Hi Everyone,
I am new on diyAudio. Played with tubes in the 70's as a teenager. I finally got back to my old hobby. I am trying to get ready to build a PP as simpler as possible without loosing the quality of an audiophile sound( like to listen to classical, vocal, rock and jazz ).
I recently bought pair of LL1660 to drive 6p3s's 😉. I haven' t decide yet what Alternative of connection to use. Inclined to use Alt V ( SE to PP). Driver tube 6p45 or ?
Anybody have exeperience using this setup?
Thanks,
Abszero
PS. Abszero means Absolute zero ( 0 K ) or anything you wish for...
I am new on diyAudio. Played with tubes in the 70's as a teenager. I finally got back to my old hobby. I am trying to get ready to build a PP as simpler as possible without loosing the quality of an audiophile sound( like to listen to classical, vocal, rock and jazz ).
I recently bought pair of LL1660 to drive 6p3s's 😉. I haven' t decide yet what Alternative of connection to use. Inclined to use Alt V ( SE to PP). Driver tube 6p45 or ?
Anybody have exeperience using this setup?
Thanks,
Abszero
PS. Abszero means Absolute zero ( 0 K ) or anything you wish for...
Take a look at some established designs first
No sense reinventing the wheel. There are a lot of designs out there. Pick something that is close to what you have in mind for a starting point. Fun stuff.
http://www.nutshellhifi.com/IT-Triode-Amp.gif
http://www.diyparadiso.com/proj/mods-upgrades/8010velleman/K8010_LL1660.gif
http://www.diyparadiso.com/proj/amps/se/845captain/schema 845light 2005.jpg
http://www.audiodesignguide.com/my/sep_mk2.gif
No sense reinventing the wheel. There are a lot of designs out there. Pick something that is close to what you have in mind for a starting point. Fun stuff.
http://www.nutshellhifi.com/IT-Triode-Amp.gif
http://www.diyparadiso.com/proj/mods-upgrades/8010velleman/K8010_LL1660.gif
http://www.diyparadiso.com/proj/amps/se/845captain/schema 845light 2005.jpg
http://www.audiodesignguide.com/my/sep_mk2.gif
I don't think the LL1660 is great for phase splitting - you could try it but the symmetry may not be optimal. They make a special version for phase splitting. One thing you can certainly do with it is make a state of the art preamp. Have a look at the "26 pre amp" thread and you'll find all the information there. You could then hook that up to a 2-stage amp of some kind. Maybe a nice SE one.
Note that the PP version is gapped for approx. 5mA. So with 10mA SE input would make sense, but so would SE output.
Note that the PP version is gapped for approx. 5mA. So with 10mA SE input would make sense, but so would SE output.
Thanks for all the info. I am still looking on the '26 pre amp' thread to find out where is LL1660. I am guessing that interstage transformers are not quite popular. My thought was that using a IT saves 2 coupling capacitors and therefore improves the signal path. But I guess the IT has its own complications of linearity and balance. Now I have them and want to get most out of them.
I spoke with Kevin Carter of K&K Audio and he recommended the connection Alt V for a SE to PP.
I agree with the remark "not to reinvent the wheel", but I am open also for perfectionism. It is who I am.
I spoke with Kevin Carter of K&K Audio and he recommended the connection Alt V for a SE to PP.
I agree with the remark "not to reinvent the wheel", but I am open also for perfectionism. It is who I am.
Hi abszero,
I have built a pp using the 1660 as phase splitter with excellent result, I've talked to kevin from k&k about my project and he have great knowledge on the lundahl products.
My is, just build it and have fun.
http://www.diyaudio.com/forums/tubes-valves/207770-pp-kt88-phase-splitter.html
I have built a pp using the 1660 as phase splitter with excellent result, I've talked to kevin from k&k about my project and he have great knowledge on the lundahl products.
My is, just build it and have fun.
http://www.diyaudio.com/forums/tubes-valves/207770-pp-kt88-phase-splitter.html
Thank you Tim. I contemplated myself Andrea's design and felt atracted by the simplicity. The less components the better... It s encourtaging to see that someone (exept Andrea) built it, and with good results. Th two monoblocks look great and I am sure sound sweet. What connection did you use for Ll1660, and what type of 1660?
I want to use fixed bias is that's going to make a difference. I have plenty of 6P3S-E's for finals. I heard good things about their sound. I am thinking to try Alt V connection driven by 6P45. I will look into trying other tubes that have the appropriate Ri and can optimally run a 18mA.( ll1660 10mA Alt V).
