Potentiometer replacement preamp Esotech PII Gradiente

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello all.
i was offered with a Gradiente PII Esotech preamp, not too mutch to read about online, but it was a part of a top of the line gradiente products, alone with his amp.

It has the central pin of the potentiometer volume broken, so i need to replace it.

I've read good thinks on Alps blue velvet, but my question comes here:
( i'm a kind of handy neophite)

I see just 6 pins on all 100K potentiometer, and the original seems to have more pins, i see 7 or 8 solder points on the board.....can someone enlight me about that?
can be the 7th point i see on the board not part of the potentiometer?
Or i'm looking for a wrong part?

Any suggestion is welcome.

Fabian
 

Attachments

  • IMG_2590.jpg
    IMG_2590.jpg
    678.5 KB · Views: 206
  • IMG_2591.jpg
    IMG_2591.jpg
    429.5 KB · Views: 197
  • IMG_2595.jpg
    IMG_2595.jpg
    835.7 KB · Views: 185
Last edited:
The one in your link is actually a standard stereo pot without the loudness taps. Zoom in and you'll see there are no leads on the extra 2 locations. Alps did make a version of the RH27 that had loudness taps but it will hard to distinguish real from fake on eBay if you do find one.

You can also make an external loudness tap by adding a few resistors in parallel with the pot terminals. I don't know the specifics off the top of my head but I have seen it done before.
 
hello all again, well after some month I'm reviving this thread because today I've discovered something interesting when I dismantle it again to start to resolve the problem.

Looking at the board where the pot is mounted, I see 2 zones, a free one marked as" VR102 Volume ALPS", and the used one as: "Volume Constanta VR102".
I'm neophyte on that but they seems to use this board to mount one or the other potentiometer...

If I bought an ALPS RK27, it will fit and works ok?
What's your thoughts?

Thanks again!
Craigtone, do you jump?
 

Attachments

  • plaqueta delante.jpg
    plaqueta delante.jpg
    1,001.3 KB · Views: 190
  • plaqueta detras.jpg
    plaqueta detras.jpg
    1,006.3 KB · Views: 116
I'm neophyte on that but they seems to use this board to mount one or the other potentiometer...
Looks like it indeed.
If I bought an ALPS RK27, it will fit and works ok?
What's your thoughts?
1. Measure footprint to make sure it does in fact match the RK27 one, although that seems likely.
2. Make sure the RK27 matches the front panel cutouts and volume knob.
3. The original pot is labeled 100KLIN, which suggests a linear law part. Analyze the circuit to make sure they didn't use the "fake log trick". That said, it is quite possible that this would have been combined with the loudness. Back in the 1950s or so, it would have been quite common for pots to have multiple taps so you could "log law" them externally. A pot without tap would definitely have to be a log one.
4. Do you happen to know someone with a metal workshop? The old pot may still be salvageable, I'm not saying it would be easy though. If the old shaft can be removed after disassembly, making a new one on the lathe would be a relatively trivial exercise. It might even be possible to stitch the old shaft back together by drilling a hole, tapping, and inserting a screw (a rather thin and long one obviously).
 
Hello Sgrossklas, thanks for your answer.
Here's the volume part of the diagram, can you discover something about loudness?

I see your point in trying to save the old pot, I like this kind of chalenges.
I guess I need to drill from the front the rivets, I'm right?
 

Attachments

  • IMG_2659 2.jpg
    IMG_2659 2.jpg
    928.5 KB · Views: 62
Hello Sgrossklas, thanks for your answer.
Here's the volume part of the diagram, can you discover something about loudness?
Found a schematic on HiFiEngine, fortunately.
attachment.php


Looks like with loudness off, there is a 15k resistor in parallel with the tap, as I suspected. They had provisions for no less than 3 different brands of pot (Noble, Constanta and Alps), with somewhat different values of components surrounding them. Somebody clearly considered second sourcing important.

Do you have any idea when these came out? The circuitry is ca. 1980, but the relatively modern transistors are throwing me off, so maybe closer to 1985?

It seems like quite a good, sensible construction, with much attention given to quiet supplies. Its weak spot may be the volume amplifier, which for several reasons isn't as low-noise as I'd like it to be (impedances are a bit on the high side, including a rather small 220n coupling cap). I'll have to simulate this.

I see your point in trying to save the old pot, I like this kind of chalenges.
I guess I need to drill from the front the rivets, I'm right?
Those usually need to be drilled out to dismantle a pot, yes. Fun.

Before you try that, I would first check the pot electrically. Not much of a point in trying to repair a pot with a broken carbon track (though that could be fixed, too...).

Doing this would have one advantage: You could clean everything properly, including sealing (leaving a thin protective film, like some contact cleaners do - people have even used vaseline for the purpose, but mineral oil based products have a habit of attacking plastics).
 

Attachments

  • gradiente-pii-vol.png
    gradiente-pii-vol.png
    45.4 KB · Views: 317
If you have any suggestion to make it sound better you are welcome!
That's not something I need to be told twice (literally)! ;)

With +/-36 V supply rails and the headroom that goes along with it, I wonder why they did not include a bit of input preamplification, like e.g. Grundig amplifiers of the time did. Maybe 6 dB there and take 6 dB away from the volume amp.

Oh.
The emitter follower after the volume amp (T113) is run very hot (~10 mA) and would have quite a lot of input current noise, and it's seeing up to 19.3 kOhms of source resistance. It's a BC548A/B, which means a beta of 110-220 or 220-450. For beta = 150, we would be seeing 67 µA of input current, for 300 it would be 33 µA. The lower bound for input current noise density is shot noise, at √(2 q Ib). That would mean between 4.6 and 3.3 pA/√(Hz). Multiplied by almost 20 kOhms and across a 20 kHz bandwidth, that potentially gets us around 10 µV of input/output noise - that's more than the entire volume amp! Granted, at high frequencies we may "only" be looking at 14.2 kOhms, but still, noise remains in the same ballpark.

This oopsie must have been an oversight. Note how hot the poor little BC548 is being run, too (~400 mW, near the limits for a regular TO-92). Ouch.

So we'll need a better replacement for T113 and its companion on the other channel before we can tackle the volume amp or gain distribution. Maybe a Darlington, a CFP or a (cascoded) JFET, and perhaps with a simple current source rather than the 3k3, or go Diamond Buffer altogether. I'll see what I can come up with. What do you have for small transistors and other components (and maybe perfboard)?

In the meantime, you could try measuring the DC voltage at T113 base, after R152 (or ahead of it - no big difference), and maybe determine whether R151 is a 56k or 68k (the schematic scan isn't of the greatest quality). With the given emitter voltage I'm getting either 20 or 24 µA as an estimate of Ib, but that still gives me an output noise voltage between 7.4 and 8 µV, only slightly less than the voltage amp. After decades of being fried, heaven knows what kind of shape the transistor is still in. (Running them warm keeps the moisture out, but hot would lead to premature degradation.) If both sides are still good, one could even recycle them.

I mean, you could always short R149 to eliminate most of this particular problem, but that would mean you'd lose the high filter functionality, and I'm always a bit hesistant when it comes to deleting features. I would have to simulate what prompted such a high current anyway, probably linearity - I bet 5 mA (or even 3 mA) would do just fine when going with a current source.

The headphone amplifier looks like a bit of an afterthought. BC548/558 as outputs (T108/109) - why? Poor little things. I would get some BC337-25 + BC327-25 for some more oomph, and a 10-100µ capacitor across D101+102 for improved linearity. I'm a bit surprised not to see a small capacitor (like 10 pF) across T107 B-C given the circuit's low closed-loop gain, perhaps C116 is doing that job already.

I've looked at the regulators again, and it struck me that with BC546 + TIP31B (or BC556 + TIP32B), they'd probably have just enough loop gain to get the job done. That said, they only have to be supplying about 100 mA anyway, at which point the TIPs would have a beta of 100+, leaving a decent amount of current for the BCs. Still, I'd prefer to see something closer to e.g. what the Sansui AU-717 has, which uses a Darlington pass transistor configuration and can thus afford to upgrade its RCR network to 10k/100µ/10k, giving higher loop gain (better regulation) and better ripple rejection due to less ripple on the zener current.
attachment.php

The Luxman C-120A used a current source to supply the zener.
attachment.php

You really want 3 transistors on a regulator like that.
Maybe one could replace R506 and R505 with simple current sources or CRDs to give ~3 mA while reducing (then less critical) R507/R503 to 1 kOhm.
Another idea would be bootstrapping the R506/7 (and R503/5) junction by connecting the 47µ cap to output rather than ground. Theory says bootstrapping increases input impedance at the price of increased output impedance, but in this case the increase in loop gain may well make up for it. Another one for the simulator.
And of course one could try following the AU-717 and hacking in additional transistors for a Darlington configuration with the TIPs, just not sure how feasible/easy that would be.

While the circuit is doing its best to suppress zener noise, one can also get much better low-noise zeners these days.

Anyway, maybe let's tackle the signal circuitry first.
 

Attachments

  • au717-reg.png
    au717-reg.png
    38.3 KB · Views: 3,232
  • l120a-reg.png
    l120a-reg.png
    27 KB · Views: 228
Last edited:
hello all again, well after some month I'm reviving this thread because today I've discovered something interesting when I dismantle it again to start to resolve the problem.

Looking at the board where the pot is mounted, I see 2 zones, a free one marked as" VR102 Volume ALPS", and the used one as: "Volume Constanta VR102".
I'm neophyte on that but they seems to use this board to mount one or the other potentiometer...

If I bought an ALPS RK27, it will fit and works ok?
What's your thoughts?

Thanks again!
Craigtone, do you jump?

It looks like the correct pin spacing for the Alps RK27. You would want one with the loudness tap however. I see them from time to time on eBay. It will have the 8 pins same as the PCB.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.