Incorrectly oriented V3205 bbd ic

[gcwills]

I have purchased a few V3205 bbd ics recently from online sellers that were quite cheap.
I was not surprised to find that they did not bias correctly or work when I substituted them for the existing V3205s in my EH DMM clone and wrote them off as fakes.
Not sure why I decided to test them some time later inserted in their sockets reversed in position, but the surprise is that they worked perfectly "backwards". So it appears that these ics were incorrectly stamped in manufacture.
It appears that this could be a more widespread issue as I purchased these devices from different sellers from Ebay and Aliexpress.
So all this to say that if you have some V3205 "fakes", try them "back to front" - they may work ok!

[bean]

Power, ground Vgg and whether the output resistors pull up or down are different between the 3205 and 3005. This is probably the source of your problem (assuming your DMM clone was built with 3005 to start).

[gcwills]

Thanks for that Brian, but my DMM clone is a Dirtbag Deluxe wired for V3205s and these replacement ics were substituted for the original V3205.

[bean]

Thanks for that Brian, but my DMM clone is a Dirtbag Deluxe wired for V3205s and these replacement ics were substituted for the original V3205.

Ahh gotcha. Well then that is pretty friggin weird!

[gcwills]

You are right there Brian. It was doing my head in - one initial thought was that these devices may have been P channel 3005s or 3008s wrongly stamped, but the pinouts don't match a reversed 3005 even in the extremely unlikely event that they were 3005s.
The only explanation I can come up with is that there were quite a few out there manufactured such that the chip in its leadframe was reversed when the IC was encapsulated.

Anyone out there with experience with chip fabrication and encapsulation that could throw light on this??

[ElectricDruid]

It's not the first time I've heard of chips manufactured back-to-front.

There was a batch off Coolaudio V2164s that had the silicon mounted back-to-front in the package, so the entire pinout was reversed, including the chip package notch. Put them in "the wrong way around" and they worked fine. I think Coolaudio sold them off cheap.

So it might happen.

[gcwills]

Thanks for that Tom - its the first time I've heard of it. Cheers

[Mark Hammer]

Are there any identifiable batch numbers or other characteristics that would earmark these chips for the user?

From GCWills' remarks, the problem is not the chips underneath the epoxy, but the epoxy covering them and numbers/info sprayed on them.  It would be helpful for folks to know up front than to wrestle with such chips, badmouth the supplier, and lose hair and sleep in the process.

[ElectricDruid]

Well, so far the only two examples we have are both CoolAudio products, so maybe watch out for them!

Perhaps they need to retrain their staff?! They probably have "L" and "R" written on their shoes...or the Chinese equivalent!

[R.G.]

What I hate is the power of suggestion applied to newbies.

From now on, every beginner who has a BBD chip that doesn't work, for whatever reason, will now reverse it in its socket, possibly killing it, and then he will have a confirmed dead chip that doesn't work either way. Worse, the logic will be applied to ALL chips, and beginners will forever be reversing chips thinking that since it didn't work, the problem is likely to be those cheap, reverse-printed chips. Never mind the soldering, wiring, placement and other errors that are far more likely to be the issue.

We gotta think about care and feeding of newbies.

[Mark Hammer]

Which is precisely why I asked about a specific batch number or other identifying mark.

But you're absolutely correct abut the power of the net to start rumours among folks who may not have enough background knowledge to avoid mistaken inferences.

The art of troubleshooting lies essentially in being able to size up the situation and conclude "It's unlikely to be that, because...", well before you even test or measure or replace something.  Internet rumours can often have the effect of increasing, rather than reducing, the search space of newbies, when it comes to troubleshooting, making it harder than it ought to be.

[gcwills]

Not sure that I was advocating randomly reversing ICs carte blanche
Anyway, here is the seller's image of the V3205Ds that I purchased - all of the samples that I purchased from two sellers had the same code of C3571306

[bean]

It might be worth contacting CoolAudio to let them know the batch #.

BTW: are there many commercial pedals using the v3205? I haven't looked into it in a long time. Kinda hoping Xvive would put out a 3205 that's cheaper than their MN3005 repro.

[Scruffie]

It might be worth contacting CoolAudio to let them know the batch #.

BTW: are there many commercial pedals using the v3205? I haven't looked into it in a long time. Kinda hoping Xvive would put out a 3205 that's cheaper than their MN3005 repro.

Off the top of my head, Behringer, Malekko, JHS, Endangered Audio, Mooer and their ilk and BOSS are currently using them.

Not sure that I was advocating randomly reversing ICs carte blanche

I don't think anything was being directly attributed to you personally, there's a Chinese Whispers effect with these sort of things and somewhere down the line someone will be spouting that you should always put BBD's in backwards on a different forum. It's a bit like everyone always assuming because the BBD is the most expensive part and that one guy had a dead BBD that one time that it's always the BBD that needs replacing.

