Anybody did compared tube vs opamp for preamp ?

xslavic

Member
2011-02-04 5:21 am
yeah right ,im buildin the 4 x lm3886 in parallel 200w +btl,and on the original pcb i could use and ne5532 but i opted for 797br,and thinking instead of JLH class A Preamp,ill try both,just need to finish the amplifier
--a preamp will give me more control and lower noise--i could opt for dac instead of normal cd player line out and i have a sub that i intend to connect someday,even if i ll use 3 way floorstanders
 
I don't know the schematic your project is based on, but I think NE5532 is wired as buffer (gain of 1), therefore I would stick with NE5532 at first...latter on you can experiment with another opamps which have to be stable at gain of 1. So, you don't need another preamp in front of BPA200. As for transformer, it is a matter of VA, secondaries AC voltage and if it is center tapped or split secondaries, but if you have only one transformer for both channel, you will have some trouble to avoid ground loops.
 
Yes I have. I have a classic PAS2 preamp which last year had one channel low, not caused by tubes. So I bought a disco mixer with 4558 op amps to get by. I replaced all the paper caps in the high gain stage of the PAS2 with plastic film, and of course it was the last one that had the burned wax shell (hidden on the bottom, of course) that had caused it to deteriorate. I then changed the PAS2 plate resistors to metal film, changed the H.V. electrolytic cap, scrubbed and scrubbed to eliminate arcovers on the PCB's, and it now sounded better than the op amp mixer on both my transistor amps, the ST120 with DJoffe bias mod, and the Peavey CS800S. A little less hiss than previously idling, and much less hiss than the disco mixer. However, the volume pot on it is stuck; a new Stereo Cost Cutters OEM pot from the Dynakit bankruptcy reject room. Has to be turned with vice-grip plers. Haven't found any stereo 250k log replacements with the loudness track.
So I bought some 33078 op amps for the disco mixer, which had great slider pots. Quieted the hiss right down, but it had maybe -40db of hum. Long war on hum ensued, grim details on http://www.diyaudio.com/forums/anal...improving-disco-mixer-mid-fi-performance.html. At the end the disco mixer has equivalent hiss, equivalent lack of distortion on top octave steinway piano, and a tiny bit more hum than the PAS2. Have to stand 20 cm from speakers to hear the hum. Have a question going on analog line level thread about changine the RIAA curve parts on the disco mixer for even less hiss (kitchen sink preamp post). Both Pas2 and Disco mixer sound the same into a dynakit ST70 with new output tubes, but ST70 mid range piano is honkier than either transistor amp. Top & bottom octaves sound nice. Looking at a triple twin triode input board for the ST70, but the triodeelectronics.com version is designed for EF86's of which they sell only a poor eastern block copy. (6pi something tube is what I got from them) I'm not into etching PCB's anymore.
 
Last edited:
I used to be a 'tube-guy' thru-and-thru, but when I needed a MC phono preamp, I decided to take a chance on Audio Sector's OPA627-based unit and I was pleasantly surprised. Very transparent and detailed, it sounds like nothing else I've heard before. So now I'm not so quick to slam SS - but I certainly give tubes the nod for ruggedness and ease to work on. In my experience, it certainly is easier to build a good tube unit than SS.
 
Looking at a triple twin triode input board for the ST70, but the triodeelectronics.com version is designed for EF86's of which they sell only a poor eastern block copy.

Yes, St. Petersburg made EF86 equivalents are not good. However, you are not restricted to COSTLY NOS to get good sounding EF86s. New Sensor's Saratov, Russia, plant currently makes 2 good varieties: the EH labeled tube is good and not especially expensive. The TungSol labeled tube is not inexpensive, but still remains affordable, and it's very good.

FWIW, I'm fond of the TE replacement ST70 driver board populated with 2X EF86s and an ECC99, as it retains Dyna's small signal topology, while using better pentodes and triodes than those found in 2X 7199s.
 
Yes, St. Petersburg made EF86 equivalents are not good. However, you are not restricted to COSTLY NOS to get good sounding EF86s. New Sensor's Saratov, Russia, plant currently makes 2 good varieties: the EH labeled tube is good and not especially expensive. The TungSol labeled tube is not inexpensive, but still remains affordable, and it's very good.
FWIW, I'm fond of the TE replacement ST70 driver board populated with 2X EF86s and an ECC99, as it retains Dyna's small signal topology, while using better pentodes and triodes than those found in 2X 7199s.
Cool, ZigZagflux's solution for triodes on a ST70 required etching a PCB. I would rather clean out my oven with lye. I got ******ed on the first modern EF86 I bought, but maybe I can do better next time with your guidance.
In case everybody doesn't already know, the Dynakit PAS2 I use for sonic reference has 12AX7's in it, with a 6x4 rectifier. I check the sounds on LP and CD against the Steinway console piano sitting between the speakers. Imaging, I don't know; I went to hear the Louisville Symphony live last summer, and couldn't really spot the instruments there either with my eyes closed.
 

M Gregg

Disabled Account
2010-06-28 11:04 pm
UK
Just for interest,

I thought I would share an experience I had when making a signal generator from a chip some time ago!

I built the generator It worked quite well, however I needed a higher output voltage swing. So I made up a board with a power Op amp with quite a good frequency spec for my purpose. I connected it up and looked at it on a scope. My jaw dropped! I said to another tech, look at this C**p waveform! (Before you say it , It was not a layout problem).

The output was good from the chip generator, however as I increased the frequency the output from the op-amp got worse and worse. I thought it was the scope chip but no. The output was good all the way up! I said I have got to try this through a small tube preamp. Guess what, the output was perfect no distortion all the way up! The other tech said let me try something, so he drew out a circuit of an op-amp with discrete components (transistors and resistors). He said build that and give it a try. Well it out performed the Op amp by a big margin. I tried other op-amps from 741 ne5534 etc etc and they all had their problems. So for the small generator I kept the transistor version in the unit because I had no HT and wanted it compact. The Op27 and 34 seemed best op-amps at the time.

Just for fun.:)

Regards
M. Gregg
 
For anyone interested,

The distortion from the op-amp was sine wave in, sine wave out- increase in frequency sine wave in -change to sine wave was more and more saw tooth (triangular).

Regards
M. Gregg


That is slew-rate limiting because you were using unity-gain-stable opamps.

It is NOT a defect! Just being used outside of its ratings.
 

M Gregg

Disabled Account
2010-06-28 11:04 pm
UK
That is slew-rate limiting because you were using unity-gain-stable opamps.

It is NOT a defect! Just being used outside of its ratings.


I agree it is not a defect, however the "sharks fin" sine wave distortion even within its operating range was interesting. There was no doubt that the sine wave performance was better with the simple tube amp. The change with sound is also evident just by changing NE5534 to OP27 which makes you wonder why manufacturers like Meridian would use such devices!

Regards
M. Gregg