MXR Delay - clever design

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

barclaycon

Well-known member
Joined
Jan 17, 2007
Messages
696
Location
London, England
I don't normally rave about aspects of analog circuit design, but whilst looking at delay circuits I came across the MXR 118 delay. (Stephen Giles very kindly upped a copy here).
I was impressed with the way they had incorporated 4016 transmission gates into the input and output filters so that they were switched by the clock. This means that they are being used as 'chopper' resistors i.e. the resistance they present varies according to the clock frequency.
It's a technique that the Pye compressor used and also the EMT 165 limiter - albeit without CMOS chips!
In the MXR the resulting sound therefore gets darker the longer the delay, because the clock frequency is switching the 4016's and lowering the cutoff of the filters. It also gives a the delay a characteristic sound.
I wonder if the Carbon Copy model usse the same technique. It's a analogue design but I haven't seen any schematics.
 
+1 yes it is a clever approach. I was aware of it but never tried to use it. Back in the '70s I used the 4016 to switch between different fixed resistors to tune my anti-alias/anti-image filters for 4 set delay ranges in a studio delay product (LOFT 440/450).

Later I was using a Reticon CCD (R5101?) that had an inherent HF loss, different from the bucket brigade devices, with frequency response in the margin that rolled off well before sampling/nyquist limits so there wasn't the same marginal improvement from pushing the filters up near aliasing.

Unfortunately that CCD went obsolete decades ago, while the BBDs are still around.

JR
 
That is clever! I am looking at their MXR dual limiter in my rack and sounds great. Super fast steamroller type attack and release, not to different than the pye, though the overall sound is a bit "weak" compared to the pye, probably just a cleaner circuit design! Thanks for sharing.
 
The Reticon R5101 was used in the 106 version of the MXR 118.
Though I believe the original version (102) used 3 x SAD1024 chips.
I've seen the R5101 chip being advertised for ridiculous amounts of money. Similarly an SAD1024 can easily set you back about £20.
There's much money to be made on obsolete chips. Even CMOS logic chips in things like BEL delays that cost about 25p a few years ago, now cost 10 times that amount.
It's even been suggested that buying old stomp boxes on eBay and harvesting them for chips can be cheaper.
Another little project, that I'm currently looking at, used the Clairex CLM50 opto as the gain element. The price for one advertised on eBay was £16 (!)
 
The BBDs that are currently being made still look like the old Panasonic/Matshusita parts. They are pretty similar to the Reticon parts but IIRC P channel instead of N channel mos, so the output source followers require pull up loads instead of pull down. (or vice versa). Not easy but possible to retro fit.

I don't quite understand the attraction for obsolete technology, but different strokes for different folks.

JR
 
JohnRoberts said:
I don't quite understand the attraction for obsolete technology, but different strokes for different folks.

JR

well...grown men are still attracted to breasts...
 

Latest posts

Back
Top