Hi there, so being new to tube amp designs I've been looking at the Berning EA 230 push-pull design, as a few of us are considering cloning it.
The schematic shows a 6.5K output transformer. I've noticed I haven't found a 6.5K transformer from transformer suppliers online but there are 6.6K and others available.
Also, when I look at the output tube data sheet (6JN6) I'm not sure how to verify what impedance should be used with those tubes. I'm confused regarding the 6JN6 as it shows output impedance 18K ohms.
Could anyone clear these questions up?:
Thanks!
The schematic shows a 6.5K output transformer. I've noticed I haven't found a 6.5K transformer from transformer suppliers online but there are 6.6K and others available.
Also, when I look at the output tube data sheet (6JN6) I'm not sure how to verify what impedance should be used with those tubes. I'm confused regarding the 6JN6 as it shows output impedance 18K ohms.
Could anyone clear these questions up?:
1. Is 6.6K acceptable to replace the 6.5K originals?
2. Does anyone know how I could determine the transformer impedance myself (6.5K for example)?
Thanks!
6.5 or 6.6K makes no practical difference.
To figure optimal load, you need a graph of plate currents vs. plate voltages at various control grid voltages. There are several good tutorials on the web on how to draw a load line on the graph. The problem in your case is that output stage is g2-driven, so you need g2 plate curves for your tube with g1 grounded, which are not provided in data sheets. So, either you draw the curves experimentally, or rely on the optimal load as determined by the amplifier designer.
To figure optimal load, you need a graph of plate currents vs. plate voltages at various control grid voltages. There are several good tutorials on the web on how to draw a load line on the graph. The problem in your case is that output stage is g2-driven, so you need g2 plate curves for your tube with g1 grounded, which are not provided in data sheets. So, either you draw the curves experimentally, or rely on the optimal load as determined by the amplifier designer.
Last edited:
Schemos and specs, please?
I don't know anything about this particular design. However, I ran a loadline for the 6JN6 and came up with a very different plate load: 3440R (P-2-P) or 3500R(P-2-P) design nominal. 6.5K looks to be twice what you need (unless this design ran an unusually shallow loadline for some reason).
Attachments
Sorry, didn't look carefully. For this particular tube, they provide g2-driven curves (very unusual!), so you are all set to draw your load line.
The 6JN6 has been used in Pete Millett's DCPP amplifier with up to 8K primary OT, and with Tubelab's mods, down to 3.3K primary I think. The 6JN6 is also the same (except for basing) as 12/6GE5 and 6GV5, and 6JM6.
The Berning amplifier likely uses the 6.5K primary OT to help with lowering the output impedance seen by the speaker some. (ie, improve the damping factor)
The present Berning ZOTL amplifiers have used 6JN6 tubes, and more recently, the slightly bigger 21JV6 tube (similar to 21/6HB5). The latest Millett amplifier is using the even bigger 6HJ5 tube with a 4K OT. All are quite suitable for grid 2 drive.
https://frank.pocnet.net/sheets/123/6/6JN6.pdf
https://frank.pocnet.net/sheets/123/6/6GE5.pdf
https://frank.pocnet.net/sheets/093/6/6GV5.pdf
https://frank.pocnet.net/sheets/123/6/6HB5.pdf
https://frank.pocnet.net/sheets/084/6/6HJ5.pdf
The Berning amplifier likely uses the 6.5K primary OT to help with lowering the output impedance seen by the speaker some. (ie, improve the damping factor)
The present Berning ZOTL amplifiers have used 6JN6 tubes, and more recently, the slightly bigger 21JV6 tube (similar to 21/6HB5). The latest Millett amplifier is using the even bigger 6HJ5 tube with a 4K OT. All are quite suitable for grid 2 drive.
https://frank.pocnet.net/sheets/123/6/6JN6.pdf
https://frank.pocnet.net/sheets/123/6/6GE5.pdf
https://frank.pocnet.net/sheets/093/6/6GV5.pdf
https://frank.pocnet.net/sheets/123/6/6HB5.pdf
https://frank.pocnet.net/sheets/084/6/6HJ5.pdf
Hi there, here's the schematic in case you still need it.
