Just a little amp advice, please...

[digi2t]

Picked up a 36w tube amp off Ebay not too long ago. Three 12AX7's and a quad of EL84's. Great little amp, for $200, I'm really pleased. The seller was even nice enough to send me all the schematics for it too. Just two little things that bug me about it though;

1) The volume difference between the Clean and Lead channels is too big. With the Clean volume 1/3 up, I need to up the Lead channel volume to about 2/3 to get a perceived evenness in volume level. Looking at the schematic, I see that the Clean stage volume pot is an A250K, and the Lead volume pot is an A10K;



I'm thinking that if I swap out the A10K for, say, an A25K, that should maybe even things out, no? My question here is, do I need to swap out any of the surrounding cast of characters, specifically the 10K (R13)?

2) The Clean/Lead switching is relay based. When switching from Clean to Lead, it's instantaneous. The problem is when switching from Lead to Clean, there's a lag of 1/4 maybe 1/2 a second. Again, referring to the schematic, I see that there's a diode, a resistor, and two caps around the relay coil. No doubt for snubbing/debouncing purposes. I'm wondering if the 22uF (C55) cap might be responsible for the lag? Is it charging up when in Lead mode, but when I switch back, the coil is being held by the charge?



So, can I reduce this value to increase switching speed? Maybe remove it completely?

Thanks folks.

[Mark Hammer]

Picked up a 36w tube amp off Ebay not too long ago. Three 12AX7's and a quad of EL84's. Great little amp, for $200, I'm really pleased. The seller was even nice enough to send me all the schematics for it too. Just two little things that bug me about it though;

1) The volume difference between the Clean and Lead channels is too big. With the Clean volume 1/3 up, I need to up the Lead channel volume to about 2/3 to get a perceived evenness in volume level. Looking at the schematic, I see that the Clean stage volume pot is an A250K, and the Lead volume pot is an A10K;



I'm thinking that if I swap out the A10K for, say, an A25K, that should maybe even things out, no? My question here is, do I need to swap out any of the surrounding cast of characters, specifically the 10K (R13)?

I would think your strategy would be to attenuate the Clean signal a bit to balance the Clean/Lead volumes.  Or is it a matter of the seeming mismatch between the respective volume-pot settings being a byproduct of the "Gain" control on the Lead channel, and where you have it set to?

[digi2t]

Picked up a 36w tube amp off Ebay not too long ago. Three 12AX7's and a quad of EL84's. Great little amp, for $200, I'm really pleased. The seller was even nice enough to send me all the schematics for it too. Just two little things that bug me about it though;

1) The volume difference between the Clean and Lead channels is too big. With the Clean volume 1/3 up, I need to up the Lead channel volume to about 2/3 to get a perceived evenness in volume level. Looking at the schematic, I see that the Clean stage volume pot is an A250K, and the Lead volume pot is an A10K;



I'm thinking that if I swap out the A10K for, say, an A25K, that should maybe even things out, no? My question here is, do I need to swap out any of the surrounding cast of characters, specifically the 10K (R13)?

I would think your strategy would be to attenuate the Clean signal a bit to balance the Clean/Lead volumes.  Or is it a matter of the seeming mismatch between the respective volume-pot settings being a byproduct of the "Gain" control on the Lead channel, and where you have it set to?

Good point, but the gain doesn't really affect the overall volume discrepancy THAT much. Yes, if I crank the Gain to the max, the volume of the Lead channel will go up a bit, but I still have a lower volume output on the Lead channel on a 1:1 basis where pot position is concerned. I'm thinking that a slightly bigger pot value will lift me from ground more, and maybe even things up. I'm just looking for more experienced eyes to let me know if there's anything else that I should consider if I change the pot.

[Mark Hammer]

I say make R8 820k, VR1 100k, and adjust C9 to taste (or remove outright).  That should drop the level of the Clean channel to be roughly comparable to the Lead channel's volume pot setting.

[amptramp]

For the switching, remove the C34 electrolytic and put a couple of more diodes in series with CR5.  The idea of the diode is to continue the current flowing through the relay coil to prevent voltage spikes when the current is suddenly shut off (V = L di/dt).  Allowing a bit more voltage drop reduces the shutoff time for the relay coil since allowing the V to go higher allows the t to get smaller.  C34 is to allow the footswitch contacts to bounce without tripping the relay.  If the switch is good enough you would not need it but some capacitance there may be necessary.

[digi2t]

Well, I started with the easiest swap. I had a A25K pot in stock, an exact type match I might add, so I replaced the A10K Lead volume. Did the trick. The Clean and Lead volumes are fairly equally matched now with the Gain at 50%. If I nudge the Gain up, I nudge the Lead volume down, and vice versa. Before, with the 10K pot, the volume was stacked way too far over to one side, which meant to get to equal volume between the two channels, I just about had to have both the Gain, and Lead volume close to dimed. Much better now.

I'm still wondering though how the 25K pot has affected the RC network around it though. The .022u, .015u caps, and 10K resistor around the pot MUST be playing some kind of role in conjunction with the pot (maintain certain frequencies as volume increases?), but I'm too dumb to know what. 

Will I have to to compensate for the new pot by adjusting the value of the 10K resistor to maintain the same response?

For the switching, remove the C34 electrolytic and put a couple of more diodes in series with CR5.  The idea of the diode is to continue the current flowing through the relay coil to prevent voltage spikes when the current is suddenly shut off (V = L di/dt).  Allowing a bit more voltage drop reduces the shutoff time for the relay coil since allowing the V to go higher allows the t to get smaller.  C34 is to allow the footswitch contacts to bounce without tripping the relay.  If the switch is good enough you would not need it but some capacitance there may be necessary.

