Showing posts with label Digitech. Show all posts
Showing posts with label Digitech. Show all posts

Tuesday, 14 July 2020

DOD American Metal

Another quick one. This American Metal would power up, pass a clean signal but wouldn't pass an effected signal.


Schematics are older DOD pedals aren't too hard to come by (I believe some of them were supplied in the original boxes). I suspected a bad switching JFET was blocking signal, but they all measured fine on the oscilloscope.


I found that two of the transistors near the distortion control had very little voltage on the collectors. The schematic shows a series 1k resistor and a 47uF capacitor from the 9V supply, so there's a good chance that capacitor is shorted.




The 47uF is right in the middle of the PCB. Pulling out brought back the effected signal, so I replaced it and put things back together.

This is a very cool pedal, definitely in the HM-2 family. Maybe not as much gain and not as much EQ control, but it certainly does the buzzsaw sound when turned up.

Monday, 13 July 2020

Digitech PDS 8000

I got this one for completeness sake after looking at the other PDS series.




The insides are very similar to the PDS 2000, with 4 times the memory. The PDS 8000 schematic is online, but the quality of the scan is so poor that the 2000 schematic is probably more useful.


As arrived, no activity. The reverse polarity diode was shorted, after clipping it out I had 5V at the digital chips but no 9V at any of the analog stuff.

I knew that the 9V supply has a soft-start circuit (as in the 2000) - the JFET was also dead. With a repaired soft-start, the 9V rail was shorted to ground. There is only one 470 uF electrolytic capacitor from 9V to ground, replacing this fixed the pedal. It also got new footswitches as the old ones were prettyy unreliable.

Monday, 4 February 2019

Digitech PDS 1700 Chorus/Flanger

I picked this up out of curiosity, I'm fond of the PDS series and the off-the-shelf design. This is a digital Chorus and Flanger, where the traditional BBD design has been replaced with an 8-bit digital delay line (i.e. not DSP or modelling) and the delayed signal mixed in the analog domain. This one worked but didn't switch very well.

PDS 1700

There aren't too many surprises when comparing the insides to the PDS delays. There is the same ADC (ADC0820) but instead of DRAM it uses a single 2k 6116 SRAM. This is probably because the required delay times for a Chorus and Flanger are much shorter (this pedal maxes out at 51 ms) they could spring for the more expensive SRAM. This also means all the DRAM refresh circuitry isn't needed, and there are fewer logic chips overall.

Guts

PCB Backside

The switches were pretty flaky. The mechanical design DOD/Digitech used at the time has a poor reputation for reliability. I find that if they are maintained they work well, but I don't think they stand up to force. The classic problem is that switches that won't switch on the first try get a harder stomp the second time, and things deteriorate.

PDS hardware. Input jack looks non-original.
The paddles press down on momentary push buttons on PCBs inside the pedal. The paddles bear against two bolts that pass through slightly oversized holes in the enclosure. There is a preloaded spring on the bolts on the insides to push them against the enclosure, with nuts holding the springs under tension. There is another return spring on the switch PCB to bounce the paddle back.

I replaced the switches, I'm using these. There was some play in the paddles, the bypass one would rotate slightly in place, meaning it would not always hit the button switch. I tightened down the nuts on the inside until this play was removed, then it switched every time. You need a 3/32" hex key and a 1/4" socket wrench to make the adjustment.

The effects are cool. There is interesting play between the delay time and depth controls, the flanger side can go from covering high to lower-frequency ranges, kind of like the difference between a Boss BF-2 and a HF-2. The chorus does odd (cool) detuning effects at high delay times, modulated chorus in the middle and traditional chorus at minimum delay.

I have this one listed on Reverb if anyone is interested.

Tuesday, 14 November 2017

Digitech PDS 1000 & 2000 Repairs

I am fond of the Dod/Digitech PDS series delays. On paper there's nothing too amazing - mid 1980s 8-bit delays with 1, 2 or 8 (if you can afford the PDS 8000) second maximum delay times. There are two features that win me over.

The first is the "Infinite Repeat" footswitch that locks the delay into a repeating loop. The loop can be pitch-shifted by changing the delay time, and when Infinite Repeats is turned off the delayed signal trails out like normal again. This is like playing with the shittiest looper and is a lot of fun, and surprisingly a lot of modern delays & loopers don't (or can't) do this. Some will repeat indefinitely without oscillating if the feedback is maxed (Boss DD series) but it's nice to have it on a dedicated switch, with feedback control available and ready for switching back to "normal" use. A surprising amount of 90s and early 00s delays can't smoothly pitch-shift recorded audio when playing with the delay times, and glitch or "jump" somewhere through the pot travel.

