| analog_sa |
Over the weekend i came across an old FET 2 preamp in slight need of attention (blown MC input, malfunctioning muting relay).
This was hardly a statement product in its heyday but the quality of case, pcb and layout really impressed me. The sound also exceeded my expectations.
Fwiw, i think that it would make an excellent simple project for anyone wishing to build a nice sounding discrete preamp. I am fairly certain that its phono section is better than any of the el-cheapo, single opamp designs.
So, this morning i attempted drawing the circuit. Although trying to be careful, i may have made an error or two, so if anyone spots anything funny please let me know.
The preamp has a separate MC gain stage which, being a common gate circuit is rather unusual. It uses a parallel 2N5566 and i found that a 2SK389 works quite well. The trimpot is used to set 0v at the input which seems to be quite reliable and stable.
Predictaby, this is the worst sounding part of the unit. Most probably because of the PS - remember this pre is built for a limited budget. Entire pre is supplied from a nice but tiny 30VA 2x20v toroid followed by a bridge, 2x2000uF Mallories and separate 7824,7924 for each channel. No local regulation, just single caps per rail near the line amp (no markings, must be around 100uF)
IME a MC gain stage requires the absolute best PS with batteries or dedicated Jung type regulator to sound acceptable.
Coupling cap to MM stage is a 1uF polyester-looking Electrocube.
R1 is selectable for 3 values - 50, 100, 200 ohm. |
|
|
| analog_sa |
After getting disappointed with the MC performance i plugged a Tribute cobalt amorphous toroid into the MM input and continued listening. What a change! Dynamics, bass, separation, detail - all the audio cliches fit :)
So, the MM stage must be rather good then. Not that i can see how a two stage circuit can have sufficient gain for correction plus some extra nfb but it seems to work.
There is a bank of switchable caps at input while the output feeds a balance and volume pots (25k and 50k ?)
Output and feedback decoupling caps are 10uF Sprague polyesters which surprisingly don't sound too bad.
The unobtanium 3N172 can apparently be replaced by a zvp3310 (available even in SA :)) |
|
|
| Zen Mod |
| next part ,please ;) |
|
|
| analog_sa |
And finally there is the line amp. Exact same topology as the phono with a green 10uF ERO MKT polycarbonate cap (Threshold signature cap?) at output.
As i don't use active line amps at present i only compared this to my TVC and an ALPS pot. What can i say? I am probably not a line amp type of guy :)
But seriously, FET2 sounds quite decent, possibly because it's so simple. I wish i had these circuits in 1985... |
|
|
| analog_sa |
The diode-looking thingies are, of course CCS; couldn't find a more convincing-looking symbol in Protel. J505 - 1mA, J507 - 1.8mA, J511 - 4.7mA.
There is a muting relay in parallel to the output. It has a 10s delay on power-on and a connection to the power switch to instantly mute at power-off. |
|
|
| GaryB |
I had one of these about 20 years ago and it was my main preamp for a while until I got heavily into DIY and started replacing it with homebuilt equipment. I think my brother-in-law may still be using it.
My recollection was that the line stage was the best part of the preamp. I had a hard time bettering it with my homebrew line stages. The MM stage seemed easily bettered by my homebrew tube units but I never used it with MC cartridges and good quality step ups. Back in those days I believed that the Shure V15 was the best cartridge out there.
Thanks for drawing out the schematics. I may need to get a hold of this unit and see if it is suspeptible to some simple tweaks to get up to snuff.
---Gary |
|
|
| lineup |
| quote: | Originally posted by GaryB
Back in those days I believed that the Shure V15 was the best cartridge out there.
|
What :confused:
Isn't Shure the best I can get for my record player?
----------------------------
I like those small Keep It Simple circuits attached here.
Nice basic starting pre amplifier circuit setups
to try to modify and improve, yourself.
Thanks, analog_sa
:)
lineup |
|
|
| MashBill |
Thanks for doing this analog_sa!
FYI, my FET 2 (Series II) uses the Sprague 431P caps for the line stage as well as the phono stages.
