The official G9 help thread

[gyraf]

Yes, the TO220 house version has been eol'ed..

 

[beautgrainger147]

Hello,
I built my g9 getting on for 19yrs ago, tried it for 6 months and put it on the back burner due to hum.
Lately I decided to pull it back out and started trying to get it back to standard. I've moved the PSU bits back to the main board, stacked the toroids (both 15v rather than one 12v), and connected the ground to the case, and the input grounds to the case.
I've also installed some PCB connectors towards the front and have electrically insulated the gain pots from the front panel.

Having soldered in new connectors for the front, I can't help wanting to check that 4 of the 10 pin strips are supposed to be in a different order on the front PCBs.

I used to have sound from it when freshly assembled but am starting to figure out why I don't now have any sound years later (stupidly I didn't check before changing things to standard). For what it's worth, I also measure C15 239v, C14 250v, p48 is about 47v, and the front led illuminates. Both heat sinks get warm and the valves glow. In the past I found the phase switches to be a bit sensitive.

Some bits of the layout in the picture aren't final.

 

[beautgrainger147]

Hello,
I built my g9 getting on for 19yrs ago, tried it for 6 months and put it on the back burner due to hum.
Lately I decided to pull it back out and started trying to get it back to standard. I've moved the PSU bits back to the main board, stacked the toroids (both 15v rather than one 12v), and connected the ground to the case, and the input grounds to the case.
I've also installed some PCB connectors towards the front and have electrically insulated the gain pots from the front panel.

Having soldered in new connectors for the front, I can't help wanting to check that 4 of the 10 pin strips are supposed to be in a different order on the front PCBs.

I used to have sound from it when freshly assembled but am starting to figure out why I don't now have any sound years later (stupidly I didn't check before changing things to standard). For what it's worth, I also measure C15 239v, C14 250v, p48 is about 47v, and the front led illuminates. Both heat sinks get warm and the valves glow. In the past I found the phase switches to be a bit sensitive.

Some bits of the layout in the picture aren't final.

I finished up swapping the wires on the switch PCBs, to be as if a ribbon cable was fitted, ignoring markings. This has helped but leaves 2 other issues:
I may have the wrong type of lorin switches for the phase switches - they only work in one position and it's not position 1.
And, channel 2 has a sort of mains buzzing.

I'll post back when new switches have arrived.

 

[beautgrainger147]

The new switches have arrived, and are now fitted.. the phase can now be changed It turned out both were varying degrees of not working properly - one would work in some positions and the other had one pole down.

The unit now has a nice low level of hiss which is good, but I do get some buzzing which is a low but annoying level - kinda noticeable beyond 1/3rd gain. I'll probably try a bit of shielding and/or look at other transformer options people have used.. currently running the standard dual potted ones from the BOM.

 

[beautgrainger147]

Over the holidays I've not done as much messing as planned, but in relation to buzzing and hiss, have discovered some things.
I've tried running my G9 into a fishman loudbox, there was little hiss and a more noticeable buzzing, but it would probably be useable live at low to moderate gain levels.
I've tried running mic-G9-front input of a behringher umc404hd, this amplifies everything and wouldn't be useable for recording.

I've also tried mic-umc404hd-G9-umc404hd, with an unbalanced trs jack-xlr Y cable. This results in some slightly increased hiss (maybe from the behringher) but nearly silent buzz.
trying variations of this, mic-G9 via a balanced or unbalanced single cable, and G9-umc404hd via the Y cable results in buzzing but low hiss.

It's as if tying the cold of the input and ground together on the G9 gets rid of the buzz. Does this suggest anything specific is likely to be the cause of the buzz?

(I do currently have the front jack inputs and volume pots isolated from the chassis if that's a likely cause)

 

[innercityman]

Another G9 is alive !

This is my second G9 build. Works like a charm...

 

[Clbraddock]

After taking a break, I came back to my G9 project and now have the power supply working correctly.

I have a new head scratcher though. Position 1-4 of the gain switch work okay. 5-11 there is no further increase in gain.

Not a ton of gain overall either.

 

[gyraf]

1-4 is attenuation only (look at schematic)

5-11 is gaining by means of your cathode shunts to ground through a chosen resistor and your 10uF/100V polyester cap. Somewhere in this path you have probably lost a connection - an open here will result in fixed lowest gain like you have.

/Jakob E.

 

[Clbraddock]

Ok, I did some more testing.

Originally, I just had channel one connected to its control board. I've now connected channel two as well. Channel two is MUCH louder. Basically, the volume I would expect. I didn't measure it, but channel one is probably at least 40ish db quieter on same settings. I tried a tube swap, but it made no difference.

Both channels respond to the gain control on settings 1-5. Neither increase gain any further for settings 6-11. It's possible I made the same error on both control boards, but the fact that the gain switch acts the same on both makes me think maybe it's something on the main pcb?

One additional piece of information. I accidentally connected both headers backwards on channel ones control board (I fixed this prior to the testing mentioned above). I still haven't connected the 48 volt supply, so I didn't send 48 volts anywhere it shouldn't go. Looking at the board layout I don't think the 12 volt supply went anywhere bad either (but I could be wrong). I don't think this is what caused channel one to be quieter, but thought I should mention it.

Any guidance on what to test is much appreciated.

