EL84 driver stage

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
The plate load resistor for the triode-strapped EL84 should be at least 2x the internal plate resistance, correct? If so, ra is probably about 2.5k (maybe less), so you'd want the plate resistors to be 5k each, minimum. That would be the plate load that would deliver the most power into the load (the KT88 grids).

That 'optimal' plate load shown in the GE manual may be for operation at a low plate voltage and high plate current, where ra of the tube is near the lowest possible. Usually when running a driver stage using plate load resistors, you may not be able to draw that much plate current. Therefore the ra will be higher than what's shown in the manual. That's how I figured 2.5k for the internal plate resistance, so 5k for the (external) plate load resistor. Maybe the ra will be closer to 2k and you can use 4k plate load resistors.

At least I think so...
--
 
Last edited:
Ok l drilled the chassis and mounted big components. 10K / 14W aluminium housed resistors was choosed as load for el84 stage. Do you see any inconvenience from this layout please?
2w3xh1z.jpg
 
Differential amp driver

Hello. Educate me please. Have any sense to use a differential amp in conjuction with a split load concertina inverter supposed inverted signals are perfect balanced? I work at a project using el84 triode strapped driver. It works in common cathode mode now but just thinking if will not perform better using simple cathode bypassed gain stages
I attached a pic with versions. The splitter is 12at7. Which one will perform better in these forms.please?or why to use one against the other?Noise rejection will not be great in first instance I guess and the second one give me the possibility to perfect balance the stage. more I can benefit by low output impedance.Can you share some opinions.pls?. Thanks.

An externally hosted image should be here but it was not working when we last tested it.
 
Last edited:
My opinion is that it is not necessary to strive perfect balance before output tubes, because very seldom the output tubes are perfectly equal.
Therefore I allways have AC-balance adjustment added into the phase inverter.

What output tubes you plan to drive with such a massive pre stage?
 
Good looking layout and well built.

I don't see any benefit with separate cathode resistors and bypass capacitors.
Common 220 ohms cathode resistor should work fine, if the two EL84s have equal
idle currents and gain. And that is possible to measure quite easily.
 
Hello. Educate me please. Have any sense to use a differential amp in conjuction with a split load concertina inverter supposed inverted signals are perfect balanced?

Yes there is. A differential gain stage allows you to move the cathodyne phase splitter farther back in the signal chain. This type of phase inverter is limited in output swing. Doing this was a feature of the Williamson design.

I attached a pic with versions. The splitter is 12at7. Which one will perform better in these forms.please?or why to use one against the other?Noise rejection will not be great in first instance I guess and the second one give me the possibility to perfect balance the stage. more I can benefit by low output impedance.Can you share some opinions.pls?. Thanks.

An externally hosted image should be here but it was not working when we last tested it.

Neither one of these designs are differential amps, both simply common cathode gain stages. The problem with the first is that 220R isn't much of a tail load, especially for hollow state with its strictly limited gm. The second version bypasses the cathode resistors, removing any pretense to a tail load.

If it's a differential gain stage you want, you'll need to elevate the cathode voltage either via a negative rail, or half DC coupling to the diff amp so's you can provide an effective tail load. As for which works best, it's pretty much six of one, half dozen of the other.

Have a look at the RCA Receiving Tube Manual (RC 30), pg. 696, the "High Fidelity 50W amp". This design uses two 6CB6s as common cathode gain stages after a cathodyne phase splitter. You don't necessarily need a differential amp after the cathodyne.
 
Hello. I.m pretty ignorant in this sense. I did not catch the whole idea around the cathode elevation. There are some question through my mind. Why is necessary? How to determine how much voltage need to elevate? How to apply in practice ? The grids will be referenced at cathodes or to the ground ? Can someone provide a source to understand the whole mechanism.please ? Thanks.
 
Thanks for your kind replies.
So.now have to find a way to elevate the cathodes and then to rebias in respect with new negative reference. How much should be necessary ? Can I use -80v available from bias power supply.please ? Thank you. Catalin

From the included schemo, it looks like the plate current is 28mA, so the tail current would be 56mA. Given the 220R resistor, the cathode is 12.32V above DC ground. For a -80V rail, the difference is:

12.32 - (-80)= 92.32
Rtail= 92.32/56E-3= 1.648K (1.6K design nominal)

This would lengthen the tail considerably, and give differential performance in a balanced in. balanced out diff amp. You might have to compensate for the increased load on the negative rail with a larger filter capacitor: 220uF for example.
 

PRR

Member
Joined 2003
Paid Member
"Long tailed pair"

at Valve Wizard

If Rt is much larger than 1/Gm (say several K Ohms), you can drive one grid and get output at both plates. If you drive both grids, you get very good balance.

I'm not inclined to worry a lot about balance. Tests with totally different output tubes do not show huge problems, so near-balance is good enough for me. But with your building skill, I guess your amp deserves all the best.

One drawback of the Williamson is the many bass cut-offs, threatening subsonic stability. Most Williamsons show sub-sonic peaking, and I have seen them get up and motor-boat. A 100 Watt motorboater is a little frightening.

Here's a hack to get the EL84 stage up on top of a large cathode resistor without adding a bass-cut pole to the small-signal transfer response. Yes, below 4Hz it will not be capable of BIG power, but for audio this is not a problem.
 

Attachments

  • catalin-1.gif
    catalin-1.gif
    33.6 KB · Views: 213
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.