Proton AM-452: PS recap and upgrade

[Dr.Turner]

Just finished this nice Proton AM-452 recap/upgrade, following a detailed 2016 article on AudioReview italian magazine.
You can find a quite accurate English G-translation here.

..........

On top of the magazine suggestions I have replaced:

• all the electrolytic capacitors in the PS + bypass cap kapton tape protected
  
• four resistors, damaged from filtering cap leaking/glue mix

Since I didn't have a 820 Ohm resistor to replace R352 I serialized a couple of 390 and 430 Ohms.
The replaced 330 Ohm resistor was running quite hot (63° C) becase it proved less than 1/2 watt (chinese bulk buy from 10 years ago).
So I made a temporary heat sink from 1,8 mm solid copper wire: it was fine-hammered and flattened on the ends to follow resistor contour, bent and secured with heat shrink tubing. And it does the job lowering temps to 46° C.



I would like to share this empirical method also, to test package rating wattage which I used to check my resistors (360 mW instead of 500 mW).



Please find attached:

• a still readeable copy of the schematics (unfortunately it's impossible to find the SM online)
• cap relacement list

I didn't performed an extensive, full source test, but with Kenwood KT-990D tuner the improvement is audible to my 55 years concert-abused hearing also...



BTW the KT-990D is an outstanding tuner...slimmer than the big brother, EU only sold, KT-7020 but exactly the same circuitry and PCB, at half the price.
If you find one don't pass on it!

 

[Dr.Turner]

After several hours of playing everything is fine but I have some high temps on these four components:

• R519-520 (6.8KOhm) at 68-69° C
• Q509-510 (BC639) at 64-65° C

Do you think it's nomal or I should investigate further and maybe replace all four?
What does these components do in the circuit? I do not understand, but I would.

......

Thanks!

 

[Dr.Turner]

Follow-up after 4 years!
I'm trying to tie up loose ends and I'm back to this old thread to finally close it properly with measurements.

The open question from 2022 was about high temperatures on R519/520 (6.8kΩ) and Q509/510 (BC639) after extended listening. After some analysis I can now confirm this was not a fault — it was the components running at their correct design point for the first time, since the previously damaged PS resistors R351-354 had been depressing the auxiliary rails for years.
Regulated rail verification (idle):

• IC351 AN7818: input 26.1V → output +17.4V ✓
• IC352 AN7918T: input -26.6V → output -18.0V ✓

Rails healthy and sufficiently symmetric — the slight asymmetry on the positive rail is normal component drift on a 40-year-old regulator.


DC offset (no signal, idle):

• L: -6mV
• R: -7mV

Main PS rails: +31V / -31V (vs ±35V schematic — consistent with 230V EU mains and conservative transformer regulation)


Output power (1kHz, 7.8Ω dummy load, clipping onset):

• Vpp L = Vpp R = 48.8V
• Vrms = 17.25V
• P = 38.2W/ch

Significantly above the 25W/8Ω rated output — explained by the doubled filter capacitance (2200µF → 4700µF per rail) maintaining higher rail voltage under load, plus the conservative original rating.

Heatsink temperature at clipping: 49°C — well within limits.


Square wave response (6Vpp, 7.8Ω load):

• 94Hz, 1kHz, 10kHz: flat tops, clean vertical edges, no ringing, perfect L/R symmetry
• 20kHz: expected slew rate limiting, no oscillation

Estimated slew rate ~2-3 V/µs — typical for a discrete Class AB design of this era.



Verdict: fully restored and performing above spec. The recap + PS resistor replacement + filter cap upgrade was worth every bit of the effort.

---