# Thread: low voltage single ended el84

1. ## low voltage single ended el84

Hiya folks, it's been a while!

Been getting the itch to tinker with some parts I have salvaged. I'd like to build a low-power single ended el84 amp.

I have a power transformer that will give me a B+ of about 190 VDC. Say I throw that at the plate of an el84 and want to cathode bias that tube. I have been looking at datasheets and this page (http://www.tedweber.com/webervst/tubes1/calcbias.htm) to try to figure out how to bias that tube in such a situation. I'm kind of stuck.

I could just try stuff, but I'd rather approach this with some intelligence. Can someone point me in the right direction of how to calculate the resistor I need to bias that tube?

I don't have any other numbers, because I don't have anything built yet.

0

2. What is the impedance of your output transformer?

1

3. Welcome back! I figured you had blasted yourself into the future with no way to return...

I use NickB's calculator here: bmamps.com

2

4. My OT can be anything I want, I haven't bought it yet!

0

5. Originally Posted by eschertron
Welcome back! I figured you had blasted yourself into the future with no way to return...

I use NickB's calculator here: bmamps.com
I went to the future, but I came back. The beer is better here.

That calculator looks useful. I plugged in some numbers, but let me see if I understand them. I need to better understand how to use those graphs.

1

6. Weber calculator only simplifies dissipation math but says nothing about biasing.

Nick B´s one is way better, but apparently uses some tube models unaccessible to us.
And we still must settle on some idle current all by ourselves.

I´d do it by hand but please post one EL84 datasheet so we all use the same graphs.

That said, I think datasheets "should" include at least one example of single EL84 running about 190/200V Dc supply, simply because that would be the power amp "recipe" for a cheap European Radio or TV or record player.

EDIT: Plan 9 (from outer Space)

Start with "somewhat similar" single ended 250V +V datasheet example, as in, say, JJ datasheet, and "tweak it until it works with 190V"

Actually build the d*mn thing , the amplifier that is, using the datasheet suggested transformer (which you might already have pulled from some corpse), cathode resistor, the full Monty, and it WILL work , 190V is not *that* far away from 250V, just won´t be optimized.

Then drive it just to clipping, I´m quite certain that since idle current is not optimized and it´s very important in Class A amps, clipping will be quite unsymmetrical, meaning either top or bottom peak will clip way before the other one.
Then replace suggested 135 ohm cathode resistor with two different values, higher and lower, say 100 ohm or 180 ohm and retest.
One of them will carry you closer to desirable symmetrical clipping, meaning new idle current will be halfway between zero and maximum possible.

It might already be close enough or, worst case, adjustable by slightly varying Rk until waveform looks fine (clips symmetrical):

State of the art "seat of the pants" design

2

7. Based on my playing around with NickB's calculator (5 minutes tops, no way "due diligence") I found myself looking at about 200R cathode resistor and a 5k or 6k OT (probably a champ OT would be close enough) for maybe 2W at clipping. If I was building it, that would be my starting point. IIRC that gave me about 30ma for idle diss.

2

8. Originally Posted by cminor9
I went to the future, but I came back. The beer is better here.
Uh oh, that does not bode well for the future does it. After hearing a report about craft brewers starting to offer "lite" options, looks like you may well be right.

About the cathode resistor. I was expecting the true experts to sound off - and we've certainly heard from one - Juan - who's thinking along the same line I was. Eschertron too. Those early European hi fi sets that sported a pair of EL84's running about 275V on the plates and sharing a common 130 ohm cathode resistor. Same deal in some old small guitar amps. Extrapolating roughly from that, I was going to suggest a 200 ohm Rk for you, plus or minus 10%. As Enzo might say, "it's a guitar amp, no need for precision here." As long as you're not threatening to melt down your EL84's plate, all is good. And might I add it's good to see you back.

OT thoughts. Sure you can get a cheap & dirty Champ replacement. But they are very limited in low frequency response, dropping off precipitously below 400 Hz. Granted, with only 2 watts or so, why stress about lows. BUT if you'd like a more solid low end snoop around for a bigger & more 'spensive single ended OT. Last year I had a nice Laney SE amp in for repair with a bad OT, and luckily had a Ceria SE OT left over from a mod I'd done a couple years previous and it worked a charm. Plush tone at bedroom volume levels, quite a nice concept.