The second win is the internal design. The expected parts for a simple delay are all there - an ADC, a DAC, some memory and some logic gluing things together - but there are no custom ASICs, microcontrollers or programmable devices at all. Everything is done with standard CMOS logic. These delays could be built from scratch with new parts today, and they are very repairable. This is a little bit of a backwards approach for a 1985 era product - at this point Boss had already released the DD2 & DD3a delay pedals and they used the same ASIC from the Roland rackmount digital delays. This could have been a cultural difference, as Japanese manufacturers seemed to prefer going for custom ICs and offsetting the cost against using them in lots of products (Yamaha have also done this). Digitech (or DOD at the time) may have to bring the series to market quickly, which could have forced them to use off-the-shelf parts. Whatever the reason, nearly ever part can be sourced pretty easily, unlike finding a replacement Boss controller IC from 1985 (good luck).

I repaired 4 of these delays recently. Here is a braindump:


PDS 2000 #1



All of the pedals in this series came with snap-in plastic battery doors that are not held captive by the enclosure (i.e. they are removable). This means they get lost, and I have never actually seen one.

This PDS 2000 (mine) was fairly easy to deal with. It's a 2 second delay that adds sampling modes that can trigger the delay sample by a footswitch or an external trigger signal. It had some broken wires in the harness between the PCB and jacks/switches. I think this is due to using the pedal with a battery but without a battery door - the weight of the battery pulls on the wiring if it's free to swing around on the end of the battery clip. This just needed some soldering to bring it back.

The footswitches were also unreliable, so I replaced the microswitches. More on that below.




In an unusual move, DOD actually sprang for a silkscreen print on this PCB. Schematics are available online but can be hard to match them with a PCB when there is no silkscreen and no part designators.

PDS 1000 #1

 

 

This one was partially working when I got it. The output jack wouldn't hold a cable in, the footswitches rarely worked and the delay length maxed out about 75% of the way through the pot's travel - but it did work as a delay. I replaced the footswitches and bent the retaining lug on the output jack (I may go back and replace this) so that it was a usable effect.



I have never seen a PDS 1000 schematic, but the PDS 1002 2 second version looks like it's very similar. The delay time control circuit is shared across the earlier series. Delay time is varied by changing the frequency of the main clock signal that shifts digital samples into and out of DRAM memory. Faster clock frequencies shift data faster and give shorter delays. Delay time is adjusted by 3(!) potentiometers, the delay time pot on the front of the enclosure, a trimpot that adjusts range of the of the main delay time pot, and a final trimpot that globally adjusts clock frequency by small amounts. You don't really get a very wide range of adjustment so trying to get longer delays by tweaking pots might not work too well.

Tweaking these delay pots brought the delay back to 1 second and adjustable through the full range of the delay pot.

PDS 2000 #2






This is the only pedal of the 4 that didn 't power up at all. None of the logic chips had any sensible voltage at the VCC pins, but the opamps were getting 9V. There is a 78L05 to drop the 9-10V input down to power all the logic chips, and this has a JFET soft-start circuit that ramps up the 5V power supply, presumably to protect the digital chips from a loose or intermittent power cable. I replaced the timing cap with no luck, then swapped the FET for a new J201 and it powered up.


This a soft-start for the 9V supply on the PDS1002. The 2000 has something similar.




Everything seemed to work, except that I couldn't change delay ranges, it was stuck on the longest range (2 seconds). The sampling modes also weren't quite right, trying to trigger samples would switch the pedal into bypass mode or sometimes do nothing at all. I traced signals back from the delay range switch to a 74HC04 hex inverter, which looked dead. I swapped this chip over from the other 2000 and it worked. I ordered a new 74HC04 and got two working pedals.

This one also was not quite a 2 second delay so I did some pot tweaking. Unfortunately I managed to slip with an oscilloscope probe and shorted two pins on the DRAM chip, and killed an input pin. I replaced with a TMS4256 DRAM from eBay - works fine. With the new chip the delay rates could be carefully dialled in to 2 seconds pretty quickly.

PDS 1000 #2

 


This was in the worst condition. I got this in a box of failed repairs from a music store in the US some years ago and never did a lot with it until now. Someone had already taken a crack at fixing it, it was missing knobs and a back panel. Two of the pots were broken, the bodies would wobble freely against the legs as if they had detached internally. The 1Meg trimmer for clock frequency was also broken off.


I replaced the trimmer and set it to the middle of the it's range, based on how the other 1000 was set. The PCB mount pots were a little harder, DOD use Alpha pots marked "W" which I don't think are actually W taper (W taper is logarithmic for half the travel, then reverse log for the other half). I replaced with long-leg alpha pots. The replacements are a little taller than the originals, so I cut the legs a little shorter and soldered some bus bar to them, then soldered the bus wire into the PCB. This actually worked pretty well, but replacing these pots and running wires to the PCB is probably a better long-term solution.


