Respect to all you tube dudes from a newbie...
Too keep us busy up here in cold artic Scandinavia (except for chopping ice and chasing away icebears from the garbage bins) we have bought us an ESL kit from ER-Audio.
To make it a real challenge, did so without output transformers!!
The Viking spirit goes on....
It is now time to get a tiny little output from them. I've read a lot of designs and Morgan Jones x3 so many times that it's falling apart.... so I'm ready to come out of the closet! Regarding safety I'm scared stiff, I've been electrocuted by 220V in the shower, 10 years ago...
I was planning on doing a Tubecad/John Broskie design as attached. A 6C45PE or 5842 driver and a EL34 for output for a cheap and easy start with a +/- 300V bipolar PS.
So I have a couple of questions I can't seem to find the answer from.
-An ordinary output transformer with secondary unconnected, can act as an anode load for the output stage. How do I choose this, do I just care about the primary resistance fitting with the output tubes? Frequency response and all this will have no impact...and so on... Anything else? (I bought second version of MJ's book to find an answer but no luck!)
-How hot can we put the operating point for EL34's?
With a choke load the peak possible voltage swing is almost twice the HT, how should this affect my choice for operating point? I have 600V over the tube, so do I put operating point at 400V, 450 or 500V or something?
- Will this bipolar PS mess up the possibility for twice the HT voltage swing? There is still 600V over the tubes, but 300 of these are negative!!! It's mentally distressing!!!
Comments will be much appreciated...
🙂
I will then by parts and get back to you guys...😉
Too keep us busy up here in cold artic Scandinavia (except for chopping ice and chasing away icebears from the garbage bins) we have bought us an ESL kit from ER-Audio.
To make it a real challenge, did so without output transformers!!
The Viking spirit goes on....
It is now time to get a tiny little output from them. I've read a lot of designs and Morgan Jones x3 so many times that it's falling apart.... so I'm ready to come out of the closet! Regarding safety I'm scared stiff, I've been electrocuted by 220V in the shower, 10 years ago...
I was planning on doing a Tubecad/John Broskie design as attached. A 6C45PE or 5842 driver and a EL34 for output for a cheap and easy start with a +/- 300V bipolar PS.
So I have a couple of questions I can't seem to find the answer from.
-An ordinary output transformer with secondary unconnected, can act as an anode load for the output stage. How do I choose this, do I just care about the primary resistance fitting with the output tubes? Frequency response and all this will have no impact...and so on... Anything else? (I bought second version of MJ's book to find an answer but no luck!)
-How hot can we put the operating point for EL34's?
With a choke load the peak possible voltage swing is almost twice the HT, how should this affect my choice for operating point? I have 600V over the tube, so do I put operating point at 400V, 450 or 500V or something?
- Will this bipolar PS mess up the possibility for twice the HT voltage swing? There is still 600V over the tubes, but 300 of these are negative!!! It's mentally distressing!!!
Comments will be much appreciated...
🙂
I will then by parts and get back to you guys...😉
Attachments
Schematic
Yes schematic was used for headphones, but I'm going to use more powerful tubes... I don't aim for very high Sound Pressure Levels's so to say. A electrostatic headphone is still just a smaller capacitor...
Yes schematic was used for headphones, but I'm going to use more powerful tubes... I don't aim for very high Sound Pressure Levels's so to say. A electrostatic headphone is still just a smaller capacitor...
You need a lot of voltage swing to drive ESLs. Really a lot. A good starting point would be to use a conventional EL34 power amplifier (with output transformer) and take the output directly from the anodes. Don't worry about not loading the secondary of the output transformer. Unless you're in the habit of grossly overdriving the amplifier (like a guitarist), you won't damage the output transformer.
Resistive loading of the output stage is feasible for headphone amplifiers but not for loudspeakers. To get the same voltage swing using resistive loading as choke/transformer loading, you need twice the HT. Assume 400V across each valve, and 70mA through it, that would be 28W dissipated in each anode load resistor and require an 800V HT. You will find that the cost of an amplifier rises very quickly with rising HT voltage.
