The official G9 help thread

[Studio Mollan]

Ok thanks for your reply!
So to reduce cross talk there should be separate regulation on each PCB. Is there not already cross talk between the two channels on the main PCB? When you say basic R-C filtering do you mean an extra filter or to move the filtering, or some of it, from the PCB  to the external PSU box?
/
Emil

When I built an off-board PSU, I measured inter-channel cross talk at ~-70dB down between 2 channels on one board (L input terminated with 200 Ohm no input, R channel 1KHz tone +4dBU output = -66dBU 1Khz output on "quiet" L channel) So I wouldn't worry about that if you're running L+R for a stereo pair of mics: you're going to get way more in-room acoustic bleed than that. Really not sure how much money you're going to save with an out of box PSU (or an inter-connect between 2 boxes) by the time you buy some high quality high voltage connectors (like 7 pin C091 Tuchels) and multi-core cable

So you built one psu like the one on the main pcb and connected it directly to the audio circuits as in the schematics. Psu like the psu schematics but larger transformers. Eq pcbs as is but without any of the psu components?
I'm probably not going to use connectors from the psu. I'll hardwire it instead.

These preamps are a bit of a electronics training for me so I wouldn't mind saving some cash if needed. I bought 8kg of nos vintage caps and almost as much resistors also nos which I'd like to start experimenting with.
Thanks!
/
Emil

 

[MeToo2]

Ok thanks for your reply!
So to reduce cross talk there should be separate regulation on each PCB. Is there not already cross talk between the two channels on the main PCB? When you say basic R-C filtering do you mean an extra filter or to move the filtering, or some of it, from the PCB  to the external PSU box?
/
Emil

When I built an off-board PSU, I measured inter-channel cross talk at ~-70dB down between 2 channels on one board (L input terminated with 200 Ohm no input, R channel 1KHz tone +4dBU output = -66dBU 1Khz output on "quiet" L channel) So I wouldn't worry about that if you're running L+R for a stereo pair of mics: you're going to get way more in-room acoustic bleed than that. Really not sure how much money you're going to save with an out of box PSU (or an inter-connect between 2 boxes) by the time you buy some high quality high voltage connectors (like 7 pin C091 Tuchels) and multi-core cable

So you built one psu like the one on the main pcb and connected it directly to the audio circuits as in the schematics. Psu like the psu schematics but larger transformers. Eq pcbs as is but without any of the psu components?
I'm probably not going to use connectors from the psu. I'll hardwire it instead.

These preamps are a bit of a electronics training for me so I wouldn't mind saving some cash if needed. I bought 8kg of nos vintage caps and almost as much resistors also nos which I'd like to start experimenting with.
Thanks!
/
Emil

If you look earlier in the thread I posted detailed results. Basically I extensively tested various combinations of a single one-board 2-channel G9 based on Gustav's PCB layout with OEP transformers, and one with Lundahl transformers; B+ PSU as designed in the schematic and mounted on board, B+ PSU as designed on the schematic but built point to point off-board in the same case, my own custom B+ PSU attempting to reduce higher frequency effects; different heater wiring arrangements, different +48V wiring arrangements, different transformer placements. I came to the clear conclusion that the original design was better than my own attempt at improving the B+ supply, but physically building the B+ off-board point-to-point with the original design as per the schematic was a significant improvement over building the B+ PSU on-board the common published PCB layout with exactly the same design and components. It is also worth fiddling with cable routing, physical transformer placement of the power transformers, and shielding (especially of the OEPs mu-metal screen) to reduce hum significantly. I never tested two 2-channel boards on one B+ supply though. If you've got lots of time but little money you could just build something and report back

 

[stevan.siemens]

Okay, so I've got my preamp working. Traced a few problems down and finally have a signal being amplified in both channels. However, I have a pretty significant hum problem. The noise is clearly power supply hum. Through searching the thread, I've found some people have had success moving the power supply off the main board. Are there any specific recommendations anyone might have? Any hints/advice for shielding or specific leads which might be problematic? I'm sure this isn't a sure fire solution so any further advice would be very much appreciated if anyone has any...

 

[chrispbass]

Okay, so I've got my preamp working. Traced a few problems down and finally have a signal being amplified in both channels. However, I have a pretty significant hum problem. The noise is clearly power supply hum. Through searching the thread, I've found some people have had success moving the power supply off the main board. Are there any specific recommendations anyone might have? Any hints/advice for shielding or specific leads which might be problematic? I'm sure this isn't a sure fire solution so any further advice would be very much appreciated if anyone has any...

Look up trace cuts and screen leads for front panel jacks, quite early on in thread.  I built mine with just the B+ supply off the board from the go, so can't comment if that had an effect or not. Also, make sure you connect all pin 1's of your xlrs to star ground and only ground pin 1 of channel 1  input back to the board.