After replacing pots I found that delay time wasn't working because of broken trace, I ran some Kynar wire to restore it. The mix knob didn't work and it was stuck at 100% wet - this turned out to be another bad JFET.

At this point things appeared to work again, except the delay was very distorted and noisy. I messed around with compounder trimpot as I thought that may be distorting - no improvement. Eventually I realised that bad memory chips would give corrupted repeats which would probably sound like distortion, so I swapped the DRAM from the other PDS 1000 and it worked. I ordered some replacement DRAM (I used MK4564) and they worked just fine.

Adjusting delay time

There may be an "official" method for adjusting maximum delay length using a testpoint on the PCB but without a service manual I've come up with my own.

The capacity of DRAM can be looked up from the P/N and the nominal maximum delay in seconds is known. My reasoning is that DRAM should be completely filled in this time, so I probe the DRAM with an oscilloscope and adjust delay length until the frequency of data in and out matches the DRAM size divided by nominal delay length in seconds.

For the 2000 the DRAM is 262144 memory locations and max delay is 2 seconds, so I want data to be going in and out of memory at around 262144/2 times per second or 131072 Hz when the delay pot is all the way up. This can be measured from the DRAM Write Enable pin, and probably from the Data in and Data out pins as well. Most of the tweaking is done on the lower global clock frequency trim, usually after making sure the delay pot is working throughout it's whole range. Sometimes I had to iterate and go back and forth between the two sets of pots. Adjustment on trimpots is fairly coarse, so this can't be dialled with really great precision but I find it much easier than trying to listen to the delay and sync to a stopwatch or something similar.

Modding for more delay time

There are online discussions about modifying these pedals for longer delay times that usually involve tweaking the delay pots away from nominal positions to get longer delays. I think this will only work well for very lo-fi sounds, as the sample rate will get much lower and aliasing effects will get worse.

The delay loop in the PDS series is basically some 4040 ripple counters counting through memory addresses and resetting when the end of memory is reached. A better approach to extending delay time is replacing DRAM with a larger IC (or multiple ICs) and adjusting the counter reset logic to address the larger memory. I have some spare DRAM for both of these pedals so I would like to try this.

I haven't seen a schematic for the PDS 8000 (8 second delay). I would like to see how this was implemented, as it should have 4x the memory of the PDS 2000. Extending the 2000 to 8 seconds may be possible depending on how extensive the differences are.

The best picture I can find online shows a single 18 pin (not 16 pin) IC, but the label isn't legible. Presumably it's a 1Mbit x 1 DRAM.

Please let me know if you have a PDS 8000 schematic or high-resolution board shots.

There is a PDS 20/20 delay schematic available (http://www.experimentalistsanonymous.com/diy/Schematics/Delay%20Echo%20and%20Samplers/Digitech%20PDS2020.pdf) which uses 2 4464 DRAMs, which is twice the memory of the PDS 2000 but still only does a 2 second delay. America's Pedal has a catalog for the PDS series which explains why - the 20/20 has a delay signal bandwidth of 16 kHz, the rest of the series only has 7 kHz except for the PDS 1700. I am guessing the higher sample rate is for the chorus/flanger modes where really short delays are needed.

Footswitches

Dod used a fairly cheap mechanical design for actuating footswitch buttons on a small PCB. The plastic foot panels are cantilevered and are returned to their neutral position by a spring at the "fixed" end, not the free end which would make more sense. There is also no real end-stop to limit the force on the footswitch - if you stomp hard you will crush the button. This is problematic as most players will stomp harder if the switch doesn't work, so once they start to fail they deteriorate quickly.




There is advice online on adjusting spring tension to get a pedal to switch more reliably. Don't bother. If a DOD pedal does not switch every time, just replace the 10mm button on the PCB. They are very cheap and should last another decade of use. Play with the mounting hardware only if the levers aren't rotating correctly.

Saturday, 6 August 2016

Digitech Whammy 4

I have a bit of a backlog of stuff I've done but not found the time to describe here. It's probably best to do so before I forget everything.

Big red pitch shifter. Another 90s digital classic. I used to own one of these these but moved it along because I didn't really get on well with no polyphony and the artifacting didn't always sound too great. I think these older Whammy pedals are pieces that need to have music written with them in mind instead of just switching them on and playing the usual stuff.


A pedal large enough that the photos look terrible

Sold as not working and had had a true bypass mod done. I was hoping that it just needed calibration or that there was a simple issue with the true bypass work. Instead it just flashed all the LEDs once when powered up and then did nothing.

