Neve 1073 input impedance

[Micky Goldplatter]

Hello,

Apologies if this has been covered before.  I couldn't turn up a prior discussion with a search.

I'm trying to finish a Neve 1272 preamp project that I started by collecting parts many years ago...a bunch of BA283 cards, lundahll 1537xl's, carnhill output transformers, and Grayhill 2 pole 11 pos switches.

I've decided that I'm going to use the 2 poles for the input attenuator and setting the first gain stage in the 20-50 db range and have a 3 way switch for the BA283 second amp to bump that up stock, +5, +10.

I'm trying to make it roughly match the 1073 gain staging in the 20-50db range and adjust the input impedance to maintain 1200/300 ohm with the Lundahll's slightly different turns ratio.

What I notice is that on the 1073, the attenuator positions work out to 6k on the secondary of the mic transformer, then it looks like at 40-45 db settings there's a load resistor, 12k & 18k, then at 50db it's wide open, so whatever the input impedance of the ba283 is, looks like an external 18k, the 128k on the first transistor, and 10k bottom of voltage divider (with the 18k) on a feed back loop over two pins...in parallel appr 23k.

So, is that right that the input impedance is high on certain settings?  Seems to rise as the gain increases on the first stage.

Maybe I'm not seeing it correctly, but I'm just wondering whether that's a feature one should include on the DIY side.

Thanks!  Great site!

 

[Bo Deadly]

I was just looking into this myself. LTSpice simulations show that the input impedance is defined by a complex interaction between the feedback and gain resistors:

BA284 / BA283 AV Input Impedance
determined by LTSpice simulation
Rf    Rg    Zin
----- ----- -----
15k   91    3.7k
18k   120   4.9k
18k   330   6.8k
18k   1.8k  8.7k

The impedance is very flat up to about 100kHz where starts to roll off.

The impedance of the secondary of the microphone transformer is 4.8k.

The impedance of the secondary of the line transformer is only 600 ohms because it is wired in parallel. You can see that the 2.2k and attanuation ladder make ~700 ohms.

From this information and taking into account the attenuation resistors used with each step if any, you should be able to fully characterize the impedance of each step:

-45dB: 6.8k || 18k = 4.9k
-40dB: 8.7k || 12k = 5.0k
-35dB: 8.7k || 3.9k + 2.2k = 4.9k
...

and so on. It seems the impedance is matched quite carefully. Someone must have used a slide rule for this one.

 

[Micky Goldplatter]

Squarewave, thanks so much for posting that simulation.  It all makes much more sense now.

 

[bobober]

Does this explain why the spec for the preamp says the input impedance is 300Ω or 1200Ω? If it does I'm not sure why!

http://www.thehistoryofrecording.com/Manuals/Neve/1073_1084_User_Manual_Issue_5_1.pdf

 

[ruffrecords]

Does this explain why the spec for the preamp says the input impedance is 300Ω or 1200Ω? If it does I'm not sure why!

http://www.thehistoryofrecording.com/Manuals/Neve/1073_1084_User_Manual_Issue_5_1.pdf

Those two number are for different strapping of the primary of the mic transformer is all. 1200 is with the two primaries in series and 300 is with them in parallel.

Cheers

Ian

 

[jim-analog]

Greetings,

I'm working on some 1970s 1073 modules for a client and came across a problem that I'm trying to figure out. In the course of my troubleshooting, I used a test on my AudioPrecision PortaOne that I'd not tried before. It's the "Generator Load Function", which measures the load Z of the DUT as seen by internal generator output. The default frequency for this test is 1KHz, which is what I used (though you can set any F desired). Generator output Z is set to 600R and a signal amplitude of 200mV is used for the processor to base the calculations on. The generator output was connected directly to the mic input of 1073 modules and a data point was taken for each mic sensitivity position.

I performed this test on two modules; one that has a low gain problem (#6) and the other (#1) that is working perfectly.  I thought the data from this test may be pertenent to this tread and hopefully of some use to those working with the 1073 circuit, so I'm posting my results. Due to my findings, I went further into attempting to figure out discrepancies and performed several other tests which I will also show the data for in a subsequent post.

