transistor biasing - aclam mocker fuzz

[Rob Strand]

I've never had a real one in my hands to read voltages, but my understanding thus far has been that the first two transistor stages generally run around 9 to 10VDC, which is one reason that section has been modelled a few times now as a standalone fuzz.

I have some faith in my estimates, however I don't know how much to trust the +25V on the schematic!

I saw one Vox schematic where someone looked like they wrote 2V on the emitter of Q2.  No idea what the conditions were, low gain mode or fuzz mode.  It makes a difference.

But you may be onto something with your calculated voltages. In any event I'd imagine that makes the voltage divider at the base of Q2 possibly even more fiddly?

The actual units are what they are.  If the design depends on hFE and is running at 6V all that means is the real units will not be consistent.

The Q1 position is quite independent of the transistor due to the large emitter resistor.   In low gain mode Q2 is pretty good too.  The sound of the circuit does depend on the Q2 gain.  The original schematic shows BC107 for Q2 but the gains are unknown.  Lower gains have less overdrive and are less gated.     I can see/hear things start to change when we get down to hFE = 100 to 150. 

The key for a clone would be to ride that point precisely but with so few samples it's an unknown target.

FYI, in low gain mode we have unity gain for Q2 which supports the 68k/47k resistors - wouldn't work too well with 3k3 in there.

I've been meaning to build the whole preamp and run it at ~24VDC to get some "real world" data, but I just haven't been able to as of yet.

Yes, always a good thing.   Although it takes time/effort/parts.   Not enough time to do that to every unit we come across.

[Rob Strand]

I've never had a real one in my hands to read voltages, but my understanding thus far has been that the first two transistor stages generally run around 9 to 10VDC, which is one reason that section has been modelled a few times now as a standalone fuzz. But you may be onto something with your calculated voltages. In any event I'd imagine that makes the voltage divider at the base of Q2 possibly even more fiddly?

It looks like possibly less than 6V!

Yesterday it occurred to me there could be more loading on the fuzz circuit power rail as the schematic shows the power rail going to the footswitch.  The fuzz is enabled via the footswitch.

I found this video which explains the footswitch:

https://www.youtube.com/watch?v=H7IGh2ml_6k

As it turns out the fuzz circuit powers the 10k resistor on the base of the transistor switches.
That means there is additional loading on the fuzz-circuit power rails of 10k + Vbe in series.
The load is present when the fuzz is on.  I estimate the fuzz supply drops to about 5.1V to 5.3V.

In clean mode the fuzz supply voltage will rise to about 9.6V to 9.8V due to the removal of the 10k + Vbe load via the footswitch, and also the decrease in Q2's bias current.

[R.G.]

I've never had a real one in my hands to read voltages, but my understanding thus far has been that the first two transistor stages generally run around 9 to 10VDC, which is one reason that section has been modelled a few times now as a standalone fuzz.

I haven't had a real one either; although I did have a Conqueror head, which is quite similar. I corresponded with a guy in the UK that has one, and took quite a bit of data on the parts and schematic from photos. I put the whole schemo into a circuit simulator and came up with about 14V for the first section supply voltage. The repro PCBs also tend to run at about this voltage when populated.


My best guess is that when the circuit is built with new parts and more consistent resistors, it runs 12-14V. The real ones may well have experienced the dropping resistors drifting up in value. The circuit is  tolerant of lower voltages as well. In my repro amplifier PCBs I have frequently run into the situation where a real, original preamp or amp today is not the same operating conditions and sound as the same amplifier was when it was new, decades ago.

[Rob Strand]

I haven't had a real one either; although I did have a Conqueror head, which is quite similar. I corresponded with a guy in the UK that has one, and took quite a bit of data on the parts and schematic from photos. I put the whole schemo into a circuit simulator and came up with about 14V for the first section supply voltage. The repro PCBs also tend to run at about this voltage when populated.

My best guess is that when the circuit is built with new parts and more consistent resistors, it runs 12-14V. The real ones may well have experienced the dropping resistors drifting up in value. The circuit is  tolerant of lower voltages as well. In my repro amplifier PCBs I have frequently run into the situation where a real, original preamp or amp today is not the same operating conditions and sound as the same amplifier was when it was new, decades ago.

If I used the "incorrect" Q2 base resistor, 3k3 instead of 47k, and remove 10k loading of the footswitch I see 15V.
https://www.diystompboxes.com/smfforum/index.php?topic=132293.msg1288226#msg1288226

Things don't add-up for a 9V rail.   If we consider dropping 25V to 9V over the 12k supply resistor then the supply can only provide (25-9)/12k = 1.3mA.   

