Ludwig Phase II Clone Debugging and Miscellany

[theehman]

2) It should be corrected on the parts list that the jacks required should be; 2 mono for the outputs, and 2 mono switched for the inputs. The switched jacks should go to ground when not in use. Otherwise, you'll get all kinds of noise from the unused input. This reflects the original units.


Cheers,
Dino

Mouser BOM updated for this suggestion.  Thanks, digi2t!

BOM for ordering:  http://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=9e9fb9c47d

[digi2t]

Thanks Ron, very kind of you.

Cheers,
Dino

[Skruffyhound]

Yeah, great Ron, that's just golden.

[digi2t]

Well, the animation speed switch (Fast/Slow) being upside down was really bugging me. What can I say, I'm just anal that way. I prefer Fast being up, with the LED on, and Slow being down, with no LED. The original units are configured this way as well, Fast up, Slow down I mean. So, a quick and dirty work around. 2 trace cuts, and 2 jumpers to flip that side of the switch around. LED side of the switch is fine.

Snapshot of the mod;


I can sleep soundly now  . Besides, it gives the board a sort of "mid-production fu*k-up, engineering department fix up, we'll hire a student for the summer to mod all the boards, but the company went under anyway" kinda mojo.

Cheers,
Dino

[Keppy]

Besides, it gives the board a sort of "mid-production fu*k-up, engineering department fix up, we'll hire a student for the summer to mod all the boards, but the company went under anyway" kinda mojo.

Because this effect needs MORE of that.

[digi2t]

Because this effect needs MORE of that. 

[R.G.]

I like it. It does add that veneer of authenticity to it, doesn't it? 

[digi2t]

Hey R.G., if I wanted to lower the input volume on the High-Z, which resistor should I be adjusting around Q2? I find that in straight filter mode (no fuzz), my guitar signal is distorting. I need to turn down my guitar volume for it to clean up.

[R.G.]

Increase R11 from 10K to something larger.

I think it's probably overdriving the next transistor, not distorting in Q2. This will lower the level without messing about with the input impedance at Q2.  Let me know if that doesn't work.

On the 'fast-is-up' mod: I looked at it. I turned the switch 180 degrees and re-routed the wires. This brought up the question in my mind about which direction turns on the LED. You want that on with 'fast' right? This would be a good time to think about making it bicolor, one color for fast, one for slow. I think that's possible as well.

[digi2t]

Increase R11 from 10K to something larger.

I think it's probably overdriving the next transistor, not distorting in Q2. This will lower the level without messing about with the input impedance at Q2.  Let me know if that doesn't work.

On the 'fast-is-up' mod: I looked at it. I turned the switch 180 degrees and re-routed the wires. This brought up the question in my mind about which direction turns on the LED. You want that on with 'fast' right? This would be a good time to think about making it bicolor, one color for fast, one for slow. I think that's possible as well.

First off, thanks for the input advise. I'll try it and let you know.

As for the speed switch, like I said before, the LED is fine. As you can see, my jumpers are effectively just swapping the traces around 180 degrees. So re-routing the traces would be just what the doctor ordered. But, a bi-color LED for the Fast/Slow would be the cat's ass. That would be in keeping with the original's "Christmas tree" look.

I've posted a survey in the For Sale/Wanted section to see if another run of boards has merit. If so, we could incorporate this mod to the new run. In the meantime, if anyone is bothered with the speed switch, they can simply mod it as I did.

[Govmnt_Lacky]

...a bi-color LED for the Fast/Slow would be the cat's ass. That would be in keeping with the original's "Christmas tree" look.

Recommend a common ANODE for the bi-color LED. Less parts involved ( Okay... its only one less resistor but its still less parts!  )

[digi2t]

Increase R11 from 10K to something larger.

I think it's probably overdriving the next transistor, not distorting in Q2. This will lower the level without messing about with the input impedance at Q2.  Let me know if that doesn't work.

OK, I tried it with a 25K trimmer, so I could sweep through and see what sounded best. Ultimately, it did reduce the input volume, slightly, and then there was no more difference. I also tried a 50K trimmer to if it would bring it down some more, but no dice. It seems to hit a certain point, and stop there.

Would a different transistor type in this section would make a difference, like a 2N2222?

*EDIT* - Just a thought... Could it be that we are distorting at the filter level? Is it possible to adjust the input volume just before one, or both, of the filters? 

[R.G.]

hmmm.

