Roland AS-1 Sustainer

[VintageGear]

Got myself an early Christmas present; a Roland AS-1 Sustainer. With little information available, I was curious. However, the unit I purchased raises even more questions. When the unit is engaged, there is a volume drop. Nothing serious, I'll just crank up the amp I thought. However, the sustain also seem to act like a volume knob. I can hear some sort of compression going on, but not really any sustain as I was hoping to hear.
What ya guys think? Broken, of badly designed?

[duck_arse]

does it have 2 9V batteries fitted? I was given one that was dead, seems the ldr gave up the ghost, but it was the last part I checked, and by that time, too late.

[anotherjim]

Hard to believe this is an Roland original design -  it's way too simple.

Like Ducks says, an LDR system fail would make the sustain control a pure volume. It should, with strong signal, drive the LED bright and the LDR resistance goes low which increases negative feedback in the 2 transistor pre-amp - thus dropping the gain.
That said, it's surprising that the effect level is too low. That 220k before the output level control implies plenty of gain, so much it's cut by about a 5th.
To me that suggests it's stuck at some compression level -  either the LDR has low resistance even when dark or the LED is stuck on, maybe due to leakage in Q3/Q4?

[Mark Hammer]

Just replace the 220k fixed resistor and 50k pot with a 100k pot and 180k fixed resistor for more output.

[VintageGear]

Great, thanks for all your valuable replies. Gives me a pointer where to look. Will get back to ya'll after the Christmas festivities!

[VintageGear]

Okay, got around to troubleshoot the AS-1. Started by replacing the nasty looking elec capacitors; very close to being completely perished. No change in effect.
As for the LDR, it gives a reading of about 7k ohms when no voltage is applied. When powering it up by plugging in the cord, it drops down and creeps up from 3k ohm to around 7k ohm after a minute of powering.
I dont quite understand what it means, but it doesnt sound right, does it?

[VintageGear]

Oh LOL, turns out its just a rubber cap over the LED and LDR, easy to seperate.
So, the LED quickly blinks upon plugging in the cord, and goes dark. The LED is working. The LDR is giving proper readings upon exposure to light.
New hypothesis: Q3 or Q4 faulty? How to check?

[duck_arse]

what is driving the led, Q3 and Q4? does the initial flash suggest anything about their condition?

you could try removing the ldr and tacking in a pot, see if you get correct compressings when you waggle the pot some. then, measure and post the voltages in and around the transistors, maybe mark-up a copy of the diagram with yr readings.

[anotherjim]

You have to be careful since any light getting to the LDR will lower it's resistance. If it can, ambient light will reach the LDR from the back of the LED. Sometimes these things only work properly when sealed in the enclosure.
Still, you know the LED can light and the LDR changes resistance according to light. The LDR though, should probably reach >500k in total dark. To prove that, you might solder some temporary wires to it and put in a dark box so you can get a dark reading.
Q3 and Q4, possibly one or both of them is failed, but there isn't any obvious way they could have been blown, unless someone's been messing about and managed to put the full 18v on the base of Q3. You should be able to use a DMM on diode test to check the base-emitter junctions. Otherwise you could temporarily short out that 0.047uF cap feeding the pair and should see the LED light when you do. If it does light with the shorted cap, then the transistors are working.

BTW, LED's do age and lose brightness. As this is pretty old -  it could be that.

[VintageGear]

Thanks guys, you all got me in the right direction.
I desoldered the LDR, clamped it up and put it in my drawer with the DMM on my desk. Reading keeps around 8k ohms max in total darkness (covered up and in drawer, should do it right )
Upon that finding and with 500k ohm to go on, I wired up a 100k pot (didnt have a larger one handy) and the LED started to react onto me touching the input plug when the pot was at maximum. Promising! So, out with the LDR, in with the new.
Err, yeah, about that.. time to source that part. Tips?

Edit: the diameter of the LDR is 8.1mm

[duck_arse]

easy way - tayda sells some cheep! cheep! look on flebay, they have many many cheap options from many places.

hard way - get the part number off the dead part and start searching.

[VintageGear]

Yeh plenty on fleabay or aliexpress, but I want to keep the workings as close to original as possible.
There is no partnumber on mine. Just a metal 8mm LDR
70 ohm in full exposure as stated in the schematic. What would be the upper limit in complete darkness? 500k as Jim stated? Or was that a wild guess?