Input sw position        1073 module 1        1073 module 6

-80                                    853                              844
-75                                    858                              848
-70                                    869                              857
-65                                    914                                895
-60                                    918                                899
-55                                    918                                899
-50                                    1130                            873
-45                                      1190                            876
-40                                      1220                            890
-35                                      1190                            874
-30                                      1160                            863
-25                                      1170                            863
-20                                      1170                            865

As you can see, at input settings -50 and lower, there is a significant deviation in the input Z as seen by the generator output. Keep in mind the column for module #1 indicated a completely functioning unit, while #6 showed about a .5dB lower gain at all settings from -50 and under.

This caused me to consider what factor(s) could be responsible. For background info, I had recently replaced the entire input selector switch assembly;  I suspected there may be an error in either the wiring or resistor values.

My next post(s) will move on to the further testing I did, which I hope may provide more data for other members to use as reference.

  Regards, Jim

 

[Grant W]

Greetings,

I'm working on some 1970s 1073 modules for a client and came across a problem that I'm trying to figure out. In the course of my troubleshooting, I used a test on my AudioPrecision PortaOne that I'd not tried before. It's the "Generator Load Function", which measures the load Z of the DUT as seen by internal generator output. The default frequency for this test is 1KHz, which is what I used (though you can set any F desired). Generator output Z is set to 600R and a signal amplitude of 200mV is used for the processor to base the calculations on. The generator output was connected directly to the mic input of 1073 modules and a data point was taken for each mic sensitivity position.

I performed this test on two modules; one that has a low gain problem (#6) and the other (#1) that is working perfectly. I thought the data from this test may be pertenent to this tread and hopefully of some use to those working with the 1073 circuit, so I'm posting my results. Due to my findings, I went further into attempting to figure out discrepancies and performed several other tests which I will also show the data for in a subsequent post.

Input sw position 1073 module 1 1073 module 6

-80 853 844
-75 858 848
-70 869 857
-65 914 895
-60 918 899
-55 918 899
-50 1130 873
-45 1190 876
-40 1220 890
-35 1190 874
-30 1160 863
-25 1170 863
-20 1170 865

As you can see, at input settings -50 and lower, there is a significant deviation in the input Z as seen by the generator output. Keep in mind the column for module #1 indicated a completely functioning unit, while #6 showed about a .5dB lower gain at all settings from -50 and under.

This caused me to consider what factor(s) could be responsible. For background info, I had recently replaced the entire input selector switch assembly; I suspected there may be an error in either the wiring or resistor values.

My next post(s) will move on to the further testing I did, which I hope may provide more data for other members to use as reference.

Regards, Jim

 

[Grant W]

Hi Jim,

Just saw your post while doing some googling. I'm a "newbie" so please excuse the mix-up above.

Thanks for posting the results of your impedance test on Neve 1073 modules.

I have noticed the change in impedance from 1200 to 900 around the -50/-55 gain point. I had presumptuously put this down to an unintended error by Neve. Do you have any thoughts on this or any further information that would shed some light?

Thanks,

Grant

 

[HenryL]

I was just looking into this myself. LTSpice simulations show that the input impedance is defined by a complex interaction between the feedback and gain resistors:

BA284 / BA283 AV Input Impedance
determined by LTSpice simulation
Rf    Rg    Zin
----- ----- -----
15k   91    3.7k
18k   120   4.9k
18k   330   6.8k
18k   1.8k  8.7k

The impedance is very flat up to about 100kHz where starts to roll off.

The impedance of the secondary of the microphone transformer is 4.8k.

The impedance of the secondary of the line transformer is only 600 ohms because it is wired in parallel. You can see that the 2.2k and attanuation ladder make ~700 ohms.

From this information and taking into account the attenuation resistors used with each step if any, you should be able to fully characterize the impedance of each step:

-45dB: 6.8k || 18k = 4.9k
-40dB: 8.7k || 12k = 5.0k
-35dB: 8.7k || 3.9k + 2.2k = 4.9k
...

and so on. It seems the impedance is matched quite carefully. Someone must have used a slide rule for this one.

Hopefully not too long ago to seek a clarification on this...
Re the 'Rg' values given above, my reading of the schematics is that the 1.8k is always in circuit and additional parallel resistances of the other Rg values you list are either permanently wired or only switched in at certain gain settings. So when you list an entry for 'Rg=120' above, for example, then do you mean (120 || 1k8), but when you list the Rg=1k8 line you mean just the 1k8 (no parallel) ? Is that what was modelled?
Many thanks