 

[gyraf]

you don't break anything here by inserting connectors wrong

Check with an ohmmeter (and power off) connection from 10uf polyester capacitor to ground - from last/rightmost pin of the 5-pin connector (see schematic for expected resistances) to GND, next-to-last

/Jakob E.

 

[Clbraddock]

Tested like you suggested and figured out that I accidentally used a 1M for R11 instead of 1K on both channels. Haven’t swapped it out yet, but that has to be why the CW gain positions weren’t doing anything.

Still not sure why channel 1 is so much quieter than channel 2 though. After swapping R11 I can try to signal trace and see where the discrepancy between the channels starts.

Edit: switched R11 to 1K and now channel 2 works perfectly. Channel one is still very quiet, but now all gain positions do actually change the gain.

 

[Clbraddock]

Ive tracked the problem of channel one being quiet down to the output tube, but can't figure out what the issue is.

When testing channel one against channel two with a test signal, VAC at the grids of V2B (pin 2) measures identical. Resistance from V2B cathode to ground also measures identical. V2B plate, however, measures almost no VAC for channel one while reading a healthy VAC for channel two. VDC at both plates measures the same.

I tried swapping the output tubes, but it made no difference on the readings.

Any ideas what could be wrong or what else to test?

 

[Clbraddock]

I just can't figure it out. It doesn't make sense that with the same cathode voltage drop, same plate voltage, and same input voltage at the grid that one channel works and the other has no signal out from said tube.

Double checked that there are no open connections from the tube socket. Double checked that every component value that connects to the tube socket is the same between channels. Double checked that there are not shorted connections between the tube socket pins.

It doesn't make sense.

 

[gyraf]

look for some significant load on the circuit's output then?

 

[thelivingroom]

Hello everyone! I’ve been the beneficiary of these forums for years for various purposes, but this is my first post on here. I had done my own maintenance on the gear in my studio for years and put together a couple of full kits like the Hairball 1176, and made some DI boxes in the past, then as my confidence grew I pushed myself to build this G9 and audiox’s W492 EQ simultaneously as my first two more involved DIY builds about five years ago. Doing some of these builds on here like this, and having several others in various stages waiting to be finished, has injected a new and different love for recording for me in my 25+ years of being in my studio. It’s safe to say I now have an addiction haha.

Before I ordered any parts and got started I read every single post, printed out the important ones, took my time once I got going, and it worked perfectly on first fire up. I’ve waited this long to post my build because I initially stubbornly put the power switch in the center with a pretty neon light above it, due to my desire for panel symmetry, causing a lot of buzz on channel 2 (I know, I know haha). I made it a lot better by making an ugly metal shroud around that and the mains wiring, but always wanted to redo the front panel correctly, use shielded mains wire, and change the tubes from whatever JJ’s I had laying around at the time to get it going to something that may be a little better. I recently redid all of that and now love this thing even more. There’s still a little buzz on channel 2 if I turn my monitors all the way up, and I may eventually pull it one more time to add a divider plate next to the power switch and wiring, but for now I’m gonna enjoy it for a bit.

Thank you Jakob for your generosity and knowledge, Gustav for the pcb’s, and everyone who has posted useful information in this thread.

A little follow up to my original post last year:

1) I swapped the modern Tung Sol tubes to NOS RCA long black plates in V1, and NOS Raytheon long black plates in V2. It already sounded fantastic before, and it’s a tad better now.

2) I changed the gain switch wiring to shielded, and while it was apart my OCD kicked in and I changed out most of the mounting hardware that was too long for my liking.

3) New powder coated RAL 5012 and UV printed front panel from Frank that came out beautiful. First one I’ve done with laying out the graphics in Inkscape and being able to use whatever fonts I want. Very satisfied!

 

[sarakisof]

I've built mine since a long time ago. Started using recently (so I believe issue was there from the beginning).

I'm facing an out of phase / asymmetrical signal issue.

Did an experiment today. Passed a mono bass sample through G9 both in Phase + and - .
Here are the results comparing Original
G9 Phase+
G9 Phase-

Phase function also seems reversed.

Any thoughts?

 

[gyraf]

..it clips asymetrically, yes (like most single-supply tube circuits)

 

[BGroen]

Hi all,
I want to use the additional PSU to build one or more MEQ-5 eq's.
My question is, how many SRPP stages I can power with the PSU?
The g9 has 4 tubes, could I manage 4 SRPP stages?
Are there any pittfalls using the same PSU for multiple channels?

Hope hearing!

 

[gyraf]

yes, the psu should be able to supply four srpp stages

 

[HerbertR]

I had a pair of "Neve" style UTM 2546 input transformers (equivalent to Carnhill 9046) left over for an other project and tried these parts in the G9 at a "lower" maximum 1:4 ratio. I directly compared the UTM input transformer on one channel with the other channel equipped with the OEP 262A3E / Zobel network and R load on the secondary side to tame the resonance that rises in the upper high end. Both LL5402 output transformers are terminated at 620 ohms.
These UTMs work very well in this preamp. They were mounted on the PCB using hex spacers to keep the wires short.

My conclusions are
->due to the lower maximum ratio I lost about 3 dB of gain (~57.5 dB to ~54.5 dB max. gain)
->but gained super, super nice overtones, THD is ~0.3% with OEP input compared to ~0.7% with UTM at ~45 dB gain, 1 V output,
->significantly better low end below 50 Hz
->not as dull and "closed" sounding compared to the OEP input TX
Not scientific measured....

I love this piece, hell .....