You really made me understand more about diodes. I appreciate the way you went through that unit and described everything in such layman terms, that even I can understand to a certain degree. Thank you.
Not to be pedantic but an important distinction for those that are interested; St Ives actually made the original transformers at the same time as Marinair, Neve had a rule to never take parts from a single manufacturer. So they had St Ives (now Carnhill) make an exact copy of the Marinair which was actually an off the shelf product. Marinair then disappeared and eventually St ives stopped making the transformer as well. However it seems after much demand both companies have been employed to start making these transformers again. Many suggest neither are exactly the same, while all these gear companys like to suggest that their product are the only true exact replica, so you have to take it all with a pinch of salt. Mark from BAE explained that Carnhill (St Ives) are still using the exact machines that they used back in the day so are exact replicas. Either way it seems its as close as you can get without a time machine.
Thanks so much for doing these under the hood break downs - I find this a far easier way to learn about circuits than trying to read the schematics first and then look at the real thing... this way for some reason makes looking at the schematics of a circuit less confusing. I really look forward to more of these videos!
Interestingly, the Yamaha GX-1 uses similar diode bridges in its voltage controlled filters and some of its early monosynths, as described in U.S. Patent 4039980. It also appears in some early Korg synthesizers like the 700, 770, and MS-50.
I feel like I learn something new every time you do one of these videos. Also thank you for doing this unit! Finally someone explaining exactly what a diod comp does. On a side note, it seems many viewers admittedly say most of this flies over their head yet find it fascinating enjoy the content. You should consider making a video on teaching the very basics of electronics and schematics. Not just showing a diagram but even pointing out the parts on a on pcb board.
Great vid many thanks - i think both the RFZ 713 and later the siemens sitral U 273 were diode bridge compressor/limiters which came out before the Neve 2254 ( predecessor to this). I have used all of them ( including the neve ) and they sound great and very different . Also worth mentioning is the WSW 601430 which is also diode bridge and would give a Fairchild 660 /670 a run for its money no problem.
So where the neve 8068 console I owned. Few of them. All with 33102 module was it class AB or A. I can’t remember. Thanks man. Best sound console I owned.
Great video! Correct me if i’m wrong but does the transformer feeding into the diode bridge GR circuit need to be impedance matched so that the voltage level across all frequencies stay within the forward voltage limit? Also, how does it work when the signal is too low? Or if the incoming imepedance is too high? Thanks for the teardown! Looking forward! Would be great to see a teardown of the REALLY NICE COMPRESSOR
Great questions. I'm not positive what you mean by impedance matched, but any good transformer in a well-designed circuit will be basically flat in the audio spectrum. That is, it will step up/down all frequencies by the same ratio. So as far as I know there is nothing too special going on with the input transformer. If the input signal is too low you turn down the threshold ;) If the incoming impedance is too high (like if you connected a guitar pickup to the input) you would get lots of loss, distortion, and filtering. But this is universal to any device, not unique to the diode compressor. Thanks for the kind words and questions!
@@PetersonGoodwyn Thanks so much for the reply. What I meant by impedance matched is that, from what I understand, the incoming signal's impedance would have to be lower than or equal to the "impedance on the diode" caused by the "voltage divider" (if that makes any sense?). As the greater "voltage divider in this case would also be the transformer against the GR circuit. Just wanted to clarify if this was true or not, and if so, how would one go about determining the range of values applicable. And if not, I would gladly shove this question up my rear end lol.
Gotcha, thanks for clarifying. The short answer is that no, there is no line-level device that would have a high enough output impedance to affect the voltage divider. The longer answer is that output impedance is like R1 (series) in a voltage divider. So the only way the output impedance of the previous device could affect the GR circuit is if it added significantly to the R1s in this circuit, which are 47k according to the schematic. Most line-level devices have an output impedance around 50 Ohms, while some older ones go up to 600 Ohms. On top of that, the step-down input transformer reduces the output impedance going into the GR circuit even further. Does that all make sense? Great question!
