I decided an Alesis SR16 from 1990 should have its battery refreshed even though I had not received any error messages (maybe the unit will not give notice of low battery level). There was just enough room to mount a button battery holder with short leads and then install a CR type. The BR type has a much longer operating life but since it is now a super easy access job (five chassis screws) I opted for the CR type. Any thoughts of what downside I might expect to have happen? I might add I despise soldered memory batteries. Who makes such decisions at the manufacturing end?
Last edited:
Who makes such decisions? In fairness, engineers do. The average life of those lithium coin cells is in most cases longer than the life of the gear. I run a commercial shop, and we do very little battery replacement. The solder in type can easily be expected to last 10 years, and are super easy for a technician to change. the cells with solder tabs are readily available from common suppliers like Mouser or Digikey. The holder takes a bit more real estate inside, adds an extra part to the assembly. it amounts to one more connector to fail. Apparently yours has so far lasted 23 years and was still working, might have lasted another 10 years.
Space in a small thing like that matters more than in say a synthesizer keyboard. Sure, I like snapping a battery into a clip as much as the next guy, but soldering one is no real challenge.
Space in a small thing like that matters more than in say a synthesizer keyboard. Sure, I like snapping a battery into a clip as much as the next guy, but soldering one is no real challenge.
The current drain from the battery is probably zero (essentially anyway... just leakage and CMOS memory) Fit whatever fits 
but don't be tempted to try and solder (no matter how quickly you think you can do it) leads to a coin cell directly because it will just go open circuit instantly. Use a holder or get a coin cell with leads.
(reminds of Philips TV and video products of the 80's. They used either a 1.2 or 2.4v Ni-Cad battery for program/memory backup and failure after only months was common. I still use a Philips portable where I replaced the Ni-Cad with a coin cell some 16 years or so back and its still going strong)
but don't be tempted to try and solder (no matter how quickly you think you can do it) leads to a coin cell directly because it will just go open circuit instantly. Use a holder or get a coin cell with leads.(reminds of Philips TV and video products of the 80's. They used either a 1.2 or 2.4v Ni-Cad battery for program/memory backup and failure after only months was common. I still use a Philips portable where I replaced the Ni-Cad with a coin cell some 16 years or so back and its still going strong)
Thanks for your input. The BR2325 isn't something I can easily find in stock at even MicroCenter. I looked at putting in the holder as a challenge. I measured the clearance and then made sure there would be room for a thin neoprene cushion top and bottom. I bought a batch of holders that solder directly and altered them with leads and sealed the solder points.
The SR 16 I was given had only a few user patterns programmed. The free memory read 88% so am I to understand that perhaps battery drain ties in with having to hold programmed input and somehow is out of the ROM loop? I bid and won a Yamaha TX802 on ebay which had been extensively programmed and when it arrived it had scrambled lettering in the custom patches so I knew the battery was failing. Another TX802 I bought on ebay had no user programming(of course the guy could have cleared his work) and the battery still read 3Vs. It was from 1990. It had a CR2032. So why didn't Yamaha use the BR type like Alesis did with their SR16? Just another question about manufacturer decisions I guess.
The SR 16 I was given had only a few user patterns programmed. The free memory read 88% so am I to understand that perhaps battery drain ties in with having to hold programmed input and somehow is out of the ROM loop? I bid and won a Yamaha TX802 on ebay which had been extensively programmed and when it arrived it had scrambled lettering in the custom patches so I knew the battery was failing. Another TX802 I bought on ebay had no user programming(of course the guy could have cleared his work) and the battery still read 3Vs. It was from 1990. It had a CR2032. So why didn't Yamaha use the BR type like Alesis did with their SR16? Just another question about manufacturer decisions I guess.
The actual memory content of CMOS memory doesn't alter the current drawn from the battery. All the battery does is maintain the charge on the individual FET's that make up the memory cells in the chip. Its a bit like keeping a small cap charged (except there are 100's or 1000's of them depending how big the memory is), and so no actual current flows apart from microscopic leakage. The memory isn't being clocked or refreshed in that state, its static.
ROM is just a chip that has logic states burned permanently into it. RAM is volatile memory. it remembers whatever you put in it until power is removed. Hence the battery. The chip normally is powered by the system, but when power is off, that IC is kept alive by the battery with an isolated power circuit. I don't recall on yours, but sometimes another small CMOS IC might be kept alive to maintain a state on a control pin.
