Romy the Cat wrote:

BTW, I have a local Bostonian guy have replied brought today a very interesting DHT amp. He has a number of DHTs and this one he reports is the special one, quite expensive as well. Although the electricity was not good today but we did some listening and the results were very quite interesting. I am still contemplating about them and will post what I think what I have more organized thoughts.

floobydust wrote:

One of the reasons I no longer use a tube tester. They make a measurement at a specific target operating point(s). You don't really know what it is, so what are the chances they will be close to your actual circuit operating point? So choosing a close match of tubes that perform the same across the noted parameters (in the actual amplifier(s)) just makes more sense to me. If the pair of tubes are the same manufacturer and internal construction (and unused), then they should age well together and remain fairly matched in long-term operation.  Just for fun, I took the final output stage for the DHT amp and normalized it (i.e., subtracted the DC shift for direct-coupling) and modeled it in the SEampCAD program. I plugged in the details for the output iron as noted. Normalized values result in 1.125K cathode resistor and a 280V plate supply. Values work out to pretty much what you show in measured values.  Transfer curve is quite good considering the lower current. Maxmum power output is about 1.85 watts with pretty low distortion. I can probably create some PDFs for the output specs and curves and email them to you.  Based on the values you are currently using, you can put a 45 triode in the circuit without any change and it will be just as clean but less power... about 1.2 watts. Should be an interesting experience. In many respects the 45 is the purer tube design and many prefer it soncically over the 2A3, "IF" you can deal with 2-3dB less output power.

First, thanks that taking your time for plotting my amp. I think it good and I have no problem with output stage. If would pick fleas on its design then it would be the first stage. It run at juts 10mA and 150V – a VERY conservative operation for this driver. I can not driver it hooter as I would then need a higher voltage on the output tube’s anode. It looks like the driver tube still sound good but I have another concern. At this operation I have only 2.5V on driver grid. At my normal 200V and 20mA I have 4V on grid. The 4V on grid is good as I can swing any input voltage and have an assurance that I will not send the driver into A2. With 2.5V I still far from danger but it is closer then with 4V. So, the 3dB voltage divider at input is kind of extra protective measure to keep the drive state from overload.

Why I said about the déjà vu is because a coupled weeks back I did think about the 45 tubes. I even bought a few for test but they arrived damaged and I sent them back. I need to but more of them and finally try those 45s. You said that might have 2-3dB less output power. I usually do not understand power in decibels I understand gain in decibels, and as I understand I will have both of them less…

Anyhow, my concern about the 45 tube is the same that I have with my YO186/RE604 tubes. All my 4V tubes sound softer then my 2.5V tubes. The 45 tubes with their plate impedance 500R higher then YO186/RE604 I will have even harder loading of those tube against my 6K output transformer. Now, if I find a way to make primary with taps of 6K and 9K then I will be able to drive 2.5V tubes against 6K and 4V against 9K. Since I do not use them at the same time I will be able to keep them permanently connected. Now, the whole beauty of my current arrangement is that I need to do absolutely nothing to switch the amp from YO186 to 2A3. If I stick into the 2.5V socket a tune with 1.7K plate impedance then I need to remaps it to high impedance primary… So, I need to come up with some kind of arrangement that have flexible primaries, also to switch them and do not degrade sound.

Anyhow, I will order some 45s and will see how it goes…

Very cool results I’m get in here. I took the new Slagle’s transformer that has reconfigurable sections and put it in the Milq.  With 13V on primary I have 625mV or 520mV on secondary, or 20:1 and 25:1 (with 200-300mV normally drive the speakers at cruse volume). The lost of gain is manageable and I have enough in my LPAD to compensate it. In fact I think I will have enough gain even for 45s to try….

It is unquestionable that my 4V tube need more then 20:1 and with 25:1 they begin to sing, particularly the Klangfilm KL-71403 and YO186. They got some transient thump and sound just wonderful. What particularly is very cool is how genital they enter high dynamic – they do it with dignity and with amassing grace.  I would not say that my 2A3 are losing in something, I am not at the state what I would make those observation. As I told with DHT MF in band path I can make any tube to sound as anything but as it is calibrated now the 4-volter do very serious job in MF.