Thank you so much for this video. Believe ut or not but it really helps for sobering like me who never studies electronics to grasp some sense in all this
What a great understanding. This give confidence in your gear built. Ok so I'd like a 8 channels summing-stereo buss compressor with 1073 amp section, CL 1B tube compression, an API continuous sidechain / HP filter knob, LA2A limiter section, blue & red silk + tape or carnhill transformer, 2 mono channel out, M/S - stereo switch, stereo width just before the limiter, all that in an SPL 120v audio rail fashion. Attack, release, treshold, Main+ 2 monitors out, with a 4 band parametric passive EQ. And all that good for Max 1600$ 😁🤭
Very insightful information. The UAD 33609 is very nice but the hardware has a more spacial sound and its clearer and bigger. This is the best diode-bridge compressor out there. The unit has a limiter after the compressor so you can overdrive the transformers for more character while the limiter keeps the signal from overloading and sending out a too hot signal into your DAW.
Nice vid, Peterson - great content :) I visited Neve's factory recently to check out the 8424, and gather that their xformers are no longer made by Carnhill... not sure how recent that is, though, so that 2018 unit may or may not have Carnhills-made ones!
Do you have any idea about why the signal is kept balanced, meaning that you have to use two perfectly matched diodes instead of one ? Could it be because the voltage has to be kept under 0.7V so there is a risk of interference and the balanced signal keeps you away from this ? And same question about having the signal unbalanced/amplified then balanced again ? Is it because the BA340 can't be done with balanced input ? Thanks for the quality of your video, as always ! :)
Very insightful questions! Balancing the gain reduction stage has a couple of benefits. It allows the DC control voltage to easily be removed from the signal at the next, unbalancing stage. And it cancels out even-order harmonics in the same way that a balanced input cancels noise picked up by an XLR cable. Rod Elliott has a great explanation of it here: sound-au.com/articles/vca-techniques.html#s9 Unbalancing, amplifying, and then balancing again is the typical approach because a) then you only need one amplifier instead of two! and b) most non-linearities in the amplifiers will not be common-mode, so they won't be cancelled out by the next balanced receiver. So you can actually introduce more noise and distortion by trying to amplify the signal while balanced.
@@PetersonGoodwyn What a perfectly clear answer, thanks ! So this is a compressor using a voltage divider circuit controlled by diodes, just as the 1176 is controlled by FETs and LA2A is controlled via a photoresistance. If I'm correct, a compressor like the SSL G-Bus uses VCA to control gain. Are there others types of compressors apart from VCA or voltage divider ones ?
DIY RE do you think there is more to the sound of a transformer than it’s ratio and sheet type? Or do you think there is more shaping the sound than what science can measure? In the end music is very simple and small band of ac signal and we should - and I think we do- know everything about it. I mean in other electronics disciplines they work with GHz and THz and don’t mess around with different manufacturers for transformers.
As I've finally learned from these exceptional talks, there's no inconsistency between deep technical insight and empathy for idiosyncratic perceptual discernment. He joyously dissects his own eager bias for the warm sound of the Neve *design language* (or does he really love cool and punchy API?)! I recall Rupert Neve touting "stoved enamel"- stoic and warm!
yep, switch-mode power supplies have come a long way and are now standard in high-end gear. to be honest, I'm so used to seeing them I didn't even think to comment on it!
1. when you say the transformers weren’t quite what they were, what does that mean? what is it about these particular transformers that are “authentic?” i tend to assume that kind of cork sniffing about things like magnets and coils of copper wire is just confirmation bias because of the scarcity of the older thing. 2. when you say that the circuit is improved by careful selection of components, what does that selection involve? is it reading lots of data sheets and picking just the right component or is it more bespoke than that like checking component values to make sure that values within tolerances are closely matched to the design spec?
It depends on if a particular type of component is ESD sensitive. Most used in vintage gear are not ESD sensitive (other than transistors). The ICs in there are almost surely ESD sensitive. Surface mount resistors could be, just depends on build. Most caps of decently high capacitance value won’t be. Transformers and metal housings are not ESD sensitive. And as soon as he touched the housing of the rack chassis, he would have most likely safely discharged any charge buildup he had accumulated.
