• Welcome to Circuitbenders Forum.

DR110 memory backup battery

Started by Circuitbenders, June 21, 2012, 06:25:39 PM

Previous topic - Next topic

Circuitbenders

I've managed to install a CR2032 coin cell backup battery into one of my DR110's to keep the pattern memories alive. All you need to is wire the cell to pin 18 the RAM chip via a diode, like this:



I found that you need to use a Schottky diode or the voltage drop across the diode makes the mod very unreliable. The only drawback seems to be that if you are using an external power supply, you should only ever turn the DR110 on and off using the switch on the back of the unit. If you power the DR110 off by just unplugging the PSU from the mains leaving it plugged into the machine, the slower voltage drop seems to cause the backup battery to not switch in properly and you sometimes lose the patterns.

Can anyone explain why this happens and would a slower diode work better here? I'm currently using a BAT42.
i am not paid to listen to this drivel, you are a terminal fool

Gordonjcp

Okay, here's my conjecture:

On the CPU board, power for the CPU and RAM both come from the battery backup supply on pin 5 of the inter-board connector.  So with the battery connected, the CPU remains powered.

Now turn your attention to the analogue board, around the power supply circuitry.  The power switch either feeds 6V to the 6V power rail, or connects it to ground through a 1k resistor, presumably to discharge any decoupling caps - why, I don't know.  Meanwhile the other pole of the power switch is connected to two inverters in IC2 via a couple of resistors, a diode and a capacitor.  When the battery voltage is applied to the cathode end of D1, C14 charges through R32.  The rising voltage is "cleaned up" by the Schmitt trigger formed by IC2 and R33, and after a couple of milliseconds the output of the right-hand inverter will go high.  This releases /HLT allowing the CPU to run.

When you turn the power off, C14 will discharge quickly through D1 and R63, pulling /HLT low and stopping the CPU.


So it's my guess that if you just yank the power supply out, the CPU will continue to run as things collapse around it, and every so often this will cause it to write garbage to the RAM - perhaps if the power goes off as it is writing.
If at first you don't succeed, stick it through a fuzzbox.

Circuitbenders

In theory the /CS pin of the RAM is held high by the backup battery as soon as power is removed, which should make the RAM inaccessible according to the datasheet. Are you saying that because of the discharge of C14 when the power is just pulled out this is not always the case, and for a very short time CPU is still running and holding the /CS pin low while eveything else has shut down? So because of this sometimes as it dies it can write random data to the RAM?


I'm trying to get my head around this section in the service manual on memory backup that says:

'During the power OFF /HLT pin of IC1 CPU is kept L, maintaining all its input and output pins high impedance, isolating its circuits from peripheral circuits and thus retains all the data so far obtained. When the CPU is re-powered it initialises internal circuits but still keeps some data intact'

Has that got any relevance here do you think?

i am not paid to listen to this drivel, you are a terminal fool

Gordonjcp

Yeah, that makes sense - tristating the bus outputs would certainly stop anything from being written to memory.  I think the CPU is left powered though.  Can you 'scope the clock pins and see if it still runs?
If at first you don't succeed, stick it through a fuzzbox.

Circuitbenders

Theres nothing on the CPU clock pins when the power is off and the backup battery is in place.

One thing i have noticed though, if i look at the clock pins while i turn the DR110 off and i just switch the power off using the on/off switch, the clock signal jumps from 5v peak to peak to nothing instantly. If i unplug the external PSU either from the mains or at the DR110 end, the clock drops from a 5v to about 3v peak to peak waveform, and then sits there for about half a second before dropping to zero.

The same thing happens if i scope the data pins between the RAM and the CPU. When nothing is happening they sit on 5v. If i turn the thing off with the on/off switch they instantly drop to zero. If i unplug the power they drop to about 3v, stay there for half a second or so, and then drop to zero.

I guess its that half a second when the CPU is 'kind of running' that fills the RAM with garbage?

i am not paid to listen to this drivel, you are a terminal fool

Gordonjcp

Yeah, that would do it.  If /CE is still low and /WR is somewhat low it may be scribbling random stuff on RAM.
If at first you don't succeed, stick it through a fuzzbox.

TeeVee

can i use glue gun to attach the cr2032 holder instead of 2side tape or will that damage the board?

Circuitbenders

I don't see why not, although glue gun glue doesn't adhere very well to smooth surfaces does it?
i am not paid to listen to this drivel, you are a terminal fool

TeeVee

I have been looking around for those Bat42, so I can make this battery mod, for my Boss dr-110 and my local electronic component store is having problem with getting them home. So I have been running around on ebay and now fall over the Bat42, but I'm quiet unsure about the specifications if they live up to the requirement for this mod.

http://www.ebay.com/itm/10x-BAT42-Schottky-Diodes-FAST-DELIVERY-/110837695157?pt=UK_BOI_Electrical_Components_Supplies_ET&hash=item19ce7102b5

Does this Bat42 live up to the conditions for this mod?

Circuitbenders

That'll do the job. It doesn't really seem to matter that much what Schottky diode you use, but it has to be a Schottky diode as opposed to a normal small signal diode like a 1N4148, or the voltage drop across it is too large to keep the RAM alive.

Remember that you have to use the on/off switch on the back instead of pulling out the power plug, or the memory can be scrambled.
i am not paid to listen to this drivel, you are a terminal fool