We return to the third part of the Filmosound projector amplifier conversion project-things suddenly changed unexpectedly...
Last time, I signed with Filmosound
, But it also screams and produces a lot of unnecessary noise. After doing some research, it found that this is a known problem with Filmosound conversion, because the grid line with the green signal protrudes from the chassis and is inserted into the top of the EF37A pentode.
After the amplifier is bolted to the metal projector rack, the cable will be effectively shielded. However, the independent arrangement allows the wires to pick up the interference and send it directly along the audio path to be amplified along with the guitar signal.
A few weeks before returning to this project, I received an email from Tony Teixeira of Tex Amps asking how the conversion would proceed. I explained the noise problem and he said I should remove the green wire connected to the EF37A grid and replace it with a shielded cable. He also suggested replacing the 33K resistor on the grid connector with a 47R resistor.
The letter followed Tony’s instructions, but the results were mixed. The amplifier is still noisy and the howling sound is the same as before, but the sound quality has been significantly improved. Before, it sounded clean and a bit dull, but now I have tasted how the normal Filmosound circuit is overloaded for the first time, which is very exciting.
However, I noticed that when the volume control is turned down, the amplifier produces almost no noise, so I think it’s time to sort things out around the input part of the amplifier.
Looking at the schematic, I found a bunch of special wiring around the jack. Since this is a projector amplifier after all, I cannot claim to understand everything that happens inside the circuit, but some dense resistors seem to be related to the projector bulb circuit.
There is also a mechanical linkage from the tip connector of the input jack to the switch of the oscillator section of the amplifier. Since this is redundant, I decided to delete the mechanical parts and all unnecessary resistors.
I also installed a modern amplifier input jack with a switch label, which connects the pin tip to ground when no plug is plugged in. A 1M resistor is welded between the needle tip and the sleeve connection in the usual way, and a resistor pin connects the sleeve to the switch label. Now that I can solder the shielded grid wire directly to the jack, it is time to test the amplifier again.
What happened next was a bit shocking-but thankfully, this was not a potentially fatal electric shock. Filmosound was very quiet until I pressed a chord and got one of the most harmonious overdrive sounds I have ever heard. Its definition is also very good and very powerful. If this is the whole purpose of Filmosounds, then no wonder they are so respected –
I can understand why some restorers prefer to keep the elements of the original circuit.
Compared with tweed, pitch control is more than just high pitch roll-off. Turning clockwise makes the sound brighter, but I also heard changes in the bass response-as the sound becomes brighter, the bass gradually weakens, and as the tone control is turned counterclockwise, the sound becomes fatter and not dull. I expected the stock circuit to be disappointing, so I resigned to destroy the entire amplifier and build something from scratch. However, I was so impressed by the news I heard so far that instead of rebuilding, I plan to optimize the existing circuit with some basic maintenance and minor modifications.
So far, I have been lucky because it is very easy to install and operate the amplifier. But now I need to check whether everything is within the range close to the specifications. Unlike Fenders, Marshalls and all other classic amplifiers, it is not easy to track the Filmosound 621 schematic. Finally, I found one
.
Many component values are difficult to read, but at least it gave me some voltage test points, and I was able to confirm that my amplifier was in very good condition. Even so, I still need to do something to maintain this state, and replacing the cathode bias capacitor on the 6V6s is the first task.
I decided to leave the existing, frankly inaccessible 50uF capacitor in place, and slightly change the wiring so that a modern replacement product can be placed next to the bias resistor. Electrolytic capacitors will leak and drift values, so this is a wise precaution. After finishing, I can check the bias of 6V6.
Although this is a cathode bias amplifier, the value of the cathode bias resistor may drift, and the power supply voltage from our wall socket may be different from the 1950s. Therefore, it is necessary to check the safe operation of those valuable power valves. To measure bias,
I used a method based on the primary winding resistance of the output transformer.
If you look at the schematic of any push-pull amplifier with two power valves, you will find that the primary side of the output transformer has three connections-two of them are connected to the valve plate, and the "center tap" wire is connected to the power supply. By measuring the resistance between the center tap connection and the plate connection, Ohm's law can be used to calculate the current consumption and plate dissipation of each valve.
As with everything related to valve electronics, this is potentially dangerous and you need to follow safety regulations. When connecting the multimeter, the amplifier must be disconnected from the main power supply and all residual voltage in the amplifier must be discharged.
Clamp the test probe of the multimeter to the rectifier label connected to the center tap of the transformer and the pole plate (pin 3) of the front 6V6, and then measure the resistance between them. I did the same with other 6V6, reading 240R and 221R respectively. Leaving the test probe in place, I set the multimeter to read the DC voltage and measure the voltage
Drop across the transformer winding.