3

9. EL84s were used in European radios but typically at Vp around 250V and all the datasheet examples I found were for Vp=250.
I found one chart for Vs=210V in the datasheet of the long life version E84L (has increased plate dissipation of 13.5W):

Nick's calculator works fine but requires some iterative trial and error. I got useful data for Vp=Vs=185V, plate load 3.5k, grid bias -4V. Results: Pout=4.05W, Rk=78R, Ip at idle 51mA.

(In the real circuit the DCR of the OT primary will reduce available plate voltage and output power.)

6

10. The EL84 data sheet doesn't have plots for a Va and Vs of 190V so some interpolation is necessary.

I'm guessing the peak voltage available will only be about 160V. If the OT is 5k then peak (bias) current is 160/5 = 32mA. Drawing an imaginary curve on the graph for Vg2 =180V gives Vg1 at -5V for an Ia of 30mA, Ik will be about 35mA so Rk = 5/0.035 = 143 ohms. I'd start with 150 ohms and adjust from there as Juan said above.

Three different methods giving near enough the same result. It must be right

I used this data sheet EL84.pdf

1

11. While I'm reading this discussion, I'm wondering if the application calls for loudest, cleanest power (near 100% idle dissipation, lower Zpri) or some potentially gratifying distortion at 'home' levels (colder bias, higher Zpri, lower output power). Either option is really a resistor value or an impedance match away. An assortment of power resistors (and maybe speaker cabs/loads) for testing should be sufficient to settle this. We're in the ballpark, it's time to, well, you know...

Originally Posted by Dave H
Three different methods giving near enough the same result. It must be right

2

12. Of course, there is nothing wrong with close to 100% dissipation in a class A stage. And with cathode bias, average plate dissipation actually decreases at high output.

1

13. Of course. But while we're optimizing the power stage, we might want to know what we're optimizing it for. (sorry about the bad grammar)

1

14. Originally Posted by eschertron
Of course. But while we're optimizing the power stage, we might want to know what we're optimizing it for. (sorry about the bad grammar)
Sure, I completely understood. You raised a legitimate question (to the OP). I think the proposals/examples cover most of the reasonable options and any interpolation between should work.

2

15. Some things never change, what a wealth of info!

I'm wondering if the application calls for loudest, cleanest power ... or some potentially gratifying distortion at 'home' levels
I am definitely not looking for the loudest cleanest power. Ideally, I'd get some nice greasy distortion when I slam the amp but when I back off the guitar volume it cleans up a bit. I'd use it for recording, living room playing, or small gigs where I play with a vocalist and another guitar.

Someone else recommended a transformer, I was kind of eyeing this one: https://www.tubesandmore.com/product...w-single-ended I want that primary Z for the el84, right?

Actually build the d*mn thing , the amplifier that is, using the datasheet suggested transformer
That's a really good idea and what I'll most likely do. Flying by the seat of my pants is fun. This project is the first one where I'm not following some schematic, I'm just taking all that I've learned along the way and trying to see what comes out. I'm probably going to pair this with a tiny speaker, one of those 6" or 8" really nice alnico speakers from Weber.

2

16. With only 190V B+, a 5K primary will get you about 2 Watt. You should get one of those "Universal" OTs so you can adjust the primary impedance down to something below 2K where maybe you can coax 5W out of this thing.

2

17. Originally Posted by cminor9
Someone else recommended a transformer, I was kind of eyeing this one: https://www.tubesandmore.com/product...w-single-ended I want that primary Z for the el84, right?

That's a really good idea and what I'll most likely do. Flying by the seat of my pants is fun. This project is the first one where I'm not following some schematic, I'm just taking all that I've learned along the way and trying to see what comes out. I'm probably going to pair this with a tiny speaker, one of those 6" or 8" really nice alnico speakers from Weber.
One of the reviewers said that OT had unusually good bass response. Which is a huge surprise given its size & price, assuming the reviewer is giving us the truth. If that's the case why not let it play thru a larger speaker? Or at least have an extension cab with a larger speaker so you can make use of the bass response available.

3

18. or small gigs where I play with a vocalist and another guitar.
This sounds like a good reason for optimizing output power as bedroom level won't suffice. Meaning low Zpri (2.5..3k) and low Rk (68R). Realistically you won't get more than 3W anyway after the OT.