I appreciate everyone's replies. I do prefer to understand why a given impedance transformer was used, so I'll be able to reference the graph(s) in the future when it comes up.
Sounds like I'll be fine using a 6.6K which I can find from several transformer manufacturers/sellers.
Thanks!
I appreciate everyone's replies. I do prefer to understand why a given impedance transformer was used, so I'll be able to reference the graph(s) in the future when it comes up.
Sounds like I'll be fine using a 6.6K which I can find from several transformer manufacturers/sellers.
Thanks!
Last edited:
That is probably the anode impedance, which has almost no connection with the optimum load impedance.Travis Clarke said:
Load impedance is a compromise between power output and distortion, and is affected by chosen bias point, so different people will come to different answers. Having found an answer you like, the OPT can be 10-20% different from this without creating problems.
Having found an answer you like, the OPT can be 10-20% different from this without creating problems
Thanks, I had assumed in general it would be a compromise for power & distortion as you mentioned.
Also thanks for the info, that will help in the future, too. 10-20% is helpful to know. 🙂
Thanks, I had assumed in general it would be a compromise for power & distortion as you mentioned.
Also thanks for the info, that will help in the future, too. 10-20% is helpful to know. 🙂
The g2 graph is not quite right either...the bloody thing is a triode now with g1 tied to the cathode. It will make triode curves of some sort.
Cheers,
Douglas
The g2 curves on the datasheet are with grid 1 at 0V, so all is well with the g2 graph there for g2 drive. Both g1 and g2 drive have similar plate Z.
The amount of electrostatic field from the plate that penetrates the openings between all the grid wires back to the cathode, determines it's feedback effect at the cathode. Linearity of Maxwell's equations says the electrostatic fields of the other grids sum linearly with the plate field at the cathode to determine drive and neg. feedback effects.
The amount of electrostatic field from the plate that penetrates the openings between all the grid wires back to the cathode, determines it's feedback effect at the cathode. Linearity of Maxwell's equations says the electrostatic fields of the other grids sum linearly with the plate field at the cathode to determine drive and neg. feedback effects.
Last edited:
Hey guys, quick follow up - I've duplicated the Berning EA-230 amp and I'm using 6.5K Hammond 1650-HA output transformers. Yes, the same Hammond that makes project enclosures also makes great tube amplifier output transformers!
They work great and the sound is excellent. My first real tube amp!
Thanks for your input & information.
They work great and the sound is excellent. My first real tube amp!
Thanks for your input & information.
-= Looking For Typical Efficiency Info For Push-Pull Tube Amps (Ex.Berning EA-230) =-
Hey guys, I'm hoping someone can provide a rough estimate of typical push-pull amplifier efficiency as I'm working on power supply design at the moment.
The example amp in question is the Berning EA-230 2 x 30W stereo amplifier w/ 2x 6JN6 output tubes & a +400V supply. I've built a clone already and all is well, but I'm unsure of the power supply capability I need to account for at its rated output power.
Additionally, as it uses a -200V grid voltage, should I expect the -200V supply current to increase somewhat or does it remain relatively the same as output power increases? (I'm new to tube design as you may have guessed).
Thanks!
Hey guys, I'm hoping someone can provide a rough estimate of typical push-pull amplifier efficiency as I'm working on power supply design at the moment.
The example amp in question is the Berning EA-230 2 x 30W stereo amplifier w/ 2x 6JN6 output tubes & a +400V supply. I've built a clone already and all is well, but I'm unsure of the power supply capability I need to account for at its rated output power.
Additionally, as it uses a -200V grid voltage, should I expect the -200V supply current to increase somewhat or does it remain relatively the same as output power increases? (I'm new to tube design as you may have guessed).
Thanks!
The manufacturer specs 60-160W of draw. The tube heaters won't change depending on what the amp is doing. The -200V rail is referred to as a bias rail (but it's not only serving that purpose), though I would suspect it has 5-10mA per channel of draw for the bias pots and the cathode followers it feeds.