Something about v3205 with backwards markings is ringing a bell to me though but I can't remember where I heard it... possibly smallbear got a batch of them?

[ElectricDruid]

Here's a link to the SynthDIY list thread where I heard about the reversed 2164:

http://synth-diy.org/pipermail/synth-diy/2017-September/165867.html

It was the V2164M (the SOIC version, not the DIP), and the post claims that Cabintech sold the chips as the V2164M-R (presumably indicating "reversed"). I don't know if that was printed on them. Cabintech no longer have any listed on their site.

Totally agree that the last thing you want to be doing is reversing chips as part of a debugging procedure. For most chips, you'll just give yourself one more problem to debug.

[R.G.]

Not sure that I was advocating randomly reversing ICs carte blanche

I'm absolutely certain that you weren't advocating that at all.

My comments were about the mind state of beginners and the effect of the internet on them, not anything you said. Someone new to DIY audio stuff has no background at all in their understanding, no context in which to fit new observations about things that are not what they are labeled. They're doing all they can to get the idea that it's really, really important to connect up the pins the way the data sheets say to. They have no idea how very rare it is to get reverse-marked ICs, or the clear and present danger of reversing any IC in a socket or PCB.

Those of us with experience know how rare it is, so rare that it's like finding dinosaur footprints in stone, and is a worthy subject of discussion about the rarity. But it's very human to enter a new situation, get a little information, then think that the situation as they entered it is normal. If some of your earliest learning about ICs is that they may have the orientation printed backwards and should be rotated in a socket, that has a huge effect. It's very common for beginners to post here as one of their first posts that they built a circuit just like the vero diagram and it doesn't work, so all the components are bad, aren't they? Then we have to lead them back through being very careful about what connects to what, soldering issues, pinout issues, and wiring mistakes. If you pour on top of that the idea that ICs are commonly printed backwards and the thing may work if you just reverse them, it makes for a lot of frustration and dead ICs.

That's what went through my mind, not that you were advocating the practice.

We have to think of the newbies.

[Mark Hammer]

Thanks for the pic, GC.  Helpful.

[Mark Hammer]

Followup.

Last night, I went to dinner with the folks from Empress, Fairfield, and Chase Bliss, and all had tales of chips that arrived in a form other than what you'd infer from the datasheets.  As noted by Tom, a batch of the Coolaudio 2164 was one of them.  And if I understood correctly, they arrived from Coolaudio with an explanatory note.  Though they worked fine, it dumped the challenge of keeping track of which run of pedals used which chip-orientation in the lap of the pedalmaker.  It also created a problem whereby the legending on the board would show one thing, but the chip would show another.

At the level of the hobbyist, that's not so much of an issue.  If you know how to make the pedal finally work right, you're good to go and to heck with legending.  But if you're trying to run a business, that includes servicing and standing behind your products, keeping track of things, and communicating to repair folks, now becomes an additional responsibility that you didn't ask for when you bought the chips.

So, to reiterate RG's point, it rarely happens.  But it happens....and is a royal pain when it does.

[bean]

Well damn I wish I had been at THAT dinner, haha.

[Mark Hammer]

It was fun, and big bowl of pho was just what the doctor ordered.  And because it was Montreal, the restaurant was a bring-your-own-alcohol establishment.  So the gang sauntered over to a nearby outlet, shopped for beer, and then headed over to the restaurant.

Some time within the last year, I repaired a Diamond Memory Lane pedal for a buddy (the thread is buried somewhere here).  It had an annoying whine.  I exchanged a few notes with a support person at Diamond Pedals, and when I eventually mentioned that the whine didn't start until 5 or 10 minutes after powering on, it reminded him that they had received a batch of LM78xxx regulators for an earlier run of the pedal with thin off-spec heat fins.  He sent me a pic of what to look for, and sure enough, one of the three voltage regulators used in the pedal had an ultra-thin heat fin.  Were the regulators thermally-coupled to something that would assist in heat dissipation, it wouldn't have been a problem.  But these regulators were used in free-standing fashion, such that the whining was a result of heat build-up and the regulator simply drifting off-spec.  As recommended by the fellow at Diamond, I replaced the regulator, the whine went away, and the buddy has a smile on his face again.

These regulators were made by a major manufacturer and obtained from a major distributor.  No E-bay resellers were harmed in the making of that order.  At no point did anyone seem to think "Hmm, that doesn't look right", until customers started returning or complaining about pedals.

I mention this because it illustrates that there are many ways for semiconductor manufacturers to temporarily lose control of quality and saddle not only customers, but distributors as well, with the burden of compensating.  But again, realistically, if it happened all that often, those semiconductor manufacturers and component distributors wouldn't remain in business very long.  Word gets around.  It just shouldn't be disproportionate to the frequency of the problem.