1054 vs 7660 for higher loads

[temol]

I have a circuit that requires a bipolar supply - somewhere between +/- 9-15V so charge pump seemed a good solution. But there are three 4558 (six opamps) to feed and each IC draws around 7mA. As for now I only have 7660/1044 charge pump. With 9V supply voltage I have +/-6 V from the 7660 because 21mA current draw. From what I see 1054 is capable of delivering up to 100mA. Do you have any practical experience in this matter?

 

[jesuscrisp]

How exactly do you build your charge pump, can you provide a schematic? I have not yet used 3 dual ICs but I did a quad IC at +/-18V and that worked fine.

 

[mybud]

There’s some info on this topic over at Mad Beans. If I recall rightly, it’s under the Road Rage rubric and he talks about exactly this issue (current draw for charge pumps).

 

[bowanderror]

I had a few +-15V charge pump boards made up a while back that used the LT1054 and they've worked great so far. I haven't really pushed them to test out what kind of current they can provide, but here's the schematic:

 

[jesuscrisp]

I had a few +-15V charge pump boards made up a while back that used the LT1054 and they've worked great so far. I haven't really pushed them to test out what kind of current they can provide, but here's the schematic:

View attachment 24561

Could you explain to me what D4 does or why it's important? When I first breadboarded a +/-18V charge pump using a 7660S I had no issues when I left it out.

 

[temol]

I'm using this


Voltages under load
No opamps: +8.88V/-8.87V
2 opamps (1x4558): +8.1V/-7.93V
4 opamps (2x4558): +7.4V/-7.04V
6 opamps (3x4558): + 6.74V/-6.23V

 

[bowanderror]

Could you explain to me what D4 does or why it's important?

Not sure the exact function, but it looks to be part of the charging circuit for the negative rail charge pump. I took this circuit directly from the LT1054 datasheet:

When I first breadboarded a +/-18V charge pump using a 7660S I had no issues when I left it out.

I don't see any dual output voltage doubler designs on the ICL7660 or TC7660 datasheets, so I'm not sure if it's required or not. I've really only built +9V to +-15V charge pumps with the LT1054, because I nearly always need the extra current capability and it's been easier to source than the 1044/7660 over the last year or two.

 

[jesuscrisp]

I'm using this
View attachment 24638

Voltages under load
No opamps: +8.88V/-8.87V
2 opamps (1x4558): +8.1V/-7.93V
4 opamps (2x4558): +7.4V/-7.04V
6 opamps (3x4558): + 6.74V/-6.23V

Have you tried reducing the 100R CLR (R31) to maybe 10-22R? Also I'd just leave out C25 and C26, they seem rather unnecessary. Might also try leaving out D7 altogether, as long as you don't willingly put in more than 9V it'll be fine.

 

[temol]

For testing, I planned to reduce R31 or maybe even remove R31, C27, C28 altogether, but I haven't had time to deal with it yet.

 

[temol]

Lowering R31 to 10R helped a bit. Now, there's more than +/- 8V under full load.

Lesson learned - I should not blindly redraw schematics found on the net.

So .. If you plan to build Ampeg VH140 using schematic from dirtboxlayouts then remove 100R in the power supply section or use much lower value. This is true for 9V supply voltage. I could not (in fact I din't want to) test the circuit with 12V because of the maximum allowed input voltage for the 7660. I only tested 10.5V and there was a slight improvement. But I want the cuircuit to work properly with standard, 9V supply.

In the first place I should describe the problem caused by IC's supply voltage that's too low, here in VH140. Preamp sounds completly fine until gain is @ 14-15 o'clock. Then, when playing palm muted, short notes, there's a brief moment of silence after note dies out and after a half second it comes back to normal "operation". It's like a slow noise gate that works in reverse. When it's not gating you can hear preamp's usual noise with slight background hum, then moment of silence and noise again. With 9V supply it's quite strong effect. With 10.5V only appears at extreme gain setting.
I'm not that good to explain the nature of it but I suppose it's related to lowered headrom @ 4th and 5th opamp, maybe even 3rd one. Also arrangement of the diodes in opamps feedback loop, electrolytics connected between feedback loop and gnd.

 

[bowanderror]

Lowering R31 to 10R helped a bit. Now, there's more than +/- 8V under full load.

Lesson learned - I should not blindly redraw schematics found on the net.

So .. If you plan to build Ampeg VH140 using schematic from dirtboxlayouts then remove 100R in the power supply section or use much lower value. This is true for 9V supply voltage. I could not (in fact I din't want to) test the circuit with 12V because of the maximum allowed input voltage for the 7660. I only tested 10.5V and there was a slight improvement. But I want the cuircuit to work properly with standard, 9V supply.

In the first place I should describe the problem caused by IC's supply voltage that's too low, here in VH140. Preamp sounds completly fine until gain is @ 14-15 o'clock. Then, when playing palm muted, short notes, there's a brief moment of silence after note dies out and after a half second it comes back to normal "operation". It's like a slow noise gate that works in reverse. When it's not gating you can hear preamp's usual noise with slight background hum, then moment of silence and noise again. With 9V supply it's quite strong effect. With 10.5V only appears at extreme gain setting.
I'm not that good to explain the nature of it but I suppose it's related to lowered headrom @ 4th and 5th opamp, maybe even 3rd one. Also arrangement of the diodes in opamps feedback loop, electrolytics connected between feedback loop and gnd.

I remember running into this CLR issue on a build with that +-15V LT1054 board. I think I settled on 10R or 15R as well.