OK, so if you have 215v max then you can slide the operating point along the loadline with optional cathode resistors to center it within the range the driver can give it, peak to peak. I would set it as high up the loadline as you can get it before clipping. But you need to add @50uf of bypassing to get the loadline to reflect normal curves. Then just accept the lower power as its limitation. Good luck, enjoy the project.
I’m able to increase the plate voltage to 243V with the current mains transformer, but I have another mains transformer I can set 300V as well.
So basically the only fix parameter is the Output transformers 5k.
So basically the only fix parameter is the Output transformers 5k.
Maximum plate voltage and current are specified for the idling (no signal) condition. The valve is expected to exceed one, and then the other, of these values for large signal conditions. Full output swings are also rare for music reproduction - (but not for guitar amps and such) - giving even more latitude.
All good fortune,
Chris
As no datasheet that I'm aware of a plate current rating is given, I think you were speaking of cathode current. And that's max 65 mA, at least for European EL84's.
Best regards!
Would you consider using the UL tap of the output transformer you already have, adding some “Schade” feeeback to further linearize it? That would permit to avoid raising the B+ as you should for a triode (because you gain around 50V of swing on the left side of the loadline plot) and help you get some more power out of it. 243V will be fine with cathode biasing.
As for the driver: I would raise its current from 2,7 mA to at least 5, better 10 mA. Something like 15k on the plate and 330 Ohm on the cathode should work with a B+ of 240-250V.
As for the driver: I would raise its current from 2,7 mA to at least 5, better 10 mA. Something like 15k on the plate and 330 Ohm on the cathode should work with a B+ of 240-250V.
Thanks, I’ll try. My other tube amp (Baby Huey variant) using the UL configuration, but it is PP not SE.
There are a lot of options to try and learn
There are a lot of options to try and learn
Running at 100% plate dissipation is not a good suggestion for the reiability of the amp, especially when running at 80% means loosing one Wrms or less, but gaining alot in reliability.
As I can I runn at 100% of Pa with a good results and long life; of course a good tube must be used and with 84 it is possible to get a beautiful stuff as EL84M Sovtek (soviet)
Walter
Just use cathode current and there will be some margin under max plate dissipation as some current is passed by the SG. When music is playing the plate dissipation is reduced because the idle current wattage is transformed out through the OPT as signal power wattage.
This is just your opinion that is against any engineering rule.
EL84M it’s not exactly an EL84, and at 12 W of plate dissipation you are not running at 100% but a more reliable 85%, because its plate dissipation is 14 W ( https://www.tubesandmore.com/sites/...ted_files/t-el84-sovt_specification_sheet.pdf )
it is a 84 and it is strong
And, with a proper trafo, it is super
Years ago ( around year 2000) I done and sold a kit, MINIMONO ( stereo single end), with it. Beautiful stuff
With MINIPRE ( phono included)
Until today I haven't seen anything from you; just simulation.
Walter
Online reviews (E.G. TNT) do not say “beautif stuff”, but enhance some problems (dark sounding, picking up FM radios, voices are behind, etc…).
You have to search better on the forum.
Why are you absessed by simulations?
Gentlemen, I like you both and I have learned some things from you both. But please cool your feud.
Why don’t you call each other on the phone and have a good heart-to-heart chat in Italian. I suspect you will find that you have more in common than any differences.
Why don’t you call each other on the phone and have a good heart-to-heart chat in Italian. I suspect you will find that you have more in common than any differences.
Hi gaszto
I suggest you post the schematic of your power supply. What is your B+ when there are no tubes in your amp? Once there is a load, your B+ will drop, right? You will anticipate it dropping down to 215v... so if there is no load, it should be something like 300v or more.
You can use software to simulate the voltage drop and design power supplies. I use duncan amps PSU-II software. It's very accurate if used correctly. It does take a bit of time to learn though...
