|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
82.120.12.37
Hi OTLers , Diyers or "simple" users !
Did anybody listen to the both Futermans , the UNregulated PSU and compared to the REGulated one ? Is that worth the effort to work on a twin 140V rail SS system ? I 'm using 12 6080wa on stereo and could ad some more during this hard winter ;-) huh ??
cheerzzzzz
Pierre
Follow Ups:
I am not aware of any Futterman-type OTLs that used voltage regulation on the output tube supply. Since you mention you are using 6080 triodes in the output circuit, I wonder whether you are using one of the Fourier amps that used triodes, e.g., the Panthere or Triomphe (but the Panthere used 6C33Cs). Otherwise, all other Futterman-type designs that I know of used pentodes or tetrodes, e.g., the Fourier Sans Pareil, all NYAL, Prodigy, Counterpoint, KSS, and Futterman originals. Perhaps you can clarify. If you are in Europe, perhaps you own a pair of Graafs or Crofts, but the former amps use 6C33Cs, so I am still a bit puzzled. The Tenor amps use regulation for the input stages, but I don't think/don't know that they regulate the output supply, and to my knowledge the Tenors are not Futterman-types. Regulating the output supply would always be problematic unless you do it in a massive overkill way, because you want the output tubes to be free to draw massive amounts of current during musical transients.
Thanks for reply Lew! I'm using a DIYed OTL based upon a jap schematic with 12BY7A current source for a pair of 5687' drivers ...
adaptation with a 5814 // direct to a 6463 with a E180F current source . I'm building a pair of monoblocks (10 6080 each channel with 4700 uF each rail) with 2 5687 (drivers) and E182F (current source) and E188CC // (input). I'd like to regulate these input and driver stages with OA2 / OB2 gas reg tubes . For output stages , that's my question : do i go this way or not ?! I could // some hi current transitors as ballast but your answer make me think it would not be easy nor useful to lower the internal impedances of supply rails .
cheeerzz
Pierre
> I'd like to regulate these input and driver stages with OA2 / OB2 gas reg tubes . For output stages , that's my question : do i go this way or not ?!Assuming you are using the 0A2 / 0B2 in a simple shunt regulator, this is not possible for the output supplies. Maximum current avaialble from a simple shunt regulator is limited by the maximum steady-state current of the shunt device (gas-discharge tube, in this case).
For an 0A2/0B2, this maximum current is in the tens of mA, not the several amps that a 10-tube 6080 output section can draw.
As with any shunt regualtor, once this maximum current is breached, the output impedance of the supply rises exponentially - a simple function of the current-limiting device (resistor) used to bias the regulator.
> I could // some hi current transitors as ballast but your answer make me think it would not be easy nor useful to lower the internal impedances of supply rails .
Atma-Sphere output sections were traditionally 'cutthroat' affairs; full-wave bridge directly feeding a large bank of capacitors. This required relatively large values of capacitance, but this was fine as large values were needed to supply the large instantaneous currents required to play real music.
Even in this original configuration, the output was very quiet, partially due to common-mode noise cancellation in the circlotron output stage.
Today, little has changed, save the addition of a 1ohm 5w resistor which is used to create a simple C-R-C pi filter. The resistor provides a little additional quieting, and makes the output supplies slightly kinder to the transformer & rectifiers.
I'm not sure why Ralph was concerned with an C-L-C filter in the output; the only real concern I have, is that the DC resistance of the choke needs to be quite low.. on the order of 1ohm or less. Coupled with the large currents required, you would have some truly massive magnetics.
As for Harvey Rosenberg's writings on these subjects, I'd tend to take them with a grain of salt. I don't like to speak ill of the dead, but trying to garner any real technical wisdom from one of Harvey's articles was the proverbial needle-in-the-haystack search. HGR had some amusing and entertaing writings (at times) but I'd view his techinical pronouncements as anecdotal, not as vetted theory & practice.