Thanks for that, it got me thinking. I reminded me that my Tourmaster uses a bunch of relays for switching, so I decided to take a look at the schematic. It uses a 10u cap parallel to a diode, instead of the 22u found here. I swapped in a 10u for the C34, and it's better. The switching delay is shorter, short enough that it doesn't bother me. It's now a "click, click", rather than a "click....click" (yes... so scientific, I know). I can live with that, and I won't have to worry about adding capacitance for the footswitch later.

[reddesert]

I'm still wondering though how the 25K pot has affected the RC network around it though. The .022u, .015u caps, and 10K resistor around the pot MUST be playing some kind of role in conjunction with the pot (maintain certain frequencies as volume increases?), but I'm too dumb to know what. 

Will I have to to compensate for the new pot by adjusting the value of the 10K resistor to maintain the same response?

One way to think about this is in terms of reactance. The capacitors have reactance r = 1 / (2*pi *f *C) where f is frequency, so behave like a resistance that goes as 1/frequency. Consider the signal after R7.

At 100 Hz, the reactance of C5 = 0.022 uf is 72 Kohm, which is in series with the lead volume pot of 10K or 25K. The reactance of C16 = 0.015 uf is 106 Kohm. C16 is in series with R13, a 10K resistor and these are in parallel with the volume pot.

At 1000 Hz the reactances are 7.2K and 10.6 Kohm.

So, at 100 Hz, you're getting the output signal approximately from a voltage divider between the vol pot and 72 Kohm. If the volume is all the way up, you get about 25K / (25K +72K), while before you had 10K / (10K + 72K).  So it's twice as loud.

I said approximately because the vol pot is in parallel with the R13-C16 path. But at 100 Hz, the reactance of C16 is 106K, which is several times the volume pot, so the R13-C16 path to ground has little effect on the 100 Hz signal.

Now at 1000 Hz, the reactance of C5 is 7.2K and the volume pot makes a voltage divider with that. So now your 25K vol pot would only make it about 1.6 times as loud as the 10K.  And we have to consider the R13-C16 path. C16 has a reactance of 10.6K, so that path has a resistance of 20.6K.  This is in parallel with your volume pot. For the 10K vol pot, the R13-C16 combo is still more resistant than the vol pot, so it has a small effect.  But for the 25K vol pot, about half your current is actually going to flow through the R13-C16 path, so less voltage drop through the volume pot, so lower signal at the output.

A slightly more correct analysis would be to actually compute the resistance of R13-C16 in parallel with the volume pot, and then as a voltage divider with C5, but you get the idea.  It looks like increasing the volume pot will make the low frequencies louder but not the high frequencies, because R13-C16 provide an alternate path to ground for high frequencies.

You might be able to restore the frequency response by doubling the other components, but you might also just try replacing R7, the 1M resistor, with a 470K, and putting back the original 10K volume pot.

[digi2t]

Reddesert, thanks so much for the analysis! All makes perfect sense to me now. i just needed someone with this kind of knowledge to paint it out for me. Cheers man!

R7 to 470k? Dude...... that's like too easy. I guess it was too obvious for me to consider. DOAH!!

[digi2t]

OK, re-installed the 10K pot, and paralleled another 1M resistor onto R7. This was a good solution, since I didn't have to pull the main board to do it. It's perfect now. Frequencies are as they were before, and volume sweeps are much better. Your analysis regarding losing the higher frequencies as I rolled up the volume was spot on. It was getting a tad dark when the volume was getting above 70%. This little amp is a real keeper.

Thanks for all the help and info everyone. 

[Mark Hammer]

May we inquire what kind of amp it is?

[digi2t]

May we inquire what kind of amp it is?

[Mark Hammer]

[digi2t]

Didn't actually come like that, this is what I actually got;





I think the company might have stopped producing them (gone under?), and is/was selling off their remaining stock on Ebay. They were offering a range of "kits", which ranged from just boards to assembled chassis (like this one). It also came with the handle, corners, and feet so I can make an enclosure for it. They supplied all the schematics and CAD file for the chassis and enclosure. It was a gamble, for sure, but for $200 it paid off. It's a great little amp. Loud as hell, and the A/AB pot, as well as the Triode/Pentode switch, add quite a bit of tonal soundscape.

Now that the volume balancing is settled, along with the switching delay, I really like it a lot. They offered a set of Chinese tubes as well for additional $60, but I passed in favor of my own choice. I put a pair of JJ ECC83S in the preamp section, a Sovtek 12AX7LPS in the PI, and a set of Sovtek EL84M's for power. It rocks!

I still see some 12w, 15w, and 18w kits, as well as other odds and ends on Ebay.

[thermionix]

Wonder why the hole in the PT end bell?  Just in case you want to lick your finger and stick it in there I reckon.

I'm surprised you got good results with the EL84Ms.

[digi2t]

Wonder why the hole in the PT end bell?  Just in case you want to lick your finger and stick it in there I reckon.

I'm surprised you got good results with the EL84Ms.

Finger? Dang it dude, just get in there and go full on tongue man! Give those hot tubes a good lickin' too while your there. 

Seriously though, yeah, not bright. That's why I'll be closing this puppy up soon enough. I've already got the grating at work, I just need to get some nice hardwood for the sides. An aluminium plate will close up the bottom.