The digitech PDS1700 - digital fun

[ethrbunny]

I bought one of these off craigslist for 30$ yesterday. It needed a new PS jack so I cracked it open.. what an array of chips!

MC74HC04N (2)
MC14007 (2)
M74HC195B1
MK6116N-20
MC74HC00N
MC74HC374
ADC0820
LM358
TL062 (7)
NE570

of these I only recognize the TL062 and the 358.. sheesh. Time to do some searchin'.

[scratch]

the MC74HCxxxxxx parts are CMOS logic (digital) circuits, same with the MC14007 (Vendor Motorola?) ...

the MK6116N-20 would be a memory chip, -20 usually indicates 200 ns (access time),the ADC0820 looks like a an Analog to Digital converter (I'm guessing 8bit) and the NE570 , compander chip, used in a lot of delay based effects (Echo, chorus, flanger ...)

[Peter Snowberg]

74HC04 - Hex inverter
4007 - triple CMOS transistor pair (maybe for bypass switching?, or was that a typeo and these are 4070s?)
74HC195 - 4-bit parallel access shift register (maybe being used as a walking ring counter???)
6116 - 2Kx8 SRAM
74HC00 - Quad NAND gate
74HC374 - Octal Latch (it is probably located next to 16 or 17 resistors which make up an 8 bit ladder-style D/A)
ADC0820 - old high speed 8 bit A/D using half-flash technique
LM358 - opamp
TL062 - opamp
NE570 - compander

I'm going to guess that there is at least one more chip than specified on your list. The missing element is an address counter.

This circuit probably works by using a couple of the inverters with an RC setup to generate a clock which then feeds a counter which is at least 12 bits long. I'll guess that the bottom bit of the counter is used to determine if the SRAM is talking to the A/D or the D/A. The remaining 11 bits generare 2K of addresses. As the clock line ticks, the following sequence repeats over and over....

1. increment address counter
2. switch RAM to source data and send value in RAM to the D/A made from the HC374 and a bunch of resistors.
3. clock the HC374 latch to capture the value from the RAM (perform the write to D/A)
4. switch the RAM to accept data now
5. send the A/D output value to RAM
6. clock the RAM to store the new value (perform the write to RAM)  
7. repeat...

Depending on how the address counter is done, you may be able to easily double or quadruple the delay time by adding more 6116(s) and using an additional address bit or two with some very basic decoding logic. There is probably an 8K chip with a very similar pinout which would be the easiest to extend the delay time

[ethrbunny]

Sweet heavenly mother.  :shock: I was impressed with myself for getting the PCB out without hurting myself. You got all that from a list of chips?

Dang. I wish my brain had its own zip code like yours.

There is probably an 8K chip with a very similar pinout which would be the easiest to extend the delay time

How would I go about tracking down said chip? Is the 6116 the one you would replace?

[Peter Snowberg]

Nah.... It's just that if you've seen one 8 bit digital delay built around inexpensive parts from 20 years ago... then... well... most of them are very similar. Digital logic was also my main gig for many years.

The chip you probably want is the 6264, but you could also go for a 62256 to extend the delay time by 16 times.

You can find both those chips (or chips with the same JEDEC pinout) being used as cache on old PC motherboards.

The big question is how is the address generated?

If you can post pics of the board, I can probably tell you exactly how to do the mod. You're going to need to add 2 or 4 bits to the address counter which should be very easy using a single 4 bit counter chip.

[dminner]

Super old post I know...But I just picked one of these up and want to do some modding for fun and this is the only post I saw on the PDS1700...

Does anyone have any ideas as to how the mods here are done:  http://www.austinhotmods.com/mods-available/flanger-phaser-chorus-mods/dod-pds-1700-series-mods/