The Quad 303 160W RMS per channel. Part two.
Last Thursday, on Valentine’s Day, I was in Antwerp, as Stefaan sent me an interesting e-mail (nothing to do with Valentine). A customer sent us a Quad 303/33/FM3 set for a check up, all bought from the man whose name should not be mentioned on this blog, (only this time).
Yes, the "Gazember" 160 W 303 was on our workbench. The pictures on Ebay reveal a 303 in a sorry state, in real life as well I can say now. The output capacitor banks are "glued" to the chassis, with wood glue, so they where rattling inside the cover of the 303 after transport.
The extra capacitors on top and beside the original power supply capacitors where held together with an elastic band.
The original trimmer pots where still in place after 40 years.
The solder work is below any standards.
So far for first impressions.
Technical review:
The regulated power supply has been removed, including the 40411 transistor.
The emitter resistors are halved in value by soldering a second pair over the existing ones.
The standard number of current limiter diodes is unchanged.
All the output transistors are replaced with MJ15003, which in itself is a good idea.
There are two rectifier bridges screwed to the chassis, with more extra diodes soldered on top of it.
All the wirering and the soldering looks unprofessional.
Measurements:
Instead of measuring with constant input voltage whe measured the maximum output before clipping, the way it should be done according to IEEE definitions.
If you let the amp clip until the level of square wave output, "almost" any RMS voltage is possible.
A 1 KHz test signal was applied.
The results:
16 Ohm 24W RMS both channels
8 Ohm 42W RMS both channels
4 Ohm 62W RMS both channels
Technical analysis.
By using the wrong (in our opinion) secondary windings from the transformer, combined with the voltage selector on 240V, you will get a power supply voltage of 70 V DC without load.
With two channels at full power, the voltage drops to 62 Volts DC.
This explains the lower performance with an 8 Ohm load compared to a standard Quad 303.
As a result of halving the emitter resistances the theoretical current limit doubled. This explains the result at 4 ohm load.
There may be 3 possible reasons why "Gazember" made the mistake to state the 303 had 2x 160 Watt RMS:
It could be possible that whe are looking at a "Monday-morning" product. So check those units before you buy them, if you like what you see, buy it.
Until it is clear under which conditions the amplifier should be measured in order to produce the promised results, and it has been checked by an independent lab, I have very strong reservations.
And there is still the issue of the "misinterpreted graph" (see my comments in the first article).
Joost Plugge
The original Joost Plugge text - not for publication - can be obtained via info@dadaelectronics.eu only. You will need Joost's agreement to publish parts of the text.
Stefaan
Last Thursday, on Valentine’s Day, I was in Antwerp, as Stefaan sent me an interesting e-mail (nothing to do with Valentine). A customer sent us a Quad 303/33/FM3 set for a check up, all bought from the man whose name should not be mentioned on this blog, (only this time).
Yes, the "Gazember" 160 W 303 was on our workbench. The pictures on Ebay reveal a 303 in a sorry state, in real life as well I can say now. The output capacitor banks are "glued" to the chassis, with wood glue, so they where rattling inside the cover of the 303 after transport.
The extra capacitors on top and beside the original power supply capacitors where held together with an elastic band.
The original trimmer pots where still in place after 40 years.
The solder work is below any standards.
So far for first impressions.
Technical review:
The regulated power supply has been removed, including the 40411 transistor.
The emitter resistors are halved in value by soldering a second pair over the existing ones.
The standard number of current limiter diodes is unchanged.
All the output transistors are replaced with MJ15003, which in itself is a good idea.
There are two rectifier bridges screwed to the chassis, with more extra diodes soldered on top of it.
All the wirering and the soldering looks unprofessional.
Measurements:
Instead of measuring with constant input voltage whe measured the maximum output before clipping, the way it should be done according to IEEE definitions.
If you let the amp clip until the level of square wave output, "almost" any RMS voltage is possible.
A 1 KHz test signal was applied.
The results:
16 Ohm 24W RMS both channels
8 Ohm 42W RMS both channels
4 Ohm 62W RMS both channels
Technical analysis.
By using the wrong (in our opinion) secondary windings from the transformer, combined with the voltage selector on 240V, you will get a power supply voltage of 70 V DC without load.
With two channels at full power, the voltage drops to 62 Volts DC.
This explains the lower performance with an 8 Ohm load compared to a standard Quad 303.
As a result of halving the emitter resistances the theoretical current limit doubled. This explains the result at 4 ohm load.
There may be 3 possible reasons why "Gazember" made the mistake to state the 303 had 2x 160 Watt RMS:
- He was confused between Ac Peak-to-Peak Voltage and RMS Voltage. Power into 4 Ohm = output voltage under load (both channels) Square / 4.
- Output-power is defined as power before "clipping" or before 1% distortion
- The power-supply has to be able to deliver the total power. In the case of a Push-Pull like the 303 this is 200% to be on the safe side. A 2x 160W would need a 640W PSU.
It could be possible that whe are looking at a "Monday-morning" product. So check those units before you buy them, if you like what you see, buy it.
Until it is clear under which conditions the amplifier should be measured in order to produce the promised results, and it has been checked by an independent lab, I have very strong reservations.
And there is still the issue of the "misinterpreted graph" (see my comments in the first article).
Joost Plugge
The original Joost Plugge text - not for publication - can be obtained via info@dadaelectronics.eu only. You will need Joost's agreement to publish parts of the text.
Stefaan
The current limit is doubled by halving the emitter resistance. The current limit diodes are the original ones. So the peak current (max and theoretical) will be 2.4V/0.3ohm = 8Amp. The RMS power in to 4 ohm's will be 8 x 8 x 4 / 2 = 128 Watt. With this current limit 160W is simply not possible! What shall we call this kind of presentation of the truth?
ReplyDeleteJoost Plugge
OK, so I take it you're not really that impressed by Gazember. :-)
ReplyDeleteSo what do you guys think of NetAudio?
Hi, Anonymous,
ReplyDeleteA comparison between Gazember and Net Audio is excluded.
Net Audio are professionals.
Stefaan