Oberheim OB-8 problem

[living sounds]

I bought a supposedly perfectly cared for OB-8. It's got a new PSU (switch mode + additional passive and active filtering) and I was able to calibrate it to perfection.

The problem is with the +5 voltage, which sags down the more LEDs are turned on. It sags even right at the PSU output. At a certain point playing in unisono mode (which fire 8 LEDs simultaniously) the synth starts to act up and switch programs.

The all-LED-on test shows some LEDs fully lit and others flickering rythmicly, with a corresponding noise coming from the outputs.

It seems to be related to the procecessor board, at least the voltage drop (only 1-2 mV per LED) happens even with the other boards disconnected one by one.

The PSU is fine. The guy who did the mod (pro synth repair shop) suggested a 4051 was probably to blame - does that make sense to you? There is one on one of the other boards, but not on the processor board.

A 25W regulated supply that i used for some digital stuff and LEDs shouldn't show a voltage drop like that, should it? I couldn't find anything getting hot either.

The fact that some of the LEDs light steady while others don't suggests to me, that the circuitry switching those LEDs seems to have a problem.

Anything obvious I should check? Thanks!

 

[gyraf]

from its users manual, page 47:

Power (..) : 46Watts

Even with some lost for inefficiency, your 25W supply may not be enough

Also, check schematic if LED's drop current into a "dirty" rail of it's own, and if that is as expected

/Jakob E.

 

[living sounds]

Thanks Jakob.

Besides the +5V PSU there is also -5V and +/- 15V.

But with a regulated PSU, provided the rating is adequate, the voltage should remain steady, right?

 

[JohnRoberts]

Thanks Jakob.

Besides the +5V PSU there is also -5V and +/- 15V.

But with a regulated PSU, provided the rating is adequate, the voltage should remain steady, right?

Yes, as long as it doesn't run out of current, or unregulated voltage.

Does it need <25W or  46W?

JR

 

[living sounds]

Yes, as long as it doesn't run out of current, or unregulated voltage.

Does it need <25W or  46W?

JR

The synth repair expert who put in the regulated PSUs thought it would be enough. The 3 switch mode PSUs he put in make for a combined 75W, and I would guess that +/-15V, which powers the considerable amount of active analog circuitry, draws the most current.

I think (hard to read my copy of the schematics) the original PSU used a 7805 regulator, which is rated for 1.5A, so the 5.0 switching PSU should be plenty enough.

 

[JohnRoberts]

The synth repair expert who put in the regulated PSUs thought it would be enough. The 3 switch mode PSUs he put in make for a combined 75W, and I would guess that +/-15V, which powers the considerable amount of active analog circuitry, draws the most current.

I think (hard to read my copy of the schematics) the original PSU used a 7805 regulator, which is rated for 1.5A, so the 5.0 switching PSU should be plenty enough.

Can you find the output regulation spec for that model switching supply? ASSuming there is no resistance in series with the output it shouldn't sag noticeably.

Don't ignore the wiring and ground return path... resistance there could cause a voltage drop with current.

Are you measuring voltage right at the regulator or in the audio circuit?

It will probably be the last thing you check.

JR

 

[living sounds]

Can you find the output regulation spec for that model switching supply? ASSuming there is no resistance in series with the output it shouldn't sag noticeably.

Don't ignore the wiring and ground return path... resistance there could cause a voltage drop with current.

Are you measuring voltage right at the regulator or in the audio circuit?

It will probably be the last thing you check.

JR

There was a voltage drop right at the PSU output, those were the measurements I was refering to (voltage going down 1-2 mV per additional LED). On the panel board, with additional series resistance, the drop was quite a bit more.

Load regulation is rated at +/- 1%  from 0% to 100% rated load. So the minimal increase in load by the LED really shouldn't matter, I think.

The strange thing is that the energy needs to be burned somewhere, but I couldn't find a hotspot anywhere.

In any case, it's no longer in my hands now.

 

[living sounds]

Should there be any measureable (millivolt range) voltage drop resulting at the regulator, from adding a few LEDs here, with the PSU far below maximum load?

 

[JohnRoberts]

Should there be any measureable (millivolt range) voltage drop resulting at the regulator, from adding a few LEDs here, with the PSU far below maximum load?

1% of 5V is 50mV.

The regulation spec says 1% from 0% to 100% of load, it may not be linear... some regulators need a little load to regulate properly.

JR

 

[living sounds]

1% of 5V is 50mV.

The regulation spec says 1% from 0% to 100% of load, it may not be linear... some regulators need a little load to regulate properly.

JR

Yes, and it's +/-1%, so in reality the possible deviation would be 100mV, right? However, the original regulator is supposed to deviate 50mV max within 1.5A (not 5A), so this one should be better (though as you said, it might not be linear).

There is already a permanent load, since the +5V also serves lot's of logic ICs, the CPU etc.

The repair shop guy (got an engineering degree) who had modified the PSU and where the seller had it sent to fix it, now upped the output voltage of the switching PSU and declared the unit "fixed" (no more crashes in unison mode) and the all-LEDs-on test (which the unit still doesn't pass) non-essential. The tape interface isn't working either, but according to him it isn't necessary anyway (since there is a midi interface). All the digital calibration settings stored in the synth are only transferred via the tape interface. Yay!  :