Philips 40/25/15 watts amplifier with Universal preamplifier circuit.

Hi John Ellis,

Nice to know someone is old enough to remember the Bailey. I built one in 1971 as a first year electronic student at varsity. I think the driver transformer had much to do with the characteristic smooth as a baby's bum sound. It was my pride and joy for a few years and was hooked onto a Pioneer LP45 and pair of KEF Cadenza, PYE tape recorder and Hamerstein valve FM/AM tuner.
 
Last edited:
Hi John,

Thanks for mentioning the Bailey amp. I found the article and it was a nice read.

BTW, the Philips 40W amp schematic I have shows a 10 nF cap (not 22 nF) at the base of the VAS.

Regarding the negative feedback taken from the speaker terminal (and not from the mid-point): perhaps that was an electronic adaptation of the "motional feedback" principle Philips tried in some of their speaker drivers in those days (I could be wrong here..)?

cheers,

Reji
 
Hi Nico
I did not mean that Bailey! - I was referring to the later complementary one with 40361 input, 40362 VAS, 40361/2 drivers and MJ481/MJ491 outputs.
Hi Reji
The motional feedback used a transducer as far as I recall attached to the speaker. The reason most designs of that era had feedback from the output side of the capacitor was generally to extend the frequency response to compensate a poor design or too small a capacitor or capacitor distortion, or all, take your pick!
YOu could only extend the frequency response at small signals - the maximum output drive decreases with the degree of feedback.
Even a 10 nF capacitor will decrease the performance of a circuit. I'd limit this capacitor to 1nF at most if it is needed, at least to push the roll-off it causes to above audio band. With my proposed compensation a capacitor is not required across the base of the VAS at all (and nor was one used in the Bailey)



John
 
hello there i am new member and i have made this amp too, the actual transistors were BD182.those who are using 2n3055,they will not get good sound,please change them by 2n3773.i made the amp by giving it 60VOLTSsupply, the main error is in this amp is that the flow of high gain frequency,if it is not shielded properly.so use this amp circuit what ever is given in that circuit and the main transistors is to be mount on large heat sink.
 
Usualfellow
The BD182 was originally a "single diffused" transistor made by Mullard/Philips and would have been very similar to the old RCA 2N3055, so I doubt that anyone using an original 2N3055 would notice the difference.
The new 2N3055's are epi base and faster so may have oscillation problems the old devices did not. Depending on who made the BD182's you used might mean the difference between the performance (oscillation or not).
Oscillation can be stopped using additional capacitors very often. It depends what is causing the oscillation.
I would not recommend the 2N3773. The original was even worse than the original 2N3055 (200 kHz) but again , epi base types today may well be faster, but it depends who makes it...
Try modern devices like the MJ21193/21194 pair which really are very good.
John
 
Look this :
 

Attachments

  • ScanImage001.jpg
    ScanImage001.jpg
    63.1 KB · Views: 1,502
  • ScanImage002.jpg
    ScanImage002.jpg
    159.3 KB · Views: 1,740
  • ScanImage003.jpg
    ScanImage003.jpg
    129.9 KB · Views: 1,468
Hi

This circuit isn't the best. I have simulated it using my 2N3055 model (ft~1 MHz) and it does give ~95 kHz band width. But the distortion at 10 kHz and 35W is 0.47%, not 0.1 as advertised.

Mostly this is crossover distortion, and seems to be caused by the Miller and "slug" capacitors C7 and C6. ALso the feedback from the loudspeaker may reduce capacitor distortion but causes clipping at low frequencies, so bass will be poor (differentiated) at high power outputs. Good capacitors today with low ESR and ESL should be OK.

At 95 kHz the capacitor C9 draws 3.6 mA at 40W. If you use slow BD182's or 2N3055's the gain will be falling to around 10 and need 400 mA base drive. The gain of the BD139/BD140s drops to about 30 so the base current needs to be 13 mA, requiring a total current in the VAS of about 15 mA. Since the VAS should remain linear, the current needs to be more like 25-30 mA to meet spec.

If you use faster 2N3055's -or a 2N3055/MJ2955 pair - then the gain will be a little higher at 95kHz and maybe need only 7 or 8 mA, suggesting that the VAS current could be ~15 mA. Conveniently this can be provided by two 1k load resistors, but the VAS will need a small heatsink.

Eliminating the loudspeaker feedback resistor means that the overall feedback resistor should be changed. It could be 1.5k with a 47 ohm "grounding" resistor.

A better bias arrangement for the input PNP is to divide the supply voltage into approximately half (less a little bit for the Vbe offset) using, say, a 27k and 22k resistor pair and connecting the base through a 10 to 22k resistor. The bias tap should be decoupled with a capacitor, but this only needs to be a 40V unit instead of 64 or 100V when connected as shown using high value bias resistors. This alternative bias arrangement seems to reduce low frequency noise.

Finally I would remove the Miller and shunt capacitors C6 and C7. Replace the resistor R14 with a wire and add a 220 pF global compensation capacitor across the feedback resistor.

Finally re-route the upper output transistor 56 ohm resistor between its emitter and base to equalise the driver currents.

Simulated distortion is now 0.1% at 10 kHz and 30W output. 2nd harmonic is 0.1% and third around 0.05% with reducing crossover components below this.

The main distortion component is 2nd harmonic which is due to Early voltage distortion in the VAS. Use a cascode to more or less eliminate this and the distortion reduces to 0.03% -as it should have been for HiFi.

A small improvement can be achieved using a complementary pair with a single base bias resistor between the two bases.

John
 
Hi
A large capacitor in the output should help with low frequency performance. If you are noticing some distortion it must be coming from something else, such as mains hum; oscillation (do the transistors get hot?); crossover distortion (is the quiescent current right?); or clipping if the mid-rail voltage is wrong.

John
 
Philips 40w Amplifier with universal preamplifier

I have done what exactly stated at the book by M.D. Hull both preamplifier and Amplifier ckt but there is motorboting sound in no signal condition please give me a remedy.......
Is there any problem with BD139 & BD 140 matching hfe?????????
I use 2N3055 as output
 
It's very probably a grounding problem, you'll have to be careful on how and where you ground different parts of the amp.
One general purpose solution is using separate ground return wires, at least for the speaker, main amp and preamp, all going to the PSU main capacitor negative.
It's more complex than that, but at list this is a first step.
 
Philips 40w Amplifier with universal preamplifier

I also make an arrangement for separate grounding but the problem persist.
Finaly i notice that if i replaced the capacitor C7= 27 pf by a 180pf capacitor then there is no motorboting sound but there is a broken BASS problem in FULL BASS condition.
 

Attachments

  • PwrSchemeDIT.JPG
    PwrSchemeDIT.JPG
    139.2 KB · Views: 1,629