Once you've gained a bit of confidence with EL34 at 400V, you might want to try KT88 at 500V or PL519 at even more. After that, you're looking at the smaller transmitter valves.
Resistive loading of the output stage is feasible for headphone amplifiers but not for loudspeakers. To get the same voltage swing using resistive loading as choke/transformer loading, you need twice the HT. Assume 400V across each valve, and 70mA through it, that would be 28W dissipated in each anode load resistor and require an 800V HT. You will find that the cost of an amplifier rises very quickly with rising HT voltage.
Once you've gained a bit of confidence with EL34 at 400V, you might want to try KT88 at 500V or PL519 at even more. After that, you're looking at the smaller transmitter valves.
Hi ,
I had some degree of success driving DIY mid and top panels directly . I built a simple dedicated amp using a 7236 double triode , , driven by a 6EM7 input stage/concertina , coupling caps to the output valve formed a high pass filter for the panels . A standard PP output transformer was used and was connected as an autoformer , in other words the 7236 anodes were connected to the UL taps and the anode taps were connected to the stators . Diaphragm was biased negative , stators sat at 250V
After a few arcs and sparks and a few more self-inflicted problems with the panels , (my fault I got a bit too 'heat gun happy' !!!) , I got it to sound ok but it was just too lethal to bring indoors ! I also did not have enough decent iron to buildup a bass amp . Eventually I cap-coupled to the stators with teflon caps , changed the 7236 coupling caps to 0.22uF and ran the bass drivers (Jordan JX150) from the transformer secondary . I never got the panels to integrate well this way though .
cheers
316a
I had some degree of success driving DIY mid and top panels directly . I built a simple dedicated amp using a 7236 double triode , , driven by a 6EM7 input stage/concertina , coupling caps to the output valve formed a high pass filter for the panels . A standard PP output transformer was used and was connected as an autoformer , in other words the 7236 anodes were connected to the UL taps and the anode taps were connected to the stators . Diaphragm was biased negative , stators sat at 250V
After a few arcs and sparks and a few more self-inflicted problems with the panels , (my fault I got a bit too 'heat gun happy' !!!) , I got it to sound ok but it was just too lethal to bring indoors ! I also did not have enough decent iron to buildup a bass amp . Eventually I cap-coupled to the stators with teflon caps , changed the 7236 coupling caps to 0.22uF and ran the bass drivers (Jordan JX150) from the transformer secondary . I never got the panels to integrate well this way though . cheers
316a
It is interesting for me how output power for ESLs is dependant on drive voltage.For as i understand, twicing drive voltage (assuming bias voltage is not changed) , twices output power(3db).But , the required driving power quadruples(P=(U^2)/2 . That would mean that , for each 3db output SPL rise wee need 4 times more drive power ? Is it correct ?
Lukas
Lukas
Voltage drive
Voltage swing is needed and the more the better, but the smaller the capacitance the less voltage swing is needed. Martin Logans need f.ex. a lot less than Quad's ESL. People have reported with Martin Logans that produce 300-400V p-p swing produce an almost decent listening level.
The reasons for my questions is simply to unravel how to produce the highest possible voltage swing. ER-Audio's ESL are very low capacitance so less current is demanded for the high's. So I'll try EL34 at 400V.
EC8010: I was planning on using choke loaded output stage, except for the primary resistance I don't know which parameteres to look for in a choke (books say nothing). Also with a choke one can swing to almost twice the HT voltage as in Morgan Jones "The Beast" esl amp. So choke is going in....
316a: Interesting...
Bazukaz: Formula, as I can recall it is correct, and power requirements quadruples for each doubeling of SPL...😀
Attached is Thorsetn's setup (as posted her) for an ESL amp....
Voltage swing is needed and the more the better, but the smaller the capacitance the less voltage swing is needed. Martin Logans need f.ex. a lot less than Quad's ESL. People have reported with Martin Logans that produce 300-400V p-p swing produce an almost decent listening level.