I want to use fixed bias is that's going to make a difference. I have plenty of 6P3S-E's for finals. I heard good things about their sound. I am thinking to try Alt V connection driven by 6P45. I will look into trying other tubes that have the appropriate Ri and can optimally run a 18mA.( ll1660 10mA Alt V).
Hi!
As Andy wrote: The LL1660 dooes not have a great symmetry between phases if used as splitter interstage. That's why Lundahl has the LL1660S which is better at that. The LL1660 will work though. It is excellent as line out transformer wired 4,5:1. Hard to beat in this configuration and would work as line out with the 26.
BTW it is a common misbelief that there is no capacitor in the signal path when using an interstage. The last cap in the power supply will be even more important with transformer coupling than with other coupling methods. It is fully in the signal path.
I use interstage coupling a lot. It sounds great when implemented well and when you use good caps in the power supply.
Best regards
Thomas
As Andy wrote: The LL1660 dooes not have a great symmetry between phases if used as splitter interstage. That's why Lundahl has the LL1660S which is better at that. The LL1660 will work though. It is excellent as line out transformer wired 4,5:1. Hard to beat in this configuration and would work as line out with the 26.
BTW it is a common misbelief that there is no capacitor in the signal path when using an interstage. The last cap in the power supply will be even more important with transformer coupling than with other coupling methods. It is fully in the signal path.
I use interstage coupling a lot. It sounds great when implemented well and when you use good caps in the power supply.
Best regards
Thomas
Vinylsavor,
Thanks for your input. I totally agree about the last cap importance in the power supply.
There will be always a compromise to make. Maybe you're right that I should redirect the use of LL1660 toward places where it does the best job. Do you recommend a different IT as a better phase splitter? I still want to try LL1660 Alt V and maybe compare with other better IT. I am in the phase of finalising the design and gathering some parts. I'll post my results later.
Again, I am looking for the most simplest design and great sound.
Thanks for your input. I totally agree about the last cap importance in the power supply.
There will be always a compromise to make. Maybe you're right that I should redirect the use of LL1660 toward places where it does the best job. Do you recommend a different IT as a better phase splitter? I still want to try LL1660 Alt V and maybe compare with other better IT. I am in the phase of finalising the design and gathering some parts. I'll post my results later.
Again, I am looking for the most simplest design and great sound.
Hi!
It is certainly ok to start with this transformer then test how it performs in your circuit. You could then replace it with the LL1660S. There is plenty of choices for interstage transformers these days. Depends on your requirements and circuit which one might fit best.
Thomas
It is certainly ok to start with this transformer then test how it performs in your circuit. You could then replace it with the LL1660S. There is plenty of choices for interstage transformers these days. Depends on your requirements and circuit which one might fit best.
Thomas
Hi,
I've used the 1660S as SE to PP IT, the amplifier used PP 2A3/300B driven by a single 71A.
The 1660S was configured 1:0.9+0.9 and the air gap in that configuration was 28mA using a custom 1660S/14mA, the 71A ran at 20mA.
Drive impedance was 2k of the driver plus the impedance of the shunt regulator tubes, which were compensated for with RC bypass to ensure the drive impedance was fairly even at 3k5 - exactly what the data sheet suggests, the secondaries were loaded only by the tubes, (no shunt resistors), in cathode bias.
It sounded ok but there was an anomily with the primary inductance being less than required to get the rated bandwidth from the IT using the rated drive impedance. I'd confirmed this with a friend who was using the same IT in the same design.
The LL distibutor suggested to use a lower Rp driver to make use of the limited inductance, and suggested that any 'ringing' as described in the data sheet as a result from using a lower drive impedance would be sonically benign.
I believe they have some other units that offer higher inductance and/or are specified for use with lower Rp tubes but in the end I had a custom IT made by Onetics in the USA, for the same application and it was clearly better in way.
I would suggest that an IT used specifically SE to PP in such a simple design will probably be the overall limiting factor in the amplifier, all other design parameters being sensible.
Shane
I've used the 1660S as SE to PP IT, the amplifier used PP 2A3/300B driven by a single 71A.
The 1660S was configured 1:0.9+0.9 and the air gap in that configuration was 28mA using a custom 1660S/14mA, the 71A ran at 20mA.
Drive impedance was 2k of the driver plus the impedance of the shunt regulator tubes, which were compensated for with RC bypass to ensure the drive impedance was fairly even at 3k5 - exactly what the data sheet suggests, the secondaries were loaded only by the tubes, (no shunt resistors), in cathode bias.