Interesting to learn that Atma-sphere now uses a CRC filter for output supply. Per Speltz's Tweak, I use CRCRC, where the first R = 0.21 ohms and the second R = 0.1 ohm. But there is value as you suggest in partially reducing the tremendous load on the rectifiers that occurs when one uses only a huge capacitance after the rectifiers. I use 2000uF - 0.21ohms - 2000uF - 0.1 ohms - 10,000uF in my CRCRC filter. There is no doubt in my mind that this Tweak improved bass response AND S/N ratio. I think Ralph's argument with Harvey did center on his belief that one needs to keep the impedance of the supply as low as possible and that this would become a problem with any feasible inductor in the filter. Of course, Harvey was free to use huge free-standing monster inductors in his homebrew modified M60s. Moreover, I think Ralph and others had tested the CLC option and found that the sonics were not satisfying, i.e., not as good as with no inductor at all. I agree with you, HR seldom had any concrete advice to give; his writings were more for entertainment value.
Hi Lew and Bill,Dangit, now I've got to get the soldering iron out AGAIN! I like the idea of the CRC filter in the output sections of my M60's. I have the 1 ohm 5W resistors "in stock" for plate resistors already.
Question: When installing the resistors, do I need two resistors per side (+ to +, and - to -)? I haven't been in there since I upgraded the caps to 2 X 5600 uF per side.
Thanks,
Not sure I understand your question. If you are going to create a CRC network, you will need ONE resistor per phase, not two, but a total of two resistors per monoblock amplifier. If the factory is using a 1.0 ohm value, as per someone else's post, then I suppose that's OK. Speltz's earlier proposal was to use no more than an 0.5 ohm resistor. I finally used 0.2 ohms and 0.1 ohms, respectively in a CRCRC network. I would not recommend going any higher than 1.0 ohm. Use a "smaller" cap in the first position, e.g., 1000 or 2000uF, then R, then C = 5600uF. Part of the improvement you will hear is due to simply increasing the total capacitance.
Hi Lew,Thanks for your reply, Lew. Since the output supplies float, I was assuming that I would need a resistor for both the negative and positive sides of the caps in each phase. For the two capacitors in each phase, a one-ohm resistor would "bridge" the negative terminals together, and another resistor would bridge the positive terminals together in each phase. I am referring to Paul's power supply tweak schematic (URL attached), where he does this with four caps and six resistors per phase of the amp, three resistors on the + side and 3 resistors on the - side for each phase.
Hope I haven't confused you (and me)more!
Thanks,
William
If you follow Paul's advice, you will be doing the right thing. You (and Paul) are correct, one needs separate resistors for the plus and minus sides, which does add up to 4 resistors per monoblock for a CRC network.
Hi Lew,I contacted Bill Layer, and you are correct. He said that the factory M-60's use a 1 ohm 5W resistor only on the positive side. He said I could experiment with 0.5 ohm resistors on each polarity, but that it probably wouldn't help. Bill reminded me that the negative supply doesn't completely float, that it is coupled to ground through 600 ohms.
Thanks, Lew!
I think I am in at least up to my eyeballs if not over my head, but I don't see how it matters that the negative supply is grounded while the positive supply is floating, with respect to the CRC filter network. I yield to Bills' superior knowledge, but if it can't hurt, why not use CRC networks on both the positive and negative supplies? Were you able to check Speltz's Tweak? I haven't tried to access the old ASOG Tweaks pages, but I think they still exist. I believe that Paul did use CRC or CRCRC networks on both sides.
Hi Lew,Yes, Paul definitely uses CRC's on both polarities. I have the link to his tweak site on my post of 2/28 if you want to revisit it.
Thanks, Lew.
This sounds like an Atma-sphere type balanced, triode OTL. I am a bit over my head technically as far as giving you advice on regulating the output stage. (I see no problem regulating the input and driver stages, but it sounds like you will use the OA2/OB2 to provide set grid voltages. I don't think you can regulate the whole B+ and B- supplies adequately with just those VR tubes, but maybe I'm wrong...) The simplest answer I can give is that it would probably be best NOT to regulate the output supply. Too bad the ASOG Tweaks archives are in limbo; Paul Speltz had posted a very good tweak for filtering the output supply voltage using a CRCRC configuration, where the Rs are very small to keep the impedance of the supply low (i.e., less than 0.5 ohms). I use his idea in my amps with good results. There was once a brouhaha arising out of a Positive Feedback article by Harvey Rosenberg, in which he recommended adding chokes to filter the output supply voltage, in a CLC configuration. Ralph disagreed with this idea, vehemently. It is a great pity that Mark Gilmore, Paul Speltz, and Al Sekala are no longer contributing to OTL wisdom; any of them could give you a better response than I have done.
This post is made possible by the generous support of people like you and our sponsors: