Scaring you? Ha! I did say I wasn't trained in solid state - I'm not an EE. JFET circuits are unfamiliar to me. I've been reading a lot about op-amps and I'm a bit further along there. Tubes are where my knowledge and practical building skills are.
In terms of sound quality, would you make a case for a JFET sounding better than an op-amp? What degree of difference?
In terms of sound quality, would you make a case for a JFET sounding better than an op-amp? What degree of difference?
The RC high pass filter is really worth trying to understand.
None of this is exclusive to either tube or SS, the principle is the same.
-
RC High-pass Filter Design Tool
C is coupling cap, R is shunt resistor.
-
Best
HK
None of this is exclusive to either tube or SS, the principle is the same.
-
RC High-pass Filter Design Tool
C is coupling cap, R is shunt resistor.
-
Best
HK
I get the RC filter, that's not the issue, it's designing a JFET stage without a largish coupling cap I'm unfamiliar with in general.
The coupling cap value is more determined by the load of the stage that follows it. So if the JFET stage is driving a tube stage with, say, a 500k grid bias resistor, the coupling cap can be small like you are used to with tube stages. If the following stage is a 10k resistor, the coupling cap would have to be much larger.
Keep in mind, I have never put JFETs into a design myself so I'm no expert, but to me they look quite similar to tubes. They require negative bias normally so they even self-bias through a resistor like tubes.
A stage between the output of a DAC and a DHT power-valve driver is a better position for an opamp (or IT) than a JFET.
Have a look at the curves of JFETs (that you can actually buy). Look at Vgs against the drain current. You'll see that in most cases, the active range of Vgs is only Vgs=0→3V to get zero→max-current. Some are only 1V. Within that operating range, the gm varies greatly, and fairly non-linearly.
DACs usually give 2V rms (so, about 2.8V pk) so the distortion with a bypassed source resistor can be truly terrible.
You can address this with positive bias and degenerate source resistors; or with feedback loops, but it's not an attractive solution.
at lower levels, such as MC 1. Stage, these problems are not felt, and they are an excellent candidate, with lower noise a real possibility.
The quality of design and implementation is most important, in both cases.
Thanks Rod. Looks like I can stay with my op-amp stage. Less to think about. It seems to work just fine.
The Schematic that Hanze posted is most probably from the days of the old Line-Level for amplifier input: 300mV. That input-level might just be manageable.
A JFET with the largest possible Vgs range - like a selected J111 might be a starting point for such a design.
JFETs for such a design usually need to be selected, because the sample-to-sample variation is bad, even if you are used to valves.
A JFET with the largest possible Vgs range - like a selected J111 might be a starting point for such a design.
JFETs for such a design usually need to be selected, because the sample-to-sample variation is bad, even if you are used to valves.
Its the same for both JFET or tube - like I said. Its all about the RC filter, and nothing else (nothing else).
Cheers,
HK
I was thinking in more general terms about JFET designs. If op-amps avoid a coupling cap, then there may be circuits using more than one JFET that can achieve the same result. I really have no idea here, it's not something I've studied at all. I take your point about the size of the coupling cap, of course. Looks like there's not too much value in pursuing a JFET design when the op-amp stage works OK without a coupling cap. I'm open - in theory - to ideas that might provide a superior solution.
There is nothing that seperates a JFET from a triode and from anything else so far as coupling is concerned. I dont know how else to say this - the coupling is either RC, LC or direct. Thats all there is to it, no majick smoke.
HK
HK
Ok, I’ve been listening to the 47 driver for a few days and am delighted with it. The driver with the LL7903 works brilliantly and am driving it through Slagle’s AVC. Sound is amazing and am more than pleased now with the 300B
Going full DHT is worth the effort in my view.
If you’re interested, here is my implementation and notes:
300B SE Amp: 47 Driver – Bartola(R) Valves
Cheers
Ale
Going full DHT is worth the effort in my view.
If you’re interested, here is my implementation and notes:
300B SE Amp: 47 Driver – Bartola(R) Valves
Cheers
Ale
Good news Ale. This looks like a design well worth building.
Well done! I can easily imagine how good it must be sounding.
Well done! I can easily imagine how good it must be sounding.
HI Andy,
Thank you. Hopefully we get to meet in not a distant future so you can listen to it. I'm very pleased with the results so far.
@gmant123: No, I haven't as I'd need a different driver power supply with the current topology as the VT25/801a would need to be biased at around 300V for 200Vpp at least.
Cheers
Ale
Thank you. Hopefully we get to meet in not a distant future so you can listen to it. I'm very pleased with the results so far.
@gmant123: No, I haven't as I'd need a different driver power supply with the current topology as the VT25/801a would need to be biased at around 300V for 200Vpp at least.
Cheers
Ale
Ale is there a trick to make the 7903 sing ?
I tried it as a transformer between the soekris dac output (630 ohm) and the input of my 801A SE Dac output stage(100k), but I experiences really degrated sound, really not something I would use. Maybe its the PP to SE conversion, maybe its the 7903. the LL1671PPAM was much, much better, but still I ended up with no IT and just used half of the DAC output (i yet have to reprogram the dam1941 to output two identical signal and lower its output impedance.
But even though I own the 7903 now for years, I never found a satisfying use for it.
I tried it as a transformer between the soekris dac output (630 ohm) and the input of my 801A SE Dac output stage(100k), but I experiences really degrated sound, really not something I would use. Maybe its the PP to SE conversion, maybe its the 7903. the LL1671PPAM was much, much better, but still I ended up with no IT and just used half of the DAC output (i yet have to reprogram the dam1941 to output two identical signal and lower its output impedance.
But even though I own the 7903 now for years, I never found a satisfying use for it.
Blitz - have you tried op-amps as an input instead of a SUT?
In fact has anyone else tried an op-amp input to a 2 stage DHT amp? Surely I can't be the only one doing this?
In fact has anyone else tried an op-amp input to a 2 stage DHT amp? Surely I can't be the only one doing this?
OFF
My 1941 is "floating", neither PSUs nor DAC board "grounded". Output RCAs also floating.
I use one ladder output as "hot", other as "cold" connected to RCA, via interconnect to 801a -asymmetrical- preamp.
Try it.
ON
I have a 300B/6A3 preamp into a 3B7-2A3/45 parafeed amp. I’m just a designer/builder wannabe...these are Paul Joppa designs.
What I don’t understand is why an op-amp stage is preferable to a third DHT stage...other than the massive difference in size and $$$$$?
> why an op-amp stage is preferable to a third DHT stage...
A third DHT stage is possible, but must be (even better) implemented than the others, for various reasons. For example: the distortion must be reduced to the bare minimum. Any remaining distortion products have their harmonic order increased by the succeeding stages (because the gain of those stages applies a multiplication to their inputs). The undesirability of higher-order harmonics does not create many arguments.
Noise and microphony are other challenges that are heightened by two further gain stages.
To sum up - it's not that the extra DHT stage is a bad thing (so long as the implementation is right up to snuff), but that some modern audio op-amps are pretty good, when they are running output levels of up to 10V pp or so, into a low-capacitance, moderate-to-high impedance load.
A third DHT stage is possible, but must be (even better) implemented than the others, for various reasons. For example: the distortion must be reduced to the bare minimum. Any remaining distortion products have their harmonic order increased by the succeeding stages (because the gain of those stages applies a multiplication to their inputs). The undesirability of higher-order harmonics does not create many arguments.
Noise and microphony are other challenges that are heightened by two further gain stages.
To sum up - it's not that the extra DHT stage is a bad thing (so long as the implementation is right up to snuff), but that some modern audio op-amps are pretty good, when they are running output levels of up to 10V pp or so, into a low-capacitance, moderate-to-high impedance load.
As Rod says, modern op-amps are pretty good at small signal amplification and that's what my ears tell me. Clean sound, transparent, no coupling cap, 5.5x gain. My DAC has an op-amp output anyway, so I'm listening through an op-amp in the first place. It took me a while to figure out the values but my NE5534 stage sounds fine to me. I agree with Ale and Rod that I don't want more than 2 stages of tubes if a third stage will degrade the sound. I found the third stage veiled the sound somewhat, particularly as regards overall transparency. Like Ale, I now have 2 stages of DHTs which is what my ideal setup is (for now...).
Last edited: