SLO type build - ghost notes on clean channel?

[Lindz]

Hello all

I posted this on the Sloclone forum but got no takers so I thought I would try here

have a weird issue with a 50 watt SLO turret build that I fabbed up recently. Lame phone pic attached, sorry I do not have anything better to upload.

I have everything working in general and the lead channel sounds great, but the crunch/clean channel sounds like it has ghost notes riding an octave low when I have the boost toggle switch that channel on clean and preamp volume control low. Like an octave pedal playing a few notes lower than the primary note. It is quite pronounced on all strings and all notes if the preamp volume is low - not subtle at all. If I turn the preamp volume up the screwy ocatavy sound gradually fades and it sounds normal with the volume way up.

If I flip the boost toggle switch on that channel to crunch and dial back the guitar for a clean sound it does not seem to do it, nor does it if I footswitch to the lead channel and dial the volume on the guitar down as clean as I can. So it seems like it would be related to something in or around that clean part of the circuit but I cannot figure it out. Have reflowed and replaced a few parts but no dice.

I had some issues getting that clean/crunch channel working at first (sounded like blocking distortion or signal not switching or leaking) where that channel sounded like ass (faint, distorted, kind of cutting out) the lead channel has always worked and sounded good. I went through and replaced a few LDR's thinking that might be the issue and now it seems to work pretty well on crunch on that channel now albeit with this octave ghosting when the toggle is switched to clean and the channel preamp volume is low.

Looks like I wired the clean/crunch switch correctly and I re-flowed a few connections around that area in case that was the problem.

Normally I would think that low value or leaky filter caps ala an old plexi might do this, but if that were the case would it not do it on all channels in all modes? I used six 100uf filter caps on the power board (4 for main power, 2 for screens, roughly half the value of those used in a 100 watt SLO) so I think that should be sufficient though I suppose one could be bad. I am considering paralleling or replacing some of them to check but want input first as to other things that might cause some low octave ghosting only in part of the circuit when the volume is low before resorting to that?

The other thing I thought is that it might be related to the LDR's somehow since I was having switching issues originally. Perhaps some signal is leaking there somehow? Below are my Opto voltages switched and unswitched (vtl5c1's) and a schem if anyone wants to take a look and suggest what I might look for there. Should the switched and unswitched voltages be similar? I am not really familiar with what they should be and the schem does not have a lot of info on it about them. Right now the voltages on 1, 4, 5 differ than 2, 3 when the different channels are active and perhaps I need to change a few resistor values so they are in spec - seems like the switching works now but I have this odd issue

Switching voltages:
LDR's 1, 4 5 read 2.45v on the anodes, 2.9v on cathodes when lead channel active
LDR's 2, 3 read 2.5v on anodes 1.1v on cathodes when lead channel active

LDR's 1, 4 5 read 1.9v on the anodes, 0.22v on cathodes when crunch channel active
LDR's 2, 3 read .22v on anodes 0v on cathodes when crunch channel active

Any other suggestions? This has me kind of stumped

thanks!

[gingertube]

An QADA (Quick and dirty analysis) of what you said above leads me to think you have a problem with the V1A stage.
Before doing anything else put a gridstop resistor that stage (pin 2), try say 22K to 47K and see if that makes any diference.
Cheers,
Ian

[Lindz]

Thanks for the reply Gingertube, I will do that tomorrow when I am soldering next.

That said it seems like this post about clean channel ghost notes might really be OD channel bleed into the clean channel

I noticed today if the crunch channel is active and I dial the crunch channel preamp volume all the way off there is still some sound coming through and if I then dial the lead channel preamp volume off it interacts and will shut off some of the bleed although even with both preamp volumes off there is a bit of sound still coming through.

The only way to get no sound at all is to turn the master volumes off

Most of the weird octave stuff quits when I turn off the OD channel preamp volume indicating the low octave ghost notes seem to actually be bleed and interaction in the clean/crunch channel from the OD channel

I guess the fact I cannot completely turn off sound using the preamp volume controls when the Clean/crunch channel is active likely makes it LDR related since neither channel is being completely muted unless I have some other wiring issue I have not caught. The OD channel seems to work fine

Maybe someone can help me sort out what the ideal voltages and values should be to switch the optos correctly - my current voltage info is above and the switching schematic is on page 2 of the PDF above

Changing r13 back to 47r from 33r (33 was recommended for 5 ldrs) seems to have slightly reduced the bleed and ghosting and it bumps up the voltage a slight bit to just under 3v but I am reluctant to go crazy changing values until I know what I should be shooting for - I have not quite wrapped my head around what the voltages and resistance should be in order to properly mute and switch

[gingertube]

Lindz,
I think you are right about the channel bleed. The LDR voltages you list are SUS.

Check:
Note 3 on Sheet 1 of the schematic
Note 13 on Sheet 2 of the schematic
AND
The NOTE "Switching Supply to be grounded to Chassis at the Footswitch Jack".

The LDR 2 & 3 OFF state voltages don't look to bad (approx 0.2V ONLY)
BUT the LDR 1,4 & 5 OFF State voltages are 0.45V.

These should be 0V. Check your switching wiring and that ground connection.

Ideally the On state voltage should be +1.4V or higher anode to cathode.
The OFF state voltage should be as close to 0V as possible or even a bit reverse biased.

Cheers,
Ian

[Lindz]

Gingertube

I believe the switching is grounded OK - used a non isolated jack (I did not separately ground to the chassis and assumed the non isolated jack would be sufficient) and I get near zero ohms checking resistance of grounds at various points in the switching circuit. That said I may be missing something or perhaps one of the resistors in the switching circuit is a bit high so there is a tiny bit more resistance to ground than should be - looking at the schem can you suggest what resistor(s) or ground would affect LDR's 1, 4, 5 off resistance assuming those are the issue

I believe I stated 2.45v not 0.45v above if that changes your thoughts on the LDR off voltages unless I misunderstood your comment. I am clueless but learning fast about how the Optos and switching work on this amp - built it from the schem and layout with minimal understanding of how Optos actually work voltage wise.

Also my PT did not have an extra tap so I had to install a separate transformer to power the switching so my voltage and current are likely a bit different than expected on the schem. But schem lists 6.3v to power it and my little transformer is 6.3v so I assume I am good there... after rectification and voltage drop from r61, r62 I have 5.4v dc. I currently have r62 at 100 ohm as that seemed to eliminate the ocatavy sound vs the stock value of 62 ohms although there is still bleed.

After more carefully measuring and playing with volumes etc it seems like I have a little bit of bleed in both channels. If I switch to the OD channel and turn off the preamp volume there is still a little bit of signal getting through from clean channel. Orders of magnitude less than when its on but if I dime the amp the bleed is pretty loud - almost as loud as a small amp, not exactly subtle. If I turn up the preamp the bleed is not noticeable as it is very small vs the signal in the channal.

Same thing on the Clean channel - but that little bit of bleed is just noticeable enough that it never seems absolutely clean, there is always this tiny amount of distortion bleeding in from the OD channel that makes it seem like there is a tiny bit of breakup happening.

In fact I could almost live with it except for that bit of distortion from the bleed and the fact that I want it to work perfectly, not with a bit of signal bleed

So I am back to trying to get the LDR's to mute completely or at least better than they are - (that or rip them out and switch to some relays ha ha)

I played with a few resistor values in the circuit going up and down a bit and some made it a bit worse some made it a bit better but never really much better. But I am kind of shooting in the dark and could use some suggestions and guidance as to what ideal values would be.

On the SLOclone forum there is some posts about people trying to shut up the bleed (apparently even real SLO's get a tiny bit of bleed) where one guy increased the cap in the switching from 1000uf to 2700uf and then hooked up LDR2 to the grid at v2a instead of v2b and that seemed to help. I do not have that value cap here but can you tell me what that would accomplish vs stock - if it seems worthwhile I will order up a bigger cap and try it

Anyone else chiming in with suggestions would be appreciated

thanks!

