Vintage Ross Flanger Repair

[miketbass]

Hello all, I have a red Ross flanger purchased from an auction site that is not fully functional. I am dealing with the seller right now but I anticipate a partial refund so I will keep the unit and repair it. This is the USA red version.

With the manual and width controls beyond 50% or so there is a very loud thump/chirp at the extreme of the sweep. This can be mitigated by adjusting the sweep to a narrow band (EDIT: the thump/chirp is audible at all settings but much more pronounced CW). I have probed the circuit and am getting a sharp square wave from the clock and a sweep of 30kHz to 650khz at the extremes. It appears that the BBD and clock are functioning fine and the problem probably lies in the LFO which appears to be an arrangement of transistors. As this unit is a clone of the MXR M117 flanger that is the schematic I am working from.

My hypothesis is that the LFO is pulling down the power supply at the highest point of its voltage. There is no hum present but my best guess is that there is a failing electrolytic capacitors in the power supply, possibly tied to the LFO. When I am ready to do the repair, would be replacing all electrolytics likely solve the issue? They all appear physically fine. I know it is good practice to replace electrolytics but from my many vintage stompboxes I have not had a cap failure like this. There are some tantalum caps on board too but I suspect they are in the signal path only and are fine. Is there anything else I should look for?

[miketbass]

So I put a scope probe on the 470uf power filter capacitor and I can see it bounce up and down in time with the noise. Does this mean the cap needs to be replaced, possibly with a higher value? I'm thinking that there is simply too much current draw for it to supply and as this is an AC power plug unit I cannot see a downside to a 1000uf or so cap. Any help is much appreciated.

[pinkjimiphoton]

i would start by replacing the electros.
check the pots too.

[miketbass]

Pots have been thoroughly cleaned. I tried strapping a 100nf and a 1000uf cap across the 470uf with alligator clips with no change. Perhaps its time to shotgun the electrolytics - I would really rather avoid swapping parts on this vintage unit but this seems the next logical step. I've been reading about decoupling the opamp supplies, they also use 4558 for all of the chips (including the LFO). I guess I'm just having a hard time believing that this is a design flaw instead of a circuit malfunction though, which again leaves me hesitant to begin replacing parts.

[ElectricDruid]

I agree that a malfunction is much more likely than a design flaw, but I wouldn't start replacing anything unless you've got a reasonable belief that the part might be faulty. Replacing all the electros "just because" is a long shot and might well cause more damage than it fixes - sorry Jimi!

With the "Width" control at 0%, you should be able to take the LFO out of the circuit. Now tweak the Manual control. Does it thump at the ends of the range? Can you still hear a thumping related to the LFO rate?

How many and what trimmers does the circuit have? Most flangers will thump if they're not set up correctly. Have you got a link to the schematic?

HTH,
Tom

[miketbass]

Link to the schematic I've been using as reference:

http://s82.photobucket.com/user/bainzy007/media/flangerm117_12.jpg.html

Its the MXR M117 factory schematic, which the Ross is a clone of. Trim pots are bias, level, and clock. Unit has been calibrated and set to 35kHz which gives me just enough range to get some good sounds, the lower settings on the manual makes it thump more as does anything past about noon for the width. The noise worsens when the frequency trim is dropped down, at about 32kHz or so it turns into a bird chirp. Basically the factory calibration specs make 2/3 of the settings unusable.

So I decided to scope it after stepping away for a bit. I'm getting about 100mv of ripple on the 470uf filter cap which doesn't seem excessive. All the voltages look good, 15.3v after the regulator nice and stable, clock outputs and the SAD chip all function fine. When I put the scope probe on the output of the regulator, I can SEE the sweeping of the flanger in it. I think my clock/LFO is bleeding into my power rail, as the frequency lowers it creates a sound in the human hearing range. Heterodyning?

Would a series resistor with the filter cap low pass the noise out of the power rail? Keep in mind I'm trying to do the very minimum amount of circuit molestation possible here. Suggestions?

[miketbass]

With the "Width" control at 0%, you should be able to take the LFO out of the circuit. Now tweak the Manual control. Does it thump at the ends of the range? Can you still hear a thumping related to the LFO rate?

Forgot to address this, with width completely CCW the rate has no effect on the sound. The full sweep of the manual control does what it should as well. So this function seems OK. So I guess its really the width (or "manual/auto" on the Ross) control that causes all of the problems when engaged. While the controls are all interactive with how extreme the problem is, with the width low enough (below 50% or so) everything sounds as it should.

EDIT: Hold the phone... it's those back to back 15uf capacitors in the lfo, which appear to be tantalum in mine isn't it?

[ElectricDruid]

Link to the schematic I've been using as reference:

http://s82.photobucket.com/user/bainzy007/media/flangerm117_12.jpg.html

Excellent, thanks.

Its the MXR M117 factory schematic, which the Ross is a clone of. Trim pots are bias, level, and clock. Unit has been calibrated and set to 35kHz which gives me just enough range to get some good sounds, the lower settings on the manual makes it thump more as does anything past about noon for the width. The noise worsens when the frequency trim is dropped down, at about 32kHz or so it turns into a bird chirp. Basically the factory calibration specs make 2/3 of the settings unusable.

35KHz seems very low for a flanger, except at one extreme of the sweep. I'd expect the centre frequency to be up in the 100KHz range somewhere.
I'm not surprised about the bird chirp when you bring it right down. That sounds like clock bleed through as the clock gets into the audio range, which it will if it's swept from 32KHz.