It sounded ok but there was an anomily with the primary inductance being less than required to get the rated bandwidth from the IT using the rated drive impedance. I'd confirmed this with a friend who was using the same IT in the same design.
The LL distibutor suggested to use a lower Rp driver to make use of the limited inductance, and suggested that any 'ringing' as described in the data sheet as a result from using a lower drive impedance would be sonically benign.
I believe they have some other units that offer higher inductance and/or are specified for use with lower Rp tubes but in the end I had a custom IT made by Onetics in the USA, for the same application and it was clearly better in way.
I would suggest that an IT used specifically SE to PP in such a simple design will probably be the overall limiting factor in the amplifier, all other design parameters being sensible.
Shane
Hi Shane,
since the LL1660 had about 10hy primary inductance in your configuration, it is no surprise that it didn't reach a good bandwidth. But that cannot be called an anomaly. This is what can be derived from the datasheet.
It would have been better to use it for example 4.5:2+2 in your case and accept less step up.
Best regards
Thomas
since the LL1660 had about 10hy primary inductance in your configuration, it is no surprise that it didn't reach a good bandwidth. But that cannot be called an anomaly. This is what can be derived from the datasheet.
It would have been better to use it for example 4.5:2+2 in your case and accept less step up.
Best regards
Thomas
Last edited:
Yes, as Thomas says. And if you use a 4P1L in triode instead of 71A you'll get much more gain to compensate, and in my opinion better sound as well. And if you use filament bias, much better sound.
Hi Thomas, Andy
ALT V from the 1660 data sheet says with a drive impedance of 3k5 ohms, bandwidth can be expected to be -1dB at 25Hz, primary inductance around 40 henries. I'm not sure where 10H comes from, but it sure measured a bit that way 🙂
The 2 stage drive cct still is LL7903 1:4, 01A direct coupled to 71A both filament biased with SLA battery/series choke supplies and whilst I hear you about stepping down and using a higher mu driver, at the same time theres something satisfying to me about playing music through these older antique tubes, and 250mA filament current has its place too.
Regards,
Shane
ALT V from the 1660 data sheet says with a drive impedance of 3k5 ohms, bandwidth can be expected to be -1dB at 25Hz, primary inductance around 40 henries. I'm not sure where 10H comes from, but it sure measured a bit that way 🙂
The 2 stage drive cct still is LL7903 1:4, 01A direct coupled to 71A both filament biased with SLA battery/series choke supplies and whilst I hear you about stepping down and using a higher mu driver, at the same time theres something satisfying to me about playing music through these older antique tubes, and 250mA filament current has its place too.
Regards,
Shane
Hi Shane,
the datasheet Alt V is given as an example for a transformer gapped for 10mA (orimaries in series). If you have a transformer which is gapped for less current (yours had 28mA?) the inductance reduces accordingly.
Although I like Lundahl transformers a lot, the LL1660 is so so as interstage. It is brilliant as 4.5:1 line output
Best regards
Thomas
the datasheet Alt V is given as an example for a transformer gapped for 10mA (orimaries in series). If you have a transformer which is gapped for less current (yours had 28mA?) the inductance reduces accordingly.
Although I like Lundahl transformers a lot, the LL1660 is so so as interstage. It is brilliant as 4.5:1 line output
Best regards
Thomas
Shane,
Thank you so much for your detailed response. Indeed your 71A and LL1660S were a mismatch. If I remember correct the driver Rp must be 2-3 times lower than the primary Z of the IT. For your configuration that leads to a driver with Rp between 1166 and 1750 ohm.
In my design I want to use LL1660 10mA in configuration SE to PP 2.25:2+2 ,0.9 Tesla @ 18 mA , 3.5 k primary (recommended by Kevin of K&K). Driver tube could be 6p45 with an Rp of 1.7 k.
I plan to start this project in January and will post the results.
Radu
Thank you so much for your detailed response. Indeed your 71A and LL1660S were a mismatch. If I remember correct the driver Rp must be 2-3 times lower than the primary Z of the IT. For your configuration that leads to a driver with Rp between 1166 and 1750 ohm.
In my design I want to use LL1660 10mA in configuration SE to PP 2.25:2+2 ,0.9 Tesla @ 18 mA , 3.5 k primary (recommended by Kevin of K&K). Driver tube could be 6p45 with an Rp of 1.7 k.
I plan to start this project in January and will post the results.
Radu
Hi Radu,
No mismatch according to the data sheet, take a look.