Inside we have a large PCB that fills most of the enclosure. There's another Motorola 56k DSP56362 (same as some of the Line 6 and EX stuff), an Atmel 89C55 (8051 architecture), a Cirrus CS4224 Codec, the MIDI optoisolator and some CMOS logic chips and opamps. The DSP56k/8051/CS4224 seems to be a really common choice of components for late 90s/2000s digital, it's very similar to what was in the Line 6 pedals I looked at previously.

The mod work was pretty bad, the wiring used was way too thick and was under strain, there were ferrite beads cracked off the PCB and the 3PDT footswitch fell apart when I was removing the old wiring.

Main PCB and LED/encoder PCB

Atmel 8051 MCU under the sticker

This has the annoying 80s/90s practice of including an AC power supply to create positive and negative voltage supplies instead of using switching converters and running off 9V DC. There are 3 linear supplies, +5V and -5V for the analog stuff and 3.3V for digital. The 3.3V was down near 0V and the regulator looked scorched, so the heatsink was desoldered and I replaced it. No change, so there is excessive current being pulled somewhere. 3.3v to GND measured around 40 ohms. I removed any 3.3V bypass capacitors in case they were shorted, and then noticed that the main DSP chip was running pretty hot.

At this point I gave up for a little while. The DSP is a 144 pin QFP and in it can be difficult to find reliable suppliers of these 56k series parts these days. There are also no schematics for this pedal online, so I would have to trace things out or make guesses.

I tried contacting Digitech and asking about a schematic, and they replied with a PDF within an hour! Here it is https://drive.google.com/file/d/0ByVCt2OFhXnyZS1JZFlmb0Z3VTA/view?usp=sharing&resourcekey=0-vBilBFlfU0YChMUrFsavrg

I also found that this IC is actually still stocked at a lot of distributors, just not in huge numbers. I decided to get one from Farnell for around 18 euros, Digikey have them as well.

Removed the original bad DSP was difficult, as the chip is quite big and there is a large top-side ground place surrounding it that takes a lot of energy to heat up. After not much progress with hot air I flooded all the pins with some Chip-Quik low-temperature solder that I had. This kept everything melted and worked well but made a mess, there were blobs of solder all over that region of the board that had be wicked and scraped off. I realised afterwards that I should have tried using a larger air nozzle for a chip this big.

After DSP removal, there is solder everywhere
After wicking and cleaning the board

 The new chip went on easily enough with some tacky flux. The true bypass mod removes the original momentary footswitch, but this switch is still needed to calibrate the footpedal, so it is usually replaced with a pushbutton mounted to the back of the enclosure. In this case, the switch was connected to the board using the original connector and then fixed down with a big blob of hot glue (you can see this in one of the pictures above). Unfortunately, all the rework heated the board enough that this glue melted and ran all over the board, leaving a thin layer around the MCU. Now nothing happened when powering it. I could peel off  some of the larger pieces of glue but most of it needed to be scraped. At this point I gave up temporarily again.

DSP resoldered,Glue cleaned, C52 replaced, Y1 removed, cleaned and repopulated
I came back a couple of weeks later and used some IPA and cotton swabs to get off the rest of the hot glue. C52, Y1 (crystal oscillator), the momentary switch connector and some of the legs of the MCU were the worst effected, so I removed all of those except the MCU and cleaned the board as best as I could. Y1 was pretty badly coated in glue, so I cleaned it and repopulated. I tried poking around wit an oscilloscope probe and found that every time I touched one of the pads of C52 the LEDs would flash. The schematic says that this is the reset timing cap, so presumably the MCU was never being reset. I replaced it with a 1uF film cap I had on hand and the Whammy fired up. I can only guess that swapping the DSP was successful, but that the hot glue was preventing Y1 from oscillating and the MCU from getting a clock signal, or C52 was bad, or both. After redoing the true bypass wiring everything works.

There is also a 1k serial EEPROM on board, but it is connected to the DSP and not the MCU, so I think it for saving calibration settings and not the program code. I would like to dump the MCU if possible, but I'm going to leave this alone. If I come across a scrap board I'll pull the chip and try to dump it. There is also a JTAG debug port wired to the DSP, I haven't looked at that at all.

Was this worth it? Replacement parts ate up a good portion of what I would have gained from selling this, and I have more than a few hours invested in it, so probably not. The newer models add polyphony and have lowered the value on these older units too. I'll add a LED (mysteriously missing? Maybe someone killed this halfway through a mod) and sell it on. I'm glad to have swapped out a QFP144 successfully and found a schematic for this, so I'm still thinking of it as a win.