The official G9 help thread

[g9builder]

Thanks for taking alook Harpo. I'm ok with the primary connections  - it is just the three secondaries that I just carn't seem to get my head around.

I have redone the drawing - hope this explains it better. Hope you can help.

Cheers and many thanks

Ian

 

[Harpo]

Ian, if your posted colour scheme from your pics text block is the same as printed on the transformer, you connect T1-yellow with T2-yellow and T2-red (in-phase secondary wires), leading to one pin of the 15VAC connector at the right edge of the pcb. Next connect T1-orange with T2-orange and T2-black (out-of-phase secondary wires), leading to the remaining pin of the 15VAC connector at the right edge of the pcb.
In Franks drawing, you're refering to,  he labled your T1 as 'toroid transformer 2' and your T2 as 'toroid transformer 1'.
good luck.

 

[g9builder]

Harpo,

Many thankks indeed - thats exactly what I originally did. That means that all of the '0' v are grouped together and all the '15' v are joined together (on the three secondaries to the 15v connector). This is the way I wired it and everything seemed to work ok except that one of the transformers T1 in my drawing got very hot after only 5 min, so I thought something may be wrong? I therefore tried again as my drawing attached and something smoked immediately!!!! Lookes like I need to now check what fried and replace and retry :-(

Any ideas what would have fried? And what would be the reason for the transformer running sooo hot!! Do you think that needs replacing also?

Many thanks indeed - I'm so near completing this thing I just want to get passing some audio through it!!

Regards

Ian

 

[Harpo]

Check orientation of all polarized parts, especially diodes and look for shorts.
If you ever powered your build with T1 secondaries in parallel, D13 and maybe D15 could be blown.

You could try to isolate the area of fault by disconnecting T2 and one of the T1 secondaries and check for healthy phantom and healthy heater voltage for the other T1 winding. When both work, reconnect T2. Be careful with high voltages.

 

[zebra50]

Just a thought - is there any chance at all that you have wired the 15V secondaries reverse-phase?

 

[g9builder]

Many thanks Harpo and Zebra50,

I did test all the diodes last night and they all seemed ok ( they were in circuit though and I understand this does not always give a true sign) I will disconnect one lead from each and test again and replace anything that is blown. Strange thing is though that I cannot see any sign on any component that would suggest it has fried but there was definately smoke!!.

I will triple check all wiring before I flick the switch again....

Many thanks

Ian

 

[MeToo2]

I'm currently building a pair of G9's. One with OEP and one with Lundall transformers.

I have a question for you guys about shielding and earthing.

In the instructions it states that:
- Connect 0V/Gnd to chassis at one - and only one - point: At the input XLR's.
- Connect the power ground from the power inlet to the ground at the input XLS's also.

This I understand. I have employed a star grounding scheme back to the power inlet to avoid any chance of hum loops.

I also understand that on the G9 PCB that the middle/ground pin on the 3 pin output header (K3 & K103) is not connected to anything (it is a blank pad) so only signal+ and signal- go to the output transformer.

My question is the following: Is it good practice to leave pin 1 of the output XLR "floating" and rely on the input XLR of the downstream equipment to provide the overall shield all the way back to the G9 output? Or is it better to strap pin 1 of the output XLR also to ground via a wire to pin 1 of the input XLR? This won't affect the g9 circuit (as I said the pad on the PCB is not connected to anything) but it will ensure an earthed screen on the output cable (at the risk of forming a hum loop with the following bit of kit)

What's the standard practice? Earth pin one of the output XLR? Earth pin 1 of the output XLR and add the option of a ground lift switch? Or leave the output shield floating and rely on the next bit of kit in the chain for earthing the shielding on the output cable?

My gut feeling wants to add that wire between pin 1 input XLR and pin 1 output XLR.

 

[g9builder]

I think i may have the same problem as graph. A faulty TL783. My wiring was right and to double check this I wired the 48v up and got the correct result. Then the heater 12v and got the correct result. I then checked i was getting 220v from the last pair of secondaries and I am. Once connected to the board R34 smoked again straight away. I have checked all the diodes and resisters and they are all correct and placed correctly. I have also checked all solder joints and traces.

I therefore think that the problem must be with Tl783? Would you tend to agree with this? I have been unable to test tl783 because as soon as I turn on the unit R34 takes the hit.


Regards

Ian

 

[MeToo2]

I think i may have the same problem as graph. A faulty TL783. My wiring was right and to double check this I wired the 48v up and got the correct result. Then the heater 12v and got the correct result. I then checked i was getting 220v from the last pair of secondaries and I am. Once connected to the board R34 smoked again straight away. I have checked all the diodes and resisters and they are all correct and placed correctly. I have also checked all solder joints and traces.

I therefore think that the problem must be with Tl783? Would you tend to agree with this? I have been unable to test tl783 because as soon as I turn on the unit R34 takes the hit.


Regards

Ian

TL783 is difficult to destroy. It has inbuilt short circuit and thermal protection. It's not expensive to replace though. On the other hand, you are irreplaceable, so be careful.

