Hi Ale,
Thanks for sharing all those results. Excellent work!
Have you ever considered trying a TVC at the output of your transformer coupled line? In my experience this will bring the sound to a whole new level.
Since gain is normally not really needed in a linestage, for me the whole point for a line preamp is to provide the capability to drive a low impedance volume control
Best regards
Thomas
Thanks for sharing all those results. Excellent work!
Have you ever considered trying a TVC at the output of your transformer coupled line? In my experience this will bring the sound to a whole new level.
Since gain is normally not really needed in a linestage, for me the whole point for a line preamp is to provide the capability to drive a low impedance volume control
Best regards
Thomas
Went back to revisit the 46 triode-strapped. This is a great valve indeed. Running filaments at 1.266A, anode current at 18mA/Va=139V. THD is 0.08%. Half of my previous test at an expense of running the valve at 80% more current.
It sounds superb. Comparing to 4P1L, the 46 is not microphonic but less linear. Both sounds amazing: hard to make a difference in my workbench with the old solid state technics amp I have. Not the ideal testing scenario, but it is what it is...
It sounds superb. Comparing to 4P1L, the 46 is not microphonic but less linear. Both sounds amazing: hard to make a difference in my workbench with the old solid state technics amp I have. Not the ideal testing scenario, but it is what it is...
Attachments
Thanks Thomas. I realised that testing some of all I have learned so far is the best way of improving
Yes, I have considered TVC. This DIY hobby is dragging me down to the financial bankruptcy. TVC are expensive though, any one you would recommend for a tighter budget?
BTW: the aluminum rails you sold me are fantastic. You can build up quickly any setup and is like playing with meccano
Hi Ale,
I'm using the AVCs from Dave Slagle of intactaudio. They are not that expensive. Especially not if you consider the sound improvement.
I can send you a pair for trial, you only need to cover the shipping expense. After your trial you can report here and send them back to me. Just be a bit careful with soldering.
Best regards
Thomas
I'm using the AVCs from Dave Slagle of intactaudio. They are not that expensive. Especially not if you consider the sound improvement.
I can send you a pair for trial, you only need to cover the shipping expense. After your trial you can report here and send them back to me. Just be a bit careful with soldering.
Best regards
Thomas
That's true but keep in mind that this decreases the lifetime of the tubes
I second this, Thomas is on the highway here. For experienced listeners who like to reach the Mount Everest I recommend to expand the line stage as 2 stage design, 1st stage as voltage amplifier and the 2nd stage as impedance converter (lowering the output impedance compared to Thomas' amp).
Hi!
Definitely not with thoriated tungsten filament tubes! I'm not sure about oxide coated DHTs. It decreases liefetime of indirectly heated tubes. But we're talking directly heated only. It does decrease transdonductance and increases plate resistance. The load of the tube should be sized accordingly. I only recommend this for preamps. Certainly not in the power stage of amps.
Most systems have way too much gain, so I don't see the need for a second stage in a line preamp. Most sources have an output voltage of 2Veff. Some even more. Power amp input sensitivities are typically 0.5 to 1V for full power. A linestage with gain of 3dB is sufficient in most cases giving some reserve for source material which does not use the full level range. Of course if the whole system requires more gain in a linestage a 2stage approach should be considered.
Best regards
Thomas
Definitely not with thoriated tungsten filament tubes! I'm not sure about oxide coated DHTs. It decreases liefetime of indirectly heated tubes. But we're talking directly heated only. It does decrease transdonductance and increases plate resistance. The load of the tube should be sized accordingly. I only recommend this for preamps. Certainly not in the power stage of amps.
Most systems have way too much gain, so I don't see the need for a second stage in a line preamp. Most sources have an output voltage of 2Veff. Some even more. Power amp input sensitivities are typically 0.5 to 1V for full power. A linestage with gain of 3dB is sufficient in most cases giving some reserve for source material which does not use the full level range. Of course if the whole system requires more gain in a linestage a 2stage approach should be considered.
Best regards
Thomas
Sounds like a great plan. I will send you a PM. Would like to know its cost. thanks
Re-assembled the preamp to fit the 4P1Ls suspended with a couple of rubber bands . This works great, you have to tap the valves significantly now to get an audible ringing. No feedback perceived to me due to microphonics.
The minimum filament current to run these properly is 500mA. Below this, distortion increases, etc.
Got them now running at 29mA and 105V. At this quiescent point the THD is 0.027% - really low. The HT supply is set to 200V which is what I have on my 26 preamp supply. So could use same supply now.
Need to get the LT supply done now....
Sounds really nice
The minimum filament current to run these properly is 500mA. Below this, distortion increases, etc.
Got them now running at 29mA and 105V. At this quiescent point the THD is 0.027% - really low. The HT supply is set to 200V which is what I have on my 26 preamp supply. So could use same supply now.
