JEL phono and line preamps

For severaI years I was hacking a Dyna PAS chassis with various cascoded and cascaded phono circuits all employing negative feedback RIAA EQ. My research ended when I discovered this unglamorous 2-stage cascaded passive RIAA EQ phono preamp circuit at the back page of an RCA tube manual. The most important lesson I learned in this excercise is negative feedback chokes the sound! I did not alter any component values except for using Ni-mH batteries for cathode bias which I thought improved detail and definition. I strongly advice against the use of a cathode follower and make sure the circuit drives a minimum load of 220K as originally specified otherwise you will lose bandwidth at both frequency extremes. I use an Alps 250K but have also tried the cheap Alpha 250K dual log taper carbon pot sold by AES. It does not have the tracking precision of Alps but at normal volume settings the channels are balanced enough.

For MC cartridge use, you will need a step-up device like a Tamura TKS83 or Altec 4722 mic transformer. See my microphone transformer adapted for MC step up duty article. I am not a fan of active MC stages and rather not go there....

I have been using this phono circuit since the early 90s starting with a PAS chassis with Telefunken ECC83/12AX7. Through the years I developed fondness for octals and nowadays I almost exclusively use 6SL7s/5691s which further improved the bandwidth and depth due to its slightly greater transconductance, vis-a-vis, the 12AX7. There are people who swear by using high transconductance tubes like WE417A, 6DJ8s. To hear what this was all about I breadboarded a famous phono circuit using these tubes but did not find the sound as appealing which is probably just a matter of taste. Personally I find the superior linearity of true audio tubes like a 12AX7 or 6SL7/5691 to play a lot more records. See Diego Nardi's MONOphono preamp article in Sound Practices issue 16 [SP archive CD] which support the technical merits of using tubes designed for audio.

If you absolutely need a low-Z out buffer for CD-R archiving, graft this capacitor coupled cathode follower from my monophonic preamp with variable EQ. To my ears this circuit contributed the least amount of sonic degradation and I experiemented with quite a few topologies and operating points. To my surprise direct coupling the cathode follower sounded worst. Perhaps the sonic effect of degenerative feedback caused by a cathode follower is more apparent and significant at low signal levels such as in a phono circuit because I do not find it objectionable, in fact it is a beneficial at the high level output of my preamp.

This an adaptation of the JEL phono stage with switchable EQ for RIAA, AES, Columbia LP and Old 78. Supplementary EQ networks are available in my article on mono playback.

Another very simple topology with a single gain stage direct coupled to a cathode follower. Originally I was just using a 6SN7 dual triode for both stages but was inspired by the Berman article in SP 13 to try a type 76 as the gain stage. I use a lower B+ supply and run the tube at lower plate dissipation and also prefer the sound with the 2.2K cathode bias bypassed with a 100uf/16V cap [in this case I can justify using Black Gate]. My buddy Joel and I tried biasing the 1st stage with 2 x 1.5V Ni-Cads, at the expense of warmth.

At this signal level the direct coupled 6SN7 cathode follower is sonically beneficial and the low output impedance results in less interaction with long and/or capacitive interconnect cables. [Cable geeks take note: many times the "difference in cable sound" you hear is more a function of load/capacitance especially if your preamp does not have low output Z].

I briefly flirted with this design as a linestage and in spite of the fact that all my SETUP friends liked the sound [whose ears I respect], I eventually went back to the above linestage. The Tango NP216N is a wonderful piece of iron although I have to admit that I needed greater bandwidth than it has to offer. However this is a circuit that offers a solid 600 ohm [as used in studios] output impedance and can even be wired in balanced mode.

At some point I may try this topology again at the output of a tube microphone preamp project. Please do not take my seeming lack of enthusiasm as a dismissal of this topology. I included this schematic because it may well be appealing to other ears and/or applications.

Here are two very nice pictures of Joel's adaptation of the NP216s in his PAS chassis and Nonong's version in a custom made chassis. Both units have separate power supplies based on the PS circuit below.

During my Dynaco hacking days, I experimented with regulated HV using solid state devices which was 'in tune' to the high end sensibilities of the time. I also chanced upon a Heath tube regulated [6L6GC] Lab PS and hooked this to my preamp, there was a slight improvement towards my goal but my ears longed for a more natural sound.