The reasons for my questions is simply to unravel how to produce the highest possible voltage swing. ER-Audio's ESL are very low capacitance so less current is demanded for the high's. So I'll try EL34 at 400V.
EC8010: I was planning on using choke loaded output stage, except for the primary resistance I don't know which parameteres to look for in a choke (books say nothing). Also with a choke one can swing to almost twice the HT voltage as in Morgan Jones "The Beast" esl amp. So choke is going in....
316a: Interesting...
Bazukaz: Formula, as I can recall it is correct, and power requirements quadruples for each doubeling of SPL...😀
Attached is Thorsetn's setup (as posted her) for an ESL amp....
Attachments
Correction...
Correction, drawn by Thorsten, design by Morgan Jones and build by Jon Finlayson... Nice to be correct...
Correction, drawn by Thorsten, design by Morgan Jones and build by Jon Finlayson... Nice to be correct...
Actually, design by Thorsten, drawing by Morgan, build by Jon.
The choke: Ideally, one would like a low-capacitance choke, but as you're going to hang 1nF or so in parallel, even that isn't crucial. The other aspect that's important is the inductance; you need it to be sufficiently high that it doesn't shunt the load at the lowest frequency of interest.
The choke: Ideally, one would like a low-capacitance choke, but as you're going to hang 1nF or so in parallel, even that isn't crucial. The other aspect that's important is the inductance; you need it to be sufficiently high that it doesn't shunt the load at the lowest frequency of interest.
Operation EL34
Thanks EC8010, at least I got the builder right 🙂 So output stage anode load choke or transformer need enough inductance and correct resistance but other than that not to much to worry about, I'll just see what I can get a hold of....
Regarding EL34 operating points, a lot of tubes are designed to run close to Va Max. Max anode voltage for EL34 is 800V, but loadlines don't look like there is any point in running them close to this... In application data on Duncanamps for the EL34 there is a spec. for running them push/pull AB1 at 775V!!! What do you say to that!!! Triode connected curves are easier to analyze, but where are they!!!
Regarding voltage swing, example... lets say I have 420V HT, choke loaded output stage and 400V on the anode, I can then swing to almost twice the HT (because of choke), so I just make sure the driver stage can deliver grid voltage so that anode can swing to (lets say) 700V, right? I think so, just want to be completely, without any doubt sure, I think... 😀
Thanks EC8010, at least I got the builder right 🙂 So output stage anode load choke or transformer need enough inductance and correct resistance but other than that not to much to worry about, I'll just see what I can get a hold of....
Regarding EL34 operating points, a lot of tubes are designed to run close to Va Max. Max anode voltage for EL34 is 800V, but loadlines don't look like there is any point in running them close to this... In application data on Duncanamps for the EL34 there is a spec. for running them push/pull AB1 at 775V!!! What do you say to that!!! Triode connected curves are easier to analyze, but where are they!!!
Regarding voltage swing, example... lets say I have 420V HT, choke loaded output stage and 400V on the anode, I can then swing to almost twice the HT (because of choke), so I just make sure the driver stage can deliver grid voltage so that anode can swing to (lets say) 700V, right? I think so, just want to be completely, without any doubt sure, I think... 😀
Your project is interesting ,
Why don't you want to use transistors instead of tubes ? I have got one schematic with FETs , i could post if you are interested in.For now , i am trying to design a simple JBT/IC direct drive amp(for up to 500pF load).
Regards,
Lukas
Why don't you want to use transistors instead of tubes ? I have got one schematic with FETs , i could post if you are interested in.For now , i am trying to design a simple JBT/IC direct drive amp(for up to 500pF load).
Regards,
Lukas
And for the choke, make sure that the inductance does not form a parallel resonant tank circuit with the capacitance of the ESL at frequencies of interest.
Look at :
http://www.diyaudio.com/forums/showthread/t-42928.html
There say that choke load is not good for ESLs.
http://www.diyaudio.com/forums/showthread/t-42928.html
There say that choke load is not good for ESLs.
