FrankK said:
If I were doing a 300B project, which I'm not contemplating at this time, my preference for a grid driver would be a MOSFET source follower. The A Number One problem with the 300B is inadequate grid drive. These DHTs all have a nasty habit of pulling grid current even before Vgk actually goes positive. Add that grid current to the current pulled by Ci + Cmiller + Cstray, and you have a problem. A 6SN7 isn't going to be up to that task.
The source follower, having both an rd(on) that's much lower than any rp, and a gm that no hollow state device can touch, solves these problems handily. The Zo of a source follower is lower than anything a cathode follower can provide, and these's no lack of driving current. Both are important for low distortion operation, and will give the performance you're looking for.
I've got an 845 SET design in the works, and this will definitely include a source follower for grid drive. The 845 has all the problems of a 300B in spades.
Agree with the topology but taking a different approach with the 81x amps currently in play: a beam triode CF. A peak plate voltage of 5000 volts with a 300 ma max current (not simultaneously of course!) makes for dead simple implementation. The other advantage is, with a CCS plate load, a direct-coupled, completely unloaded voltage gain stage is trivial too. Downside, the beam triode needs LOTS of voltage.
So...the linearity of 300B it seems to be hard to beat from what I read here...
However, excellent, thd, imd, tid measurements do not always tell the whole story about how an amplifier will or will not alter the sound signals than pass through it.
So...if the 300B is hard to beat,what manufactures of 300B it seems to have the most balanced sound?...in your subjective opinion ofcourse.
Thorsten have this review but now,in this game,is coming fast the KR tubes.
http://www.enjoythemusic.com/magazine/equipment/1201/300b/
http://www.enjoythemusic.com/channel/ch300b/
However, excellent, thd, imd, tid measurements do not always tell the whole story about how an amplifier will or will not alter the sound signals than pass through it.
So...if the 300B is hard to beat,what manufactures of 300B it seems to have the most balanced sound?...in your subjective opinion ofcourse.
Thorsten have this review but now,in this game,is coming fast the KR tubes.
http://www.enjoythemusic.com/magazine/equipment/1201/300b/
http://www.enjoythemusic.com/channel/ch300b/
The TubelabSE amplifier uses a circuit that I call PowerDrive. It is a CCS loaded driver tube (5842) coupled to a mosfet follower driving the grid of a 45, 2A3 or 300B tube. Dozens of these amplifiers have been built and the reaction has been 100% positive.
http://www.tubelab.com/AssemblyManualTubelabSE/Schematics_TSE.htm
The 300B maintains its linearity over a wide range of operating points. In short it has a large sweet spot that is easy to hit. It is harder to make a bad amplifier using the 300B compared to other tubes, leading to a high success rate.
I have not tried any of the expensive brands. My experience has been limited to Chinese and Russian versions. Last weekend I ran across some WWII vintage 307A - VT225 tubes and some older Raytheon RK-75's. These are DHP's. I tried them in a TubelabSE wired as triodes. They easilly better the 300B tubes that I have been using. I just can't stop listening to it.
Design your driver such that it can swing the 845's grid up to +75 volts and you can get 40+ WPC without violating the tubes specs. See the third schematic on this page:
http://www.tubelab.com/845SE.htm
The voltage on the gate of the mosfet and therefore the grid voltage on the output tube is derived from an unregulated negative supply. The plate voltage on the output tube is derived from an unregulated positive supply. These two supplies tend to move together when the line voltage changes to afford some degree of stability of output tube current. The effect is not much different from any other fixed bias (not cathode bias) amplifier that doesn't use a regulated power supply. My TubelabSE amplifier that uses NX-483 tubes (a 5 volt 45) has required no bias adjustments since I built it (the tubes are 75 years old, so they have been thouroughly burned in) even though my line voltage varies over an 8 volt range daily. I have checked it a few times ant the current is between 29 and 31 mA.
I do not use the 845SE very much, so I can't say whether it varies. It currently has no tubes in it, and hasn't been used in several months.
As with any other vacuum tube amplifiers the following guidelines should be followed:
In an amplifier that uses a triode (or a triode wired pentode)output tube you should not use a regulated plate supply with an unregulated bias supply. The plate current will decrease as the line voltage rises. You should not use a regulated bias supply with an unregulated plate supply. The plate current will increase as the line voltage rises.
In an amplifier that uses a pentode output tube the plate voltage has little impact over the plate current, but the screen voltage does. If the bias supply is regulated, the screen supply should also be regulated.
I do not use the 845SE very much, so I can't say whether it varies. It currently has no tubes in it, and hasn't been used in several months.
As with any other vacuum tube amplifiers the following guidelines should be followed:
In an amplifier that uses a triode (or a triode wired pentode)output tube you should not use a regulated plate supply with an unregulated bias supply. The plate current will decrease as the line voltage rises. You should not use a regulated bias supply with an unregulated plate supply. The plate current will increase as the line voltage rises.
In an amplifier that uses a pentode output tube the plate voltage has little impact over the plate current, but the screen voltage does. If the bias supply is regulated, the screen supply should also be regulated.
tubelab.com said:
I spent most of yesterday parting out, and I don't think I'm gonna do it. What with the ouput and power iron, that gets $UBER PRICY. For the same power supply, and the same cost, I could get 200W of speaker busting power from a PP pair of 814s. Doesn't look cost effective, so it's onto something else. Dunnow what just yet, may be time for some SS designs.
felixx said:So...the linearity of 300B it seems to be hard to beat from what I read here...
However, excellent, thd, imd, tid measurements do not always tell the whole story about how an amplifier will or will not alter the sound signals than pass through it.
So...if the 300B is hard to beat,what manufactures of 300B it seems to have the most balanced sound?...in your subjective opinion ofcourse.
Thorsten have this review but now,in this game,is coming fast the KR tubes.
http://www.enjoythemusic.com/magazine/equipment/1201/300b/
http://www.enjoythemusic.com/channel/ch300b/
i have been lucky enough to own a pair of vintage british 300a (300b) valves...and done a direct comparison with the kr 300xls tubes in the same amp..
the old tubes beat the krs by a margine, the bass is very taught and deep, good top end, more detailed and lush ( as in real) sounding..
the kr's seem to have less bass problems than many, but it is still loose compared to the genuine artical...
the issue with new most new production is that just that.. new production.... they are not as good as the old stuff..
that goes for all new production valves in my experence.when used as driver or output
the kr's are still quite good, to get them into perspective.
I use mostly british late 1930's and 40's 4 volt valves
px4, px25 etc in SE, where 2 watts will do 105 dbs and not lack in any department ...
steve
I got a pair of OPT's from Handwound Transformers before Mr. Lucas decided to just steal peoples money. The pair of OPT's cost about $130 and they sound pretty good. The OPT's came first, then I decided to build the amplifier. The HV power is easy and cheap. Use a voltage doubler off of a 480 volt industrial control transformer. The reality is that it is still a big (2 chassis) heavy amplifier that radiates 400+ watts of heat. It is too ugly and dangerous to be allowed into the living room, and I don't need 80 watts of audio in my 10 X 10 foot lab, even with 87db speakers. Someday I will rebuild it into a nice looking amp, but for now it sits on a shelf, unused.
I have no experience with the 1638SEA.
I have used both the 1628SE and the 1628SEA.
I bought a pair of 1628's from AES. Their catalog showed the 1628SEA so I ordered a pair. They shipped me a pair of 1628SE's that were two years old. I complained and they agreed to take the 1628SE's back (I had never soldered the wires), but I would have to pay the shipping which was over $40. They agreed to order me a pair of the 1628SEA's for a fair price, so I wound up with both sets of transformers.
The original 1628SE has problems. Most SE OPT's have a notch in their response where the winding capacitance and the inductance resonate to create a notch where the frequency response can dip 10 to 30 db. On a good OPT the notch is usually far above the audio range. The problem lies when the notch falls near the audio range. This is the case with the 1628SE. There is a notch that is 35 db deep that falls at 38 KHz. The notch is wide enough to encroach into the audio band. It also causes high distortion above 10 KHz and response is 3 to 5 db down at 20 KHz. It is different from transformer to transformer and varies with loading and secondary connections. I have been using my set for experiments, since I wouldn't put them in an amplifier that I might sell to someone as "HiFi". They will wind up in a SE guitar amp some day.
The 1628SEA has a single tapped secondary instead of the two winding arangement in the 1628SE. This appears to cure most of the problems. The 1628SEA still has the notch but it is further away from the audio range. The problem is that the notch still intrudes on the audio band causing some rolloff of the high frequencies. It is about 1 or 2 db down at 20KHz. I put these in a SimpleSE amp. These transformers have the best bass I have ever heard in an SE amp. I can shake my neighbors walls with a KT88 SE using 15 inch speakers. These tested as the best SE transformers based on frequency response, and distortion. They have the lowest efficiency also, giving up more that a watt of output over the little Edcor (the best). The testing reveals no saturation effects at full power at 20 HZ. No surprise here these are the largest of the SE OPT's that I have.
Miles Prower said:
What about parafeed, Miles? If size is not an issue, might be easier to get a big choke and an ungapped OPT.
And switching gears completely: We often see discussion of design trade offs in amplifiers here, but with the unstated assumption that one amp will drive the entire frequency range (except maybe the sub). What with all the amp builders here, I'd expect to find more multiamped systems where some of those design trade-offs can be swapped between amps in the system.
Sheldon
Hi Ashok. Hours of simulations strongly suggest the largest distortion creating components of your circuit will be the coupling cap and grid stopper. To my surprise it has nothing to do with tubes and is solely related to a non-zero impedance driving a diode. That's when the 'tubelab' light came on in my head. If the sims are representative of reality, ditch the grid stopper and go direct coupled CF if at all possible. The distortion reduction in A2 isn't subtle.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.