anyone familiar with R5101 CCD Delay chips and has anyone used 2 x mn3005

[njkmonty]

anyone familiar with R5101  CCD Delay chips?
and has anyone used 2 x mn3005  instead?
   

[Mark Hammer]

1) CCD = BBD

2) The R5101 is 4096 stages, equivalent to a single MN3005/MN3205/SAD4096, though not pin-for-pin compatible.

3) Use of two MN3005/MN3205 chips to extend delay time is actually quite common.  There are two issues to be addressed when doing so:

• whatever clock is driving the chips has to be able to get past the combined input capacitance on the BBDs (700pf per 3x05 chip, summed over chips)
• sometimes you can get lucky, but one generally has to rebias the input to each BBD separately

[njkmonty]

so its potentially doable?

I have a friend who wants a smaller boss dm1 delay as his original is too big, after looking at schematic most of it seems not too difficult except implementing an alternative delay chip

https://elektrotanya.com/boss_dm-1_delay_machine_effects_.pdf/download.html

https://www.gearslutz.com/board/attachments/geekslutz-forum/307657d1346634244-boss-dm-1-analog-delay-reticon-r5101-ds-1-bw.pdf

the dm1 in service notes says its delay goes up to 500ms    and the r5101 can achieve up to 1000ms

so i assumed you would need 2 x mn3005 to get around 500ms?

[Mark Hammer]

I see that the DM-1 uses the same trick employed by the MXR Analog Delay, where the filtering is harnessed to the HF clock, so that the filter rolloff moves downward as the clock frequency decreases.  That results in less bandwidth for longer delay times rather than fixed filtering that takes a stab at a one-size-fits-all rolloff.

The service manual states that the unit can provide up to 500msec of delay time, which is quite a bit more than one is used to with a single 4096-stage chip.  But with  13 poles of lowpass before the BBD, and an additional 11 poles after, now I can see how they can achieve it.  Of course, with all that lowpass filtering, not only is the clock noise kept at bay, but so is the intelligibility.  I would imagine that at max delay, the repeats sound like they are passing through a padded wall.

Although the DM-1 could be replicated with a smaller footprint, simply by moving the transformer outside the pedal and using an external wallwart, as well as omitting the input-level switch (largely irrelevant for a guitar pedal), the adaptation of an R5101-based circuit to an MN3205 or even MN3005 would require some other changes.  The R5xxx series of chips does the flip-flopping internally.  You will note that there is only one clock line going from the 4013 to the R5101, because the 5101 divides that clock into a tick and tock, internally.  Were you to sub an MN3205 chip the supply voltage would need to be different.  And if you wanted to go with a pair of MN3x05 chips in series, you'd likely want to figure out what the optimal rolloff would need to be to get greater bandwidth at respectable delay times.

In short, I think your friend is likely asking for more than you can deliver for a while, and at a cost he would likely balk at.  Make him a Rebote 2.5.  They sound fine, are dirt cheap to make, have a small footprint, are a known quantity (I've made several) and can be modded.

[njkmonty]

im up for the challenge    ive got through a ce1  phase 5/  jetphaser  so im ready to have a go!

(I remember you advising a similar easier route a few years ago when i asked if a ce 1 clone could be done! )    

Re: help with ce 1
« Reply #3 on: August 13, 2011, 11:56:14 AM »
Quote
You know what?  The BBD is nothing special.  Indeed, much of the entire audio path is nothing particular special.  And the clock generator  is far more trouble than it's worth.

So why do people like it?  Easy.  The LFO/modulation section.  Why don't we simply whip up a daughter board with the CE-1 LFO section for vibrato and chorus, and connect that to an MN3101 or 3102?

bipolar power is no issue here

i was thinking up translating the schematic into eagle cad  and seeing what input i can get with modifications,online

i am aware there will need to be some modifications however if i can stay as close to the original as feasible   then that would be great .

my understanding is there is a dry signal going into the r5101   and comes out into the intensity pot.

perhaps if i can use a madbean "Deluxe memory man 2 x mn3005)  as inspiration for this
however im not too sure how to implement  a 4013    and 4069   as the clock section (if im understanding the schematic correctly!:) )

[Mark Hammer]

Why the fascination with the DM-1, if I may ask?

[njkmonty]

Why the fascination with the DM-1, if I may ask?

ask away!!

firstly my friend was really happy with the ce1 clone i made for him , and he said
"Tried the unit out and was very impressed. It is scarly similar to the original CE1 (which I have).

I would love a DM1 clone, but have never seen one.

Those two units are incredible sounding but way to heavy and bulky for an old fart like me!

Thanks again, I was very happy with your unit."
this has given me confidence and motivated to help him

i feel like ive been  working through boss/rolands old catalogue of pedals recently!..

besides i enjoy the challenge , and ive learnt a lot along the way

If i come up with a working pcb, I will send you one!

ive made many dm2 and deluxe memory man variants, but hadnt heard of this pedal before

i also just found your delay chip zip files   !   very helpful

[Mark Hammer]

You're most welcome.  I suspect I probably need to add to it.  I've gathered more useful snippets here and there over the years.

[njkmonty]

thanks for your help
have manged to get info how to implement the mn3005 !
cheers for pointing me in right direction

[StephenGiles]

I have a Yamaha E1010 that I built back in the days when we wrote letters to manuafacturers asking for service manuals!! Here is a bad scan

http://31.41.209.193/RemoteDownloader/download.php?hash=d600a0c04ffea9bac60c571ebcec9fc57b9f2395&file=E1010+YAMAHA+Service+Manual++-+EN.pdf

You will see that it uses 4 x MN3005 with enough filtering for a Trump sponsored brothel!!

[njkmonty]

you mean the white house???

[StephenGiles]

you mean the white house???

Ha ha yes - possibly!

[Mark Hammer]

I can understand why you'd want one, Stephen.  Understanding why you would build one is another matter.   

Sixteen and half thousand stages and it ONLY gets you up to 300msec delay?  Mind you, with 8khz bandwidth on the delay, that would have been very close to what we expected of some of the better digital delays of that era. I see that Yamaha used the paralleled-invertors strategy for being able to clock those BBDs as high as they did.

1) DID you ever finish it?
2) How long did it take you to complete?
3) Did you have a wife or gainful employment at the time? 

[StephenGiles]

1 A friend of mine made some boards and I finished it perhaps 90%. I didn't bother with the modulation, and used standard EH 4047 clock with the buffers.

2 I don't remember how long it took me, quite a few hours probably.

3 I was in between marriages at the time and fully employed!

One thing I remember I did was to mix the outputs from each MN3005 to simulate a Watkins Copycat and it worked quite well. It's still up in our attic, I'll be getting our winter clothes down next weekend (2C this morning!) so I'll dig it out.

[Mark Hammer]

Hmmm, you've given me some ideas.  I made a Tonepad Rebote 3.0, which uses a trio of PT2399s.  The basic objective of the design (thanks Francisco!) is to be able to achieve delay times out to 1 second, using cascaded 2399s.  An admirable goal, but now you've got me wondering whether the outputs of the first two delay chips could be tapped to provide the functional equivalent of 3 tape heads, when combined with the final output.