[anotherjim]

Hee,hee! It was a wild guess. The ones I've used have all been about 1M dark. It seems to me that the light resistance at 10 lux is the main measure and the biggest variable between types and that's a LOT higher than 70ohm in the spec's I've seen -  more like 8k at 10Lux. I have no idea what Lux an "incandescent source" might be 

[duck_arse]

The LDR really isn't too critical - I've used ORP12, RPY58A, and some cheap Chinese unmarked things I got from Hong Kong on Ebay all with equal success. The colour of the LED doesn't matter either - just don't try to use a modern low current / high brightness type.

I'm working on the 18v version supasustain and I've found that different LDR's do change the tone of this pedal quite a bit. I've got an original supasustain to compare with and I haven't been able to find an LDR that hits sweet spot like the original.

both the above quots come from "the other place", in a thread on the AS-1, not surprisingly. I thort I'd seen a number for the ldr (nsl something?) and I looked on my original part, but it is also unmarked.

(the colorsound supasustain I have is not the same circuit as the AS-1, but fet based. just so's you know.)

[VintageGear]

Great, thanks for the information! I will find some cheap LDRs and see if I can find some at 1M dark. Will report back after they arrive!

[PRR]

> around 8k ohms max in total darkness

That's pathetic. You got a reject. Either dampness got in, or someone spilled coffee in the PbS vat.

The "dark" resistance is hard to specify, because it keeps going up and up the longer it is in the dark. But I never seen one not go over 10K in a second and 100K in less than a minute. Over 1Meg is typical after 10 minutes in total dark, but we rarely need anything that high (other parts in our circuit typically leak more).

Get 'em while you can. They have Lead, so are no longer made for New Production in Europe, which means it isn't worth making production lots for the Americas either. Small Lead is still legal for "repair", but nobody gives a hoot about the repair market. They may still be in large production for the China internal market (PbS is still king for dusk/dawn light switches and furnace flame detectors), but as they switch to other technologies for the World market they will be quitting Lead for their own needs. 

[VintageGear]

The LDRs arrived. Bought like 20 for couple of bucks. Of course they were smaller than the original 6mm LDR, so I neatly fitted it with some electrical tape, ensuring I covered up the back as well. Fits snug in the rubber casing along with the LED.
Yep, it sprung to life! However, if the sustain knob was set to 0, it started to self-oscilate / overdrive, resulting in nasty feedback, even when not playing. So, I started measuring. Because my DMM only goes to 2M resistance, I couldnt tell the max it was at (well > 2M at darkness), but I expected it was too big for the circuit. As a quick fix I installed a 2x 1M resistor (serie) in parallel with the LDR, capping the LDR to ~2M at dark (according to the DMM). Squeal gone, sustain audible.
Still, being the perfectionist I am, kinda curious if this is the correct approach. Wotchathink?

[anotherjim]

I think you hit on the correct solution.
I guess you found an unstable state of the circuit with the new LDR, but if it's happy now, we all can be

[Souleyman]

Have the same problem. Trying to use PerkinElmer Optoelectronics VT3ØN1 and VT33N1. They are in TO-5 housings and installed with no problems in the original rubber ring.
Same problem with selfoscillate when no signal. I selected the parallel resistor using pot. Stay now with 4m resistor for maximum volume and no selfoscillating. But seems I have some overload. Need more tests.
Still interesting to know what are the parameters of working original LDR...

[Souleyman]

All tests are ended for today.

Results:
Those little distortions are the artifacts of LED ripple.
The same situation I met with my BOSS CS-1.
It may be decreased with installation of electrolytic cap in the LED circuit but it changes that cool fast sound of an optical compressor.
Roland engineers made the place on the pcb with holes for this cap but not used it in original.
Me too.
The character and volume of those little distortions depends as well from the model and material of LDR.
I have only two now, as I wrote VT30N1 and VT33N1 PerkinElmer made.
And the VT33N1 sounds better. And both of them needs parallel resistance about 4m to prevent self oscillating when no signal on input.

But I've ordered some another LDR' for trying.
Silonex NSL-7530 and Luna NSL-5510 and NSL-6510.

Will see [emoji846]