You can basically take any impedance you want, as long as you respect maximum ratings and change the Va-c and bias current accordingly.
Va max for an ECL805 is 300V, for the ECL85 it's 250V.
Although it depends a little what datasheet we are using.
Max peak rating is 550V for both.
With 7k @ 250V and 32mA, the max peak will be 481V.
Plate dissipation will be 8W ~ 90% of total dissipation (quite normal for class-A)
So all still within spec.
A lower impedance is possible yes.
For 4K perfect center bias, Va-c = 190V , Ibias = 42.6mA
B+ will be 190 + 19 (and a bit) = 210V
In general tubes are more linear with higher (screen) voltage
Last edited:
The same distortions in numerical amount (in percentage) do not always sound the same to the ear. Thus, when using a pentode with the same output power, the one using a lower impedance output transformer will have a more favorable distortion structure for the ear. Here, the second and third harmonics will prevail at maximum power, with a very low level of higher odd harmonics that sound extremely uncomfortable. At higher load impedances, the third harmonic will predominate at the same power, with a stronger presence of the fifth and seventh, which are very unpleasant, and the pair harmonics will be minimal. Measurements should be made with equal output power and distortion (in this case it is ~ 10%).
The situation is significantly different for triodes or pentodes / tetrodes connected as triodes, and for UL junctions each case should be considered separately because the distortion structure is highly dependent on the percentage of windings where the lead for G2 is on the output transformer and tube type.
The situation is significantly different for triodes or pentodes / tetrodes connected as triodes, and for UL junctions each case should be considered separately because the distortion structure is highly dependent on the percentage of windings where the lead for G2 is on the output transformer and tube type.
At higher voltage makes the tube more linear, so especially 3rd order would go down.
(which is also shown when using simulation models)
With this particular tube I know that 2nd order is very dominant
But anyway, yes, different arrangements can sound different, you're right.
Totally subjective and up to the user what to pick.
Something that was also very rarely considered in datasheets
Although, most of the time the spare-parts bin will decide.
Point I was making, is that 7k could still be used (or I guess 3,5k when used at 4ohm)
When NFB is applied this is all not really an issue anymore.
(which is also shown when using simulation models)
With this particular tube I know that 2nd order is very dominant
But anyway, yes, different arrangements can sound different, you're right.
Totally subjective and up to the user what to pick.
Something that was also very rarely considered in datasheets
Although, most of the time the spare-parts bin will decide.
Point I was making, is that 7k could still be used (or I guess 3,5k when used at 4ohm)
When NFB is applied this is all not really an issue anymore.
Well, thank you for all demo calculations. One question about using the ECL85 at high voltage is the maximum rating of Vg2. Will that be a problem?
Thanks all for all suggestions about using the ECL85.
Comparing the curves of both tubes at Va around 250-300V it is obvious that the ECL86 has a lot more gain. Especially in the output section: about u=20 for triode-connected pentode?
The ECL85 being the more low-impedance option but with less gain. About u=7 for triode-connected pentode? With a -Vg1 of over 30V it needs about 3 times the voltage for the same output as the ECL86.
Therefore it seems not very attractive to use the ECL85 now. I think I will stick to the ECL86 for this little project and save the ECL85 for other projects.
Comparing the curves of both tubes at Va around 250-300V it is obvious that the ECL86 has a lot more gain. Especially in the output section: about u=20 for triode-connected pentode?
The ECL85 being the more low-impedance option but with less gain. About u=7 for triode-connected pentode? With a -Vg1 of over 30V it needs about 3 times the voltage for the same output as the ECL86.
Therefore it seems not very attractive to use the ECL85 now. I think I will stick to the ECL86 for this little project and save the ECL85 for other projects.
@davorin: Do you know the primary impedances of the OPT's in #32 and #35? And you might notice that the data in #40 relate to a plate voltage of 170 V instead of 250 V.
Best regards!
Best regards!
The lower supply voltage G2 can be easily provided by Zener diodes connected in series to the supply. By using 1.3W Zener diodes such as BZY12, BZY15, ...., BZY24 and their series connection (already how many volts should be lowered) a sufficiently stable voltage drop can be ensured to avoid overloading the G2. The Zener diodes are all bridged together with a foil capacitor ~ 0.1-0.47uF / 250V.
A stabilized lower voltage can also be provided with a simple stabilizer for both channels, and as these are relatively small currents, stabilization using a FET will not require a particularly strong FET and a large heatsink.
A stabilized lower voltage can also be provided with a simple stabilizer for both channels, and as these are relatively small currents, stabilization using a FET will not require a particularly strong FET and a large heatsink.
lol, I didn't realize you had both laying around.
Yeah, I would normally always take the ECL86 as well, if I would have the choice.
since it's a better tube overall.
Better as in more linear, less heater current, more linear triode section etc
Yeah, I would normally always take the ECL86 as well, if I would have the choice.
since it's a better tube overall.
Better as in more linear, less heater current, more linear triode section etc
Have a loock here, PCL86 4€ + postage.
Electronic Valve (TV) PCL86 / 14GW8 Egro #17520 | M-Ware
Filament voltage is 14V/03A but if you have 2x6,3V just put a rectifier bridge + a capacitor (470uF?) and you can get 14V rms. I bought 10 parts and I will test them tomake paires.
Electronic Valve (TV) PCL86 / 14GW8 Egro #17520 | M-Ware
Filament voltage is 14V/03A but if you have 2x6,3V just put a rectifier bridge + a capacitor (470uF?) and you can get 14V rms. I bought 10 parts and I will test them tomake paires.
Note that this appears to be a used tube, and there are only 2 pc. in stock. Egro wasn't a manufacturer, btw, just a rebrander.
Best regards!
Best regards!