One tip that I read elsewhere and found useful was to wire up a 60W bulb in series with your mains supply. If your g9 is sucking too much power, the bulb will light thus limiting the current to less than 250mA and hopefully protect whatever is causing the problem. Otherwise if all is OK it will glow momentarily (1/2 second or so) and then go out.

I personally would not have thought R34 could be damaged easily. A 100K 2W resistor would not burn out even if it had a full 400V across it (v^2/r = 1.6 watt). R37, perhaps, r33 also. But R34?

Obvious questions. Did you check the values of your resistors with a multimeter before inserting them in the board? It's relatively easy to make mistakes with (modern) 4 and 5 band colour coding schemes.

Suggest taking things step by step from the transformers onwards.

Four potential checks:

1. Remove the TL783. Check that the bridge rectifier is working correctly and that you are getting HT DC out of it at the top of C14. Even without the TL783 in place, C14 / R33 will smooth pretty effectively under no load conditions. You should be seeing more than around 280V DC here.

2. If that is OK, reinsert the TL783. There's a jumper marked "HT" right next to the TL783. Have you tried powering up with both the 'ht' jumper and c15 removed? Then there's virtually nothing that can draw current on the TL783 side of the circuit. So at least that would point you which side to look. If that's still a problem look very carefully for shorts around the TL783 and double check the orientation of the TL783.

3. Also check c15 is still acting as a capacitor and is not short and that it was connected the right way around (it is an electrolytic and polarity sensitive). If that checks OK then reinsert C15 and test again.

4. If that's OK then try removing the valves and then reinserting the "HT" jumper. Again there should be nothing that draws significant HT current.

The HT supply in all cases should power up at somewhere more than 240V DC (I found that the TL783 doesn't actually regulate the voltage very well until it has quite a load on it)

Of course, be very careful to properly discharge the HT capacitors C14 & C15 using say a 10-100K 2W resistor and check the DC voltage on them with a meter before handling them or any other part of the board at all [I actually permanently soldered a 1Meg bleeder resistor across them so that at least they'd always discharge themselves after a few minutes even if I forgot, so have a cup of tea after every test and take it easy].

 

[g9builder]

Many Thankd MeeToo2 - I will  work through your list.... and report back.

 

[g9builder]

My Mistake earlier - it is R33 not R34 that is getting really hot and smoking.

I tried some stuff including removing tl783 and ht jumper and c15 and still no joy - whatever I did R33 smoked. I therefore rebuilt the 245 power section again with all new components and double checked everything diodes/zeners/resistors etc that a they were the correct values and b were not shorted and working ok and that their orientation is correct.

I flicked the switch and R33 smoked again

I really am at a loss with this. I also checked all traces and they appeared ok. I am getting power ok from the transformers. The 12v and 48v work fine.
The transformer is giving me 240v to the 220 connection  - would this have anything to do with it? They are 30va 230v 2x15v.

What am I missing here? My head is starting to bleed from all the scratching...

 

[mikeyB]

Hi G9 builder - not seen any mention of heatsinking/mounting of the TL783 -  -  have you checked isolation of TL783 - ie. make sure that the metal tab is not connected to ground/0V via the mounting bolt - you have got it heatsinked with an insulation pad - yes??
I've done this a couple of times where i haven't de-burred the holes and when tightening the reg to the heatsink/case the sharp metal edge has pierced trough the insulating washer to the regulator, hence short to ground!!
Note from Gyraf about isolating the 12V reg - even though the tab is grounded - otherwise you can cause earth loop resulting in hum.

Hope this helps

 

[g9builder]

Hi Yes the heatsink is isolated using an isolation kit  - silicon rubber pad  and washer and polycarbonate screw.

Could there be a problem with the pcb?

 

[g9builder]

http://www.groupdiy.com/index.php?topic=36682.0

In doing some more searching in an attempt to sort this out I came across this thread. Now my R33 isn't smoking anymore but It gets extremely hot after a few seconds and gives off a smell - therefore I keep turning off the unit. (I think if I left it on it would begin to smoke again)

Is it right that this is fine to be running hot anyway? Could I change it to another value to help?

 

[Kir_For_Pleasure]

is it a bad way to add a insertpoint into the g9 after r44 with switched jack then to xlr ?

 

[gyraf]

- R33 will get hot when charging the reservoir cap. Should cool down again in a couple of seconds.

- There are no obvious place for an insert point in the G9 because of impedance issues. Attaching to R44 is the same as putting whatever after the G9, dosn't make sense as insert.

Jakob E.

 

[MeToo2]

My Mistake earlier - it is R33 not R34 that is getting really hot and smoking.

I tried some stuff including removing tl783 and ht jumper and c15 and still no joy - whatever I did R33 smoked. I therefore rebuilt the 245 power section again with all new components and double checked everything diodes/zeners/resistors etc that a they were the correct values and b were not shorted and working ok and that their orientation is correct.

I flicked the switch and R33 smoked again

I really am at a loss with this. I also checked all traces and they appeared ok. I am getting power ok from the transformers. The 12v and 48v work fine.
The transformer is giving me 240v to the 220 connection  - would this have anything to do with it? They are 30va 230v 2x15v.