If I just can buy one. 33609 or MBP Portico Ii. I have used both on Analog Access, great results, but the latency and offline process, make the work flow not fun.
Some what of a myth with regards to the transformers, but hey myths sell. Reality is this, the original input transformer is an off the shelf part from Mariner and from time to time it would be sourced out to different companies, such as St Ives, when Mariner couldn't keep up with demand. Now the original output transformer was a different story, it was a custom design by neve and was only ever made by Mariner and one other company, which was not St Ives. Since the 60's people have come and gone and neve has been though multiple hands, not just siemens group, and over that time the original drafts for the output transformers have been long lost. As a consequence the output had to be reversed engineered, which happened decades ago now, and is the real reason people say they don't sound the same, there pretty close though but hey like said myths sell.
You really made me understand more about diodes. I appreciate the way you went through that unit and described everything in such layman terms, that even I can understand to a certain degree. Thank you.
Not to be pedantic but an important distinction for those that are interested; St Ives actually made the original transformers at the same time as Marinair, Neve had a rule to never take parts from a single manufacturer. So they had St Ives (now Carnhill) make an exact copy of the Marinair which was actually an off the shelf product. Marinair then disappeared and eventually St ives stopped making the transformer as well. However it seems after much demand both companies have been employed to start making these transformers again. Many suggest neither are exactly the same, while all these gear companys like to suggest that their product are the only true exact replica, so you have to take it all with a pinch of salt. Mark from BAE explained that Carnhill (St Ives) are still using the exact machines that they used back in the day so are exact replicas. Either way it seems its as close as you can get without a time machine.
Thanks so much for doing these under the hood break downs - I find this a far easier way to learn about circuits than trying to read the schematics first and then look at the real thing... this way for some reason makes looking at the schematics of a circuit less confusing. I really look forward to more of these videos!
I'm glad you find them helpful!
That's a beast of a compressor! And you're a beast for doing this, cant thank you enough!
Thanks very much, appreciate it.
I like your style of narration. You get right into it, without a lot of fluff! So I liked, and subscribed!😊
Great!Finally here's a guy gave a deep explanation of diode compressor.Thanks!
Glad it was helpful!
Absolutely loving these teardown videos, thanks for doing them!!!
Glad you like them!
What a brilliant explanation of diode bridge compression. Thanks for the video!
Great explanation for diode compressors
One of the best explanations, full stop!
Glad you think so!
Thank you Peterson for another mind opening video.
You’re an amazing teacher.
Thanks so much Izhar I appreciate it.
Amazing video. Love the detail and technical breakdown. Liked and Subscribed!
Awesome, thank you!
Interestingly, the Yamaha GX-1 uses similar diode bridges in its voltage controlled filters and some of its early monosynths, as described in U.S. Patent 4039980. It also appears in some early Korg synthesizers like the 700, 770, and MS-50.
Excellent video! The designer of the 33609 is Geoff Tanner.
I feel like I learn something new every time you do one of these videos. Also thank you for doing this unit! Finally someone explaining exactly what a diod comp does.
On a side note, it seems many viewers admittedly say most of this flies over their head yet find it fascinating enjoy the content. You should consider making a video on teaching the very basics of electronics and schematics. Not just showing a diagram but even pointing out the parts on a on pcb board.
Amazing video man!
Appreciate it!
This was incredibly cool! 👍
Thanks glad you liked!
Great vid many thanks - i think both the RFZ 713 and later the siemens sitral U 273 were diode bridge compressor/limiters which came out before the Neve 2254 ( predecessor to this). I have used all of them ( including the neve ) and they sound great and very different . Also worth mentioning is the WSW 601430 which is also diode bridge and would give a Fairchild 660
/670 a run for its money no problem.
So where the neve 8068 console I owned. Few of them. All with 33102 module was it class AB or A. I can’t remember. Thanks man. Best sound console I owned.
excellent video.
Great video! Correct me if i’m wrong but does the transformer feeding into the diode bridge GR circuit need to be impedance matched so that the voltage level across all frequencies stay within the forward voltage limit?
Also, how does it work when the signal is too low? Or if the incoming imepedance is too high?