CMOS mainly only draws current - and not much even then - when it CHANGES logic states. To just sit there maintaining all its logic states takes almost no current at all. If you look at a RAM IC, there will be one R/W (read/write) pin for the whole thing. The IC has no idea how much data you have stored. Each cell in it could hold data or just 1111 or 0000. So the entire map of memory is kept alive. So if you were inferring that the amount of stored memory affected battery drain, it doesn't. it is like a book, it weighs the same whether there is print or the pages are blank.
Engineers have options, and in many cases it isn't a matter of being "wrong" it is just a choice. It could even be something like they got a better price on the BR ones. Maybe the pc board footprint was different and worked better for their layout. Sometimes a lead engineer simply gets an idea in his head.
DOn't get me wrong, I have zero objection to you doing it the way you did, I was just responding to why they make such decisions.
CMOS mainly only draws current - and not much even then - when it CHANGES logic states. To just sit there maintaining all its logic states takes almost no current at all. If you look at a RAM IC, there will be one R/W (read/write) pin for the whole thing. The IC has no idea how much data you have stored. Each cell in it could hold data or just 1111 or 0000. So the entire map of memory is kept alive. So if you were inferring that the amount of stored memory affected battery drain, it doesn't. it is like a book, it weighs the same whether there is print or the pages are blank.
Engineers have options, and in many cases it isn't a matter of being "wrong" it is just a choice. It could even be something like they got a better price on the BR ones. Maybe the pc board footprint was different and worked better for their layout. Sometimes a lead engineer simply gets an idea in his head.
DOn't get me wrong, I have zero objection to you doing it the way you did, I was just responding to why they make such decisions.
I've noticed when removing a soldered battery that sometimes only one side of the two pronged pin is soldered. It is usually coming from the positive contact. It isn't easy trying to remove such a soldered battery. Other times -even though there are two holes for soldering only one prong of the connector is soldered. On the negative prong it is usually only a one prong insert. The Alesis SR16 with the BR2325 battery had both prongs of the connector soldered (ugh!). So again it is all about decisions of not only the type of mounting but whether to conform to the mounting configuration. Thanks for all of your input. On my tombstone I guess I'll have to put: CR button battery resurrection only please!
Last edited:
My desoldering station usually sucks the joint free. But with just an iron, typically I will het one joint until it melts while applying some up or side pressure to lever the lead from the hole. Then at the other end, if two pins are soldered, I pick one and do the same thing. Of if only one, then melt and pull straight out.
BR2325 replaced with CR2032
My approach may be completely unorthodox. I carefully separate the positive connector from the top of the battery and fold back the battery to give me access. I snip one of the prongs near the solder point. From then on it is fairly easy. The negative connection is usually just a one pin solder point. I've actually been able to remove a soldered battery without having to access the bottom of the PCB. I then solder 24 gauge stranded wire to the battery points taking care which is positive and negative. Peeling that little bracket from the top of some batteries is often very iffy. Often a razor blade is needed to get a start it is so firm a connection.
My approach may be completely unorthodox. I carefully separate the positive connector from the top of the battery and fold back the battery to give me access. I snip one of the prongs near the solder point. From then on it is fairly easy. The negative connection is usually just a one pin solder point. I've actually been able to remove a soldered battery without having to access the bottom of the PCB. I then solder 24 gauge stranded wire to the battery points taking care which is positive and negative. Peeling that little bracket from the top of some batteries is often very iffy. Often a razor blade is needed to get a start it is so firm a connection.
Last edited:
Electrolysis and decomposition is the usual causes of that phenomena. The way the electron moves from negative to positive and takes some base metal with it.
Last edited:
You can deconstruct the old battery. Maybe because I am used to it, it seems to me just pulling the board for a few second solder jobs is easier.
If I had to remove the old one from up top, I might find the pin on the side and snip it, then wiggle the remaining one back and forth until metal fatigue snapped it off. A single pin might be twisted until it was "wrung" in two.
If I had to remove the old one from up top, I might find the pin on the side and snip it, then wiggle the remaining one back and forth until metal fatigue snapped it off. A single pin might be twisted until it was "wrung" in two.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.