What am I missing here? My head is starting to bleed from all the scratching...

240V AC sounds perfectly normal for 2 back to back connected 15V transformers on a UK supply.

When you say 2*15V, are they two separate & fully isolated secondary transformer windings 0-15 + 0-15 like this one http://nl.farnell.com/jsp/search/productdetail.jsp?SKU=1419541, or is it a single 15-0-15 winding with a centre tap? It would help if you linked to the manufacturers data sheet. Have you thoroughly checked the wiring of your transformers to make sure there isn't an unexpected earth somewhere? Have you also checked the phase of the windings as you mentioned earlier they were getting hot? [the 220v should be floating relative to ground until you actually connect it to the board, you can test that with the power off and buzz it out to make sure there is no unexpected connection to the chassis or earth, or between windings. Check you're seeing just the expected winding resistance. No more. No less]

It would also help if you posted a zoomed photo of both sides of the HT power supply area so that we can have a look at component orientations. Back lighting the board from the top whilst photographing the trace side from underneath would help identify shorts.

If the TL783 & HT jumper & C15 are removed then there is no way that R33 should smoke. Smoking indicates that too much current is being drawn. If you read the schematic and remove the TL783 and HT jumper & C15 completely, there's no path to anywhere for current to flow through R33 of more than around 1.2mA via D3-5 plus R34-R36 (240V-117V dropped across the 3*39V zener diodes via 100K+470R+47R to earth = 123V/104K ~ 1.2mA). You've got a short either on the board or through a blown diode (e.g. D14) or a short capacitor (C14 C15) or a component connected in the wrong orientation, or something odd in the wiring of your transformers. In other words something is basically wrong in the construction of your HT supply. That's probably not news to you. As it is almost certainly a basic fault you need to apply appropriate basic debugging and not look for anything too complicated. Most likely scenario seems to be a structural mistake that you've repeated also in the rebuild e.g. zener diodes in backwards. IMVHO you're trying to change too much between tests and then expecting it all to work.

The HT supply is where you should always start a build (as there is much more potential for frying stuff downstream.) Get that stable and tested first. Suggest you temporarily remove the 15V connections to the transformers for the heater supply and the 48V supply. You can leave the components for the 48V supply and the 12V supply in place. Now remove all components around the HT supply up to and including the HT jumper, check their values individually on a meter, and also buzz it out the pads individually checking for a short to earth somewhere (with no power applied). Then build & test the HT supply step by step, single component by single component, starting with just the transformer connected to the board, test with power connected for correct voltage (approx 240V AC, but it could be higher due to no load), then build the bridge rectifier D17-D20, test again (approx 240V*1.414 peak out of bridge rectifier), then add R33, testing again (same as previous). Add in D2, D3-5, and retest. Up until this point there should be near zero current through R33. Without wishing to insult you in any way, you can of course check for current across R33 by measuring for any voltage drop across it with a well insulated meter....... current (mA) = Vdrop (mV) / 470. It should be as near as 0V as makes no difference up until now. If there's any current at all through R33 you've done something wrong, so go back a step. Now connect in C14, test. There should still be near zero steady state current through R33, except now you'll see an initial burst to charge C14 before it drops back [remember to discharge C14!]. There should be around 300V DC at the top of C14 relative to earth. Then add in r34-36 and test that. The steady state current through R33 should now be around 1.2mA after the initial surge. Then add in R37, C15, and the TL783. Again there should be very little significant current flowing in steady state, and then finally add the HT jumper. At no point should you see smoke. If you don't go step by step you'll probably continue to guess where the problem is [and inhale smoke].

 

[g9builder]

MeToo2 - what can I say other than many many thanks.

I have now solved my problem. I'm a little embarrased to reveal but my problem was R34. I had a 100R and not a 100K resistor there. :-[

I checked my order and I did order 100k but was sent 100R instead. When I dismantled the 245 power section and rebuilt it I even checked this resistor and missed it again. ( what a fool!)

So everything was wired correct - all components were right except this one. I have burnt out 2 tl783's because of this and lost lots of sleep. If there is anything I have learnt, and I have learn't alot building this is to always check components before soldering them to the board and I mean really check them!!!

Thanks to all that have tried to help me on this issue. Now to go and pass some audio through it and hopefully smile!!. (here's hoping I make no more schoolboy errors)

Cheers

Ian

 

[MeToo2]

MeToo2 - what can I say other than many many thanks.

Cheers

Ian

Glad to be of help. Good luck with the rest of your build!

 

[g9builder]

Dropped in a new tl783 and 100K resistor and everything is now fine. Tried it out last night with a a bass and it sounds great!!. Only two small problems left.

Very slight hum when gain fully cranked. I am using a Perusha case and Zebra50 mentioned something about this so will try what he suggests.

DI on channel 1 doesn't pass any audio - channel 2 is fine. Need to look at this but otherwise I'm very happy.

Once again many thanks to all that have answered my queries and helped me get my first project completed.

Regards

Ian