After blowing them up after a day of full volume...
Across nearly seven years of... " Oh F.F.S, when I'm I gonna get those on the bench and fixed ?! ".
Looking deeper into what Peter Walker had originally designed...
The goal:
Rock solid reliability, 20Hz to 20Khz. Match two 303's,
Utilising the best developments in componentry available to me.
Semi conductors were revised, updated where required, along with capacitors. Front panel connections updated.
I didn't get images of both amps complete, or the main heatsink transistor refit- too busy doing it!
Matching the resistors in all four modules was not completed, another time- along with a paint job and new front panel.
I would have probably gone for Takman carbons. Maybe Vishay Dale RN60D/65D metal films in the input driver stage.
Best manual download: search using: Quad-303-Illustrated-Service-Supplement.pdf
Transistor Outline types https://eesemi.com/to-types.htm
I've tried to find the best resolution, most up to date images of the old manuals. Numbering is on the image jpg:
After several iterations:
The input line sensitivity was increased to 1.5V the circuits design max, before interfering with feedback.
C100 increased to 1uF (.68uF limited the lower frequency range to 30Hz due to the result of the overall limitations of components of the 1970's),
lowering the frequency range handling. Using the highest quality PP Film cap available; MKP1839.
Increasing speaker output caps capacitance for similar reasons.
Looking closely to the Hfe of transistors, their gain structure throughout the input driver circuit and for low noise versions/ shielding.
These are in my opinion the best transistor replacements for the 303, old and new.
The (BCxxx) below are listed on the net for the 303, but IMO have too much gain or can't be easily sourced.
Presenting the new list of components for the 303 Power amplifier:
Quantity per PCB amp per module:
TR104 1 x BC556B T.O.92
TR102/ TR103 2 x BC546B T.O.92
TR101 1 x BC109 T.O.18; lower noise shielding / TR107 1 x BC184B (or K) T.O.92; better stability for TR107. 2 x BC184B fine also. (BC550C)
TR100 1 x BC560B T.O.92 Original design Hfe, 'A' versions available, worked fine (BC214C unobtainable)
TR106 1 x 2N5320 T.O.5 Metal shielding, better vibration resistance when heat sunk. Also MJE182G T.O.126
TR105 1 x 2N5322 T.O.5 Metal shielding, better vibration resistance when heat sunk. Also MJE172G T.O.126
The bias trimmer can be 1K along with a change of R132 to 2K4.
Those bourns 3296Y top adjust 25Turn trimmers are good, more stability.
Quantity per main heatsink:
All T.O.3
TR1 TR2 4 x MJ150003G. Legendary power transistors 250W; 20A per device! The originals are 117W.
TR3 1 x MJ21194G regulator
Quantity per PCB PSU module
TR200 1 x BC556 T.O.92
TR201 1 x MJE 182G 3A T.O.126. Or MJE243 4A. Not BC441
TR3 see heatsink
All diodes 4 x 1N4148
MR201 BZX79-C3V and BZX79-C9V1 connected in series, cancelling temperature gradients- voltage stabilisation results.
The PSU bridge rectifier array has the super quiet NXP BYV29X 600 diodes. They required new heatsinks.
Caps:
Cornell-Dubilier CD15 silver mica
Panasonic FM/ FR electrolytics
Wima MKS4 P.E.T
MKP1839 Polyprops
Electrolytic 'T' network capacitor
T-power BHC Aerovox ALN20S1108DF now Kemet. Rated to 100V,
Only one required, uprate to 10,000uF with noise suppressing MKS4,
On a single rail PSU connect the negative through both pins.
Speaker Output caps:
Kemet BHC slitfoils ALP20S1009DD 10,000µF 50 Vdc
Add noise suppression on cap pins, again MKS4.
---
PSU board increase the 6.8K to 10K
The only other genuine improvement that I didn't get around to was the use of two semiconductor current sources,
in place of R112 and R117 to improve noise and bias stability. Class D regulators i think.
If someone else figures that out, please post it here so there is a definitive list.
The 'before' images.
The 'after' images:
Words after the refit?
Alive. Engaging. Open. Seismic.
It's like a heavy cloth has been removed from the speakers.
Across nearly seven years of... " Oh F.F.S, when I'm I gonna get those on the bench and fixed ?! ".
Looking deeper into what Peter Walker had originally designed...
The goal:
Rock solid reliability, 20Hz to 20Khz. Match two 303's,
Utilising the best developments in componentry available to me.
Semi conductors were revised, updated where required, along with capacitors. Front panel connections updated.
I didn't get images of both amps complete, or the main heatsink transistor refit- too busy doing it!
Matching the resistors in all four modules was not completed, another time- along with a paint job and new front panel.
I would have probably gone for Takman carbons. Maybe Vishay Dale RN60D/65D metal films in the input driver stage.
Best manual download: search using: Quad-303-Illustrated-Service-Supplement.pdf
Transistor Outline types https://eesemi.com/to-types.htm
I've tried to find the best resolution, most up to date images of the old manuals. Numbering is on the image jpg:
After several iterations:
The input line sensitivity was increased to 1.5V the circuits design max, before interfering with feedback.
C100 increased to 1uF (.68uF limited the lower frequency range to 30Hz due to the result of the overall limitations of components of the 1970's),
lowering the frequency range handling. Using the highest quality PP Film cap available; MKP1839.
Increasing speaker output caps capacitance for similar reasons.
Looking closely to the Hfe of transistors, their gain structure throughout the input driver circuit and for low noise versions/ shielding.
These are in my opinion the best transistor replacements for the 303, old and new.
The (BCxxx) below are listed on the net for the 303, but IMO have too much gain or can't be easily sourced.
Presenting the new list of components for the 303 Power amplifier:
Quantity per PCB amp per module:
TR104 1 x BC556B T.O.92
TR102/ TR103 2 x BC546B T.O.92
TR101 1 x BC109 T.O.18; lower noise shielding / TR107 1 x BC184B (or K) T.O.92; better stability for TR107. 2 x BC184B fine also. (BC550C)
TR100 1 x BC560B T.O.92 Original design Hfe, 'A' versions available, worked fine (BC214C unobtainable)
TR106 1 x 2N5320 T.O.5 Metal shielding, better vibration resistance when heat sunk. Also MJE182G T.O.126
TR105 1 x 2N5322 T.O.5 Metal shielding, better vibration resistance when heat sunk. Also MJE172G T.O.126
The bias trimmer can be 1K along with a change of R132 to 2K4.
Those bourns 3296Y top adjust 25Turn trimmers are good, more stability.
Quantity per main heatsink:
All T.O.3
TR1 TR2 4 x MJ150003G. Legendary power transistors 250W; 20A per device! The originals are 117W.
TR3 1 x MJ21194G regulator
Quantity per PCB PSU module
TR200 1 x BC556 T.O.92
TR201 1 x MJE 182G 3A T.O.126. Or MJE243 4A. Not BC441
TR3 see heatsink
All diodes 4 x 1N4148
MR201 BZX79-C3V and BZX79-C9V1 connected in series, cancelling temperature gradients- voltage stabilisation results.
The PSU bridge rectifier array has the super quiet NXP BYV29X 600 diodes. They required new heatsinks.
Caps:
Cornell-Dubilier CD15 silver mica
Panasonic FM/ FR electrolytics
Wima MKS4 P.E.T
MKP1839 Polyprops
Electrolytic 'T' network capacitor
T-power BHC Aerovox ALN20S1108DF now Kemet. Rated to 100V,
Only one required, uprate to 10,000uF with noise suppressing MKS4,
On a single rail PSU connect the negative through both pins.
Speaker Output caps:
Kemet BHC slitfoils ALP20S1009DD 10,000µF 50 Vdc
Add noise suppression on cap pins, again MKS4.
---
PSU board increase the 6.8K to 10K
The only other genuine improvement that I didn't get around to was the use of two semiconductor current sources,
in place of R112 and R117 to improve noise and bias stability. Class D regulators i think.
If someone else figures that out, please post it here so there is a definitive list.
The 'before' images.
