• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Need a bit more power :)

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I've built for a friend a tube amp made from stock parts . The output section is made from a pair of EL84 tubes in PP cathode bias with common resistor , Phase Inverter is a cathodyne using 5814 tube ...the amp sounds good but a bit underpower , putting out 8W rms with sine input .
I tried fixed bias but only minimal change , and the PI is not near clipping .
The problem seems related to the low output ht voltage of 250Vac , about 280Vdc with maximum load.

I wonder if there's a power tube with same socket of EL84 capable of a bit more power with this kind of supply voltage.

EL86 maybe ? I've a new pair of EL86 NOS from valvo
Sounds like you need to redesign your power supply. If you are getting 280v ac that means after rectification you are only getting like 150v on the plates. I think you need a power transformer that puts out more voltage in order to drive more watts out of your 6bq5's they should have (guessing) 250 to 350v DC on the plates. What are the specs on your power tansformer? Voltage of secondaries and weather they are center-tapped?
I may be out of line here but the fact that your power supply drops to below 90% of it's open circuit voltage under load, that suggests to me that you are drawing too much current from the trans. Do you possibly have a shorted tube somewhere? Do you have a bigger power tranny that you can swap out to see how that affects the issue?

An old timer once told me that to see the amp rating that a power trans is good for, load it down til the voltage is 10% below open circuit voltage and measure the current with that load. that is the maximum safe current to draw from an unknown transformer. Using this formula and the numbers you provided I am going to suggest a bigger (higher current) power supply. Do you have a schematic?
Before fixing the problem, first you have to find what the problem is. This might seem obvious, but it is often overlooked. Exactly where is the limiting happening?

An output stage can run out of current or voltage, for a start - depends to some extent on reflected load, so depends on OPT and external load.

A driver stage can run out of headroom, or can hit the grid current or grid cutoff limit of the output stage.

Find the problem, then think about a solution.
Adding just a wee bit to DF96's advice ... also, make measurements.

Measure the cathode voltages of all the tubes. Measure the plate voltages. Summarize what values are in use for resistors, capacitors. Post a schematic here, to get more of us to think on it with you. Have an oscilloscope? Use your PC output jack to feed a 1V input to the amp at some frequency or another. Convenient signal generator, and it works. With oscilloscope, measure at various points in the signal chain. Are the A/C volts where you expect them? Are they enough?

The reason for all this, is this: using Ohm's law [E = I R], one can determine I from volts and resistance. Its very useful to know. It also will show operating points of the tubes, and their biases. Very useful in turn. There comes some predictability (or at least "analyzability by us Tube Clowns") of gain from the quiescent (no signal) measurements you make and the specs on the tubes (assuming they're behaving in normal ways).

So, i recommend, start with the schematic. Print it out on a big piece of paper (or two sheets, taped together after printing), so that you have a work-page to record values, draw squiggles, margins to make notes of questions and observations. Use pencil, not pen. Prepare to ERASE things.

Then, whatever "problem" you have should become basically trivial to ferret out.


Sorry for not reply , i'm very busy from work , i'm an electrical engineer ... electronics is an hobby !


Some update . I stated a wrong voltage in previous post , the HT winding is at 200Vac , and i get 280V dc at idle , 260Vdc under load.

Schematic is quite simple , a Push Pull of EL84 with 200 ohm cathode bias , 1 resistor for each tube , 1K screen resistor from B+ , 5K grid stopper , driver is a cathodyne PI with 47K plate and cathode load.

I discovered the output transformer is for 4 Ohm and 16 Ohm load , not 8 ohm .

With EL84 seems it work really bad and the maximum power output is 8W on 8 ohm using the 4 ohm output , every other combination lead to less power .
4 ohm would run the tubes out of current , giving less power , same for a 16 ohm load on the 16 ohm tap.
Anyway the 8 ohm load on the 4 ohm tap lead to distortion due to primary saturation , at 8W i get 320Vac plate to plate , don't know TU ratings but i'm near TU saturation.


Tried EL86 , two NOS NIB from Valvo , 5,00 € a pair from a local repair man.
Swap is easy because they're pin to pin compatible , i've no time to change the cathode resistors so applied a negative voltage at grids , about -25V to get 45mA each tube.
Not really matched BTW

Now i get 14W rms continuos sine on resistive load @4 and 16 ohm, 16W rms with hotter bias , no red plating anyway. I also get 12W rms on 8 ohm using 16 ohm tap.
Also the 2 tubes are not matched (i read 8V and 10V on the cathode resistors) , so one tube start distorting earlier meaning a well matched pair would lead to a bit more power :D
EL86 are great !!

