Most JFETs cannot withstand more than 30V between gate and source/drain, but most tube preamps can deliver audio signals much larger than 30Vpk. Your circuit will work, but you still need to add protection Zeners (e.g. 2x15V in anti-series to ground) or your JFET will be destroyed if someone turns up the volume too high!Please look at how it works.
140V?
WHICH 140V?
Gentlemen
Here's updated status after installation of 1uf capacitor and 12Komh resistor.
The situation got much stable. I can increase input volume limiter from 10% to 50%, that provides volume to break windows
.Whoever said that this pre-amp has a very high gain was right.
My notice that also that power output is flex about +30v on and off on the 275v point. Is that normal for tubes rectifier or indication of problem?
What can stabilize it? Bigger capacitor ?
There is a potential problem with using only an R-C - consider what happens if the amplifier is turned off and then quickly turned back on while still 'hot', such as during a power line brownout. The time constant for the tubes may be much faster, and high voltage impulse may appear on the AC coupled output.
How I would solve this problem is to put a current limiting resistor (1k) in series with the output, followed by a bidirectional TVS diode to ground. This way you can have your cake and eat it too as it no longer matters if the time constant of the R-C filter is <1Hz. Tubes will likely not care if their bias current is doubled for a second or two. A 15V TVS should prevent damage to the inputs of most amplifier. Under normal operating conditions (<<15Vpeak) the TVS conducts only nano amps or less, so won't do anything to the signal. Back to back zeners will work too, but have a softer turn on than TVS diodes so have a tendency to start conducting and distorting the signal below the clamping voltage.
Combining this with raising the R-C cutoff to something reasonable (20Hz or so) can help prevent the potential thump from harming speakers.
edit: seems like that is not the OP's primary problem though
How I would solve this problem is to put a current limiting resistor (1k) in series with the output, followed by a bidirectional TVS diode to ground. This way you can have your cake and eat it too as it no longer matters if the time constant of the R-C filter is <1Hz. Tubes will likely not care if their bias current is doubled for a second or two. A 15V TVS should prevent damage to the inputs of most amplifier. Under normal operating conditions (<<15Vpeak) the TVS conducts only nano amps or less, so won't do anything to the signal. Back to back zeners will work too, but have a softer turn on than TVS diodes so have a tendency to start conducting and distorting the signal below the clamping voltage.
Combining this with raising the R-C cutoff to something reasonable (20Hz or so) can help prevent the potential thump from harming speakers.
edit: seems like that is not the OP's primary problem though
Are you measuring this with a multimeter or oscilloscope? If a multimeter and it is occuring slowly (seconds) then either you have another capacitor which is leaky or a problematic tube. The 160V rating on the 10uF capacitor in the feedback loop is marginal - under certain circumstances more than 160V can be placed across it. I would replace it with a good quality 10uF 400V rated electrolytic.
Last edited:
Sure, good idea
, just try 2 x 30V so as not to clip sound prematurely.In any case the FET unmuter + 30V Zeners (or 25V ones if that´s the Fet spec) is way better than plain 2 x 15V zeners and nothing else, for 2 reasons:
1) the MAIN one, it fully addresses the turn on 140V peak, which is the original problem affecting the OP, let´s not forget that because many threads tend to wildly digress and even not answer the original question.
2) the plain 15V Zeners will still let a 15V "squarewave" hit the power amp front end, a hell of a Thump if you ask me.
3) bonus reason
... 30V peak is 6dB better headroom than 15V peak 4) not even a reason
, 15V peak or 30V peak or unmitigated 140V peak may even destroy the power amp front end.I suspect that does not happen only because its input has some form of protection (which is to be expected from a good designer)
Don´t have the schematic but would not be surprised at all to find a series resistor plus back to back Zeners (say 4.7V or so) or small signal diodes clamping input to +/-15V or even a couple humble but useful small signal diodes wired base to base at the input differential pair.
Some so called (or self called) "Purists" have cramps when thinking of adding this "sand" to protect amplifiers and avoid them like the plague, truth is you may never need them in a Home Hi Fi setting, they are DEFINITELY needed in PRO equipment, where a drunk or stoned assistant may connect a speaker out to a power amp input any day of the week.
Now answering
(sorry OP, you didn´t ask about this), thanks for your concern, it certainly LOOKS so, but in practice it does not seem to be a problem.
No time to dig for Tube preamps using Fet switching (rest assured they are out there, I vaguely remember Acoustic tube amps using them, the 80´s polished wood combos) but just on my desktop I have the Marshal JCM800 2205 Preamp schematic and they use an even weaker channel switcher, same configuration as my unmuter: grounding hairy Tube signals.
They use very humble specs CA3046 transistor array, where transistors stand meager 15Vce
Yet they worked flawlessly, for years.
I think we can agree that a Tube Guitar preamp is THE place to find hairy signal voltages, you deal with them all the time.
Again: sorry OP for the digression.
Although it IS related to your problem after all, you ARE dealing with a hairy 140V peak.
Attachments
Gentlemen
Here's updated status after installation of 1uf capacitor and 12Komh resistor.
The situation got much stable. I can increase input volume limiter from 10% to 50%, that provides volume to break windows
Whoever said that this pre-amp has a very high gain was right.
My notice that also that power output is flex about +30v on and off on the 275v point. Is that normal for tubes rectifier or indication of problem?
What can stabilize it? Bigger capacitor ?
Hi this design has too much gain as most of tube circuits do. The schematic and the tube type give it away clearly just as the fact that it has 2 stages. Nowadays we have 2 Veff line signals and this often is adapted to the power amplifiers sensitivity therefor gain of "anything in between" should be low. Check the sensitivity of the solid state amplifier you use and the output signals of your sources. So essentially the design is unfit for the purpose and anything in between sources and power amplifier is superfluous. This is often, many tube products are merely a nice exercise to build but they are unnecessary when both sources and power amplifier are solid state. You are adding additional contacts, switches etc. and risk. However you might need a buffer (a preamp amplifying 1x) to drive a long cable or something like that. If you examine the situation you will probably find out that you don't need any gain at all
A nice example of adding something without a purpose that only creates more issues and questions. This is the cruel aspect : when you need gain there is not enough gain and when you don't need gain there is always way too much gain
About the output pulse and using too small output caps (when driving a semiconductor power amplifier as you do) or loading the device too much you can read the previous posts for solutions.
Last edited:
2x 30V is too much. In your circuit the gate is pulled up to about +15V; if the source swings down to -30V, there will be 45V between gate and source and the JFET may be destroyed.Sure, good idea
There is no problem using a timed relay ( except for it might not protect for turn-off pulse)
The real cure imo is to reduce the output cap so it will discharge faster then it can be
charged by the tube ( remember it's in series with the tube). A lower limit of 20hz is
perfectly fine with me as speakers at one end and recordings at the other has no music
information below that.
As read in the thread there is other opinions
The real cure imo is to reduce the output cap so it will discharge faster then it can be
charged by the tube ( remember it's in series with the tube). A lower limit of 20hz is
perfectly fine with me as speakers at one end and recordings at the other has no music
information below that.
As read in the thread there is other opinions
Last edited:
I using Multimeter and it rises slowly, 5 sec cycle. Is that possible the rectifiers-tubes 6Z4 are the first to since it starts on their exit?
Any recommendation of upgrading them?
Yes. It just a FYI project but I like it's sound additions and would not mind to find a way to lower his amplification and maybe to add sound controls.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.