![]() ![]() |
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
I was going to say in the title "Tube used as active load for Foreplay" but that really sounds bad. Anyway, have you tried using a tube configured as a constant current source for the foreplay? Looks like it could be done to the first stage replacing the 120K resistor. I think the 12AX7 would be a good tube as it has high plate resistance. By my rough and probably inaccurate calculations, the effective AC load would be around 5.5Meg. I'm not too sure a tube would be a good choice as a constant current source at the cathode follower; not enough voltage to work with.Just curious,
Tom S.
Sure, it can be done, this is what SRPP and mu followers are all about. We use a mu follower to load the voltage gain stage in the ParaS.E.X. They can sound very nice. Of course, you now have more heater current to supply, another socket to add, etc.But - forget 5.5 meg. You're gonna shift the decimal point to the left on that figure if you use a tube.
That's why we use the C4S, a tube can't begin to give the kind of impedance the cascode PNP setup can, nor can it give the low noise figure. So yeah, with a tube as a CCS you've got another tube in the circuit instead of yucky transistors, but is the tube you're trying to help ( the one below it)operating in as optimal an environment as it would with a faster, higher impedance current source?
Perfect intro for my new soapbox- Don't get sucked into the idea that a tube is always better, regardless of its application in a given circuit.
Another ss vs tube example-
I get asked two or three times a week why we use solid state rectifiers in our stuff. The common assumption seems to be we are compromising here, cost cutting, or whatever.
To quote an infamous president, poppycock.
Impedance is the reason. The higher impedance of tube rectifiers makes it much more difficult for an SE amp to have any kind of acceptable bass speed.
At the VALVE meeting yesterday we listened to the newly completed Ultrawhamodyne speaker design, built by Ed Fallon. very cool, 7.5 feet tall, 16 5" woofers per channel, and Raven tweeters.
Ed has 300B SE parafeed amps with C4S loads on the driver stage. Very sweet amps, and with parafeed and C4S those little woofers should have had kickin bass on 98 dB speaks, but the bass was like soggy oatmeal, very ample quantity, but flabby.Next we put a more or less stock ParaS.E.X. online. Should have been working much harder, less power, way smaller output iron, but suddenly the bass tightened up markedly.
Why? Ed has GZ37s in his power supply, the ParaS.E.X. uses UF4007s with a reverse recovery spike filter, working into about the same amount of capacitance in the PS as Ed's amp. Once again we are using a SS device where it does the job best.
I'm putting UF4007s in the mail to Ed today.
Other observations from yesterday's meeting:
-Two different tube preamps both sounded better than a passive in Ed's system.
-That new Tequiza 'beer' stuff tastes like perfume after a couple of swigs.
-Ed and Brainiac cook a mean burger.
-Quest is a cool guy. Great to meet you, bud.
-And once again a meeting full of ideas for small changes took a nice but not stellar sounding new speaker setup forward quite a bit - toe in, distance from the back wall, minor padding adjustments on the tweeter, the addition of "wings" to the narrow front baffle, all helped improve the setup quite a bit - one small step at a time. Like the racer's say, you want to lighten your race car a pound, find 16 ways to remove an ounce.Now I gotta go over to Ed's sometime and help him pull all those 5" drivers and reline 32 individual woofer chambers with felt instead of foam - yeow...might as well shave the hair off my arms before I get started.
I'm fairly new to tube audio, but decided to give it a try with an ST-70 and Pilot tube preamp. The sound is much superior to my JVC receiver with one exception - sloppy bass. Much as you described, plenty of it, but slow and muddy on fast passages. I've posted on this elsewhere and can generally group the answers into two camps: 1-excessive negative feedback due to too small coupling caps. 2-weak power supply. Being on a budget, I'd love to try the tweaked UF4007's, but lack the experience or knowledge to know what else needs to be changed to compliment that change. Specifically, would I retain the choke? The LCR 40-20-20-20 cap that I used in the rebuild is rated at 500 volts. Would I need to go with a cap of higher rating due to the instant unloaded turn-on? Should I increase the capacitance of any or all sections? I know this isn't exactly on subject with the original post, but I'm hoping that someone might take the time to share some knowledge with a newbie. Thanks very much. Tom Kuhn
Sorry... That message number is not valid.
If I were you, I would not change the power supply to solid state diodes. The GZ-34 has a slow warm-up feature to protect the caps. If you want to add capacitance after the choke, then you might improve things. The bass problem is mostly due to the output transformer and the lack of feedback available at low frequencies. The distortion below 50 Hz is very high. I have increased the coupling caps, but it doesn't help much to fix the bass problem.
I knew there was a good reason why you were going with the SS devices for CCSing. I wuz just curious if you had tried using a tube and found the C4S to perform better.The only good thing I can think of for using a tube rect. is slow start up. Then again, there's SS devices for that too.......
> > UF4007s with a reverse recovery spike filter < <
Capacitors?
Tom S.
> > >
The only good thing I can think of for using a tube rect. is slow start up < < <How many tube rectifiers have you listened to? Which one would be easier to fix if it blows, a tubed rectifier, or ss one? just what IS your experience with rectifiers, tubed or otherwise?
