Phabio, the one that got away

1000009908.jpg
This is a passion project of mine that I've wanted to do for years. I used my original pedal from the late 80s as the bench mark and test subject that I took voltages from and checked values. Unfortunately, this project is currently on the back burner and is likely not going to be released any time soon which is a real bummer because we put a lot of effort and time into getting it right. So what is the phabio and what went wrong?

The phabio is a faithful recreation of the boss PH-2 super phaser. The PH-2 is a 12 stage phaser with 8 stages +4 fixed stages. Two modes and a very wide range on all controls. It is OTA based and the pedal uses two IR309 ICs. It is also the first phaser I've seen using a compander much in the same way an analog delay uses one. It's a fairly complex circuit that gives a lush and smooth sound with an LFO that can go from super slow to super fast. At times it can border on cusp of faux flanger territory. So what went wrong?

The pedal uses the proprietary Roland IR3109 chip that was made for use in some Roland synths of the late 70s and 80s. As far as I know, the ph-2 was the only pedal to use this chip. The 3109 is a complex IC that houses 4 ota's as well as dedicated buffers and what not. The chip was EOL in the late 90s. The 3109 was brought back to life by alfa as the as3109. The problem is that the original 3109 never had a data sheet released so when alfa made their version it was an estimated guess based on a circuit diagram in the Roland Juno synth service manual. Plain and simple this chip does not work in the ph-2 circuit. It clips, sounds off and does not give the lush phasing of the original chip. I tried many different component changes on the prototype board but when I put the original ICs in it was perfect. The Alfa chips in the OG boss pedals gave the same disappointment.

So in the end since I already pulled the chips from the boss pedals and put sockets in for testing I decided to box this one up. Keefe did an amazing job with a fantastic layout and no noise, such a shame this is being shelved. I did do the lower noise mods in the service manual as my original pedal had them as it was made after 1987. If you saw my recent workbench thread then you saw the broken ph-2 I pulled ICs from to make another one, the saga continues.

For the enclosure I used a matte black tayda 1590BB2 and UV print. For the phabio theme I went with THOR, the god of margarine. The calibration did require my scope to get the timing right on the phasing. Maybe one day we can revisit this or a newer reissue chip that functions better will be released. Until then this is likely to be only 1 of 2. I tried to get this to y'all but it's gonna be a while longer😩
1000009930.jpg 1000009928.jpg 1000009927.jpg
 
Wow! Very cool!

How does the specs of the 3109 compare to a 13700? Wonder what the magic is, maybe a redesign around something more common would do the trick?
 
Wow! Very cool!

How does the specs of the 3109 compare to a 13700? Wonder what the magic is, maybe a redesign around something more common would do the trick?
We thought about it but without the data sheet for the 3109 it's just throwing stuff against the wall to see what sticks. Either way, it would require a major redesign and that's above my knowledge base. Plus more chips would mean 1590XX.
 
I'll also add a non-factual thought:
I wonder if Alfa screwed up the buffers. I suspect the AS3109 may work perfectly fine at higher/split rail voltages. It would still be quite a redesign to incorporate a -9v rail but I have to wonder if it would work fine(and possibly better than the original) with the higher voltage rail.
I remember reading a thread on Gdiy or modwiggler a while back where it was pretty much determined that the OTAs were identical or near identical to 13700s and the Roland BA662. There are 2 pairs of 2 different types of buffers on board. That's the real conundrum/mystery.
I'll see if I can find that thread for anyone interested.
Anyhow, awesome effort and execution as always!
 
regardless of the chip issue this is phenomenal work! Hopefully one day that will get sorted out, but until then everyone just started scouting for broken ph-2s 🤣
Thanks Rob. It's a really slept on circuit and it was a real learning experience. Hopefully one day this will get an actual release. Until then, let the FOMO commence 🤣
 
I reached out to them and asked multiple questions in regards to the ph-2. But that was weeks ago and they are in Europe and said they are slow to respond to emails. I would love to test them in curcuit
Well well

$60 a pop though
 
Back
Top