The output tubes are biased at 3mA each (according to the schematic) which is a very shallow class B operating point. You'll want to design your power supply to have good regulation for when the amp actually gets up and runs (could be ~50mA or so per channel at rated power).
The output tubes are biased at 3mA each (according to the schematic) which is a very shallow class B operating point. You'll want to design your power supply to have good regulation for when the amp actually gets up and runs (could be ~50mA or so per channel at rated power).
Thank you I appreciate that!
I don't know how the thread I started ended up here in this thread again, even though it's a separate topic.
I don't know how the thread I started ended up here in this thread again, even though it's a separate topic.
You should always keep your questions about a build in one thread. For example, this is for a car, right, which you have said in other threads. But audiowize answered your question assuming a line PS.
Ok, thanks I'll remember that in the future.
I'm subtracting roughly 18W for the heater power consumption so I'm going to estimate perhaps 125W or so would be required?
I thought somewhere that I saw a push-pull amp would be about 65% efficient, giving me about 2*(30W/.65) = 92W minimum pow. supply capability required (so let's say 100W). Which would still be below what he mentioned.
(The Berning EA-230 heater supplies were wired in parallel unlike the car version, which I've verified is right at 18W.)
I'm subtracting roughly 18W for the heater power consumption so I'm going to estimate perhaps 125W or so would be required?
I thought somewhere that I saw a push-pull amp would be about 65% efficient, giving me about 2*(30W/.65) = 92W minimum pow. supply capability required (so let's say 100W). Which would still be below what he mentioned.
(The Berning EA-230 heater supplies were wired in parallel unlike the car version, which I've verified is right at 18W.)
Very very few tube amps much exceed 50% efficient.
6L6GC @ 450V, fix bias, 5.6K load, makes 55W out from 104.4W of DC input: 52.7% efficient. The 26W output condition wants up to 53W input: 49% efficient. (Not counting heaters and drivers.) EL84 will do similar: 50% @ 17W out. 6550 @ 60W-100W runs 48%-55% efficient (a hair better with separate G2 supply). 6F6 in AB2 does 48%. HK257 will do 63%, but this is a 1,500V 315 Watt condition. 6146 can do similar at 750V 120W level.
This is for "low THD". In guitar and car-sound markets, it is common to push well into clipping. "RMS Power output" rises somewhat faster than DC consumption. Some time on the test bench and you can publish "65% efficient" numbers.
But 65% or 50% efficient is only 1.14dB different, a teeny amount.
6L6GC @ 450V, fix bias, 5.6K load, makes 55W out from 104.4W of DC input: 52.7% efficient. The 26W output condition wants up to 53W input: 49% efficient. (Not counting heaters and drivers.) EL84 will do similar: 50% @ 17W out. 6550 @ 60W-100W runs 48%-55% efficient (a hair better with separate G2 supply). 6F6 in AB2 does 48%. HK257 will do 63%, but this is a 1,500V 315 Watt condition. 6146 can do similar at 750V 120W level.
This is for "low THD". In guitar and car-sound markets, it is common to push well into clipping. "RMS Power output" rises somewhat faster than DC consumption. Some time on the test bench and you can publish "65% efficient" numbers.
But 65% or 50% efficient is only 1.14dB different, a teeny amount.
So are you comparing with regards to the bias consumption? I'm wondering because I measured a pretty low "steady state" power consumption at 0W output (idle) which was about ~2W per channel. Somewhere today I saw an estimate of push-pull being 50-60% efficiency so I'm using the 60% factor.
I know that other push-pull designs including the EL84 like the Dynaco ST-70 consume more at idle due to higher bias current.
Thank for the reply. I'll try to account for 50% then I suppose. I didn't realize they were quite that inefficient. :/
I know that other push-pull designs including the EL84 like the Dynaco ST-70 consume more at idle due to higher bias current.
Thank for the reply. I'll try to account for 50% then I suppose. I didn't realize they were quite that inefficient. :/
Let's see a pic of your amp.🙂 A guy over at Audiokarma built a copy of the 230 many years ago now using 5k toroids for OPT's. He said that the idle current for each output tube was 3ma.
jeff
Last edited:
- Status
- Not open for further replies.