Upgrading Counterpoint SA-7.1:
I bought recently little cute preamplifier Counterpoint SA-7.1. Need to say, Michael Elliott did a great job with this preamp.
- First, as he admitted in his website AltaVista.com, he designed it with “original, troublesome cathode-follower (it likes to whistle)”, “phono stage that is always connected to the line stage, so its noise is always in the background, and it presents a difficult load for your line-output components to drive” and some other questionable design decisions.
- Second, he went into multiple modifications, like “originally had two 12AY7's in the line stage. Later versions used a two 6DJ8's in the line stage to overcome a tendency of the older tube to "whistle." All versions (12AY7 line stage and the first versions of the 6DJ8 design) have a gain stage followed by a cathode-follower. Later units have a MOSFET source-follower after the 6DJ8 gain stage” without any reflection in the model number/version/release. Now he charges up to $100 just to provide you with schematics to the particular model you have.
- Third, now he offers to fix the mess he created through the upgrades for additional amount from $1000 to $3000.
You don’t like it? So what, it’s a successful business model
implemented by Microsoft with Windows operating system, and we all live with
it. But despite all that, you have what you have: four tubes,
power transformer, and decent hardware to host it, so this preamp is a perfect
candidate for upgrade/modification. I didn’t have a schematics for
preamp, and feel like embarrassment to pay AltaVista $100 for copy, so I
started with schematics posted here:
and later restored mine from PCB and components on it.
First group, common recommendations:
1. Replace all resistors to the better ones, like Holcos,
Caddocks, Holcos or Vishays.
2. Replace all capacitors with the better ones, like Musicap, Aircaps and likes.
3. Replace all tubes and rectifier diodes.
4. Replace rectifier diodes with FREDs (probably Hexfreds diodes, HFA04TB60PBF-ND diode hexfred 600V 4A TO220AC from Digikey, or Schottky diodes up to 1200V 6A for plate and 600V 4A for Bias).
5. Bypass each of the electrolytic with some film capacitor.
6. Add Jensen paper in oil into the Power Supply
Second group, specific to this model suggestions:
7: Get rid of “bad” line stage follower
8: Change input selection, disconnect phono stage from line stage
9: Get rid (bypass) of balance pot
10. Replaced the volume control with a better one (TKD makes excellent sounding volume and balance controls.)
11. Remove Line Stage Input Capacitors
12. Change the RIAA Equalization Curve
Final. Get rid of this preamp and buy a better one, b/c implementation of 1 – 12 is equivalent to building a whole new preamp.
My intention is to try all or most of them.
a. flat frequency response at least up to 50 KHz
b. noise level and composition
c. THD, transient response, etc.
For evaluation and testing of preamp I need reliable repetable sources, typical would be:
- CD with test tones (Rotel CD player),
- square and sine wave generator,
- and pink noise generator.
Sourse connected via splitter to preamp A in question (SA-7.1) and reference preamp B (ARC SP-6B, or SS like Adcom GFP-555).
A-B comparison done by switching outputs from preamps using AudioSource 10-band Spectrum Analyzer.
Power amplification is done by tube ARC D-76 power amp. Speakers are Thiel 2.0.
Listening my favorite reference LPs, CDs and test tones showed that that my preamp has a big rollover in frequencies greater than 12 KHz in the line stage. Other than that it worked acceptable. Phono stage had a reasonable noise with some excessive component at 60 Hz. Both reference preamps had flat noise spectrum.
Original Counterpoint already has Wondercaps, which some considered as an upgrade recommendation. No needs to upgrade in the first approach.
The original resistors looks like Dale (same as Vishays), these should be fine.
Problems and issues:
After inspection I found that my SA-7.1 rev. B has two Zener diodes 1N5368B (47 VDC cut off) in series, cathode to cathode, in parallel with 35 pF capacitor connected between grid and cathode in the follower stage.
The purpose of zeners is to protect the grid of V4 from positive voltage during the warm up. After first minute I measured a negative 20 V between cathode and grid, so they shouldn’t affect sound in any way. As for the capacitor, my guess is that capacitor was added to fix “troublesome cathode-follower (it likes to whistle)” how Michael Elliott put it. And this capacitor shunting the grid is partially responsible for the HF roll over.
1. I removed those capacitors shunting the grid.
2. Power supply. Replaced electrolytic and bypassed with high quality film capacitors
3. PCB had a provision for bypassing PS electrolytics and add additional electrolytics.
Even more bypassing and additional electrolytics in the phone stage filters.
4. Mechanical upgrades. Circuit Board damping in critical points using thick wool felt damping pads between the circuit boards and the chassis to damp all (possible?) resonances.
5. Must to do. New golden plated RCA connectors in the critical paths Phono in and Line out.
What do we have in line stage? Very basic (classic) topology with a single gain stage direct coupled to a cathode follower. What is wrong with it? RCA used it, Marantz used it, McIntosh used it, ARC used it, and maybe I can keep it also.
What I would actually like to do is to try a little bit of negative feedback in the phono stage to kill excessive gain and improve THD. It looks like that excessive gain was designed to support both MM and MC cartridges. As a result when using MM you have to keep volume on 9 AM position unless you want to blow your speakers. Simple swith to toggle gain can be used as the actual MM/MC switch.
- Remove the Mute Switch from the Circuit
- Remove the Balance, then Volume Control from the Direct Signal Path
Those ideas do not excite me in any way.
Options 3 -4 make sense, but implementation would require effors comparable to building a new preamp.
A. Tubes replacement:
I. Line stage – original 6DJ8, Sovtek 6922/E88CC, or Russian 6922 / 6H23n-EB.
II. Phono stage - 12AX7 of all sources.