Need to get the LT supply done now....
Sounds really nice
Attachments
Mogliaa, what load are you using for your THD measurements? I am trying to gauge/baseline wheat moving from a 5842-SE OPT headamp to the 4P1l would give me. With the 5842 I get around 0.125% THD after the transformer which is the triode swinging around 17Vrms across the primary (with practicaly no load just the soundcard input impedance.).
My problem is IMD, with this tube it is 4x the THD (and its not from noise), I am right on the edge of not enough clean power for my headphones. My goal is sort of the "first milliwatt" concept, but instead the first centiwatt is what I need for these phones, I can get more like ~2-3 milliwatts before the IMD takes its toll.
Have you taken any IMD measurements of your DHT's, does 4x THD sound normal ?
Thanks for the rubber band photo's. you may want to try some of the small parachord that the M4 rifle enthusiasts love to wrap on everything.
My problem is IMD, with this tube it is 4x the THD (and its not from noise), I am right on the edge of not enough clean power for my headphones. My goal is sort of the "first milliwatt" concept, but instead the first centiwatt is what I need for these phones, I can get more like ~2-3 milliwatts before the IMD takes its toll.
Have you taken any IMD measurements of your DHT's, does 4x THD sound normal ?
Thanks for the rubber band photo's. you may want to try some of the small parachord that the M4 rifle enthusiasts love to wrap on everything.
Last edited:
Hi Regal,
I measure the THD with the amplifier connected so I can see the impact in the THD of the stage due to the AC load. You can notice the difference with triodes with a higher anode resistance (e.g. CX112, 01A). I'm using Pmillet's interface to measure with my PC using audiotester. Haven't measured IMD and am only looking at H2 and H3 as have the PC's output video signal filtering in at a very low level around 30kHz.
0.125% is pretty good have to say. I would suggest playing around with the quiescent point to see if you can reduce the THD? That's what I did in all my tests so far.
I could easily get the 4P1L THD to 0.20% if biased in other operating point, so need to check this.
Re the small paracord of the M4 rifles, no idea which ones you;re referring to. Do you know where can I get it?
Cheers,
Ale
I measure the THD with the amplifier connected so I can see the impact in the THD of the stage due to the AC load. You can notice the difference with triodes with a higher anode resistance (e.g. CX112, 01A). I'm using Pmillet's interface to measure with my PC using audiotester. Haven't measured IMD and am only looking at H2 and H3 as have the PC's output video signal filtering in at a very low level around 30kHz.
0.125% is pretty good have to say. I would suggest playing around with the quiescent point to see if you can reduce the THD? That's what I did in all my tests so far.
I could easily get the 4P1L THD to 0.20% if biased in other operating point, so need to check this.
Re the small paracord of the M4 rifles, no idea which ones you;re referring to. Do you know where can I get it?
Cheers,
Ale
I think my problem is not so much THD of the tube itself but a sharp SS like jump in distortion when I go from moderlate/low listening levels to "rocking-out" levels, I'm somewhat limited in moving the OP point with my OPT. Probably just need to jump full speed ahead into the DHT's, have 1 pair of coleman regs, 4x4P1L just waiting on another pair of colmans kit to get here from the UK, course I need a better set of OPT's.
I recommend "real" military grade 550 parachute cord, with white strands inside a colored outer sheath, rather than the braided stuff you get at WalMart (I'm in the USA.) I don't know of a source in Europe but will ask around. I am afraid the rubber bands won't last long with the heat.
edit here is the part number at least:
http://www.bestglide.com/550_cord_info.html
Last edited:
Please provide me more info why thoriated filaments don’t decrease the lifetime when they are heated with lower voltage. I know that we are talking about direct heated tubes
I only agree of input sensitivity and need of gain in the line stage. But I don’t agree why only using 1 stage. If you add a 2nd stage you can take the tube’s amplification factor as transformation ratio and will get a much lower output impedance to drive the load.
Anyway Thomas, I respect your work but we think here different - there are many ways to Rome
Hi!
I got this information from talking to S. Bench who did some research on lower filament voltages. I don't have a specific source. But in addition to that actual experience from using thoriated tungsten DHTs like that over many years.
I already have output impedances in my linestages of less than 200 Ohms. Some even only 50 Ohms. Lowering much more would not bring any significant advantages IMHO.
But as you say: many ways to do things. I prefer as few stages as possible
Best regards
Thomas
I got this information from talking to S. Bench who did some research on lower filament voltages. I don't have a specific source. But in addition to that actual experience from using thoriated tungsten DHTs like that over many years.
I already have output impedances in my linestages of less than 200 Ohms. Some even only 50 Ohms. Lowering much more would not bring any significant advantages IMHO.
But as you say: many ways to do things. I prefer as few stages as possible
Best regards
Thomas
With those output impedances you are be using an output transformer with a high ratio 15k:600 ?