So I decided to scope it after stepping away for a bit. I'm getting about 100mv of ripple on the 470uf filter cap which doesn't seem excessive. All the voltages look good, 15.3v after the regulator nice and stable, clock outputs and the SAD chip all function fine. When I put the scope probe on the output of the regulator, I can SEE the sweeping of the flanger in it. I think my clock/LFO is bleeding into my power rail, as the frequency lowers it creates a sound in the human hearing range. Heterodyning?

That all sounds good; good voltages, clean clocks, etc etc. All good.

When you're scoping the regulator output, where have you got the probe grounded? I've often been caught out think I was seeing some noise in some rail or other, but actually I just had a bad ground.

You also say the "clock/LFO" is bleeding into the power rail. Which is it? Those would look pretty different on the scope, with one being a low frequency sweep and the other being a very high frequency square wave.

What range are you getting from the clock oscillator? Lowest freq? Highest freq?

Would a series resistor with the filter cap low pass the noise out of the power rail? Keep in mind I'm trying to do the very minimum amount of circuit molestation possible here. Suggestions?

I doubt a series resistor would help, because it'd filter noise *from* the power supply, but in your case, the noise is getting in *after* the power supply.

I don't really know what to suggest. It's not simple.

[miketbass]

The sweeping in the power rail appears to be a sine wave, not a square. It could totally be a bad grounding... As both the scope and the flanger are plugged into the same power strip I have been relying on the 3 prong ground they have in common.

About the frequency calibration. 35kHz is the lowest point in the sweep measured when all knobs are CCW. The MXR units have been measured at 29kHz to 32kHz. Mine was set at 32kHz when I got it so I just nudged it up a bit to get some more usable range in the knobs. This keeps it out of major bird chirp territory. Once I get it fixed I will go back to factory settings.

Answering your question about the manual mode/rate got me looking at the schematic more. I have done almost no work on flangers so I am not bringing a lot of knowledge to the table here. Anyways the width knob appears to simply pan between the LFO and the manual control. At one extreme the LFO is cut out, at the other its 100% in. As I turn the width clockwise I am favoring the lfo more and more until it is completely in circuit. This makes me think that the problem lies in that area. The only things I can see that would likely fail is the 15uf capacitors which look like tantalum in mine. The circuit rate and range of the width seem fine so maybe not, but it seems that the problem must be in the lfo.

[anotherjim]

... it's those back to back 15uf capacitors in the lfo, which appear to be tantalum in mine isn't it?

Oh, I hate tant's...
Probably, a single 6u8 NP electro could be a better choice... or 8u2. Wants to be 7u5 to equal those two 15u tant's.

[ElectricDruid]

... it's those back to back 15uf capacitors in the lfo, which appear to be tantalum in mine isn't it?

Oh, I hate tant's...
Probably, a single 6u8 NP electro could be a better choice... or 8u2. Wants to be 7u5 to equal those two 15u tant's.

Yeah, I hate Tants too. They _are_ a reasonable bet for a failure, but if it oscillates ok, then there's not much chance the caps are bad.

Alternatively, could the LFO bias have got way off, so that as you increase the width, it shifts the frequency range somewhere it shouldn't be? I guess that'd be set by R21/R22 the 15K/16K pair on the integrator op-amp.

[Scruffie]

I'd be looking at the 10uF off the voltage divider from the width pot.

[ElectricDruid]

Yeah, that's a good thought too.

[miketbass]

Alright... Did a little more digging, I decided that it was time to check all of the caps on board - specifically the tants for shorts. Guess what? Found a shorted tantalum between pin 7 of the 4013 tied to pin 5 of the 4558 in the LFO. It doesn't seem that this is on the MXR schematic, at least that I can see. The color bands are (from the top) brown, green, blue with a yellow stripe on one leg. This color code matches the back to back caps in the LFO.

The online calculators tell me this is a 1.6uf cap but I have a hard time believing it is that small. Seeing as this cap is tied from pin 5 of the 4558 to ground it would seem like any large-ish value would work here. Any recommendations before I proceed?

[miketbass]

Here is a picture of the capacitor in question. It seems these tantalum caps tend to short when they fail which makes them nice and easy to him down. Here's to hoping this will get it up and going!

[ElectricDruid]

Sounds pretty likely. That'd be across R21 in the LFO, so it's making that divider into more of a "proper" Vbias like we're accustomed to.

A bit sneaky to have faulty components that don't even appear on the schematic though, don't you think?! That's cheating!
I suppose Ross added a few bits over the MXR version after all.

[miketbass]

Welp, pack it in boys, nothing left to see here. Replaced the cap and rigged it all up to the amp..... Set it to full settings.... and plugged the power in. I just about crapped myself because I didn't hear anything. Gave a strum on the guitar and WHOOOOOSH there it was. I was so used to hearing the background noise that I thought it was toast! I'm off to calibrate and enjoy the rest of the holiday weekend playing this beautiful flanger. Thank you to all for your help!

[miketbass]

Alright, I'd like to leave a couple of notes in this thread to help anyone in the future looking for info on this flanger. The factory clock frequency sweep with the manual at the ends of the rotation are 32kHz CCW and 625kHz measured at pin 8 of the SAD1024. BBD bias is standard procedure. To trim the level, set the trim pot (upper left, between the auto and rate pot) for its mid point. Feed pink noise in to the input, controls set for midrange, regeneration CCW. Slowly turn the level trim clockwise until the flanger effect breaks through and sounds full. I use a pair of headphones and turn up loud, you will hear the effect suddenly stronger. This is as close to 50/50 mix as you can get by ear - this technique has worked well for me on many flangers.

All caps are rated for 25v,  all tantalums are 16uf, all opamps 4558.

[Fender3D]

Thank you for the clock numbers

I'm posting them @ the other site's BBDs archive