Turn off the amplifier, consume the voltage, and then clamp the multimeter probe to another 6V6 board. Turn on the amplifier again, and now my voltage drop readings are 10.1V and 9V. Since the current (I) is equal to the voltage (V) divided by the resistance (R), I can deduce that the before and after 6V6s consume 42mA and 40.7mA respectively.
After closing and draining again, I clamped a test probe to the chassis ground and left the other probe on the valve plate to measure the plate voltage. I did the same with the other valve, and each valve got 343V. Multiplying the plate voltage by the current reading (in milliamperes), you can get the plate dissipation results of the 250R cathode bias resistor, which are 14.4W and 13.9W, respectively.
For the cathode-biased 6V6, I want to shoot at 12 watts or less, so the operating temperature of the 6V6 is too high. After some experiments, I chose 340R cathode resistance. This gave me a plate voltage of 352.5 and a voltage drop of 8.2V and 7.2V. Calculating these numbers, I get 12W and 11.5W plate dissipation readings, which I call "close enough to jazz".
The audio circuit is connected to the first cathode of the ECC35 double triode and the negative feedback loop. I had expected to cut off the negative feedback to make things lively, but this would make the tone control not work properly and no negative feedback.
Filmosound sounds a bit too familiar.
Even so, I suspect that the scope of control is still somewhat limited. Knowing how the value of the capacitor has tended to drift over the past few decades, I decided to swap out the tone circuit capacitor to make sure everything is working properly. This greatly improves the tone control function. I hear an extreme sound with very clear treble enhancement/bass attenuation, while the other end hears treble clipping and bass enhancement. Compared with the earlier Fender Tweed which had a separate tint control, the default position of the Filmosound control is near the center instead of near the maximum.
At lower volume settings, I still feel that some sparkling frequencies are missing. I have a simple solution for this. I am a big fan of tweeter capacitors because they allow you to extend the high frequency response of your amplifier in a non-intrusive and fully reversible way. Fender uses them in multiple amplifier models, usually with a bright switch.
Bright capacitance values can range from 50pF to 500pF, and they are connected across the input and output labels of the potentiometer. The smaller the capacitor, the higher and narrower the frequency range. The larger lid will extend the "treble bleed" to the upper and middle ends. When the volume control is turned down, high frequencies can "jump" in the pot unimpeded, which keeps everything bright and clear.
When the volume control is turned up, the resistance between the input and output of the potentiometer will decrease, so the influence of the upper limit will decrease. At maximum volume, the upper limit has no effect at all.
I use silver mica capacitors for this application and it is fun to try to find the value that best suits your pickup and sound taste. For this Filmosound, I chose 120pF. All that was left was to drill a hole near the power switch on the front of the case, and I just managed to squeeze it into an indicator light.
You may be wondering why I did not replace the multi-section filter capacitor in the power supply, because I always recommend you to use an old-fashioned amplifier. Well, first of all, the amplifier is running without noticeable power buzzing, so the capacitors in the container are clearly working. Secondly, I am familiar with this brand of Leak Hi-Fi amplifiers, and my pair of 1963 TL/12 Plus monolithic speakers are still running silently with their original filter covers. I'm probably lucky with this amplifier, because if it "dead on arrival", I certainly wouldn't like the prospect of debugging it.
I'm not surprised that most of the technicians who use Filmosounds simply understand the interior, but it's a shame, because the stock circuit sounds so good that all the Mods I plan now are not worth it.
The tone circuit is particularly interesting because it behaves more like a shroud control, which can lower the midrange or lower the bass to enhance compactness and brightness. Compared with tweed, when the tone control rolls back to push down, overdrive will increase rather than increase the brightest gain.
Filmosound combines the chewy fat of Fender tweed in the early 1950s with the extra gain and aggression of late 50s tweed, with the lower lows and smoother speeding of Gibson GA-40. It always sounds clear and bigger than life, and the way speeding responds to playing dynamics is definitely old school. I found the edge of clean tone and partial tone produced by this amplifier as exciting as the original metal overload. Those professionals who restore and convert these amplifiers emphasize the quality of transformers, and I can certainly understand why.
Most of them seem to be using Filmosounds with 5Y3 rectifiers, but I prefer 5Z4. It provides a soft start, and the HT is higher. As a result, the amplifier has more vivid dynamic effects, and has cleaner headroom and clarity. But be careful, because the metal-cased valve may become very hot, and if you want to experiment, you may need to change the bias resistance value together with the rectifier.
The premise of the project is to find out whether it is possible to get an excellent 1950s amplifier tone for less than £200. The clear and unambiguous answer is yes, in this case 180 pounds will be closer to the mark. All that remains is to choose the ideal speaker and decide whether to use Filmosound as a standalone speaker or build a combined speaker for it. Pay attention to this space, because you are likely to see this amplifier again in the near future.
It is the leading authority and resource in all guitar fields in the world. We provide insights and opinions on gears, artists, technology and the guitar industry for all genres and skill levels.
©2021 Guitar.com is a member of the media department