Here goes.......
as a complete novice and have only worked with solid state stuff...can someone help answer this question.
A friend has a Conrad Johnson ART MK2 preamp. It uses parallel ECC88s. The manual says the tubes will last a mere 3 months if left on at all times.
The question is - why 3 months? Is this to do with the B+ voltage which is 160v? The heater voltage is 12.6v. I've looked at the spec and the ECC88 heater is rated at 6.3v. If you tie pins 4 and 5 together (heater pins) do you then run them at 12.6v?
thanks for not killing me in advance
Rob
as a complete novice and have only worked with solid state stuff...can someone help answer this question.
A friend has a Conrad Johnson ART MK2 preamp. It uses parallel ECC88s. The manual says the tubes will last a mere 3 months if left on at all times.
The question is - why 3 months? Is this to do with the B+ voltage which is 160v? The heater voltage is 12.6v. I've looked at the spec and the ECC88 heater is rated at 6.3v. If you tie pins 4 and 5 together (heater pins) do you then run them at 12.6v?
thanks for not killing me in advance
Rob
Just the number of hours of operation before the tubes start to degrade audibly in their performance. FWIW I do not recommend any tube component be left operating continuously unattended or when not in use for more than a couple of hours. This is both for safety and longevity of components including tubes.
A lot of electrolytic capacitors are rated at only a couple of thousands of hours at their rated temperature, and even at a fraction of that in continuous operation might last only a few years.
Finally there is a real safety issue as well if a component fails and the power fuse does not blow for some reason. I have had more than one instance where had I not been in the room or nearby a tragic outcome may have been the result.
For critical listening normally less than an hour of warm up should be required for the maximum glory
of which the particular device is capable. Longer than that and there is probably a design issue IMHO.
I like the simplicity, and elegance of the ART preamplifier. So I decided to build something like the ART. I did some simulation, and the result is quite good:
4pcs. of E88CC (8 triode parallel), with 6.2mA bias each. Distortion is 0.00075% with 100kohm load, and 1V output! Output impedance is about 800ohms. The distortion with 800ohm load, and 1V output it 0.1%. With 100kohm load, and 10V output is 0.05%
I know that is only simulation, but it looks promising!
Sajti
4pcs. of E88CC (8 triode parallel), with 6.2mA bias each. Distortion is 0.00075% with 100kohm load, and 1V output! Output impedance is about 800ohms. The distortion with 800ohm load, and 1V output it 0.1%. With 100kohm load, and 10V output is 0.05%
I know that is only simulation, but it looks promising!
Sajti
If B+ has a dedicated switch keeping filament on would prolong tube life. For power saving purposes filament in standby mode can be one fourth of normal operational voltage.
B+ especially if it goes on right away with filament that wears off cathode the most.
Amprex PQ 6DJ8 claimed to be the best ECC88 very hard to find NOS however. 6922 is another option.
B+ especially if it goes on right away with filament that wears off cathode the most.
Amprex PQ 6DJ8 claimed to be the best ECC88 very hard to find NOS however. 6922 is another option.
It is a typical myth from Russian online forums. Typical myth from American online forums is that all tubes feel better when B+ is delayed by relay...
Now all myths are international!
I mean,
1) B+ relay delay saves lives of tubes;
2) Keeping cathode hot with no B+ saves lives of tubes.
Actually, both are bad for them. First cause big transient currents and cathode poisoning, second causes cathode poisoning only, but of greater degree.
Last edited:
Yes, according to the Telefunken datasheet, the E88CC is free from cathode interface problems. I guess this does not mean that all valves labelled as E88CC (or 6922) will be OK.
BTW, my understanding is that cathode poisoning and cathode interface are two different problems. Cathode poisoning is caused by ionic bombardment when the cathode is not protected by the space charge e.g. when the heater is not quite hot enough so the valve is temperature limited. Cathode interface comes from full hot heater but insufficient anode current. It seems that as a valve heats up, if HT is present you get poisoning. No HT causes interface. So which ageing mechanism do you prefer? Fortunately, a couple of seconds of danger does not use up too much service life.
BTW, my understanding is that cathode poisoning and cathode interface are two different problems. Cathode poisoning is caused by ionic bombardment when the cathode is not protected by the space charge e.g. when the heater is not quite hot enough so the valve is temperature limited. Cathode interface comes from full hot heater but insufficient anode current. It seems that as a valve heats up, if HT is present you get poisoning. No HT causes interface. So which ageing mechanism do you prefer? Fortunately, a couple of seconds of danger does not use up too much service life.
OK I would not argue. Lets consider it like me be a believer in B+ delay to be a "proper audiophile stuff"Now all myths are international!
BTW from my practice BlackGates take most time to warm up not tubes imho.
Just another international link FYI
Soft-Start for Power Tubes
Soft start and abrupt delayed start are totally different kinds of fish.
Here is my own soft start version: reference voltage for B+ is taken from bias shunt regulator. Since both regulators are in series, time constant depends on current drawn from B+, so the stabilizer literally waits for tubes to heat up to start drawing current, then slowly brings up B+ voltage.
Here is my own soft start version: reference voltage for B+ is taken from bias shunt regulator. Since both regulators are in series, time constant depends on current drawn from B+, so the stabilizer literally waits for tubes to heat up to start drawing current, then slowly brings up B+ voltage.
I found anoother solution: Use 6H30, with higher current. Something like CT-5. I did some simulation, and it looks promising. The output impedance is less than 800ohm, if the bias is 25mA, and I connect parallel the 2 half of the 6H30. If I push up the bias to 40mA/triode (80mA together), the output impedance goes down to 500ohms, and the distortion is 0.02% at 1V output level, and 100kohm load.
Sajti
Sajti
yeah - really the reference
in any case - with so many triode systems in parallel , cascoding is first what's on my mind ....
If You ever hear the ART, the cascoding will leave Your mind...
Sajti
Last edited:
I read about ART years ago , read more to complete enough information about working points used , make my own version and had it for year or so
later I got original schematic ( even with one or two original mistakes ) , conclusion is that I was somewher in 10% of working points with my own version .
so - years later , and few preamps in between , I say - I'll cascode it
but that doesn't mean that you must do the same
btw. your schm is , also , not original ART one - just principle
I'm not saying that is worse or better .....
later I got original schematic ( even with one or two original mistakes
) , conclusion is that I was somewher in 10% of working points with my own version .so - years later , and few preamps in between , I say - I'll cascode it
but that doesn't mean that you must do the same
btw. your schm is , also , not original ART one - just principle
I'm not saying that is worse or better .....
Last edited:
Thanks to share Your experience!
I know that my schematic is not same. I never did exact copy of any amplifier. I just like the simplicity, and I want to make my version.
So I really like the E88CC version, but I can build the dual 6H30 shorter signal path. The only question is the bias I should use. I don't want to try lot version because the high quality resistors costs lot of money.
Sajti
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.