Thanks for the teardown! Looking forward! Would be great to see a teardown of the REALLY NICE COMPRESSOR
Great questions. I'm not positive what you mean by impedance matched, but any good transformer in a well-designed circuit will be basically flat in the audio spectrum. That is, it will step up/down all frequencies by the same ratio. So as far as I know there is nothing too special going on with the input transformer.
If the input signal is too low you turn down the threshold ;) If the incoming impedance is too high (like if you connected a guitar pickup to the input) you would get lots of loss, distortion, and filtering. But this is universal to any device, not unique to the diode compressor.
Thanks for the kind words and questions!
@@PetersonGoodwyn Thanks so much for the reply. What I meant by impedance matched is that, from what I understand, the incoming signal's impedance would have to be lower than or equal to the "impedance on the diode" caused by the "voltage divider" (if that makes any sense?). As the greater "voltage divider in this case would also be the transformer against the GR circuit. Just wanted to clarify if this was true or not, and if so, how would one go about determining the range of values applicable. And if not, I would gladly shove this question up my rear end lol.
Gotcha, thanks for clarifying. The short answer is that no, there is no line-level device that would have a high enough output impedance to affect the voltage divider. The longer answer is that output impedance is like R1 (series) in a voltage divider. So the only way the output impedance of the previous device could affect the GR circuit is if it added significantly to the R1s in this circuit, which are 47k according to the schematic. Most line-level devices have an output impedance around 50 Ohms, while some older ones go up to 600 Ohms. On top of that, the step-down input transformer reduces the output impedance going into the GR circuit even further. Does that all make sense? Great question!
Is there any chance that one day this comp will be available as DIYRE kit? .... )))
Thank you so much for this video.
Believe ut or not but it really helps for sobering like me who never studies electronics to grasp some sense in all this
That's so great to hear, thanks
What a great understanding. This give confidence in your gear built. Ok so I'd like a 8 channels summing-stereo buss compressor with 1073 amp section, CL 1B tube compression, an API continuous sidechain / HP filter knob, LA2A limiter section, blue & red silk + tape or carnhill transformer, 2 mono channel out, M/S - stereo switch, stereo width just before the limiter, all that in an SPL 120v audio rail fashion. Attack, release, treshold, Main+ 2 monitors out, with a 4 band parametric passive EQ. And all that good for Max 1600$ 😁🤭
Haha thanks very much. I'll get to work on that right away...
Very insightful information. The UAD 33609 is very nice but the hardware has a more spacial sound and its clearer and bigger. This is the best diode-bridge compressor out there. The unit has a limiter after the compressor so you can overdrive the transformers for more character while the limiter keeps the signal from overloading and sending out a too hot signal into your DAW.
WOW!! This is so cool.
Thankyou for sharing knowledge. There was great to see these diodes in the hardware. Maybe 0,1% of the area (and cost) for the main work :)
Thanks for this, so interesting!
Thanks as always for following!
Refreshing to hear the TRUTH and not some silly forum warriors
Nice vid, Peterson - great content :) I visited Neve's factory recently to check out the 8424, and gather that their xformers are no longer made by Carnhill... not sure how recent that is, though, so that 2018 unit may or may not have Carnhills-made ones!
amazing vídeo! thanks for that
This was FUN!
Do you have any idea about why the signal is kept balanced, meaning that you have to use two perfectly matched diodes instead of one ? Could it be because the voltage has to be kept under 0.7V so there is a risk of interference and the balanced signal keeps you away from this ?
And same question about having the signal unbalanced/amplified then balanced again ? Is it because the BA340 can't be done with balanced input ?
Thanks for the quality of your video, as always ! :)
Very insightful questions! Balancing the gain reduction stage has a couple of benefits. It allows the DC control voltage to easily be removed from the signal at the next, unbalancing stage. And it cancels out even-order harmonics in the same way that a balanced input cancels noise picked up by an XLR cable.
Rod Elliott has a great explanation of it here: sound-au.com/articles/vca-techniques.html#s9
Unbalancing, amplifying, and then balancing again is the typical approach because a) then you only need one amplifier instead of two! and b) most non-linearities in the amplifiers will not be common-mode, so they won't be cancelled out by the next balanced receiver. So you can actually introduce more noise and distortion by trying to amplify the signal while balanced.