Of course, roughly calculating the voltage drop is very helpful too. I calculate below that with your schematic, you are looking to get something like 200V A-K, running around 50mA and the OPT is 5k right? Well if the current flow is suddenly a lot lower, say 20mA, then your power supply needs to supply a higher voltage! How much higher? Easy, ohm's law tells me that 5kOhm and 20mA gives 100V - so I need to ADD this 100V to my 200V = 300V supply (at least). I think you have been figuring this out in previous posts though, so I will let you work on that separately. 😉
(Just an aside, when the current drops, the OPT primary impedance will not be 5kOhm either... This 'static' value is only valid really for the specs of the transformer, right? And your Toridy OPT has a nominal anode current of 150mA noted in the online specs sheet...)
Let's look at your loadline though. When I look at your schematic in post #1, I see that the cathode of your EL84 is at 5.18V, and the cathode resistor is 2x220 ohm in parallel (= 110 ohm). This means 5.17V/110ohms = 0.0471A = 47 mA.
I prefer to work on a screenshot of the actual loadline. If you can't measure the actual load lines for a tube, then the datasheets are usually a good place to start. For the EL84, we have the Philips datasheet, and it even has triode curves. 🙂
Take a screenshot of that, post it into MS Paint, or any other SW you prefer, then start to draw that ideal load line. Your OPT has a 5k primary impedance, so let's draw a line from say 300V to (300V/5kOhm)=60mA). Then MOVE that line around in MS paint until it intersects that 5.2V, 47mA point.
See how I did that? I can move that red loadline around all over the place until it hits the desired intersection point. The blue line that intersects 200V suggest to me that you should have A-K voltage of 200V. kmaier is right - this amp is not going to have much power... I hope this is what you were expecting. Otherwise a design change may be in order...
Ok, so then I was thinking why are you not quite measuring 200V A-K? Sometimes people make the mistake of doing measurements without a load on the secondary of their OPT. Did you have a resistive load on the secondaries of your OPT? I am sure you did. Of course tubes can vary as well... In any case you must have covered that. Or it could be that the OPT is designed for 150mA nominal load and you are only drawing 50mA.... I'm not too worried about this though.
Now let's consider the design. I think the biggest problem is that you don't have any cathode by-pass cap on your EL84. You're going to need that or you won't have any bass response and your amp. I pretty much think this is the problem you are facing here.
Here is the standard formula to calculate the cathode bypass cap: C= 1/((2π*Frequency)*R) - Let's use a cut-off frequency of 20Hz - as you go lower you will see you need a bigger cap.
C= 1/(2π*20Hz*110Ohm) = 7.23e-5 Farads = 72uF (use 100uF instead)
Put this 100uF cathode bypass cap into your LT Spice simulation and see if it helps. Of course you could use a string of LED's instead, or use fixed bias instead...
I suggest you post the schematic of your power supply. What is your B+ when there are no tubes in your amp? Once there is a load, your B+ will drop, right? You will anticipate it dropping down to 215v... so if there is no load, it should be something like 300v or more.
You can use software to simulate the voltage drop and design power supplies. I use duncan amps PSU-II software. It's very accurate if used correctly. It does take a bit of time to learn though...
Of course, roughly calculating the voltage drop is very helpful too. I calculate below that with your schematic, you are looking to get something like 200V A-K, running around 50mA and the OPT is 5k right? Well if the current flow is suddenly a lot lower, say 20mA, then your power supply needs to supply a higher voltage! How much higher? Easy, ohm's law tells me that 5kOhm and 20mA gives 100V - so I need to ADD this 100V to my 200V = 300V supply (at least). I think you have been figuring this out in previous posts though, so I will let you work on that separately. 😉
(Just an aside, when the current drops, the OPT primary impedance will not be 5kOhm either... This 'static' value is only valid really for the specs of the transformer, right? And your Toridy OPT has a nominal anode current of 150mA noted in the online specs sheet...)
Let's look at your loadline though. When I look at your schematic in post #1, I see that the cathode of your EL84 is at 5.18V, and the cathode resistor is 2x220 ohm in parallel (= 110 ohm). This means 5.17V/110ohms = 0.0471A = 47 mA.