That's odd. Increasing that resistor should drop the volume pretty readily.  Q3 can be thought of as a kind of current-input gain stage. Increasing the series resistors with its input should drop the signal level from the source going into its base. Obviously, if R11 gets to being a few tera-ohms (an inch or two of air  ) I'd expect the signal level to drop a lot. What happens when you sub in a 1M pot and diddle with it?

Q3 has a quite low input impedance, and the signal currents from the lo-z and hi-z inputs are mixed by the currents through the series resistors coming into Q3 base.

We can always split the emitter resistor of Q2 and dial the voltage down, but that's clumsy.

What happens with an even bigger resistor? BTW - it may be that the loading on the signal from the mixing resistor coming from the low-Z side is causing the plateau. If you don't get good results, I'll see if I can run some sims to see what gives.

[digi2t]

hmmm.

That's odd. Increasing that resistor should drop the volume pretty readily.  Q3 can be thought of as a kind of current-input gain stage. Increasing the series resistors with its input should drop the signal level from the source going into its base. Obviously, if R11 gets to being a few tera-ohms (an inch or two of air  ) I'd expect the signal level to drop a lot. What happens when you sub in a 1M pot and diddle with it?

Q3 has a quite low input impedance, and the signal currents from the lo-z and hi-z inputs are mixed by the currents through the series resistors coming into Q3 base.

We can always split the emitter resistor of Q2 and dial the voltage down, but that's clumsy.

What happens with an even bigger resistor? BTW - it may be that the loading on the signal from the mixing resistor coming from the low-Z side is causing the plateau. If you don't get good results, I'll see if I can run some sims to see what gives.

Hey!!! I just had an idea!

Looking at the Fuzz footswitch, the com feeds the filter circuit, with one side providing the clean sound, and the other the fuzz. The real problem is the clean side. Not only with the distorting, but the fact that I can't get unity with the bypassed signal, even with the bypass pot at max. So, I took a 10K pot, and inserted it between the fuzz footswitch clean pole and ground. This way I can turn down the clean signal hitting the filters, by going to ground. Well Holy Crow!, it did the trick. Not only did it tame the distortion, but I can achieve unity with the bypass signal now. Now, when the fuzz is engaged, it's much louder, whereas it was the opposite before. I guess the same trick could be used on this pole to balance between the clean, fuzz, and bypassed signals.

So how about this for another "quick and dirty" workaround; Since one of the poles on the Fuzz footswitch is the ground for the LED, I can bridge resistors from the clean and fuzz poles to lower the input level to the filters. This allows me to find unity with the bypass mode as well, with the pot maxed.

For the clean side, I found anywhere between 390 ohms to 470 ohms gets me into the right zone. For the fuzz side, 1.2K to 1.5K seems to be pretty good. I realize that there's not much resistance on the clean side, but like I said before, the clean side is REALLY loud compared to the bypassed mode.

Can you sim that R.G.? It sounds alright to me, but maybe I'm messing with something downstream.

[digi2t]

OK, I decided to forego the quick and dirty fix above, and have taken a look at the transistor voltages. Placing the original and clone side by side, I've looked at the differences between voltages in the input/fuzz section.

For the record, the main voltage of the original is at 33.61vdc, and the clone is sitting at 35.91vdc.

All the transistors in this section (Q1, Q2, Q3, Q4, Q5, Q6, Q21, Q22), but one, are fairly close to each other, taking into consideration the main voltage difference. The one that seems to be off is Q22. From the way I understand it, this is the last tranny in the clean signal chain feeding the filters, connection point 17 in the original schematic. The fuzz'd signal comes from connection point 16.

As far as the voltages between the two Q22's are concerned, here they are;

Original -
E - 0.877
B - 1.495
C - 15.73

Clone -
E - 0.761
B - 1.369
C - 20.57

As you can see, big difference between the collector voltages. All the other trannies are very close to each other, considering the almost 2.5 volt difference in the main voltage supply.

Could this be what's overdriving my signal? If so, how do I go about toning down Q22?

[R.G.]

Took a bit. Sim kept crashing and I had to so some system reconfig.

The simulation matches the measured from the clone pretty closely, both with 35.9V and 33.6V for a power supply.

I'm guessing that the original has some drifted resistor values or some other issue.

On the other side of this, Q22 does present a good place to change gain, I guess.

Try putting a 5K pot in series with the 470R emitter resistor on Q22 and dink with it. In sim, this changes the gain from as much as 26db down to 1.8db.

[digi2t]

Try putting a 5K pot in series with the 470R emitter resistor on Q22 and dink with it. In sim, this changes the gain from as much as 26db down to 1.8db.