So thats a voltage ratio of ~ 2/3. Line level is ~0.7vrms. A #26 has a gain of 8. So about 3.5Vout rms. So to match a standard DAC output of 2V, the #26 is only swinging a couple volts and the reflected load with a worse case 10k input amplifier is only putting a 250k load on the #26. For the 4P1l things are even better. Can't see the need for two stages here.
The only question I have is it is even possible to get 50 ohm output impedance with either of these tubes because the 600 ohm secondary DCR is probably at least 50 ohms on its own in order to have enough primary inductance to deal with the high Rp of the #26?
Also do you have issues with the transformer mechanically ringing the notes with the secondary so lightly loaded? I've seen this when running RMMA and you hearthe damn transformer running thru the notes 5 feet away with the output hooked to the soundcard (no speakers hooked up.) This really couples with the tubes and causes all sorts of interactions.
Summary for those who hate long posts.
Those are just two things I've encountered with "preamp" output transformers:
1. high secondary dcr in order to have proper primary inductance
2. the OPT turns into a speaker with the lightly loaded secondary.
So thats a voltage ratio of ~ 2/3. Line level is ~0.7vrms. A #26 has a gain of 8. So about 3.5Vout rms. So to match a standard DAC output of 2V, the #26 is only swinging a couple volts and the reflected load with a worse case 10k input amplifier is only putting a 250k load on the #26. For the 4P1l things are even better. Can't see the need for two stages here.
The only question I have is it is even possible to get 50 ohm output impedance with either of these tubes because the 600 ohm secondary DCR is probably at least 50 ohms on its own in order to have enough primary inductance to deal with the high Rp of the #26?
Also do you have issues with the transformer mechanically ringing the notes with the secondary so lightly loaded? I've seen this when running RMMA and you hearthe damn transformer running thru the notes 5 feet away with the output hooked to the soundcard (no speakers hooked up
.) This really couples with the tubes and causes all sorts of interactions.Summary for those who hate long posts.
Those are just two things I've encountered with "preamp" output transformers:
1. high secondary dcr in order to have proper primary inductance
2. the OPT turns into a speaker with the lightly loaded secondary.
Hi!
Secondary DCR has nothing to do with primary inductance.
I use for example the 801A into 4.6:1 step down which yields about 250 Ohm Zout. Add to that the 70Ohms DCR results in 300 Ohms. The 200 Ohm number above was a bit low. The transformer used is the Lundahl 1660
If I want lower Zout, I use for example the Lundahl 1689 in 9:1. This one has much lower secondary DCR.
My line out transformers are loaded with a autoformer volume control.
No issue with mechanical ringing at the typical line levels
Best regards
Thomas
Secondary DCR has nothing to do with primary inductance.
I use for example the 801A into 4.6:1 step down which yields about 250 Ohm Zout. Add to that the 70Ohms DCR results in 300 Ohms. The 200 Ohm number above was a bit low. The transformer used is the Lundahl 1660
If I want lower Zout, I use for example the Lundahl 1689 in 9:1. This one has much lower secondary DCR.
My line out transformers are loaded with a autoformer volume control.
No issue with mechanical ringing at the typical line levels
Best regards
Thomas
Vinyl, you need to study up on this (no offense intended). More turns on the primary equals more inductance (given same core/topology) to keep the ratio that means more turns on the secondary which means more secondary DCR, the two are related very much so. In other words one can wind a 15000:00 OPT with 5H inductance and a nice low secondary but when we want a reasonable bass response with a givien Rp we need more turns on both sides to keep the ratio and the result is high secondary DCR.
Ah auto former, great solution and you have avoided the issue, no harm done and this is of course just discussion.
I love the Autoformer concept but haven't found one that that can handle more than a 10Vrms across the coil, have to be out there
Last edited:
Hi!
I don't think I have to study on this topic, simply because I'm relying on experts winding my transformers. Still the secondary DCR is not necessarily related to the primary inductance. While your statements are true, there are other parameters which influence the primary inductance which is core material, air gap etc. Even if all those are given you can still choose a thicker wire for the secondary which will yield different DCR. Which Lundahl apparently does if you compare the DCRs of various transformer types
Thomas
I don't think I have to study on this topic, simply because I'm relying on experts winding my transformers. Still the secondary DCR is not necessarily related to the primary inductance. While your statements are true, there are other parameters which influence the primary inductance which is core material, air gap etc. Even if all those are given you can still choose a thicker wire for the secondary which will yield different DCR. Which Lundahl apparently does if you compare the DCRs of various transformer types
Thomas
Yes but its always a good idea to take a few frequency response measurements, and measure the dcr, maybe even measure the true output impedance. I've been burnt before (not Lundahl) but I always aproach preamp and headphone amp OPT's with caution, sometimes the experts aren't so much as they think.
I do need to look at Lundahl, did they make your Autoformers ?