Now, about the mistake. I did not think about it before I have done it but not I see that it would be very good idea to have an automated compensation for gain lost when I flip a switch from20:1 and 25:1. It would not be default to do but as not it is gone, so I would need to find a way to fine-calibrate my LPAD, probably getting one of those old knobs from old scoops….

There was a very funny moment. I was listening one MF channel just with YO186 and restarted the amp. The Helsinki Philharmonic was crashing through the reticules Samuel Feinberg's Piano Concerto (if it were LP then I would swear that it was played backward) and then suddenly someone knocked in my door.  I went to open but no one was there. To my surprise it was how the YO186 played the bass drum. There is some sense of “aluminum insanity” in the way how YO186 stress the bottom of S2. It sounds “distorted” but so … lucrative and attractive! I would LOVE to hear a full-ranger with YO186 with a “right” driver. It would be nice to know what the “right” driver might be. Perhaps I need to convert my Sun Audio crap into full-ranger YO186 and for an assault trip?  I wish I have ego to it… but it is pity as it might be a very interesting full-range DHT SET.

Funny for in my DSET case it is mostly about a primitive equalization in MF. Well, almost.

I god a few other versions of 2A3, one more 604 and a few 45. I in fact like the 45s, even though with my attenuator wide open I have ~2dB less gain then I would like when I use 45s and 25:1 loading. So, I need to research if I can remove the divider from the grid of the driver stage (where I burn 3db). I can remove it but I need to learn it with max signal I would not drive the 5E6P into grid current.

Then I kind of know what kind tube I would like to use in output stage. I would like to use the soft sounding tube but I would like to drive it very contrasty (higher loading ratio, fast core, perhaps higher current to drive impedance lower). What I would like to accomplish sonically is following. I like generally how DH Sound and some of the DHT has VERY interesting HF. They are not strong and in way rolled off (here is where most of the people feel the vintage sound) but they are also very lucrative and flavorsome. The character of the upper range of some DHT is intriguing and I can only describe it with effect of a butterfly landing on you shoulder. It is there but it is also almost “not there” or at least it is not defined in the terms of HF event. So, what I would like to have is a higher amplitude of that butterfly effect of I need to learn how to drive “some” soft-sounding 45, 2A3, 604 and 186 harder to accent their HF region.

I found an interesting flow in the sound of my current OPTS and while it will be certified I am experimenting with park of amorphous core transformers that left over from my former projects. I am trying to remap my Tribute but it is just 40mA. It might be enough for RE604 and YO186 as the run at 35mA but it is not good for 2A4 as some of them go up to 40mA. I have 50mA Lundahl 1623A that I used for my failed headphone project and how use for my fundamental channel. The leading battle is a very interesting amorphous transformer. It is 5-6H, make custom for me with limited turns, specifically foe my DSET configuration. It uses a large core and can handle even 100mA. I kind of slowly am experimenting with them.

At this point I very much know what I need I need – I heed the HF extensions, not the response itself – all of them are good for couple hundreds of kHz according to measurements but I am looking for the subjective auditable HF extensions while the transformers are running the “soft” DH tubes. I almost looking for a “bright” output transformer, sure if I found one then I will be bitching that it corrupted my DH’s “butterfly effect” (read above). Still among the few transformers that I have in THIS specific configuration I like more those that give me more HF from my DH tubes. Can I blame myself?

The key I see in this that with more “punchy “extension of HF from my MF channel I can implement the Macondo’ drivers integration model that I would like to. Otherwise it will be not the best case in my application. I always have my full range Milq (injection channels) with amorphous 6C33C that give me an idea how fur the S2 driver might be taken…

I got few inspirations recently that made me to dedicate today for building another DH MF for second amp. It is end of the day and the conversion of the second amp is done, including the PS, marking and drilling the holes, driver stage and the DH stage – everything. In fact the amp was working perfectly fine right after assembly and each voltage and current was right spot on.

I will keep it for another day or two open and I am experimenting with removing the voltage divider from input filter and then the saga with Melquiades 2009 modification will be over. The final decision about the output transformer I made is that the transformer I need 25:1. No taps, no re-mapable coils, no loading change. The strait coils 25:1 and it will be it. Perhaps there is a way to do it without loosing anything but I am not in business to write the audio theory and make more experiments. The 25:1 will be it. It might be a bit rough for 2A3 (remember, I load to 16R) but I can shunt my LPAD with extra résistance and it shall not have any negative effect, not like shunting the S2 driver…

The friends of mine who observe this project advised me a kinky techniques how to make my Nickel core transformer to sound much “expedited” then it is, so the VERY ideal 25:1 with additional “expediting” will be plenty to suck all HF out of my DHT.