Thank God you mentioned that. Just before going to bed last night, I took and educated guess, and while the DMM was connected to the collector I just touched a 47K resistor in parallel with the 470R. When I saw the voltage fluctuate, I thought I might be on to something. I was going to try the pot test tonight, starting with a 1K though. I`ll report my findings thereafter.

[digi2t]

OK, I messed about with it, and playing with this resistor doesn`t help. Now I`m debating if it`s the effect that`s distorting, or whether it`s overdriving the input on my amp?

Sidebar: After doing some research, I found that the transistor used in the Q1, Q2, Q3, Q4, Q21, and Q22 spots, are much closer to 2N5088 spec`s, so I swapped them out. My voltages are much closer now to the original, especially for Q22`s collector, which is now sitting at 16.10vdc with the 2N5088. I found this info going back in Jimi`s original Ludwig thread. There was an EV amp manual link, which spec`d the original numbers, and equivilants.

Anyway, since I know that turning my guitar volume down was the only sure way to tame the beast, I looked at the input side of things. Reducing R6 seems to produce the best result. I can crank up my guitar volume now, and by reducing the resistance here, I`m basically reducing the input volume, and the distortion as well. Also, this now puts the Bypass Balance volume, at max, higher than unity, which is great, since I don`t have it dimed now. This effect is just plain loud. Much louder than the bypassed signal, especially in clean FFM mode. I think I`ll just shoehorn a trimmer in place of R6, adjust it, and button it up.

Does this make sense to you R.G.?

[R.G.]

OK, I messed about with it, and playing with this resistor doesn`t help. Now I`m debating if it`s the effect that`s distorting, or whether it`s overdriving the input on my amp?

Hmmm. That resistor makes dramatic changes in the gain of Q22.

My take? It's oscilloscope time. Nothing like seeing the waveform.

Sidebar: After doing some research, I found that the transistor used in the Q1, Q2, Q3, Q4, Q21, and Q22 spots, are much closer to 2N5088 spec`s, so I swapped them out. My voltages are much closer now to the original, especially for Q22`s collector, which is now sitting at 16.10vdc with the 2N5088. I found this info going back in Jimi`s original Ludwig thread. There was an EV amp manual link, which spec`d the original numbers, and equivilants.

Good bit of knowledge. However, there is a reason I violated my own Second Law and didn't recommend the 5088. The 5088 is rated at only 30V for Vceo. It'll probably be OK - which means no decent engineer would ever use it there. Probably doesn't work in engineering situations. I would recommend the 2N5210 (Vceo = 50V) for that, as it has a high gain and low noise, although neither is as good as the 5088. The 2SC1815 would be a really good choice as well, although it doesn't have the same EBC pinout - it's ECB.

Anyway, since I know that turning my guitar volume down was the only sure way to tame the beast, I looked at the input side of things. Reducing R6 seems to produce the best result. I can crank up my guitar volume now, and by reducing the resistance here, I`m basically reducing the input volume, and the distortion as well. Also, this now puts the Bypass Balance volume, at max, higher than unity, which is great, since I don`t have it dimed now. This effect is just plain loud. Much louder than the bypassed signal, especially in clean FFM mode. I think I`ll just shoehorn a trimmer in place of R6, adjust it, and button it up.

Does this make sense to you R.G.?

It does. If your tinkering produces good results with modern guitars and amps, I'll probably put in an alternate position to make R6 either fixed or a trimmer, user's choice, in any future layouts.

The Ludwig is an example of how engineering, even from a knowledgeable audio company, can go awry in the face of the music industry's unspoken "standards". It's big, complicated, special purpose, and - did I mention complex? - probably hurried to production with minimal testing.

And it's a relic. Today, we'd do that with some opamps, maybe a uC, and shoehorn all the electronics into one footpedal enclosure.

Good detective work, Dino.

[digi2t]

It does. If your tinkering produces good results with modern guitars and amps, I'll probably put in an alternate position to make R6 either fixed or a trimmer, user's choice, in any future layouts.

I`ll run some more tests with this over the weekend and report my findings. I wasn`t able to get up to really loud volume levels yesterday, but it seems to be the best solution. Also, I want to try it with two other amps as well, just to validate the mod. The alternate resistor/trimmer setup would be great though.

Good detective work, Dino.

All thanks are owed to you, and Mark Hammer for where I am today in my stompbox education thus far. Who says that you can`t find things, walking in the shadows.