Switching scheme with muting and LEDs

[Gregg]

What I see on the schematic at first glance is even with the channel switch in the middle both Q1 and Q2 that are already 'on'. One of them will be always on of that's why your crunch LED will also be always on.
Take some measurements at Q1 and Q2 bases and collectors in all 3 positions of the channels switch. Also check if you have 12V at the relays positive side and let us know.

[frusciante89]

What I see on the schematic at first glance is even with the channel switch in the middle both Q1 and Q2 that are already 'on'. One of them will be always on of that's why your crunch LED will also be always on.
Take some measurements at Q1 and Q2 bases and collectors in all 3 positions of the channels switch. Also check if you have 12V at the relays positive side and let us know.

Ok... by the way, I updated the PCB drawing with the component names for better reading and debugging:

[nickb]

I shoulda held out for the PNPs ...it was so much simpler as there were less inversions to think of

First, using the NPNs the relay operation is inverted i.e. the function is selected when the relay is NOT energized. Hope you were aware when the board got laid out.
Second I should have shown D3 and D4 cathdodes going to the 'lead' and 'clean switch' lines, duh!
There should also be a couple of diodes shown in parallel with each relay, cathode to +ve rail to stop the back emf from the relay blowing up the transistors. but I see you did that anyway

So, if you were aware of the inverted relay operation then you just need to mod the diodes D3 and D4. If not then can you swap the wires to the NC and NO contacts of each? There are other approches but the obvious might be the easiest.

[frusciante89]

I shoulda held out for the PNPs ...it was so much simpler as there were less inversions to think of

First, using the NPNs the relay operation is inverted i.e. the function is selected when the relay is NOT energized. Hope you were aware when the board got laid out.
Second I should have shown D3 and D4 cathdodes going to the 'lead' and 'clean switch' lines, duh!
There should also be a couple of diodes shown in parallel with each relay, cathode to +ve rail to stop the back emf from the relay blowing up the transistors. but I see you did that anyway

So, if you were aware of the inverted relay operation then you just need to mod the diodes D3 and D4. If not then can you swap the wires to the NC and NO contacts of each? There are other approches but the obvious might be the easiest.

So if I understand correctly, you're saying to keep the BC546s but invert the direction of D3 and D4?

Also, I took the measurements requested, and:

1) the positive rails of the relays are all at 11.8V, so everything good there.
2) the voltages at the EBC for the three transistors, in the clean, crunch and lead positions are:

[Gregg]

These readings are very odd. If looks like your transistors are not working and are in between on and off state because if they are completely off you should have around 12V at the collector and around 0.6V if they are on. If you have 0V at Q3 base the LED shouldn't be on. The bases must also have some voltages different than zero at all times.
Check again your transistors' pins orientation, component values and do NOT invert D3 and D4. And make sure you take the measurements properly at the right locations.

[nickb]

So if I understand correctly, you're saying to keep the BC546s but invert the direction of D3 and D4?

No. What I said was "D3 and D4 cathdodes going to the 'lead' and 'clean switch' lines". So they are the correct way round but connected to the wrong place.

[frusciante89]

No. What I said was "D3 and D4 cathdodes going to the 'lead' and 'clean switch' lines". So they are the correct way round but connected to the wrong place.

oh ok! So I'm assuming the correct wiring will look something like this? I kept the orientation of the diodes as it was

PS: thanks again for the help

[nickb]

oh ok! So I'm assuming the correct wiring will look something like this? I kept the orientation of the diodes as it was

PS: thanks again for the help

Yup

Going back to debugging your voltages it might throw some light if you note the voltages at the anodes of the clean and lead LEDS.

[frusciante89]

Yup

Going back to debugging your voltages it might throw some light if you note the voltages at the anodes of the clean and lead LEDS.

Ask my brother to this mod today since I'm away for the week and apparently it now works. The LEDs switch correctly and there's only one LED on at a time. However the relay don't get switched, and the amplifier is permanently stuck in the CRUNCH mode, even with the LED indicating the channel change. He did some quick measurements to the relay voltage drop and they are on the order of 5V when turned on, which is not enough to excite the coil. I don't see who's in charge of this massive voltage drop in the circuit. The supply voltage is exactly 12V...

[nickb]

What is the DC resistance of the relay coil? I expect it to be 960 ohms for a G5V-1 12V.
What transistors did you use? If some uncertain stuff off eBay they could be low gain or the wrong way round ( E and C swapped).
Are R2/R6/R4/R7 the correct value?

It's just a combination of the HFE of the transistors, relay coil resistance and choice of those resistors that's not turning he transistors on hard enough.

If all is correct make R2 / R6 2.2k ohms to compensate for the low gain transistors and (on 2nd thoughts) leave R4/R7 as is.

[frusciante89]

So, took time to do some more debugging. I measured the relay resistances and the clean is exactly 960R, while for the other 3 I measure 360R (guess it's the measurement getting into parallel). Then, by changing R2 and R6 to 2.2k, the amp remains stuck on the clean channel (it was stuck in the crunch before). When the switch is in the clean mode, the clean relay measures 11.7V (which is good), and in the other two positions it drops to 5.4V (which should turn it off). The other relays, however, are stuck at 4.5V, with the voltage increasing of only around 1V when I act on the channel switch. Any ideas?
I just feel I want to desolder the transistors and eat them at this point, I'm clueless

[nickb]

So, took time to do some more debugging. I measured the relay resistances and the clean is exactly 960R, while for the other 3 I measure 360R (guess it's the measurement getting into parallel). Then, by changing R2 and R6 to 2.2k, the amp remains stuck on the clean channel (it was stuck in the crunch before). When the switch is in the clean mode, the clean relay measures 11.7V (which is good), and in the other two positions it drops to 5.4V (which should turn it off). The other relays, however, are stuck at 4.5V, with the voltage increasing of only around 1V when I act on the channel switch. Any ideas?
I just feel I want to desolder the transistors and eat them at this point, I'm clueless

Having three relays in parallel is a twist I wasn't aware of until now.

Have you done the steps asked i.e. verified the transistors are good quality and checked that they are in the right way round? Part numbers, PCB photos and PCB layout images will allows us to verify.


Still waiting for the voltages on both sides R2 and R6 (see #38). It would be best in a little table to have the corresponding collector voltages too.

[Gregg]

Here's a tested working solution for your current problem.
1/ Ditch transistors Q1 and Q2 and surround resistors.
2/ Run each LED in series with a 1k-4k7 resistor and in parallel with the respective relays' coils.
3/ Q3 circuit stays the same.
4/ Connect D3,4 cathodes to the relays' negative ends.
5/ Connect 3 way switch to respective relays' negative ends.
As far as switching from the front panel via the 3 way switch your problem is now solved.
If you want a separate button for each channel on your footswitch and muting relays' pops and clicks that makes two more problems for you to solve but more on that later.

[frusciante89]

Thanks everyone this was super helpful and I learned something new