The 'after' images:
Words after the refit?
Alive. Engaging. Open. Seismic.
It's like a heavy cloth has been removed from the speakers.
Attachments
Last edited:
If you increase either of C100 or C104 you must increase both in the same ratio. Otherwise you are upsetting the two-pole LF HP filter.
Neat looking and sensitive upgrade. But nothing, other than the physical design of the 303, was due to Peter Walker.
Reading an interchange between Peter Baxandall and Douglas Self:
"I agree with your comments on triples. I was instrumental in introducing them to Quad,,,I had much to do with the 303 design. The input part pf its circuitry departed from my original scheme. and was due to Mike Albinson....The use of a regulated power supply was also my contribution, and had nothing whatever to do with any notion that it might be "sonically superior"! It was done purely because in 1967 we were all scared of busting the power transistors!"
So the overall electronic design of the amp is mainly due to Peter Baxandall, acting as a friend of Peter Walker, and as a paid consultant to Quad, and little to do with Quad staff.
I was lucky to have half a day talking to the late great Mike Albinson at Quad during the Summer shutdown in the mid 80's A good friend had joined Quad, and was responsible for the hybrid protection module in the 405, and he gave me a tour, "So what audio system do you have at home, Mike?" Expecting him to say this and that "No personal interest at all. Audio amps and FM radio design is the day job. I really like restoring old motor bikes in my spare time"
Reading an interchange between Peter Baxandall and Douglas Self:
"I agree with your comments on triples. I was instrumental in introducing them to Quad,,,I had much to do with the 303 design. The input part pf its circuitry departed from my original scheme. and was due to Mike Albinson....The use of a regulated power supply was also my contribution, and had nothing whatever to do with any notion that it might be "sonically superior"! It was done purely because in 1967 we were all scared of busting the power transistors!"
So the overall electronic design of the amp is mainly due to Peter Baxandall, acting as a friend of Peter Walker, and as a paid consultant to Quad, and little to do with Quad staff.
I was lucky to have half a day talking to the late great Mike Albinson at Quad during the Summer shutdown in the mid 80's A good friend had joined Quad, and was responsible for the hybrid protection module in the 405, and he gave me a tour, "So what audio system do you have at home, Mike?" Expecting him to say this and that "No personal interest at all. Audio amps and FM radio design is the day job. I really like restoring old motor bikes in my spare time"
Thanks guys, that means a lot to me.
It was chaos trying to write this up, all the info scattered between four files on paper and computer on a 'back burner' priority project over many years, where little was written down and project sharing component lists were unavoidable... with nothing more than a multi meter!
I had the pleasure of having a number of lengthy chats with Mr Bunting before he departed Quad, his knowledge was formidable.
He had valuable info about the progenitor to the 303's, the 50's amplifiers, i hope to upload when that project gets completed.
I love the case design, three of them can be fitted side by side to fit into the rack format for studios.
The caps values! I didn't assemble the list in full, will do when the chance presents itself, there were other subtle value adjustments.
The small electrolytics fitted on the amp modules are:
C 101 470uF 16V
C 106 100uF 25V
C 104 22uF 25V
It was chaos trying to write this up, all the info scattered between four files on paper and computer on a 'back burner' priority project over many years, where little was written down and project sharing component lists were unavoidable... with nothing more than a multi meter!
Interesting.
I had the pleasure of having a number of lengthy chats with Mr Bunting before he departed Quad, his knowledge was formidable.
He had valuable info about the progenitor to the 303's, the 50's amplifiers, i hope to upload when that project gets completed.
I love the case design, three of them can be fitted side by side to fit into the rack format for studios.
The caps values! I didn't assemble the list in full, will do when the chance presents itself, there were other subtle value adjustments.
The small electrolytics fitted on the amp modules are:
C 101 470uF 16V
C 106 100uF 25V
C 104 22uF 25V
Last edited:
Nice work. One question though - you have used smaller "star" heatclips on the drivers. These never seemed to me to have a low enough thermal resistance. Were the original TO-5 heatsinks not usable?