I've been thinking about doing some mods to it. I have already replaced most of the electrolytic caps. I still need to change the 2000uF PS caps. (Large axial caps are hard to find.) I plan to switch out the Sprague 431Ps too.
Bill |
|
|
| vdi_nenna |
| quote: | | And finally there is the line amp. |
Where would you put the pot on this circuit on the input or output? |
|
|
| steenoe |
Thanks for posting, analog_sa.
Any chance to see some close-ups of the boards/ internal?
Steen:) |
|
|
| MashBill |
| quote: | Originally posted by steenoe
Thanks for posting, analog_sa.
Any chance to see some close-ups of the boards/ internal?
Steen:) |
Here's 1. |
|
|
| MashBill |
| And 5 (the last). |
|
|
| analog_sa |
| quote: | Originally posted by vdi_nenna
Where would you put the pot on this circuit on the input or output? |
Where would i put it? It is a 50k pot after all. I would certainly do as Mr Pass and put it at the input :) |
|
|
| analog_sa |
| Huh. My PCB is quite different - also a series II. |
|
|
| MashBill |
| quote: | Originally posted by analog_sa
Where would i put it? It is a 50k pot after all. I would certainly do as Mr Pass and put it at the input :) |
Hmm. My pot is a 25k Noble.... |
|
|
| vdi_nenna |
| quote: | | Where would i put it? It is a 50k pot after all. I would certainly do as Mr Pass and put it at the input |
That's why I asked because I thought he liked to attenuate all the noise from the amp at the output of the circuit. I'm just not sure if it applies to this type of ciruit. Is it "fielder's choice"? Sorry baseball term. Does it make a difference or just preference? |
|
|
| MashBill |
| quote: | Originally posted by analog_sa
Huh. My PCB is quite different - also a series II. |
That is different! My S/N is 185060365. Is mine newer or older than yours?
Bill |
|
|
| analog_sa |
Mine is clearly older - 184030014. Phono connections use much longer internal wires instead of pcb tracks. Funny thing is these wires and the outputs are silver in teflon and all other inputs use standard pvc copper. Mr Pass didn't like cd sound 20 years ago :)
Vol pot is Noble and may well be 25k - already put the lid back on. |
|
|
| steenoe |
MashBill and Analog_sa, thanks for posting the pictures. Nice:)
| quote: | | Does it make a difference or just preference? |
It does make a difference. When wiring the pot at the output you raise the output impedance of the preamp, throwing away some of your signal in the process. The pot is often wired at the output of a BoZ or BosoZ, but those preamps has enough output, so you can afford to throw some of it away. The advantage of wiring it at the output is, as you mentioned, that the circuit noise is attenuated when turning down the volume. Furthermore the pots at the output of the BoZ/BosoZ family is only 5-10K, which makes it a bit more acceptable. The "normal" thing is to wire the pot at the input of a preamp, though.
Steen:) |
|
|
| vdi_nenna |
All the parts are available at Mouser, except the 3N172. It's described as a P-Ch enhancement mode mosfet. Diagram shows 4 leads. Any substitutes?
The 2N5566 FETs are dual FETs. They are over $14 a piece. :att'n: |
|
|
| analog_sa |
| quote: | Originally posted by vdi_nenna
All the parts are available at Mouser, except the 3N172. It's described as a P-Ch enhancement mode mosfet. Diagram shows 4 leads. Any substitutes?
The 2N5566 FETs are dual FETs. They are over $14 a piece. :att'n: |
Earlier in the thread i mentioned easily available replacements - 2sk389 for the fets and zvp3310 for the mosfet. Both have been endorsed by Mr Pass in previous threads.
Btw, in post #1 i described the PS transformer as being 30VA. Wrong, it's just 15VA. |
|
|
| analog_sa |
It's interesting to compare the phono section to an Aleph Ono. It's very similar with some of the CCS being replaced with cascodes and the complementary source follower at output more capable of driving long cables and the scaled down riaa network.
Looking at the circuits, the FET 2 may well sound more transparent. In practice my Ono clone sounds better (with the exception of the bass which is deeper but not as tuneful) but it's a really unfair comparison with all the top grade components in the Ono.