Hm, I was playing with options to get rid the 3dB divider in the MF channel input and I realized that I was a moron somewhere. I was operating under presumption that I was getting 2V on the driver cathode in MF channel that made the swing voltage on the grid valuable to drive the input tube into gird current. This is where the idea of the input voltage divider comes from. I person who observe my project informed me a couple day back that it might not be 2V.

I just measured it on this second amp and it is 3.25V. That is VERY BIG difference as my signal at full volume does not go over 2.2V behind the filter and 20K resistor.  So, bye-bye the input divider on MF. I have no idea why O flipped with the cathode voltage measurement. Perhaps what I did it I had a current-dead driver. Not I have 332R resistor and 3.25V. That make on the tube 10.8mA, which is right where, is shall be. Now, I need to figure out where my out stage will be clipping with more voltage from driver…

The Cat

It is it, I eventual got what I was striving for a while and now my MF DHT channel is done and the most important is that it now sounds exactly how I need it to sound. The new 25:1 loading turned out to be the key. It is “fast” enough and I did not play with distribution of gap – it is good as is.

I call it a “few steps back” solution because it is how I used my S2 driver before I went for single stage. Before I used a half of 6C33C loaded to 1150R, which is very-very light for that tube. I know very well the sound of under-loaded S2 and it was pretty much how I used it since 2002. Then I went for damping S2 with plate, ending with 550R load and with 15:1 on the 6E5P plate. Now with DHT and 25:1 I somewhere near my 1100R on ½ of 6C33C and – boy this driver sings again.

With today configuration the “bright” and “bold” DHT tube are hardly useable – the single-plate Chinese tubes, some version of 2A3 and Valvo LK460 for instance are too brutal and too amusical. However the softer sounding 2A3, the 45 with plate of 1.7K and my 4V tubes are singing like canary – the very direct hit.

Actually my YO186 and Telefunken RE604 and Klangfilm KL71403 do the very best in the new configuration. The Telefunken and Klangfilm need one and a half dB less help from tweeter the YO186 needs 1.5 more but even with no tweeter they do amazingly pleasant and VERY VERY different. I matched a pair of YO186 and when I first time put them with both channels and the full system it was just an amassing gratification. It is not the go things and not the “my playback” complex – with the YO186 my playback do sound differently… I just went over a deep 2 hours listening meditation, thinking about this new sound, listing a few CDs of ….Jimmy Scott. I know, it is not my music but I always liked him and it is very interesting to see how the bottom region of YO186 handles the Jimmy Scott’s unique phrasing. The simplicity of Jimmy Scott is a great help in what I was interested to hear and boy – you need to see what YO186 did with Scott’s inimitable articulation!

Anyhow, it is fan in here and I am planning on this weekend my becoming ritual ceremony of trashing my Weller solder station. Since the 2002 it will be the third one….

Yes, the people in the beginning of the thread were right – the DHT two stage solutions do sound more interesting than one stage amp. Still, it does not suggest that if I have 12dB more gain on my single-stage triode I would not be able to load my single-stage to 9K-10K ten I would have worst result.  Well, where to get a low-voltage triode with gain of 60-80 and able to dissipate 5-8W on plate? The reason I still have doubts because loaded to identical 15:1 the DH and single-stage amp, having a virtually the same plate impedance sounded remarkably similar.

Still, we are not taking about the theory but the practical rendering of tangible solution and in practicality the two stages MF with DHT in output, idly relatively running against S2 truly turned out to be victorious over my former single-stager. For a last could day I am like a pig in shit, experimenting with different settings trying to put the benefit of the more prominent output from my MF into a full use.

The MF with my new amp is different and it needs to be used differently in Macondo. I did not figure out how yet and I would say that all together Macondo does not sound now as I would like it to. However, I do see some very tremendous potentials and I very much look forward to caponize on some advantages the I go from putting the DHT double-stages into the game.

Once again, thanks for all supporters and clique people who stayed along with this project.

I need to finish posting in this thread as the subject of this thread is practically over. Still, here are a few comments that I would like to make.

