Roland SDD 320 Calibration !

Hello, I come to you because I ran into a problem.

I bought an old SDD-320 (Roland Dimension D) and I wanted to calibrate it, I was wrong and suddenly I unregulated the machine.

The first point that has misled me is that my machine does not have the same chips of amplification as on the doc (I have the chips of origin but it is a MAJ of the brand roland)

The chips of the first version are single amp amp, but I own MIL8 dual amp (an amp is not used a priori)

So I was referring to the documentation that told me to take me out of these chips (pin 1) to make the various trimming, but of course the Pin 1 on my chips are not the output, so I measured anything and totally unregulated my machine.

I suddenly have less effect, the chorus disappeared ...

My dual amp chips are JRC (I do not have the exact model sorry, but I could find the pin out on the internet diagram, so this is good)

But now I must confess that I do not follow on the trimming to perform for the calibration, the documentation is not clear, I have to calibrate the BBD Bias and the BBD Clock. During my first trimming I made sure that my signal (sin 1khz) does not vary, and I guess that's why I no longer have chorus (more modulation of the sound)

But what to do?

Of course, I enclose the roland memo.


Thank you

cordially

Bah. You trim a BBD by ear.

(And there is almost no reason to "calibrate"; the system is unlikely to go far out of bias even after decades.)

Look at what a  BBD accepts. Say you have a garage 15 feet wide. There is a 3 foot door, near the middle, but not exactly ON the middle (due to how much beer the carpenter had). You come home on your bike in the dark and try to glide through the door. Even if there is a full moon behind the garage, so you can sight-in on the center, you may miss, or bang one shoulder.

You want to "bias" your bike-path so the tires run through the center and wide shoulders brush equally on both sides.

Set the Bias pot to the center. Feed a steady music signal. AM radio is great. Compressed pop will work. Listen to the output.

Pot centered is likely to pass "some" signal, garbled. Like your bike hits the door-frame but your shoulder flies through the door.

If you really have NO signal with bias pot centered, gently turn the bias pot one way or the other. I doubt you need even 1/4 turn (that pot's range is more than needed).

Keep trimming for best sound (least distortion).

If sound is still distorted at best seting, reduce input level until barely-distorted. Trim again.

If signal gets very clean, increase signal to some distortion and trim again.

A BBD will never reach NO distortion. But you can find a bias and level where the distortion is musically mild.

The Balance trim is very subtle. While I could find an optimum on 'scope, to the ear anywhere near the center sounds all the same. Way off balance and long delay times leaves a high-pitch whine, trim for the center of the least-whine zone.

Next time you "calibrate", MARK the factory settings and verify your document against the actual equipment before you pick up the screwdriver.

PRR said:

Bah. You trim a BBD by ear.

Click to expand...

I guess you could but I always used a scope and trimmed for largest (mostly) clean sine wave.

(And there is almost no reason to "calibrate"; the system is unlikely to go far out of bias even after decades.)

Look at what a  BBD accepts. Say you have a garage 15 feet wide. There is a 3 foot door, near the middle, but not exactly ON the middle (due to how much beer the carpenter had). You come home on your bike in the dark and try to glide through the door. Even if there is a full moon behind the garage, so you can sight-in on the center, you may miss, or bang one shoulder.

You want to "bias" your bike-path so the tires run through the center and wide shoulders brush equally on both sides.

Click to expand...

The optimal DC bias can shift a little with clock frequency so depending on the effect you may want to check at short (high clock frequency) and long (low clock frequency) delay.

Set the Bias pot to the center. Feed a steady music signal. AM radio is great. Compressed pop will work. Listen to the output.

Pot centered is likely to pass "some" signal, garbled. Like your bike hits the door-frame but your shoulder flies through the door.

If you really have NO signal with bias pot centered, gently turn the bias pot one way or the other. I doubt you need even 1/4 turn (that pot's range is more than needed).

Keep trimming for best sound (least distortion).

If sound is still distorted at best seting, reduce input level until barely-distorted. Trim again.

If signal gets very clean, increase signal to some distortion and trim again.

A BBD will never reach NO distortion. But you can find a bias and level where the distortion is musically mild.

Click to expand...

eyeballing a sine wave with a scope was always easier for me.

The Balance trim is very subtle. While I could find an optimum on 'scope, to the ear anywhere near the center sounds all the same. Way off balance and long delay times leaves a high-pitch whine, trim for the center of the least-whine zone.

Next time you "calibrate", MARK the factory settings and verify your document against the actual equipment before you pick up the screwdriver.

Click to expand...

The clock frequency trim again is easier with a scope. The trim balances the DC level between pairs of consecutive output samples (same sample repeated 2x) to avoid a full scale voltage step between samples, occurring at the clock frequency.  You "might" be able to measure for HF noise at a low clock frequency, but again much easier to see with a scope, since there will be LPFs trying to filter out clock noise.

BBD delay was only high fidelity compared to passing signal through a long garden hose (google "Cooper time cube"), but you can make decent efx with them for modest amounts of delay.

JR

I would think the issue is not with biasing the BBD's, but rather making the clock do what it should. A large part of the circuitry is devoted to making the modulation circuitry drive adequately the clock generator. An oscilloscope is de rigueur then.

Hello, thank you for your answer. Of course I have an oscilloscope, my mistake was to ensure that the signal that I perceived the oscilloscope was frozen .... that's why I no longer have modulation, the signal must be in fluctuation !!

But the documentation does not help me because by changing the action of the potentiometer, my signal did not move as I hoped.

I will replace Caps because of the age of the machine and I will have to re-calibrate it ....

If I can not do it, I'll try as you mentioned, at first to do it with ears, although this must be part of a first approach.

I would do tests this week, maybe I would get better ...


O and merry christmas!

Hello, I found the problem  , I mention it if it can help someone. For the adjustment of the machine the doc mentions chips for the measurement, but there is an error:
- CMR chip 103 for R and 203 for L
- BDD chip 109 for R and 209 for L

there is an invertion, it's chips 109 and 209 for the CMR and 103 and 203 for the BBD.

note also that being double opamps IC103 and IC203 only use one of them, and so for 103 the output pin is 1 but for 203 the output pin is 7.

here it may help somebody.

Thanks for the update.