A FET 2, built with nice parts and good regulators, preferably separate for each stage, might sound really nice. |
|
|
| vdi_nenna |
| quote: | | Earlier in the thread i mentioned easily available replacements - 2sk389 for the fets and zvp3310 for the mosfet. Both have been endorsed by Mr Pass in previous threads. |
I reread, but missed the replacement parts, sorry. :o
I think I'm going to try the line amp section. |
|
|
| wrenchone |
I've been doing some simulation on a similar circuit using 2SJ74 and ZVP3306, with current mirror load and tail current generator slaved to a master current source/reference. The big difference is that I cascoded the input differential stage to get rid of the Miller capacitance, as I'm looking at using the circuit in an RIAA stage.
Due to the cascode and the increased gm of the new devices over the originals, I find the new circuit to be pretty slippery and requiring lag compensation on the output stage to prevent HF rail to rail oscillation. The P-channel lateral output MOSFET used in the original circuit undoubtedly had a much lower transconductance than the newer vertical relacement devices. Both the 2SJ74 and 2SK389 have at least twice the transconductance of the 2N5566. A pair of matched PN4393s would probably be a closer replacement for the 2N5566.
I'm going to continue messing around with the simulation until I get some results that suit me. One thing that I found out right away is that it is necessary to very closely match the current source load current to the expected device current in the differential stage in order to minimize distortion.
A possible replacemet candidate for the original output MOSFET might be one of the new FETs from Advanced Linear Devices (available from Mouser). |
|
|
| analog_sa |
| quote: | Originally posted by wrenchone
Both the 2SJ74 and 2SK389 have at least twice the transconductance of the 2N5566. |
not quite. 2N5566 7.5-12.5mS 2sk389 8-20mS |
|
|
| wrenchone |
| I would assume transconductance on the high side for the 2SK389s unless one is deliberately using GR ranking and current starving them. I found some source degeneration helped to tame the oscillations somewhat, as the models I'm using for the 2SJ74 appear equivalent to approximately BL rankng. |
|
|
| Nelson Pass |
| quote: | Originally posted by wrenchone
The big difference is that I cascoded the input differential stage to get rid of the Miller capacitance, as I'm looking at using the circuit in an RIAA stage. |
One production version used IT505 cascoded matched JFETs.
:cool: |
|
|
| wrenchone |
IT505 - I haven't seen a part number like that in ages. Intersil, wasn't it? I used J174s for the cascode FETs in my simulation, as I have both models and devices in hand.
BTW, I was interested in doing the simulation (though not necessarily the real thing right away) to check out the open loop gain and closed loop distortion available from such a simple amp. Doing a feedback-style RIAA preamp asks a lot from the circuit in question. |
|
|
| AMV8 |
Hi
I was looking at the picture in post 16 and wondered if anyone could tell me what the grey box next to the transformer was ( I think it says Midland cross on it ). Is it a servo or a reed switch?
Don |
|
|
| analog_sa |
| It's a 24v DPDT relay shunting the output at power-up/power-down. |
|
|
| ANALOG GUY |
| Line-amp schematic only |
|
|
| DJG |
Hello,
I'm wondering if it would be possible to use 2sk170 or 2sk30 instead of 2n5566? |
|
|
| HKC |
| quote: | Originally posted by analog_sa
The diode-looking thingies are, of course CCS; couldn't find a more convincing-looking symbol in Protel. J505 - 1mA, J507 - 1.8mA, J511 - 4.7mA.
There is a muting relay in parallel to the output. It has a 10s delay on power-on and a connection to the power switch to instantly mute at power-off. |
Would anyone tell me where is the diode J507? I can not found it on any of the 3 circuits posted. There are only J505 & J511 diodes.
Thanks |
|
|
| HKC |
| quote: | Originally posted by HKC
Would anyone tell me where is the diode J507? I can not found it on any of the 3 circuits posted. There are only J505 & J511 diodes.
Thanks |
Sorry should be J511. There are only J505 & J507 on the circuits no J511. |
|
|
|