ALT V
2.25:2+2 (=1:0.9+0.9)
0.9T with Ip = ~1.9x Ip on the label
42H for a -1dB of 25Hz assuming 3k5 ohm drive impedance.
Note that LL do not specify a primary impedance for these transformers. That 3k5 is the maximum recommended drive impedance for full bandwidth. Its not a 3k5 load, per se.
What I am saying is that I was using 71A (Rp 1.8k) and a shunt regulator to set the drive impedance at 3.5k ohms, for the entire passband, and within the current ratings of the air gap, so the situation met the criteria completely.
If I were you, I'd be measuring BW in circuit (your drive impedance could well be different from the Rp of the tube by itself, and if you're using that 3k5 figure as a maximum to still get -1dB at 25Hz, well if your samples are like mine and my friends, in short you wont).
Hi Thomas,
It was a custom version gapped for 14mA, which in ALT V is good for 26mA. Just as the 10mA version is good for 18mA in ALT V. I was running 18-22mA, depending with consistant LF measurements.
I agree with your thoughts about so so for an IT. I used if for some years this way as the LF wasnt an issue in the application. It didnt take a lot of money to get a much much better result, which brings me back to what I said about me thinking the IT is crucial for overall performance.
*edit - Think I might have just registered what you're saying, Thomas. .. any idea what impact moving from a 10mA to 14mA custom type should have on primary inductance, all else being equal?.. regardless, the original poster should disregard my comments, I was not using a 10mA version.
Happy Heinekens,
Shane
Last edited:
Hi Shane,
inductance scales about linearily with the air gap. So you have about two thirds compared to the 10mA version. These are not custom models. Lundahl offers most of his transformers with airgaps selectable over a certain range and in certain steps.
I never found the 1660 to work well in step up configurations. In your case I would have preferred for example a 18mA version wired 4.5 : 2+2 or the other way around 4 : 2.25+2.25
Best regards
Thomas
inductance scales about linearily with the air gap. So you have about two thirds compared to the 10mA version. These are not custom models. Lundahl offers most of his transformers with airgaps selectable over a certain range and in certain steps.
I never found the 1660 to work well in step up configurations. In your case I would have preferred for example a 18mA version wired 4.5 : 2+2 or the other way around 4 : 2.25+2.25
Best regards
Thomas
Hi Shane, Hi Thomas,
Agree with your assessment. When you said "it didn't take a lot of money to get a much better results" what were you referring at? What type of IT would you recommend? I will try the LL1660 as IT, but I guess that will be a so so sound result. Can't wait to try.
I would like to try also what Thomas suggested 4.5:1 line output. Can you refer to a thread that approaches this configuration? Can you suggest better choices for an IT SE to PP, or I should just try a transformer less phase splitter?
Agree with your assessment. When you said "it didn't take a lot of money to get a much better results" what were you referring at? What type of IT would you recommend? I will try the LL1660 as IT, but I guess that will be a so so sound result. Can't wait to try.
I would like to try also what Thomas suggested 4.5:1 line output. Can you refer to a thread that approaches this configuration? Can you suggest better choices for an IT SE to PP, or I should just try a transformer less phase splitter?
Hi Abszero,
I had them made by Onetics in the USA on an M3 core, PN# 10386, they have an email address, a quick search should drag it up.
Hi Thomas,
Custom gap, ok. 2/3 of 42H is 28H, which is about what most 5k primary transformers seem to offer. I wonder at what frequency and voltage drive level this is calculated at. For example, the replacement units offer 29H at 20Hz with 108vAC drive and 22mA standing current, which falls to 23H at 20Hz with 20vAC drive.
I still have the the 1660s.. Figure gap for 12mA with 4.5:1 and ~80H.. what would you use them with, indirectly heated medium to high mu tubes with low Rp? 5687, 6922, 6C45, triode D3A etc?
Regards,
Shane
Hi Shane,
what were the probelm you had with this L1660 when you used it with the 71A? How far as the bandwidth off ?
As for usage with other tubes, this depends on what you want to achieve. It is difficult to recommend a tube without knowin what the requirements are. hat kind of gan is needed, what is the max signal they will see at the input and so on. Is this for phono, line or any other application?
Thomas
what were the probelm you had with this L1660 when you used it with the 71A? How far as the bandwidth off ?
As for usage with other tubes, this depends on what you want to achieve. It is difficult to recommend a tube without knowin what the requirements are. hat kind of gan is needed, what is the max signal they will see at the input and so on. Is this for phono, line or any other application?
Thomas
- Status
- Not open for further replies.