@@PetersonGoodwyn What a perfectly clear answer, thanks !
So this is a compressor using a voltage divider circuit controlled by diodes, just as the 1176 is controlled by FETs and LA2A is controlled via a photoresistance.
If I'm correct, a compressor like the SSL G-Bus uses VCA to control gain. Are there others types of compressors apart from VCA or voltage divider ones ?
@@arthurlauth1380 yes the final type is pulse-width modulation which is what the Pye used that Ruper Neve was hired to replace
Tube-tech CL1B teardown next!!!!!
DIY RE do you think there is more to the sound of a transformer than it’s ratio and sheet type? Or do you think there is more shaping the sound than what science can measure? In the end music is very simple and small band of ac signal and we should - and I think we do- know everything about it. I mean in other electronics disciplines they work with GHz and THz and don’t mess around with different manufacturers for transformers.
As I've finally learned from these exceptional talks, there's no inconsistency between deep technical insight and empathy for idiosyncratic perceptual discernment. He joyously dissects his own eager bias for the warm sound of the Neve *design language* (or does he really love cool and punchy API?)! I recall Rupert Neve touting "stoved enamel"- stoic and warm!
Do you have an EE degree or did you just pick up everything in the field and experimenting?
yes!!!!
Mind... blown!
is that a switching PSU? kinda interesting
yep, switch-mode power supplies have come a long way and are now standard in high-end gear. to be honest, I'm so used to seeing them I didn't even think to comment on it!
1. when you say the transformers weren’t quite what they were, what does that mean? what is it about these particular transformers that are “authentic?” i tend to assume that kind of cork sniffing about things like magnets and coils of copper wire is just confirmation bias because of the scarcity of the older thing.
2. when you say that the circuit is improved by careful selection of components, what does that selection involve? is it reading lots of data sheets and picking just the right component or is it more bespoke than that like checking component values to make sure that values within tolerances are closely matched to the design spec?
Can’t a static discharge damage the circuit’s when touching it in this manner?
It depends on if a particular type of component is ESD sensitive. Most used in vintage gear are not ESD sensitive (other than transistors). The ICs in there are almost surely ESD sensitive. Surface mount resistors could be, just depends on build. Most caps of decently high capacitance value won’t be. Transformers and metal housings are not ESD sensitive. And as soon as he touched the housing of the rack chassis, he would have most likely safely discharged any charge buildup he had accumulated.
So is this where beringer got the idea to make their 369
I thought I would buy this as my first diode compressor .. finally do you think I should buy this? Because this is my first analog experience
This is so great.
Is that a switching power supply in it though? If so, why in the world not to have a proper transformer based one?
Yes good eye! Switch mode supplies can be made to have just as good specs as linear nowadays, with the added benefit of accepting any AC wall voltage.
Great. The only problem I see is possibly damaging the unit because of ESD (Electrostatic discharge).
Nar Diodes are hardy.
If I just can buy one. 33609 or MBP Portico Ii. I have used both on Analog Access, great results, but the latency and offline process, make the work flow not fun.
It is a pity that they produce significant distortion at low frequencies! (Due to a 'ripple' on the control voltage.)
so this design is considered "diode bridge?"
@20:20 it's a 3630?!
oops
liked and subbed
Some what of a myth with regards to the transformers, but hey myths sell. Reality is this, the original input transformer is an off the shelf part from Mariner and from time to time it would be sourced out to different companies, such as St Ives, when Mariner couldn't keep up with demand. Now the original output transformer was a different story, it was a custom design by neve and was only ever made by Mariner and one other company, which was not St Ives. Since the 60's people have come and gone and neve has been though multiple hands, not just siemens group, and over that time the original drafts for the output transformers have been long lost. As a consequence the output had to be reversed engineered, which happened decades ago now, and is the real reason people say they don't sound the same, there pretty close though but hey like said myths sell.
Class A Single Ended, not push pull output
Talking about sound without playing any sound is the definition of insanity.
Second
first