Stacking BBDs

[xieliwei]

I have two issues I need help on here:
1. How do I stack BBDs, specifically the MN3008. I am trying to replace a MN3005 with two MN3008s since the 3005s are nearly impossible to obtain. My guess is that I can connect the output of the first 3008 to the input of the second one and allow them to share the same clock lines.
- Is there a potential problem to that? Clock sharing issues?
- What are the two outputs of the 3008 for? Are the two outputs the same or some form of inverted or direct output? I can't seem to find anything about it from the specs sheet.
- If the two outputs are the same, should I connect both outputs to the input or only connect one?

2. I'm quite sure the MN3005 has only 8 pins but on this schematic, it seems to be using 10 pins each, am I reading it wrong?

 

[Diver300]

If you look at that circuit diagram, you can see that the two MN3005s share the same clock lines, so I don't think there would be any problem connecting several together.

Also, the MN3005 labled A is wired as a reverb with the output connected back up to the input. I think that you can do the same to wire the output of one MN3008 to the input of the next.

The pin numbering on the diagram is odd, but it seems consistant. The MN3005 has 8 pins and the data sheet numbers them 1 to 8. However the IC is the same shape as a DIL-14 IC and the diagram uses those numbers.

Data sheet    Circuit     function
   1                 1         GND
   2                 2         CP2
   -                 3         no pin
   -                 4         no pin
   -                 5         no pin
   3                 6         Out1
   4                 7         Out2
   5                 8         Vdd
   6                 9         CP1
   -                 10        no pin
   -                 11        no pin
   -                 12        no pin
   7                 13        input
   8                 14        Vgg

What I don't understand is why the connection to non-existant pins is shown on the circuit diagram. Pins 4 (on one of the MN3005s) and 11 (on both) are not there. I guess that the connections go to those pins for layout reasons.

 

[xieliwei]

Ah... Thanks!

You're right, the clocks seem to be shared between the two 3005s.

Hmm, I don't seem to get this one. The input pin is pin 13, but if I'm reading it correctly, the output of the A 3005 is connected to the input of B 3005?

And it seems that the output of the 3005, even if it is to be reconnected back into the input of itself or another 3005, needs to be passed through an op-amp. Does that mean I need to do the same for my stack of 3008s?

If not, is it suffice for me to just directly connect the outputs to the input?

You're right, the unconnected pins seem to be used to pass signals through the underside of one 3005 to the other one (or in the case of pin 12, to the op-amp). I managed to find a picture of the PCB.



I also hit another few problems with the schematic.

1. Look at the NE570, specifically, pin 10. Notice that the 7.5kOhm resistor is totally bypassed with such a weird connection and also notice that pin 8 does not seem to be connected. My feeling is that one of those pin 10s is actually a mislabelled pin 8. Which one?

2. Also, can anyone figure out the unknown valued capacitor?

3. I see a lot of electrolytic caps wired to act like an unpolarised cap throughout the schematic. Any reason why they chose to do that over using a ceramic or other types of unpolarised caps? Their values seem to be small enough. Can I just replace them?

4. What specs should the diode to the right of the A MN3005 be?

Thanks again!

 

[Hero999]

Why are you using a BBD?

It's no longer the 1980s, use a digital processor IC like the PT2399.

PT2399 Datasheet pdf - Echo Processor IC - Princeton Technology Corporation

 

[Speakerguy]

Are there higher fidelity parts like the PT2399? (For hi-fi audio use). Thanks!

 

[Diver300]

Ah... Thanks!

And it seems that the output of the 3005, even if it is to be reconnected back into the input of itself or another 3005, needs to be passed through an op-amp. Does that mean I need to do the same for my stack of 3008s?

If not, is it suffice for me to just directly connect the outputs to the input?

That circuit mixes the outputs of both of the MN3005s and feeds it into the B MN3005. The opamp is needed for the mixing.

I think that you can connect directly.

I have no idea what the NE570 does.

You should be able to use a 1 :mu:F ceramic cap instead of the back-to-back electrolytics. It will probably be better.

What specs should the diode to the right of the A MN3005 be?

1N4148 would be a good start.

I agree with others that a BBD is very old technology. Those parts look like they will get more and more difficult to obtain.

 

[xieliwei]

Thank you all for the replies.

I agree with you. However, I am building this for a friend who believes in analogue sounding better than digital. It is a great experience trying this anyway.

But the MN3005 and MN3007 both have two outputs and one input. So I do have to mix the outputs? Or are the outputs identical?

Good idea, I will work on replacing the parts. Thanks!

 

[Mikebits]

1 uF would not be ideal for audio. About 1500 ohm at 100 Hz. You would lose low freq response.

 

[audioguru]

The NE570 was a Compander. It acts like a noise gate to reduce the horrible noise from the old bucket brigade echo ICs.

I don't think any of those old ICs are still available.

 

[Mikebits]

**broken link removed**
2006 data sheet https://www.electro-tech-online.com/custompdfs/2008/06/NE570-DPDF.pdf

 

[audioguru]

Small Bear and probably lots of other guitar accessory sites have lots of old parts.

Signetics invented the NE570 IC a long time before Philips bought them. I didn't know that ON Semi makes it now in a surface-mount package.

 

[Hero999]

I agree with you. However, I am building this for a friend who believes in analogue sounding better than digital. It is a great experience trying this anyway.

Tell him that a digital processor IC gives lower distortion than a BBD IC.

Like a digital signal processor, a BDD needs to sample the signal at fixed points so it isn't any better in this regard.

BDDs store the samples in capacitors, unfortunately capacitors discharge so there's attenuation and not all the capacitors are exactly the same value so there's a high degree of quantitisation error which distorts the signal.

A high quality digital processor will have a lower quantitisation error and therefore lower distortion than an old BDD IC.