Mines really quiet now, although initially I had hum due to a slight build error of my B+ supply (look back a few posts).
best of luck
Chris

 

[stevan.siemens]

Look up trace cuts and screen leads for front panel jacks, quite early on in thread.  I built mine with just the B+ supply off the board from the go, so can't comment if that had an effect or not. Also, make sure you connect all pin 1's of your xlrs to star ground and only ground pin 1 of channel 1  input back to the board.

Mines really quiet now, although initially I had hum due to a slight build error of my B+ supply (look back a few posts).
best of luck
Chris

Thanks for the reply. You gave me some things to think about and try. It occurs to me that the current setup has the HT ground (the negative connection on my filter caps) connected to the board. If I moved this ground connection off board, the ground traces on the board will then be mostly a ground buss for the signal. I suspect this will cut down significantly on some ground currents inducing noise on some signal connections... Thanks.

 

[Studio Mollan]

Ok thanks for your reply!
So to reduce cross talk there should be separate regulation on each PCB. Is there not already cross talk between the two channels on the main PCB? When you say basic R-C filtering do you mean an extra filter or to move the filtering, or some of it, from the PCB  to the external PSU box?
/
Emil

When I built an off-board PSU, I measured inter-channel cross talk at ~-70dB down between 2 channels on one board (L input terminated with 200 Ohm no input, R channel 1KHz tone +4dBU output = -66dBU 1Khz output on "quiet" L channel) So I wouldn't worry about that if you're running L+R for a stereo pair of mics: you're going to get way more in-room acoustic bleed than that. Really not sure how much money you're going to save with an out of box PSU (or an inter-connect between 2 boxes) by the time you buy some high quality high voltage connectors (like 7 pin C091 Tuchels) and multi-core cable

So you built one psu like the one on the main pcb and connected it directly to the audio circuits as in the schematics. Psu like the psu schematics but larger transformers. Eq pcbs as is but without any of the psu components?
I'm probably not going to use connectors from the psu. I'll hardwire it instead.

These preamps are a bit of a electronics training for me so I wouldn't mind saving some cash if needed. I bought 8kg of nos vintage caps and almost as much resistors also nos which I'd like to start experimenting with.
Thanks!
/
Emil

If you look earlier in the thread I posted detailed results. Basically I extensively tested various combinations of a single one-board 2-channel G9 based on Gustav's PCB layout with OEP transformers, and one with Lundahl transformers; B+ PSU as designed in the schematic and mounted on board, B+ PSU as designed on the schematic but built point to point off-board in the same case, my own custom B+ PSU attempting to reduce higher frequency effects; different heater wiring arrangements, different +48V wiring arrangements, different transformer placements. I came to the clear conclusion that the original design was better than my own attempt at improving the B+ supply, but physically building the B+ off-board point-to-point with the original design as per the schematic was a significant improvement over building the B+ PSU on-board the common published PCB layout with exactly the same design and components. It is also worth fiddling with cable routing, physical transformer placement of the power transformers, and shielding (especially of the OEPs mu-metal screen) to reduce hum significantly. I never tested two 2-channel boards on one B+ supply though. If you've got lots of time but little money you could just build something and report back

Hi again and many thanks for your help.
After your answers a have decided to build separate PSUs for my preamps, but of board as you did. I guess the advantage with this design is noise right?
I've got a question though? Where did you hook up the different voltages to the main PCB and how did you solve the grounding? Did you star it from 0v of all the voltages to the star ground point or only one from the PSU PCB? 

 

[MeToo2]

Hi again and many thanks for your help.
After your answers a have decided to build separate PSUs for my preamps, but of board as you did. I guess the advantage with this design is noise right?
I've got a question though? Where did you hook up the different voltages to the main PCB and how did you solve the grounding? Did you star it from 0v of all the voltages to the star ground point or only one from the PSU PCB?

I only built the HT supply point to point off-board. 12V heater and 48V can stay on board.

Advantage was lack of mains hum (and PSU switching harmonics)
This was the last remaining noticeable noise (after eliminating a load of other things from my build that were layout related)
This was my 3rd G9 and I've gone back and retrofitted this to my first 2 builds.

Some pictures are worth a thousand words.....

http://s958.photobucket.com/albums/ae69/MeToo2_Prodigy/G9_PSU/

 

[Studio Mollan]

Hi again and many thanks for your help.
After your answers a have decided to build separate PSUs for my preamps, but of board as you did. I guess the advantage with this design is noise right?
I've got a question though? Where did you hook up the different voltages to the main PCB and how did you solve the grounding? Did you star it from 0v of all the voltages to the star ground point or only one from the PSU PCB?

I only built the HT supply point to point off-board. 12V heater and 48V can stay on board.

Advantage was lack of mains hum (and PSU switching harmonics)
This was the last remaining noticeable noise (after eliminating a load of other things from my build that were layout related)
This was my 3rd G9 and I've gone back and retrofitted this to my first 2 builds.

Some pictures are worth a thousand words.....

http://s958.photobucket.com/albums/ae69/MeToo2_Prodigy/G9_PSU/

Thank you!
This will help me a lot!
/
Emil

 

[bruce0]

Can I use a transformer like this http://www.jensen-transformers.com/datashts/123slpc.pdf as an output transformer for the Gyraf G9?