Err.. two questions. Did you experience any instability using the MJ15003's? These have a higher frequency response than the "good ole" (original) 2N3055's. I've built several 303 clones, all using modern 2N3055's but a slight instability is curable with some adjustments to the HF response. Which also had a bonus of increasing the frequency response to over 70kHz.
Err.. two questions. Did you experience any instability using the MJ15003's? These have a higher frequency response than the "good ole" (original) 2N3055's. I've built several 303 clones, all using modern 2N3055's but a slight instability is curable with some adjustments to the HF response. Which also had a bonus of increasing the frequency response to over 70kHz.
Q1 Those are twice the height that Dada use on their upgrades. They're convenient for other components too. I checked the W/m·K at the time, but what that is and where they were sourced, i have no record, this was a intermittent project that had years between sessions on the bench.
No problems so far, but its not been blasted at full for the better part of a day at the height of summer, yet!
Q2 No. I don't have scopes to confirm. They are a well known recommendation, there are those who reckon them to the one of the best transistors ever made.
The power supply now has greater reservoir and transient demand delivery, the head room is noticeable, better dynamics is an understatement.
There was a slight rise and fall after the bias was set, that would half cycle over 7 seconds.
The only other genuine improvement appeared to be the use of two semiconductor current sources in place of R112 and R117 to improve noise and bias stability. Class D regulators. I assembled an availability list but didn't get further than that, the prices put me off, being between £10 and £40 per device, eight required in all. If someone else figures that out, please post it here so there is a definitive list.
Is there someone out there or perhaps someone you know of who has a dead/ carcassed/ doorstop 303 amp that has the main heatsink intact?
Please let me know It's the only thing left to complete this refit, one spare main heatsink. I love to know of one thats genuinely available.
Please let me know It's the only thing left to complete this refit, one spare main heatsink. I love to know of one thats genuinely available.
Last edited:
Best I can find in ancient RCA data books is that the original integrated heatsinks were around 50C/W in free air. In a brutal sine wave signal, the base emitter junction of the power transistor will be around 0.68V, across the 68 ohm resistors R123/R127. So 100mA peak half cycle in the heatsunk transistors TR105/TR106, so 70mA RMS at a voltage maximum (assuming no voltage loss in transistors) of 67/2V = 2.3W. Divide by 2 because it is half cycles, so about 1.2W absolute maximum, or a junction temperature of about 60C. Taking transistor saturation etc into account it will probably be closer to 50C above ambient. For a sine wave signal. For regular music reproduction the junction temperature will be much lower.
The smaller DADA clip on ones are 63C/W, and the taller ones are 44C/W. Either of which, based on the original heatsinks, will work perfectly.
To put these temperatures into context, the maximum junction temperature for silicon is 150C.
The smaller DADA clip on ones are 63C/W, and the taller ones are 44C/W. Either of which, based on the original heatsinks, will work perfectly.
To put these temperatures into context, the maximum junction temperature for silicon is 150C.
Baxandall's role in the 33 and 303 design is also mentioned in an interview with Ross Walker (son of Peter Walker) on page 25 of Ken Kessler's book "The Closest Approach"
And again from the Ross Walker interview "Oh Peter did all the industrial design...including the 33/303, did it all and the office was littered with mock-ups of what it might look like. The 303 was easy because the Quad II was this shape, and we had a cabinet that it fitted right in"
It's not a question of whether I believe Baxandall or not. I believe every word of it. It's a question of whether 'nothing, other than the physical design of the 303, was due to Peter Walker.' can be substantiated from those quotations. I don't see it. Historical Method 101.
Peter's real strength was his work on loudspeakers. First the corner ribbon, the so-called ESL57 was a collaboration between PW and DTN Williamson, launched in 1957. The ESL63 development was started by PW in 1963 and finally launched in 1981, 18 years after PW started on it.
But electrical design, other than early valved/tubed gear, was done by others, certainly in the transistor era with the late greats Peter Baxandall and Mike Albinson. But since they, and Peter Walker have long shuffled off this mortal coil, we only have their writings and interviews left.
It is generally true that QUAD products had long gestation periods, both electrical and appearance, but once launched had long product lives.
