutely 0 benefit to use DC. However with modern PCB traces going AC is asking for trouble. So the designers power the tubes with DC to save cost, keep using pcbs and no need for twisted wires.
To me, beyond measurement, the tubes are underpowered with DC and it results with a lot of harmonics and bloom sound. A lot of people like this 'euphoric' sound which could be almost all attributed to no feedback + DC heaters. Enough feedback could easily get rid of this DC heater distortions.
To me, beyond measurement, the tubes are underpowered with DC and it results with a lot of harmonics and bloom sound. A lot of people like this 'euphoric' sound which could be almost all attributed to no feedback + DC heaters. Enough feedback could easily get rid of this DC heater distortions.
What test jig? So far I could not see any.
Again, that measurement cannot be seen anywhere.
Maybe I am missing something?
Please, just say the page, I cannot find those measurements, anyway, from that pdf
i) Valves with double helix heater: 7025 and EF86, not as many as you suggested...
ii) On page 5, cannot see the main solution for all, except EMI, it is called DC...
iii) AC mains is full of garbage, even on the secondary of a transformer, as figure 21 clearly shows.
iv) On page 21 "Appealing forms of hum" we can see that you like hum.
From (iii) it is clear that a realistic measurement must be made powering heaters with AC from a transformer, as most people do.
If you will present your measurements as you did so far, do not bother, to me is enough.
That's your personal opinion.
This is the second time you make that statement, and it is time to present measurements that support it, a detailed calculation should be also OK.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.