They can be wired 2:1 (or 1:1 or 3:1)

Would the DC resistance per winding be too low? Would this cause loss of the low end due to the 4u7 output cap?



The DCR of the primary on my transformers would be 40 Ohms (wired 2:1) or 60 Ohms (3:1).

I have a bunch of these 4 winding quadfilar transformers that spec out pretty much exactly like these Jensens, and aside from the issues of fit ( I can mount them off board  or on an adapter card) I was wondering if they would work.

 

[gyraf]

In general, I won't recommend substituting the LL5402 in the G9 output circuit. It works pretty well, and few others will do the job as good without changes to the circuit (specially output capacitor value).

That said: if you want, go ahead and try it - let us know how it turns out.

Jakob E.

 

[phelar]

It seems like everything working but after moving the HT psu of board to get rid off the noice it was making i still got the 100hz hum -80db. Just like when everything was nicely hooked up in the case and the B+ was on board. Its practically the same readings as before.

 

[phelar]

What to do know! No more ideas :-\ Yes the small HT board goes to chassis ground. I really thought i should get rid of the hum by moving the psu. What is wrong! Pulling my hair out here.
Her it is, the B+ is outside the case when this reading is done.

 

[gyraf]

There is no problem in general with the on-board power supplies.

If you run into hum problems, you should look elsewhere.

Jakob E.

 

[gyraf]

phelar,

Do you have the same hum, same level, in both channels? Then probably noisy supply lines

More in R than in L, then probably electromagnetic stray from power transformer. Rotate it till hum cancels out.

Jakob E.

 

[bernbrue]

What to do know! No more ideas :-\ Yes the small HT board goes to chassis ground. I really thought i should get rid of the hum by moving the psu. What is wrong! Pulling my hair out here.
Her it is, the B+ is outside the case when this reading is done.

Hi,
I didn´t follow your troubleshooting but I would check a few things before going on:

1) There are a few ground bridges on the main pcb. Check that you soldered in all the bridges (piece of wire)
2) Touch the switch board where the wires from the main board are soldered in. The solder pads are quite close next to each other. Check for fault contact.
3) Check whether the switch board is properly grounded (Compare schematic with wires going to the switch board)
4) Check contact between star ground/chassis and main/switch pcb ground
5) Check contact between XLR ground and main pcb ground/star ground (Chassis)

This might help.
regards
Bernd

 

[phelar]

Thanks!

Jakob,

Its the same on both channels. How do i solve noisy supply lines? I´ll absolutely try the on board psu again if i get rid of these hum levels. It sounds awesome by the way, thanks!


Bernd,

Thanks for the check list! I will look it over. There is a couple of your bullet points i haven´t checked

 

[bernbrue]

Hi,
I zoomed into your picture posted above. It seems, that the DI mod is not done correctly. The DI jacks should have 3 wires connected. First is ground from the main pcb to DI jack. The second one from LL1528 to DI jack, the third goes from DI jack to C2. The jack functions like a switch. With no DI instrument inserted the switch is closed (contact from LL1528 to C2). With a Bass connected to the DI input the switch is open (no contact from 1528 to C2). The Bass goes straight to C2 and grid of V1. It looks like you haven´t got the right DI jack as well. The schematic makes it quite clear.
Good luck!
Bernd

 

[MeToo2]

It seems like everything working but after moving the HT psu of board to get rid off the noice it was making i still got the 100hz hum -80db. Just like when everything was nicely hooked up in the case and the B+ was on board. Its practically the same readings as before.

-80dB relative to what?

Judging by the photo earlier in the thread, you also don't seem to have your inputs terminated with a dummy 200 ohm mic load when you're measuring o/p noise.
That can also affect the result quite a bit if the inputs are just floating and there's some gain.

And you don't seem to have the lid on. [remembered to scrape off some paint under the screws on each panel so that it forms a good faraday cage?]

 

[phelar]

The DI jack looks like this. From LL1528 to C2, and the ground wire. 3 wires, is this right? I´ve put the DI cables on the underside of the pcb. I couldn´t find good places to drill with the lundahls in the way. If someone got good explaining pics how to place the cables from the component side it would be great.

The faraday cage is a check. Yes you're right MeToo2, no dummy load. When i did the first test run ( B+ on board , lid on and with a dynamic mic on the input ) same readings as picture i posted earlier. I will put everything back in the case, psu off board and with the lid on and make a new measure on the output, with dummy load.

Still wondering how to fix noisy supply lines.

...and thanks for your help. I really appreciate it.

I´ll be back

 

[phelar]

Hi again.
I got some new questions. How do i messure the preamps noice level in a proper way? When i got the g9 knobs turned down and the input gain on my mbox1 turned down i read -105db at 100hz and -110db at 200hz. This is with a dynamic mic on the input. How do i make a 200 ohm dummy load and exactly how do i connect it to the input. I want to compare readings with other g9´s.
I dont really know where to turn the gain knob on my mbox1 to get fair readings :-[.