I leave you to draw your own conclusions of how things were in Huntingdon UK.
But electrical design, other than early valved/tubed gear, was done by others, certainly in the transistor era with the late greats Peter Baxandall and Mike Albinson. But since they, and Peter Walker have long shuffled off this mortal coil, we only have their writings and interviews left.
It is generally true that QUAD products had long gestation periods, both electrical and appearance, but once launched had long product lives.
I leave you to draw your own conclusions of how things were in Huntingdon UK.
I was only querying whether the original heatsinks could have been re-used. The reason I ask is that they appear to have been constructed from two parts - probably aluminium- with the transistor clamped onto the base plate by the upper plate. Bolting a TO-5/TO-39 onto a metal plate is the best method of removing heat from those devices.
The original RCA heatsink which was attached to TO-5/39 devices was a rectangular folded sheet of steel, welded to the TO-can perimeter. That would have taken a bigger footprint than the sinks Quad used with fins over the top of other components.
I'm not sure what you are implying by the 0.68V of the power transistor- across 68 ohms that is 10mA to be added to the base current of the output devices. Base-emitter voltages vary considerably with current in power devices (greater than the theoretical 60mV/decade), so will the current in the 68 ohm resistor.
The driver power has to take into account the gain of the output transistor as that determines the demanded base current. The gain of the MJ15003 is higher than the old 2N3055 (25 at 5A compared with 20 at 4A) and taking 3.5A as the peak for 50W into 8 ohms that makes the base current around 140mA+the current in the 68 ohms, peak, in the worst case.
Those old RCA heatsinks were rated at 50C/W. The ones you mention are like the old Fischer SK510 at 44C/W. I tended to use the Fischer KK510 which were larger in diameter (20mm) star design which came in two parts held together by a spring clip. They were rated 38C/W, but the only To-5/39 heat sinks available now are typically those you mention or similar at up to 60C/W. Since device reliability is improved with lower temperature excursions, I recommend using lower thermal resistances where possible, thus the taller version is preferable not the shorter one, merely being one that "works perfectly".
And to be pedantic, silicon junctions of devices packaged in metal cans operated up to 200C. It was when most devices were shipped in plastic that the 150C limit became a standard.
The original RCA heatsink which was attached to TO-5/39 devices was a rectangular folded sheet of steel, welded to the TO-can perimeter. That would have taken a bigger footprint than the sinks Quad used with fins over the top of other components.
I'm not sure what you are implying by the 0.68V of the power transistor- across 68 ohms that is 10mA to be added to the base current of the output devices. Base-emitter voltages vary considerably with current in power devices (greater than the theoretical 60mV/decade), so will the current in the 68 ohm resistor.
The driver power has to take into account the gain of the output transistor as that determines the demanded base current. The gain of the MJ15003 is higher than the old 2N3055 (25 at 5A compared with 20 at 4A) and taking 3.5A as the peak for 50W into 8 ohms that makes the base current around 140mA+the current in the 68 ohms, peak, in the worst case.
Those old RCA heatsinks were rated at 50C/W. The ones you mention are like the old Fischer SK510 at 44C/W. I tended to use the Fischer KK510 which were larger in diameter (20mm) star design which came in two parts held together by a spring clip. They were rated 38C/W, but the only To-5/39 heat sinks available now are typically those you mention or similar at up to 60C/W. Since device reliability is improved with lower temperature excursions, I recommend using lower thermal resistances where possible, thus the taller version is preferable not the shorter one, merely being one that "works perfectly".
And to be pedantic, silicon junctions of devices packaged in metal cans operated up to 200C. It was when most devices were shipped in plastic that the 150C limit became a standard.
Yes. But they get in the way. The Fischer Elektronik 'star' heatsinks SKK 510 are an elegant solution.
Also Fischer elektronik heat sink goo has better thermal transfer properties than most.
I reduced the gain in the input driver circuit, thats more noticeable.
Originally fitted were: RCA 38494 T.O.3 + TR3 RCA16112, which when you turn it upside down reads as the modern replacement MJ2119 4G !