Hi Guys, I received my GA73 about a week ago. Since I have serviced several Neve 1073 modules and Neve consoles I was excited to open up the GA73 and have a look inside.
What I saw impressed me and I am sure it would also impress Rupert Neve himself with how closely the circuit follows the original drawings that I have from 1973 of the Neve 1073.
It also sounded really good out of the box and you would have to get up toward nearly $1200 a channel to get something that compares in sound and features I think.
The only difference I notice right off was that the GA73 used electrolytic coupling capacitors where Neve use tantalum but this would only increase the price about $15. The tantalums are a little more accurate and generally considered more “HiFi” than electrolytics.
The heart of the Neve 1073 and a good part of the sound is derived from the 23 position gain switch which varies in 5db steps from -20db to -80db. This switch just like the Neve has 3 sections that are controlled together. The first section, adds 5db of attenuation to the -20/-25 and -30 positions after the mic transformer but before the first gain stage. The second section increases the first gain stage in 5db steps of up to -50db. Here the 3rd section is engaged after the OFF position and the second stage’s gain is increased from -55 to -80db. This elegant gain staging maximizes the signal to noise and minimizes distortion.
This is the part of the circuit where other Neve clones usually fall short and use a pot and often a single gain stage yielding 65db of gain max. The repeatability of the input gain settings in 5db steps is a marvelous feature when doing dialogue replacement and going from a “whisper to a scream” and back again.
Because there is no pre-transformer attenuation like API, the Tridents “A” range, “B” range, WBS M460/M470 and other discrete preamps from the 70’s. The “sound” of the transformer is more prominent in the Neve “sound” especially from high SPL sources that drive in the input transformer hard. A kick drum or snare drum can produce -10dbu at the input to the mic transformer on peaks.
I have included a picture of the stock GA73 transformer on the left, the Neve/St Ives in the middle and the Neve/St Ives with the case removed on the right. The next two pictures compare the stock GA73 transformer laminations to the Neve/St Ives laminations. When I measure the ratio of the stock transformer I get 1:2 in Hi Z and 1:4 in Low Z. I also noticed the stock mic transformer was more susceptible to hum and noise interference than the St Ives. The Neve /St Ives VT22670 has a ratio of 1:2.75 in Hi and 1:5.25 in Lo and the bi-metal laminations are much more sophisticated.
I concluded that changing the input transformer would make the biggest “improvement” to the sound signature. When I replace the stock transformer with the VT22670 I noticed an extended clarity in the low end and in the lower midrange as this area seemed to punched through more like the Neve. It was really evident on high level signals reaching the input transformer as the bigger transformer with its more sophisticated bi-metal laminations was able to handle the transients and low end better.
Tom at Cinemag makes an excellent shielded mic transformer for under $50 that would work well in the GA73 and be able to handle the full output level of microphones without any attenuation. It was a bit of a squeeze to fit in the VT2260 and the Cinemag is slightly smaller.
The output circuit feeds 24 dc through the windings of the output transformer to the 2N3055 output transistor configured in a class “A” circuit. Because class “A” output stages are not very efficient this transistor like in the Neve is mounted on a serious heat shink. Feeding the power supply through the output transformer causes it to slowly “choke” a bit on very large transients when the output level is high. This is another part of the Neve sound. This output transformer has a ratio of 1:2 so it increases the output level 6db giving Neve its enviable headroom. I am not sure if replacing this output transformer will make a noticeable difference to sound. The output transformer in my experience would not be a critical to the sound than the input transformer. If any difference could be perceived it would be at very high SPL levels that would cause any A/D converter to sound horribly distorted. The best of the A/D converters can only handle and input level of +23dbu before clipping distortion. This transformer is a bit smaller than the original in a Neve but no smaller than those used in a API or MCI back in the discrete transformer coupled era.
Cheers, Dave Thomas
www.aamicrophones.com