[roberto]

The lead of the slo sound that way also because of the bleed.
Try to measure the resistance of the opto instead of the voltages.
Should be around 100 Ohm when on and some MOhms when off.

[gingertube]

What is the Part Number of teh LDRs you are using?
Cheers,
Ian

[Lindz]

optos are Perkins Elmer vtl5c1

I am not sure sure if I trust my meter(s). I have 2 super cheap meters and they give different numbers and neither really seem close to ideal. Its time to spend more than $35 I think

The better of the 2 meters reads

switched to OD -
ldr5 over scale (1)
ldr1 1 meg
ldr2 1.5k
ldr3 400 ohm
ldr5 15k

switched to crunch
ldr5 580 ohm
ldr1 750 ohm
ldr2 1.5 meg
ldr3 48k
ldr5 600 ohm

[gingertube]

The datasheet for these is here:
http://www.excelitas.com/downloads/dts_vtl5c1c2.pdf

See the graphs on page 2.

For low ON resistance then there should be a minimum of say 10mA, 20mA would be better, thru' the LED. The forward voltage drop (anode to cathode) at 20mA would be 1.65V typical.

For High OFF resistance then you want no more than 100uA through the LED.
Better if you can get that LED current off completely, due to the LED forward voltage threshold, then as long as the forward voltage (anode to cathode) is below 1V then it should be OK. Zero bias or slight reverse bias would be so much better.

If I can blow my own trumpet here - I'm a Professional Electronic Design Eng. in the day job, I do reviews of other guys designs just like my designs are subject to review by other engineers. If that switching design had been presented to me for a design review I would have failed it and sent the designer back for another go. There are too many things interacting, it seems to fail the old Einstein Test of "Make it as simple as possible but no simpler".

You have some hope in that it presumably has beent built by others and made to work but it could be VERY much better without having to be much more complicated.

Cheers,
Ian

[Lindz]

Gingertube

That circuit is the Soldano SLO. If you are not familiar with it the SLO is a pretty famous amp introduced in the late 80's and some of the high gain techniques Mike introduced have been "borrowed" by a lot of big brand high gain designs since in one way or another (some blatant rip offs from some pretty big brands in fact)

that said there have always been lots of comments about how the switching is not so hot, but there are a lot of SLO's out there.

In any event it supposedly works OK in actual SLO's though even those amps sometimes have bleed issues. I have another clone I built using a clone PCB that switches better than this one.

My issue is weak knowledge in electronics theory in general and perhaps some mistakes, grounding or soldering issues in this build that may exacerbate any problems - I have Merlin's books, Oconnor's books and others but am still pretty green.

perhaps you can suggest some reasonably simple changes to help it along? I would greatly appreciate it. I have attached a layout to go with the schem so you have an idea what I have in there and where it is on the board which might help you visualize simple changes that would not require too much surgery. I'm thinking I may pick up a good meter so I feel I can trust the numbers a bit more than I do now as well (since my 2 different meters read different numbers on the same parts)

Even if I stick with the existing LDR switching but can vary a few components and or mute or switch from different locations in the circuit to shut up the bleed a bit I would be pretty happy with this build. Its my first completely scratch build making my own boards, chassis, cabinets, faceplates etc.

Unfortunately it looks a lot better than it works right now at least where the switching is concerned .. ha ha

[gingertube]

A quick and dirty (10 minute) design where the LED currents are all individually controlled by their own current set resistor.
So guys its up for the design review. Yes its is superficially more complicated but in operation its actually more simple/predicable.
Easy to add or subtract LDRs

It has 2 transistors and 2 diodes as active switching elements.
Diodes are not critical 1N4148/1N914 anything like that.
Transistors not critical either any general purpose NPN such as BC337, use the 'C' or 'B' variants ie. BC337B or BC337C to guarantee enough Beta (current gain).

Cheers,
Ian

[Lindz]

Thanks!!!!

I think I can figure that out...