When I found my first copy of the Radiotron 3rd edition, I "digressed" back to 1930s technology. This is the PS circuit I've been using for the past couple of years for deluxe versions of my preamp. I've deleted the .33uf/630V 'tuning cap' at the output of the rectifier turning it into a full choke input followed by another choke/capacitor stage in order to get a very well filtered B+ supply.

I use a bridge rectified and highly filtered DC filament supply for the heaters without having to use silicon regulators. If you pay close attention to the schematic you will notice that there is a voltage divider network [120K series connected to a 47K and 68uf/160V combo shunted to ground] at the output of the B+. This network is mounted on the preamp chassis and keeps the filament floated above ground to further minimize hum and noise and also to keep the heater to cathode voltage on the 6SN7 cathode follower below maximum limit.

An external power supply is mandatory to get your hum and noise down. Since the phono and line stages have no global feedback the circuit will be sensitive to spurious noises from AC hum fields induced by the power transformer. The numbers in boxes pertain to the terminals of a 4-prong connector I use at the end of the umbilical cord and terminal [4] which is missing in the schematic is for ground.

If your power transformer has a 5V, 3A rectifier winding, you can experiment with rectifiers. My favorites are the 5V4G and 5Y3, but have also used GZ34/5AR4 and 5U4. If you know how to use a DVM, you will note that different rectifiers will produce higher or lower B+ voltage which alter the operating point of the circuit. In general, the different voltage is what changes the sound not the rectifier itself.

Under the hood pictures

preamp main chassis

and power supply.

Variations on a theme

Through the years I built several versions of the above circuits starting with 9-pin and eventually octal and 5-pin based tubes. Here is a picture of my original protoype using octals.

Ding's preamp

This is an internal shot of the above octal [5691 and 6SN7GTB] preamp I built for William which used premium parts [Shinkoh tantalum 1% resistors, Angela/Jensen copper foils, Kimber AGSS silver wire and Black Gates] and the power supply. This preamp has been modified by Joel in recent years to use 76s and Tango NP216N preamp opts just like he did to his PAS preamp

which uses 12SL7s and 12SN7s. Nonong grafted the JEL phono circuit to a similar transformer coupled linestage in his preamp using a type 76 as the gain stage.

 

Boutique quality parts aren't necessary to achieve good sound. After Mike insisted on getting my preamp along with the Stereo SE300B/XE60-3.5s amp, I built two almost identical preamps, unit no.1 [bottom view] which sports a deluxe power supply, Angela/Jensen copper foils signal caps and Cerafine PS caps,

and unit no. 2, [bottom view] built with LCR PS caps, Sequa paper in oil coupling caps and a modified Genrad type 1203-B unregulated PS [6X5GT rectifier]. I replaced the filter caps with an LCR 50uf + 50uf/500V, inserted a 7hy/50ma. choke in a C-L-C type circuit and DC rectified the 6.3V filament.

In both units, I used Allen-Bradley 1W and 2W carbon composition resistors matched to 1%. Although there is a perceptible difference in sound [unit no.1 has a slight edge on detail retrieval maybe due to the copper foils and tighter PS regulation], I'm hard pressed to choose which one is actually better, preferring just to listen to music.

JEL Mini preamp

I even built this simpler version [click here for a bottom view] using 2 x 6SL7s for phono and a single 6SN7GTB with each triode section functioning as a plate loaded amplifier [same plate load and cathode bias resistor values as the 76 in the 76/6SN7 circuit] sans cathode follower for the linestage, so the final .22uf coupling cap will be connected to the plate of the 6SN7 and the other end of the cap connected to the preamp output RCA jacks. Without a cathode follower the output impedance is higher so cable length should be kept to 2 meters maximum. The external PS is a C-L-C, rectified by a 5Y3 and bridge rectified DC filaments. I use this preamp for my bedroom audio/video system.

JEL Mini preamp in vintage grey finish

JEL Mini preamp in a single chassis

Built during the the Holidays '07 for my brother. I finally succeeded in building a one chassis preamp with minimal hum and noise without the raised ground potential voltage divider network. The trick is to ground the 6VDC filament at the last tube - in this case I ran the filament wiring from the 6SN7 and then 6SL7s. The last 6SL7 is where the negative 6VDC supply is grounded not at the bridge rectifier!

These are really very simple and versatile circuits which can even be implemented on a Dyna PAS chassis using 9-pin 12AX7s and 6SN7s or 12AU7s that do very little interference with signal sources.

Last revised: 01/06/08 JE Labs all rights reserved.

Back to Main Page or Shop Talk