Jack
> > How many tube rectifiers have you listened to? < <None.
> > Which one would be easier to fix if it blows, a tubed rectifier, or ss one? < <
Neither as both the tube or SS rectifier would have to be replaced. It'd be easier and cheaper to source a SS rectifier though.
> > just what IS your experience with rectifiers, tubed or otherwise < <
No hands-on experience with tube rectifiers. I played with plenty of SS rectifiers in tech school and a more at work when I fixed weather radar units.
> > > The only good thing I can think of for using a tube rect. is slow start up < < <
Let me rephrase: The only electrical advantage of a tube rectifier over a SS is that it provides slow start up for the HV supply. The disadvantages of a tube rectifier (electrically speaking, not sonically) is that you need another filament supply (more energy usage) and a power trannie with more voltage output (at the same current) as the tube rectifier gobbles up alot of power supply voltage (energy) in the rectification process. The tube rectifier is nowhere near as efficient as a SS rectifier. This is an electrical fact that one does not need direct experience with. I cannot comment on what tube rectification does for sonics as compared to SS rectification. It was not my intention to imply that SS diodes are sonically superior to tube rectifiers.
Tom S.
> > > This is an electrical fact that one
does not need direct experience with. < < <
be carefull with electrical facts, or else one will be using only Sovtek tubes, since it is an electrical fact that they are superior.> > .It was not my intention to imply that SS diodes are sonically superior to tube rectifiers. < < <
Understood.
For the record,IMO a *good* tubed rectifier sounds as good or usually better than most ss rectifiers. FWIW, theres nothing more depressing than having a nice $7K amp go belly up because it blew a 50¢ diode in its rectifier. It happens more often than most people will admit. At least, if a tube blows, it can be replaced in a few seconds.
Jack
> > go belly up because it blew a 50¢ diode in its rectifier. < <It's strange that some amp designers will take meticulous precautions to ensure proper heat disipation of the tube or SS amplifying device yet neglect proper ventilation or heatsinking of their rectifier diodes.
"Nah, we'll just use diodes with a higher power rating."
Tom S.
I'm kinda stumped here guys. What is this deal with SS rectifier reliability and heat dissipation? With tube amp HV supplies it's pretty negligible. We've had one rectifier that was bad on all the kits we've sold over the years, and it was apparently bad before it ever fired up.In an SE tube amp a typical SS rectifier on the B+ supply drops maybe .7 to 1.7V. If the amp is drawing say 75 mA, typical of one of my own amps, anyway, the dissipation is about an eighth of a watt.
Power down a running amp with an SS rectifier, let the HV supply bleed down, and unplug the amp from the AC mains. Then touch the rectifier. It won't be very hot. If a diode goes, it is usually taken out by a shorted filter cap that is drawing way more than the rectifier's rated current.
Now that failure mechanism may be more likely to occur if the first cap in the filter string is not rated to handle the peak V that the cap charges to before the tubes warm up and draw current. This is more likely to be an area where a designer may miscalculate the ratings required. In this case a tube rectifier will have an advantage as it limits voltage surge during its slow warm up and usually the peak V the first cap charges to is lower than with a SS rectifier.
But - There is also a limit to how much capacitance that first filter stage can have with a tube rectifier ( usually around 40 uF), so one can screw up a tube rectifier/filter just as easily by putting in say 100uF at the first node. The current surge created by the increased capacitance will greatly increase the wear and tear on a tube rectifier, but a SS rectifier won't care as long as it's current rating is not exceeded.
Basically, Jadis amps are notorious for blowing their rectifiers. the importer(frank garby? Northstar) says its aonce in a lifetime thing, but they seem to blow alot-may he meant a mayfly's lifetime. Other rather expensive amps(wont mention names), of the SET persuation have been known to blow their diode-also claimed to have been a "fluke", but I hear about more and more "flukes", just like I hear about more and more KR tubes arriving DOA that are "flukes". Maybe I (and others) have bad luck. Who knows?
Jack
My point is that it's not the diode's fault if the design around it is inadequate. If certain amps are blowing SS rectifers or tube rectifiers, it is up to the designer gain the knowledge to be able to correct the fault in the design, and then to take care of the problem for the customer.
Ò¿Ò
When I said heatsinking and ventilation for diodes I was thinking of SS designs where voltage is low and current is high. I don't think my statement applies to tube designs using high voltage and low current.I'd like to say where that quote came from and the context in which it was said but I've already dug a big enough hole for myself.
I need more sleep. Where's the author self-delete button at?
Tom S.
For the UF4007 in particular a 10-12 ohm 2W wirewound resistor in series feeding each diode, and .01uF 3kV caps shunting across each end of the resistor pair kills the reverse recovery spike and gives a nice smooth sine wave output, actually cleaner than a HEXFRED.
The resistor by themselves, without the caps, work pretty well, too.
Hy folks!First of all I have to thank Dr. B> for his complements..That VALVE meeting sure is cool, WAY cooler than what I was thinking of initially...can't wait till the next one :-)
Yes! I've been playing with some UF4007 before (still have 4 in my stock bin) but without the "so-called" "snubber circuit" (a cap stringed with a resistor on it) to kill the spike noise you will still hear the same kind of SState rectifier harsh sound as many discontent board memebers would refer to.
I was very impressed with the "battle Royal" yesterday between Ed's parafed 300B with GZ-37 and the other paraSEX parefed with UR4007 (snubbed of course). The bass speed of "snubbed" UF4007 is taut and plenty yet retains much of the midrange "sweetness" that a tube rectifier is famous for.... I almost knocked off my chair!!!!!
Oh yeah!! with Ed's XXXWhamodyne (reaching extraterrestrial level)speaker array in place, Stevie R. Vauhn's guitar and voice suddenly sounded so intimating, so fluid.......so sweet!!Great meeting! COOl guys! great experience I have so far... Way to go Doc! Big cheers for you :-)
One question: Will "snubber" circuits works well on Motorola MURs as well as HEXFREDs? Or does this scheme mainly works on high speed soft recovery diodes like UF4007?
regards,
Quest
Yo Quest Dude,ThanX for the little Parcel. A very pleasant surprise. Nice to hear you like yor new Amp....
As for the MUR Diodes I personally use the following scheme:
Before the Rectifie I connect a moderate Value Resistor usually a 39 Ohm Allen Bradly Carbon Comp in each leg. After that a 10nF Paper/Oil Cap (a few kV) across the input and the output of the Rectifier to sink noise....
This works well and has all the good sides of a Valve Rectifier without the bad ones.... That said, my recipy for Bass-Slam and speed is to use a regulated PSU with a twist....
My Design of PSU deliberatly reducecs the Open Loop Gain of the Reg dramatically above a few 100Hz (if it uses Feedback) and each reg is designed to hand over smoothly (witout any reactive peaks) to large value Polypropylen Capacitors that essentially deal with the PSU from the upper bass upwards (I aim for a 100-200Hz "crossover"....
It does require a lot more parts tha a simple unregged supply but allows for unweighted > 96db S/N (re. 1W) in a 300B SE Amplifier....
I can't hear a thing with my head stuck into the Front-Horns of the Lowthers I use nowadays.... That is no audible noise with a 103db/W/m Speakers at a much closer range than 1m....
Ciao Thorsten
Quest:What Thorsten describes as his diode treatment is essentially the snubber network that I built into the ParaSEX amps that you heard at the last Valve meeting. Instead of the 39R Carbon resistors, I used 10R 1W Wire-wound. .01uF 2KV ceramic caps between them, on each end.
-.--- 10R 1W WW ---.---- To diode
| |
.01uF .01uF
| |
-.--- 10R 1W WW ---.---- To diodeAs well, Doc had this little trick in his pre-amp.
My two cents
Regards to all...
.
Just one more question:In J.C. Verdier's design he calls for the use of 1 cap after the regulation which he claimed to be sufficient in suppressing almost all the hums and buzz without the need to resort to "pi" filters again.Is that true?
If so would you recommend my use of SState rectifiers (with caps of course) --> 10Watt 50ohm (I'm actually planning on bypassing it with a 3K resistor coupled with a 30sec. time delay switch, as to short the resistor when the time is up) --> regulator (valve/FET, haven't decided yet) --> PSU filtering caps? (initially calls for Cerafines but if my internal space permits I'll probably stick in some Solens 47mfd for the time being .Will 47mfd enough you think??)
Or do you prefer the addition of an additional filtering foil cap before regulator?
Kind regards,
Quest
Kind regards,
.
The particular high ( on the order of 10-20 mV)amplitude but very quick single reverse recovery spike that the UF4007 produces is actually easier to squash than the higher order spuriae from the HEXFRED. I think Buddah is going to look into the MUR, but I haven't heard any result, nor tried them myself.With regards to listening to tube rectifiers vs. solid state-
I imagine I've listened to and used more types of both than a lot of people. I think that if a person wants to use a tube rectifier because it makes the sound the way they want it to be, more power to them.
The issue in the case I made was that a guy came to me and said, "Hey, why is the bass so soft?" I suggested a reason, a change was made, and the result seemed to support my suggestion, that's all.
To further this line of thinking, I will throw out the challenge, "How many of you have looked at the waveform of the various devices on your scopes, as well as listened to them?"
Assuming that a smooth waveform is what we want to see, devoid of any higher harmonic stuff, the results of such a comparison seem to show that the filtered UF4007 looks very smooth clear up to the tens of megahertz at least, quite similar to a typical tube rectifier. Lynn Olson and Matt Kamna did an interesting study of the waveforms coming off various unfiltered tube and SS rectifiers and found that the one which showed the least "glitch" ( they say even tube rectifiers have some form of "notch" )was in fact a TV damper diode.
Sooo, may be time for tube rectifier lovers to reconsider on those GZ37s and Mullard 5AR4s and add an extra tube socket for a pair of damper diodes... My own experience has been that the damper diodes are smooth, but still suffer from high impedance, and that the bass from an amp using them, while voluminous, is still on the slow side. YMMV.
Who is the manufacture of the UF4007?
General Semiconductor, it's available as Mouser PN 625-4007.
This post is made possible by the generous support of people like you and our sponsors: