I just got a box of 12v 7551 beam tubes, datasheet says they’re used for class C RF equipment but it also mentions AF use. Looks like it’s not very far from EL84 power wise, around 15-20W for a PP pair with a 5K output transformer. Has anyone used these tubes, or the 6.3v 7558, for audio?
The curves look fairly linear for a pentode. With just under half the gm of an EL84, you'll need like twice the grid drive volts of an EL84 Amp. And twice the gain in front for any global N Fdbk loop.
I got a couple of PhilipsECG 7558 tubes in, so I took some curve traces. Both are at 50V/div Horiz. and 10 mA/div Vert.
1) pentode 149V on grid 2, -1V steps from 0V on grid1 (0V is top curve, almost off screen). Sharper knees than I expected from the datasheet. About 55V knee Voltage for 149V on grid 2. Normal for a beam tube.
2) triode mode -3V steps from 0V on grid 1.
1) pentode 149V on grid 2, -1V steps from 0V on grid1 (0V is top curve, almost off screen). Sharper knees than I expected from the datasheet. About 55V knee Voltage for 149V on grid 2. Normal for a beam tube.
2) triode mode -3V steps from 0V on grid 1.
Last edited:
Some similar tubes. All at 50V/div Horiz., and 10 mA/div Vert.
1) RCA 6EM5 147Vg2, -1 V steps on g1
2) GE 29GK6 (nominal equal to 6BQ5) 185Vg2, -1 V steps on g1, the higher gm here is making for bigger spacing between curves.
The 6EM5 looks fairly close to the 7551/7558 tubes.
Interesting that the 29GK6 takes more Vg2 to reach the same currents. 185V versus 147V or 149V. So using 7551/7558 in a 6BQ5 amplifier design would likely need the screen voltage dialed back by a 149/185 ratio from the nominal 6BQ5 screen V.
1) RCA 6EM5 147Vg2, -1 V steps on g1
2) GE 29GK6 (nominal equal to 6BQ5) 185Vg2, -1 V steps on g1, the higher gm here is making for bigger spacing between curves.
The 6EM5 looks fairly close to the 7551/7558 tubes.
Interesting that the 29GK6 takes more Vg2 to reach the same currents. 185V versus 147V or 149V. So using 7551/7558 in a 6BQ5 amplifier design would likely need the screen voltage dialed back by a 149/185 ratio from the nominal 6BQ5 screen V.
Cranking thru the datasheet info (gm1 and mu factor), the gm2 for 7558/7551 should be about the same as the 6BQ5/6GK6/EL84 tubes. So screen voltages should be nominally the same. Probably just a tube to tube or manufacturer variation that showed up above.
Thanks for the info! I got 14 of these tubes so I might try a PP amp with them when I finish my other projects 🙂 Which tubes would you recommend for the front end/phase splitter? Anything usable in the 12A_7 family?
For 6BQ5/EL84 Amps, the usual front end is a 12AX7. One triode is used for the input gain, 100X mu, and the other used for the splitter (Concertina).
But 7551/7558/6EM5 have half the grid1 gm. So you need 2X the gain. Not going to find a 200X mu tube easily. Either have to split the gain up into two tubes consecutively, or go with a pentode gain stage followed by a triode splitter. Could try some combo tube like a 6KV8 or 6F12P or similar. Another approach could be to parallel up 7551 tubes to get 2x the gm effectively on each P-P side, and use the usual 6BQ5 front end, if you just want to get something running quick. Combined tube Watt ratings will pass the 6BQ5/EL84 then.
Another approach would be to use 6HB6 output tubes instead. Just like a 6BQ5 but twice the gm. (except 10 Watt plate) That would be a cheat obviously. But they are relatively cheap tubes too.
But 7551/7558/6EM5 have half the grid1 gm. So you need 2X the gain. Not going to find a 200X mu tube easily. Either have to split the gain up into two tubes consecutively, or go with a pentode gain stage followed by a triode splitter. Could try some combo tube like a 6KV8 or 6F12P or similar. Another approach could be to parallel up 7551 tubes to get 2x the gm effectively on each P-P side, and use the usual 6BQ5 front end, if you just want to get something running quick. Combined tube Watt ratings will pass the 6BQ5/EL84 then.
Another approach would be to use 6HB6 output tubes instead. Just like a 6BQ5 but twice the gm. (except 10 Watt plate) That would be a cheat obviously. But they are relatively cheap tubes too.
Last edited:
Another common design to get sufficient drive for the 7551/7558/6EM5 tubes is to use a triode gain tube up front, followed by an LTP like splitter stage, which has gain. Here is an example:
https://www.diyaudio.com/community/...lled-tube-fanatics.293419/page-3#post-6365449
This type design gives considerable flexibility as to tube choice to get the final gain product needed.
Speaking of relatively cheap output tubes, 10/6JA5 is another linear Vertical Amp TV tube (beam pentode), just like 6EM5 and 6HB6, with 19 Watt plate diss and 10300 gm1 (move over 6BQ5! ). 10/6JA5 has curves that are near identical to the long lost EL506. Then there is 6JC5 which gives you a "modern" beam pentode version of the 6V6.
38HE7 is another hidden gem, with dirt cheap prices and very good linearity. They can usually be configured for operating the beam pentode alone, uprated to 15 Watts Pdiss without the damper tube running. (heater 21V across pins 10 and 12 for the pentode only, most 38HE7 allow this due to an internal series heater string)
https://www.diyaudio.com/community/...lled-tube-fanatics.293419/page-3#post-6365449
This type design gives considerable flexibility as to tube choice to get the final gain product needed.
Speaking of relatively cheap output tubes, 10/6JA5 is another linear Vertical Amp TV tube (beam pentode), just like 6EM5 and 6HB6, with 19 Watt plate diss and 10300 gm1 (move over 6BQ5! ). 10/6JA5 has curves that are near identical to the long lost EL506. Then there is 6JC5 which gives you a "modern" beam pentode version of the 6V6.
38HE7 is another hidden gem, with dirt cheap prices and very good linearity. They can usually be configured for operating the beam pentode alone, uprated to 15 Watts Pdiss without the damper tube running. (heater 21V across pins 10 and 12 for the pentode only, most 38HE7 allow this due to an internal series heater string)
Last edited:
I’m not opposed to using pentodes, that could be fun to try. Seeing as I have a couple of EF86s, maybe a Mullard 5-10 front end (or a variation thereof) could have enough gain to drive these things.
Or, as you say, I could parallel the output tubes. I have enough of them, and it would look impressive 🙂
Or, as you say, I could parallel the output tubes. I have enough of them, and it would look impressive 🙂