1

19. Fixed bias to the grid may work out simpler than faffing around with cathode resistor values

2

20. Originally Posted by pdf64
Fixed bias to the grid may work out simpler than faffing around with cathode resistor values
Well, yes

To begin with your "effective" +B goes higher since you are not wasting Volts across a cathode resistor and to boot bias is easily adjustable, so you can drive your amp *just* to clipping and adjust for symmetry.

1

21. To begin with your "effective" +B goes higher since you are not wasting Volts across a cathode resistor
..by maybe 4V. (The OT primary DCR will "eat" up to 15V at idle and about twice as much at full power.) Easy adjusting is certainly a benefit.

1

22. With only 190V B+, a 5K primary will get you about 2 Watt.
If that's the case why not let it play thru a larger speaker? Or at least have an extension cab with a larger speaker so you can make use of the bass response available.
This sounds like a good reason for optimizing output power as bedroom level won't suffice.
These are good suggestions, just not really in line with my goals for the project. I left some context off my original post so as not to bore y'all with the details but that might have been a mistake on my part. But basically, 1-3 watts through a small speaker and loud bedroom levels are ideal here.

I have many low powered amps and cabs I've built over the years. I can plug this little guy into a cab if I want to move more air. I am talking about maybe using this in situations where even a 5f1 is too loud, like with an acoustic guitar and a singer who is singing through just enough PA so she doesn't have to strain. If it gets loud enough to swamp the acoustic guitar that's ok, but I'd like to have at least some usable tones in that ballpark. If I only use it for recording and living room jamming, that's ok too.

Every build I do, it seems like I always justify bigger and bigger speakers. On the 5f1, I just couldn't bring myself to use an 8" so I used a 10" speaker. I'd really like to hold the line here. I'll try a 6", but if I really hate it I'll make the cab large enough to fit an 8".

I have some nice black walnut and cherry laying around, and this will be a woodworking project just as much as it will be an electronics project. I'm want something 12"x12" or even smaller, and going for an antique radio vibe. I'll glaze and stain the wood, cut dovetails, etc, so something smaller is a lot more manageable. I see myself putting it on the credenza in my living room and not hauling it around to just any gig. This one place we play is a *tiny* room, though, and it might be fun to use it there.

As for more info... I am pretty open to how it sounds. It'd be a bonus to come up with something funky and unique. I was going to use that power transformer, a SS rectifier, and a 47uF/250v solen I have on the shelf for filtering. The preamp will be pretty standard 12A_7 fare, a volume and tone knob, and a recovery stage with the other half of the 12A_7. So that leaves the power amp. I have some noval sockets laying around, so I'd just as soon use those and the el84 for low power. So I'm really just trying to design a power stage for that setup. I know that sounds an awful lot like a valve Jr or something.

There was some talk of fixed bias vs cathode bias. I was leaning toward cathode bias just because it's easier and I don't have any negative voltage from the transformer unless I build a rectifier that'll get me some. But I am open to options.

I guess I just don't know as much about how to design the power amp. So if I go to fixed bias, how would I get some negative voltage to apply to the grids? What would be the advantage there?

0

23. It seems like I have some gaps in my knowledge because I don't really understand the advice I'm getting. I'm trying to fill them in. I didn't know how to ask the question I wanted to ask. I was pretty much trying to figure out how to design the power amp.

I did some digging and found this thread here: https://music-electronics-forum.com/...ad.php?t=48879. That person seems to be building something similar to what I want.

It linked to the same page eschertron linked me to: http://bmamps.com/ivds.html but I still didn't know what to do with that.

So I read this: https://robrobinette.com/Drawing_Tube_Load_Lines.htm and this: https://wtfamps.com/load-line-calculations/ Those seem to fill in some gaps for me.

I think I plotted a load line for the el84, but I don't know what to do with that? I know my PT voltage, and I know the tube I want to use. How do I know what values to fill in on http://bmamps.com/ivds.html?

0

24. Buy a used Valve Jr off of CL for \$80. I gigged with one for years. Hint.. use a compressor..

1

25. Merlin’s page on output stage design may help? http://www.valvewizard.co.uk/se.html

1

26. Don´t reinvent the wheel

Build a Champ or Gibson GA5 or VOX AC4 or basic AX84 or similar which will start working (which is fine in my book ), then tweak to adapt it to 190V +B instead of datasheet suggested 250V .

Shortest path to build "just one".

PS: or the "Champ with tone controls" ... is it the Princeton?