A few years back, driving the S2 with 6C33C I was pitching the harder loading, trying to damp the S2 resonances with 3c33c PLATE. The last month with my DHT I obviously went to opposite direction. Do not forget that the plate of my DHT does not see the S2 driver and loaded to 16R LPAD. I discard the changing out impedance of LPAD as the S2 has no exertion and therefore no returning current.

One of my major fears that I had is to over-idle the plate, making sound to barbarically-impressive. I do have some signs of it with 2A3 as my 25:1 is too low load for it. With RE604, 45 and YO186 that have higher plate impedance the 9.5K looks like absolutely perfect load (in my situation). The driver has the very right amount of “zippiness” and “color” where it needs to be “zippy” and “colorful” but it turns very disinterested and almost monotonously-flat when the music is called upon.  I made quite a few control listening and I am very pleased with result. I think if I take a few more turn from the transformer secondary then it will be too light load and it will “fracture” the sound.

Now is the most important thing. The high contact and high transient capacity of the new MF driver kind of cracked down the imaging. However, here is what the wonder of the very capable MF drivers come to surface: the imaging might be moderated by…. attenuation of MF level.  Get this: an attenuation for 2.5dB of MF driver is practically not auditable in term of volume but it useful in order to settle down the proper imaging. Here is where a very few smart readers who have ears and an experience in systems setup would say: Romy over-idled the MF out tube if the change of volume for 2.5dB is not auditable. Well, trust me – I am an academic in it: I did not over-idled the plate – it is the unique benefits of S2 driver by soft tube. For whatever reason the S2 now does not dive into “collision mode”, it go very fast and super transient but still holding the sound together with no artifact of  the typical under-loaded plate. I do not know what is responsible for this: the special quality of my DHT of the special character of the David Slagle’s Nickel transformer? I have no idea but I certainly very welcome the results…

I assure you that my experience has very little meaning to others as I do not look for “better tube” but look for a tube that would go the best in my specific DSET application and in very custom application of Macondo MF channel. Also, a local guy snatched my Sun Audio amps and I have no way to play my DHT tubes full range…

With all lower MF range magnificence of YO186 it a bit roll of at it’s HF and this is exactly what I would like to punch in my MF channel. The best German RE604-type and other 4V tubes are more balanced. Marconi are strange, Philips sounds like cotton, one Mullard and one PX4 that I got turned out to be very much used  and not as fresh as I would like them to be.

Here is what I got a tube that so far answered my play - it was type 45. I bought a few RAC and Tang-Sol used and one pair of brand new Philco from 30-40s. this 45 are very good, particularly the new Philco. They are more extended then my 2A3 and my 4V tube and they are more EXCITING I would say. Also, they are incredibly clean.  I will stay for a while with 45s and I do like what I am getting with them. Any, specific 45 that you might suggest me to try?

Thanks, KM.

I did read a few days again all that you told about the 45 before and have order a two pairs of new (promised) Sylvania from 30s and 40s. From what I tried the Sylvanias never wee bad.  I did not try 50sa or 245, perhaps I will. I am interesting is my specific application – an ability of the tube to push very strong into HF but very clean and very discriminatively. Some of my 2A3 very much punch but they do all HF in the same way. I do not know if it is a sign of coloration or absent of coloration but I like very much what this Philco 45 does – it very high, not thin, not lean and has still where it operate very high some integrity how it show favoritism to tones. It is very much reminds me how my single-state sounded but this time it is a bit more contrasty (thanks for loading). I do not want a tube to be stupidly extended (like Valvos for instance) but I would like it to be “stupidly able to be extended” if it called upon.

Interesting that from RE604 is very extended and superbly clean but the 45has some “Zippines” (positivly speaking) that ads some "excitement". Perhaps it is because my RE604 are used buy my 45 are new? The RE604 run at 36mA but 45s at 41mA. I do not know… Still, the 45 are not expensive and I have plenty of gain to put in there whoever can birth. Oh, if I would not be too lazy I would try my brand new 10Y… BTW, while listening the tubes I do EQ then by gain with .25dB precision, so the gain is not a factor at all that might the a case in a multi-channel installation…

Paul S wrote:

I know zip about using the 50, except it is another very "clear" tube that seems to have more "qualities" than the 45, and some folks attach +/-