Next step is to change cathode resistor and find a more matched pair
So you have 320Vac anode to anode with 8Vac on the 4 ohm tap, right? (8W into 8ohm).
That means a turns ratio of 40.
With an 8 ohm load on that tap, the output valves see an imedance of 12800 ohm.
For EL84's that is very high, resulting in low power output.
Putting 4ohms on that tap gives 6k4, which is too low.
A pair of EL86's are happy with 5k5, so 6k4 is not that far off. Check the voltage rating of the second grid. I think it has a limit of 200V.
I'm searching something about these EL86, and I think it could be used with 300V on plates watching for bias. Datasheet says 250V on plates with 5.5k a-a and 14W RMS output, maybe 280V on plates and a max screen voltage of 220V could make some more Watt. I'm looking for 20W output.
Tesla, you should be able to check the screen voltage at idle and prepare a loadline and then assess whether your screen resistor(s) are appropriate or causing some slide as you approach grid-current limiting of output power.

Are you bypassing the cathodes?

How much ripple is on your B+ at max output power?

How much is the plate-to-B+ OT primary winding resistance?
Uhm , i don't have he amp on bench , i can check during week end. B +in my amp is at 280 without load (EL86 removed) and 260 at load , screens are connected to 250V trough a 1K/1W resistor , 250V comes from B+ with a R-C , 250V also supplies the PI stage.
Tried also differente cathode bias : Separate resistors , common resistors , CCS with lm317.
Actually the best is 2 resistors (one for each tube) by-passed by a 100uF cap.
El86 are biased at 45mA each with a 390ohm cathode resistor for each tube(10W dissipation). At full power current sourced to the center tap of Output transformer goes to 110mA at clip , 90mA at 0 signal.

EL86 seems to work well (and sound damn good and loud with master volume at full)
I Think EL84 works bad in this amp due to the low B+ voltage and impedance mismatch
Raa is near 4K

For Slideman , 20W sinus power from a EL86 pair is challenging , 16W are ok (biasing my amp at 55mA per tube i get 16Vrms @16 Ohm before clipping) and you don't notice the difference between 16 and 20 W !
I'll tell ya this: If cap-bypassed, individual 390 ohm resistors are (so far) the "best", then the plan seems obvious. Since you've got relatively low plate voltage, increase amperage! That's so easy, that it just should be done.

How? Cathode resistor change. 330 ohm. "Next obvious number down" so to speak on the old fashioned (but perfectly suite) "10% sequence". Your bias becomes a bit less (hence higher plate current). The tubes still are in their design spec. The bypass cap (100 uF) is fine for each. Decent response down to 5 Hz. The WATTAGE of the bypass caps does not need to be 10 watts each. Use P = I²R = V²/R formulas. The 390, for instance: 45 ma = (0.045A)²x390 = 0.78 watt. 10 watts... is overkill.

Anyway - here's a nice datasheet that'll allow you to get the most from these babies.


When i wrote 10W dissipation , i was speaking about tube dissipation , not resistor , i used 3W resistors for cathode...I tried also more current using 330 Ohm resistors , the current was about 55 mA per tube or more , but not so much power increase , so i choose the 390 ohm to stay a bit low on dissipation meaning longer life of power tubes.
Well (thanks for reply) - then it seems that 2 things are left: according to the PDF, a 2500 ohm plate load in PP seems to optimize output power. But the second thing is, what is the voltage swing on the DRIVE (the grid #1) for the tubes? [I think about that: you say you do go into clipping, so this really shouldn't be the problem. But worth the ask]
You right , i'm not sure if drive is enough , due to the fact the power is now double compared to el84 i haven't check drive signal anymore and el86 require more drive voltage .
I'm really surprised by the 25W at 1% stated in the datasheet , i'm quite sure to get that power you have to go fixed bias , so this worth a try ! I'll check during the WE

Anyway , are PL84 same as El86 , except heater voltage ? I have a Tunsgram quartet of PL84 NOS ... i would like to make a stereo amp capable of 20-25W .

Also , anyone has got the Klein and Hummel TS60 schematic ? They use EL86 getting 25WPC.
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.