2

27. Originally Posted by J M Fahey
Don´t reinvent the wheel

Build a Champ or Gibson GA5 or VOX AC4 or basic AX84 or similar which will start working (which is fine in my book ), then tweak to adapt it to 190V +B instead of datasheet suggested 250V .

Shortest path to build "just one".

PS: or the "Champ with tone controls" ... is it the Princeton?
yeah, that might be what I have to do here. I really do like the idea of figuring it out as I go, but then I started trying to geek out

0

28. Originally Posted by pdf64
Merlin’s page on output stage design may help? http://www.valvewizard.co.uk/se.html
I've been reading and re-reading that that, trying to digest it. I'm not quite there yet, but I'd like to try and learn more instead of painting by numbers. I'll probably just end up taking J M Fahey's suggestion and start with a known model, but I'd also like to understand what I'm doing.

Thanks for bearing with me, folks.

0

29. Both are very valuable goals, but in interest of practicity pick one first

Also the time scale is different: starting with known and tweaking is a short path with relatively little unknowns, and satisfaction guaranteed in, say, within a week.

Learning the full deal may take from Months to Years.

Compare it to learning a foreign Language.

1

30. A single ended output stage is not hard. Just copy the basic circuit of any similar amp. There's really one one good way to do it. The GA5 is as good an example as any. The only real unknowns are the load impedance that will work for you, and the required bias.

So here's the recipe:

1. Buy a universal OT that will allow you to try different loads and choose the one that sounds best to you.

2. Build the circuit using your PT and OT.

3. Adjust the cathode bias resistor to get 65mA or so cathode current. That's less than 80% max dissipation considering screen current, so you can safely go higher (or lower) than 65mA.

4. Play the amp. Try different taps on the OT to see how they sound.

5. (optional) Try adjusting the cathode bias resistor to see how it changes the sound.

6. Report back here to tell us how happy you are.

1

31. 3. Adjust the cathode bias resistor to get 65mA or so cathode current. That's less than 80% max dissipation considering screen current, so you can safely go higher (or lower) than 65mA.
Don't forget that the max. (average) cathode current for a real (PH, Valvo, Mullard, TFK) EL84 is 65 mA.

2

32. Alright! I will report back when I get something. Gonna build this with a friend who is interested in learning, so that’ll be fun too.

Much appreciation for all of the advice!

0

33. The Application Report has details of B+ of 200V operation - that will do.

Cheers,
Ian

2

34. Originally Posted by cminor9
I
So I read this: https://robrobinette.com/Drawing_Tube_Load_Lines.htm and this: https://wtfamps.com/load-line-calculations/ Those seem to fill in some gaps for me.
(Warning - Rob drew a 4k load line for one 6V6 tube and has it idling at 10.88W, and the load line crosses Pmax curve.)

For PP load line lessons - see http://www.valvewizard.co.uk/se.html then, http://www.valvewizard.co.uk/pp.html (each Class A load line is drawn as per a single-ended load line, and then you flip it around and invert it.)

To draw the grid curves at the desired screen voltage you want, you can use a transfer characteristics chart to eyeball the height of the grid curves on the plate characteristics chart. (The JJEL84 datasheet has these two charts conveniently plotted side-by-side for doing this). The dotted red lines in the attached chart project some eyeballed g1 voltages onto the plate characteristics chart and the point where these intersect the plate voltage abscissa (green dotted line), is the horizontal reference point for each grid curve (which you copy the shape of from the existing grid curves, and smudge things a bit - its all ballpark stuff after all)

1

35. Also... If you go with a universal primary OT this point is moot, but if you go with a fixed ratio OT you can always adjust the speaker impedance to half the reflected ratio/primary impedance. If you used an OT for an 8 ohm secondary you could use a 4 ohm speaker like you normally would for a Champ and there are plenty of good sounding models available. This does have the down side of reducing the relative inductance and raising the LF knee, which is already pretty poor with most readily bought, fixed ratio Champ-ish OT's. If you just want the right OT for the job rather than a universal primary type, on my last single ended build I went with the Edcor XSE line and was very happy. They are 10W units. The price was good, but it cost as much as any other anyway because of their stupid, high shipping charges. Their GSXE line offers some 5W units if that aesthetic appeals to you.

https://www.edcorusa.com/xseseries
https://www.edcorusa.com/gxseseries

EDIT: P.S. Nice to see you and welcome back

3

There are currently 1 users browsing this thread. (0 members and 1 guests)

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•