I have to say that the very premise of this thread is wishful thinking: "small and easy ESL amp" is an oxymoron if I ever heard of one. Maybe for ESL headphones, but certainly not for full range speakers. I'm speaking as one who has built huge 833-based ESL amps that crank out the THOUSANDS of volts needed for normal volumes and also the hefty currents required at high frequencies. Not easy and not small.
Maybe not for full range , but for a sensitive mid/high panel , it should work.Not too load , but the quality may be very good , assuming that amplifier has low distortion.
Oxymoron
😀
Thanks Lukas for the link, I saw it before once. Interesting... So you have a project going too, cool!!! Regrading choke loading of output stage, I think I will just have to make my own experience regarding this, some do it some don't... It's always interesting to look at schematics, so if you want to send me one, thanks!!!
Brian, thanks for reply, wishful thinking it is..... I don't expect more than hopefully medium or a little low listening levels. Some report decent sound levels from 4-500V with low capacitance speakers... Mine are low capacitance... Which ESL's do you use by the way?
Anyways it's just kind of like a dangerous adventure and feasibility study at starts... I'm planning on using a sub for lows, and I can anyways not hear 20kHz. (my limit goes on 17kHz, so every little bit counts) 😀
Remco at Ultranalog, Morgan Jones and some others have design that swing around 1,5kV and are pleased with that, so if I can build a EL34 that swing 5-600, I'll hear some faint screams from the ESL's at least. I guess they'll be begging for more volts
I also have a backup plan, kind of, I'll rebuild and voltage double the PS and use 211's if I get the taste for it...
😀
Thanks Lukas for the link, I saw it before once. Interesting... So you have a project going too, cool!!! Regrading choke loading of output stage, I think I will just have to make my own experience regarding this, some do it some don't... It's always interesting to look at schematics, so if you want to send me one, thanks!!!
Brian, thanks for reply, wishful thinking it is..... I don't expect more than hopefully medium or a little low listening levels. Some report decent sound levels from 4-500V with low capacitance speakers... Mine are low capacitance... Which ESL's do you use by the way?
Anyways it's just kind of like a dangerous adventure and feasibility study at starts... I'm planning on using a sub for lows, and I can anyways not hear 20kHz. (my limit goes on 17kHz, so every little bit counts) 😀
Remco at Ultranalog, Morgan Jones and some others have design that swing around 1,5kV and are pleased with that, so if I can build a EL34 that swing 5-600, I'll hear some faint screams from the ESL's at least. I guess they'll be begging for more volts
I also have a backup plan, kind of, I'll rebuild and voltage double the PS and use 211's if I get the taste for it...
I use Quad ESL-63s for my basic system, although I'm playing with DIY ESLs. Quads are pretty high impedance ESLs (high voltage, low current) at the step-up secondary. Full rated "power" equates to 4800 Volts RMS at the secondary of the step-up transformer. That's 13,600 volts peak-to-peak! KT-88s just won't cut it. By the way, all else being equal, a low capacitance ESL will need more voltage, not less (but it needs less current).
Hey, some people are quite happy with 3 watt SE amps, so some people will also be happy with ESL amps delivering just a few hundred volts, but to really make most ESLs sing, you need voltages that are in the range of transmitter tubes. Hence my use of 833s in a push-pull bridge. I have watched the musical waveforms on a scope (with a high voltage probe) while music was playing at reasonable levels on my Quads. It's impressive to watch how many throusands of volts are seen.
Yeah I'll just have to settle for what I get regarding output level....
but a bit back to topic, how high in volts can I load the output tube (EL34 or KT88). Given reasonable distortion and given a choke or transformer as anode load to allow swing to almost twice the HT...
EL34 are normally run with 400, KT88 closer to 500.... I don't understand these as absolute values..
Anybody have any experience with this?
🙂
but a bit back to topic, how high in volts can I load the output tube (EL34 or KT88). Given reasonable distortion and given a choke or transformer as anode load to allow swing to almost twice the HT...
EL34 are normally run with 400, KT88 closer to 500.... I don't understand these as absolute values..
Anybody have any experience with this?
🙂
- Status
- Not open for further replies.