From the circuit's perspective with a 9V supply:
- Assume the first transistor stage biases to Vc1=4.5V, IC1 = (9-4.5)/22k = 0.2mA
- Assume the second transistor bias to Vc2=4.5, IC2 = (9-4.5)/3.3k = 1.36mA
  We know this transistor is biased with a low collector voltage so more realistic is (9-3)/3.3k = 1.8mA
- The loading through the footswitch via the 10k resistor and transistor switch bases:  (9-0.6)/10k = 0.84mA.

If we add up our low current draw estimate (0.2mA + 1.36mA + 0.84mA) = 2.4mA.  Even that exceeds the 1.3mA "target" load for a 25V + 12k resistor source to produce a 9V rail.    The more realistic bias estimate sums to (0.2mA+1.8mA+0.84mA) = 2.84mA.

Whatever way you look at it the circuit at 9V draws more current than a 25V + 12k resistor source can provide.  The bias point of the second transistor has a significant effect on the supply current;  that's also why we need to use to correct base resistors on Q2.

Notice 25V - 12k*2.4mA = -3.8V!! 

With a lower supply voltage the transistor stages will draw lower currents, the droop on the supply will be less and 25k - 12k*Iload will end-up at a positive voltage; which I'm seeing as 5.2V or so.

[pacealot]

Thank you, Rob and R.G. – I fear that I may have contributed to derailing SSC's Mocker-specific thread, but I think this info is critical to understanding the original circuit and how it was "supposed" to work to begin with.

My only potentially valuable contribution to any of this might be identifying that 3.3k as "wrong" (relative to the original circuit), and in case it helps, I did find at least one of the pics which helped me determine that at least one amp really did have a 47k in that spot. I did legitimately trace out the tagboard at one point, but it's been a while and I can't really find all my work to show it – hopefully you can trust me to have identified it correctly:



You can at least see that this 47k is coming off the base of Q2 (the insulated tranny on the far left). Note that the insulation on the BC10x's doesn't help us narrow down the gain ranges, sadly.

Also, I did a screengrab from Steve Walsh's video that Rob linked to above, and it sure looks to me like he wrote "47k" just above where the schemo reads 3.3k:

[Rob Strand]

Thank you, Rob and R.G. – I fear that I may have contributed to derailing SSC's Mocker-specific thread, but I think this info is critical to understanding the original circuit and how it was "supposed" to work to begin with.

Yes, absolutely the case.

My only potentially valuable contribution to any of this might be identifying that 3.3k as "wrong" (relative to the original circuit), and in case it helps, I did find at least one of the pics which helped me determine that at least one amp really did have a 47k in that spot. I did legitimately trace out the tagboard at one point, but it's been a while and I can't really find all my work to show it – hopefully you can trust me to have identified it correctly:

You can at least see that this 47k is coming off the base of Q2 (the insulated tranny on the far left). Note that the insulation on the BC10x's doesn't help us narrow down the gain ranges, sadly.

A very important contribution to say the least.  Sure looks like 47k is correct.

The thing is, the circuit has to work in both clean and fuzz modes and 47k seems to do it.  If I was sitting in isolation and saw that 68k/3.3k base circuit on the schematic, I'd probably think the 3.3k should be 33k.  Then after doing an LTspice sim I'd probably think they should have used 39k (for clean mode).  For fuzz mode the choice isn't clear since we need to know the transistor gain and have a very good idea where the circuit is riding on the *slightly* sputtery bias point.

Also, I did a screengrab from Steve Walsh's video that Rob linked to above, and it sure looks to me like he wrote "47k" just above where the schemo reads 3.3k

I noticed that as well.


The PDF schematics on the VOX site have got 2V and 0.6V marked on the emitter of the second transistor.  However it is unclear if that is clean mode or fuzz mode.   I can't get 2V for either mode unless I manipulate the transistor gains.

[R.G.]

I have learned not to trust Vox schematics too much, and to always simulate as well as reasoning about how the circuit works. Thomas Vox was really bad.

One form of "copy protection" is creative mistakes in official schematics.

Trust, but verify, and verify the sanity of what happens as well.

[SprinkleSpraycan]

Thank you, Rob and R.G. – I fear that I may have contributed to derailing SSC's Mocker-specific thread, but

  

I was able to get the help i needed AND enjoy watching the topic spiral.