I prefer to work on a screenshot of the actual loadline. If you can't measure the actual load lines for a tube, then the datasheets are usually a good place to start. For the EL84, we have the Philips datasheet, and it even has triode curves. 🙂
Take a screenshot of that, post it into MS Paint, or any other SW you prefer, then start to draw that ideal load line. Your OPT has a 5k primary impedance, so let's draw a line from say 300V to (300V/5kOhm)=60mA). Then MOVE that line around in MS paint until it intersects that 5.2V, 47mA point.
See how I did that? I can move that red loadline around all over the place until it hits the desired intersection point. The blue line that intersects 200V suggest to me that you should have A-K voltage of 200V. kmaier is right - this amp is not going to have much power... I hope this is what you were expecting. Otherwise a design change may be in order...
Ok, so then I was thinking why are you not quite measuring 200V A-K? Sometimes people make the mistake of doing measurements without a load on the secondary of their OPT. Did you have a resistive load on the secondaries of your OPT? I am sure you did. Of course tubes can vary as well... In any case you must have covered that. Or it could be that the OPT is designed for 150mA nominal load and you are only drawing 50mA.... I'm not too worried about this though.
Now let's consider the design. I think the biggest problem is that you don't have any cathode by-pass cap on your EL84. You're going to need that or you won't have any bass response and your amp. I pretty much think this is the problem you are facing here.
Here is the standard formula to calculate the cathode bypass cap: C= 1/((2π*Frequency)*R) - Let's use a cut-off frequency of 20Hz - as you go lower you will see you need a bigger cap.
C= 1/(2π*20Hz*110Ohm) = 7.23e-5 Farads = 72uF (use 100uF instead)
Put this 100uF cathode bypass cap into your LT Spice simulation and see if it helps. Of course you could use a string of LED's instead, or use fixed bias instead...
Last edited:
The power supply question is difficult 🙂 If there are no tubes then the B+ is around 300V little more.
Originally the power supply of the EL84 was a CRC filter 330uF - 470R - 470uF after the EL84 anode voltage connection TL783 based regulator dropped the voltage to the appropriate level. When I built it there was some humming noise with my Grado headphone. (quite high sensitivity) with this configuration the B+ for EL84 are exactly 220V when the tubes are installed. Because of the noise, one of my friend recommended to remove the 470R change the order of the capacitors (first 470uF the 330uF) and between the capacitors place a gyrator.
This solved my humming noise issue, it was dead silent. Additionally there was no voltage drop on the 470R resistor, therefore the B+ is 243V
When I rebuilt the amp, for driving speakers instead of headphones, I realized TL783 based regulator not working optimal, not easy the adjust because the ECC88 only a few mA and for TL783 15mA is the minimum. So I removed all of the component what is neccessary for TL783, and put Ale Moglia's hybrid mu follower, and use only the b+ supply of that pcb. I like it, sound really good, highly adjustable (I can set anything let's say between 100V-240V for ECC88)
This is the current configuration. More or less I have the complete power supply in LTspice.
As I mentioned I have an other transformer, with 250V secondary, it is around 350V after the diode bride rectifier. Probaly it would be better, but I'm not sure it's voltage drop under load, maybe it will too much. Anyway, I have Salas HV reg as well, I can use, it instead of the current gyrator, and then the B+ for EL84 will be also adjustable
Yes, I have a speaker with 101dB/W/m sensitivity so 1.5-2W should be enough.
I also planning to remove the one of the 200R resistor from the 2x200R and put a bypass capacitor in the place. It's an easy change.
Why do we have to reinvent the wheel? There are so many schematics of amplifiers that work very well, but still there is someone out there on the world wide web who has to reinvent the wheel. Instead, use constructions that others have built after? The Little One, is my el 84 se favourite. Built it 4 to 5 years ago and it plays beautifully
Ciao Francois,
can I be offended by this statement? 😁 Just jocking.
I can understand that it can be seen as an argue, but I’m not attacking anyone: as in other threads I’m just writing what is technically obvious to avoid other people make mistakes and as a reply I get the same old Schopenhauer-style personal attacks.
From Schopenhauer to Mr. Wolf, I like you too and I appreciate the good intentions of your post. Thanks.