I picked this one up from a factory in an industrial area - apparently, it had been sitting there as long as anyone can remember, and it was actually used to play music at one stage.
I know one thing for sure - they never opened it up to dust it.
PHILIPS EV4437 PA AMPLIFIER
- Likely mid to late sixties.
- ~100w, maybe 120w
- 6 x Philips 6CM5 valves (it was missing two of them)
- 12AX7 preamp & phase inverter
- Solid-state rectifier
It has the standard three knobs - Mic, Pickup and tone control. The Mic volume has a push/pull switch for bass cut, which just selects between a small and a larger value coupling cap. The tone control is a bass cut / treble cut, so something a little different there.
The 6CM5 valves are beam power tetrodes, which were actually designed for TV tubes - something about horizontal deflection, which I know nothing about... one tested OK, the rest are dead.
What I do know is that these appeared in quite a few Australian guitar amplifiers, likely due to price and availability. You can still find these at very reasonable prices online. I'm not a fan of the cap on the top of the valve - it has over 300v on it, and it's not insulated. Ouch.
2 x additional inputs had been added, all attached to the triode used for the mic input.
OUTPUT TRANSFORMER
- Primary impedance ~1.15k P-P
- Secondary impedances 25 Ω, 50 Ω & 100 Ω
- Use 50 – 100 for ~7.9 Ω
- Use 25 – 50 for ~4.7 Ω
There is also a 5w 470 ohm resistor in series with a 22n capacitor across the speaker output, which was not on the schematic below. I believe this is a Zobel network, which is used to dampen highs around the 15khz mark when using the 100 Ω output.
Calculations below:
Primary: 14.89 Vrms plate-to-plate at 440 Hz
Secondary:
0 – 25: 2.189 Vrms → Vp/Vs = 6.80 → (Vp/Vs)² = 46.3
0 – 50: 3.141 Vrms → Vp/Vs = 4.74 → (Vp/Vs)² = 22.5
0 – 100: 4.379 Vrms → Vp/Vs = 3.40 → (Vp/Vs)² = 11.6
Effective primary impedance:
With 25 Ω load on 0 – 25 = 1.16 kΩ p-p
With 50 Ω load on 0 – 50 = 1.12 kΩ p-p
With 100 Ω load on 0 – 100 = 1.16 kΩ p-p
50 – 100: Vs=1.235 → Vp/Vs = 12.057 → (Vp/Vs)² = 145.364 → Zs needed ≈ 1150/145.364 ≈ 7.9 Ω.
25 – 50: Vs=0.949 → Vp/Vs = 15.690 → (Vp/Vs)² = 246.182 → Zs needed ≈ 1150/246.182 ≈ 4.7 Ω.
6CM5 OPERATING CONDITIONS
Given the operating conditons list 3.5k p-p for a pair of valves at the same voltages,and I'm planning on running at lower power (not 6 output valves):
Pair of 6CM5: Use 16 Ω on 25–50 → ≈ 3.9 kΩ p-p
Quad of 6CM5: Use 8 Ω on 25–50 → ≈ 2.0 kΩ p-p
I think they put the wrong plate on this amp, as the 4417 is a slave amp from this series.
It's a fixed bias amp, but for some reason, there are 22ohm resistors to ground from the cathodes of the 6CM5 pentodes. People sometimes put 1ohm resistors here to measure bias, I’m not familiar with this size resistor being used for this purpose. A 10ohm would make a little more sense, as the mental math would be easier.
PHILIPS EV4437 SCHEMATIC
I haven't completely checked mine against this schematic yet - but there are definitely some differences. Some I know are mods, some may be factory.
The power supply is interesting, as it has a number of different feeds for the B+, screen voltages and bias. Screen voltages run at half of B+
CONVERSION PLANS / REPAIRS
Conversion went pretty well - a bit more work than usual, but the results are decent. I was getting some pretty nasty distortion in the cathodyne phase inverter, so I made some adjustments there. The preamp also had quite a lot of gain, which was not helping matters.
- Needed new output valves - only came with 4 out of 6, and 3 of them were toast.
- Filter caps were not great - enough hum to be concerning. Replaced with some JJ's, 2 x 100uf, 500v
- Removed extra inputs that were tacked on the preamp
- Rewired output transformer for normal speaker impedance
- Removed some extra resistors and grounding on the bias supply
- Replaced 47 ohm screen resistors
- Preamp - reduced 270k resistors down to 100k.
- Tone controls removed and modded to suit guitar (Fender one knob tone control).
- Added a power switch and replaced the giant orange power cord.
- Replaced 4amp fuse with 1amp.
- Dropped bias voltage down from about -30 to -25 with a 27k resistor to ground.
- Changed the cathodyne phase inverter - increased resistors and removed negative feedback.
FURTHER READING
Full evaluation by Tim Robbins - Dalmura site, although this is for the 4437a, which has some additional features
Aussie Guitar Gear Heads
Amplifiers with Valves - EV44 series
Don't know why I even took this photo - this is the shed in an industrial area of Brisbane where the amp was found.
No comments:
Post a Comment