Re: [AH] TR 707 repair

From Kenneth Balys
Sent Fri, Mar 7th 2014, 02:36

Ok, I looked at the data sheet for the Hitachi 6303X CPU and it does 
have 3/4 page of RAM on chip (192 bytes) but that does not mean the 
Roland engineers didn't code some low level dependency into the RAMs.

The -RD and -WR lines from the CPU go straight to the 2 RAMS and the 
program ROM unbuffered so there is nothing to get in the way there.

When the machine crashes do you still see -RD -WR activity? Or is the 
CPU just catatonic? You can also check the address lines and/or data 
lines. The CPU always needs to be doing something so if its not locked 
up, the evidence will be on these external points.

Its looking like a bad RAM but I really am hesitant to advise you to 
start popping hard to get CMOS RAMS without a lot more evidence. They 
are, far and away, the most ESD sensitive device in your 707.

PS: thanks for the nicer schematics

On 07.03.14 1:15 AM, Kenny Balys wrote:
>
> If the machine came up enough to do anything at all with the two -CS
> lines bridged, then your CPU is most likely fine. Same with the ROM.
>
> (most likely)
>
>
> On 07.03.14 12:48 AM, luke aldred wrote:
>> Hey Kenny,
>> appreciate your input. HAhah I was a bit confused when you said low, I
>> thought maybe this was the issue after all! But yes the reset line
>> goes high and stays high, around 5V so I think it is ok. I get the
>> feeling it is something a bit more complex as I have been
>> troubleshooting it for a good few hours already.
>>
>> If you have a look through the earlier posts, I said how the CS input
>> on pin 18 of IC14, Static RAM, is staying high on power up. It is
>> around 4.5V, looks the same as the "BU" voltage which powers IC16.
>> Contrary to static RAM IC15 which is receiving the chip select data.
>> Would this be normal for only one to receive it at any given time? The
>> strange thing is that when I bridged pin 8 and 9 of IC16, which is the
>> output for the CS data, the unit powered up and I could trigger sounds
>> etc, but could not write patterns and it was still pretty glitchy.
>> however IC16, and 11 are both OK as I have swapped them out.
>> What else could cause the output of IC16, pin 8, to be held high when
>> it should be outputting data??  This is what brings me to the P-ROM or
>> CPU but as I said my understanding is not great.
>> Cheers,Luke
>>
>>
>>
>>> Date: Fri, 7 Mar 2014 00:05:15 +0000
>>> From: kenny@beatkamp.com
>>> To: laproductions@live.com.au
>>> CC: lhammond@uvic.ca; wackyvorlon@me.com; analogue@hyperreal.org
>>> Subject: Re: [AH] TR 707 repair
>>>
>>>
>>> Duh! I mean is the Reset line stable and high after power up?
>>>
>>> (why does my brain always take off right when I need it?)
>>>
>>> On 06.03.14 11:58 PM, Kenny Balys wrote:
>>>>
>>>> The reset line is held low by the "power good" circuit until it hits
>>>> the
>>>> operational threshold. This is done so the CPU does not kick off during
>>>> the chaos of power on.
>>>>
>>>> The reset line is NAND'd to the -CS (chip select) of the two static
>>>> RAMs
>>>> as they are battery backed and contain your pattern data. So you do not
>>>> want random stuff happening. The -RESET prevents anything from getting
>>>> to the RAM chips.
>>>>
>>>> I don't know the pinout of the 707 cartridge but the -RESET to the
>>>> buffer IC7 is going to be protecting something as well.
>>>>
>>>> The 138 Mux just takes a 3 bit number on one side and activates the
>>>> correct line on the other (0 to 7) when the 3 enable lines are
>>>> SET/CLR/CLR
>>>>
>>>> Is your reset line stable and low after power up?
>>>>
>>>>
>>>> On 06.03.14 11:15 PM, luke aldred wrote:
>>>>>
>>>>> Hey Lorne,
>>>>> I don't know much about these CPU circuits, but I assumed the reset
>>>>> signal here turned on the chip select data outputs from IC16 so that
>>>>> RAM can be written only after reset signal. Not sure though as I say.
>>>>> Thanks again
>>>>>
>>>>>
>>>>> From: lhammond@uvic.ca
>>>>> To: wackyvorlon@me.com; laproductions@live.com.au
>>>>> CC: analogue@hyperreal.org
>>>>> Subject: RE: [AH] TR 707 repair
>>>>> Date: Thu, 6 Mar 2014 06:57:38 -0800
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Paul:  I noticed this off the block diagram, any thoughts on
>>>>> which pin for the other resets?
>>>>>
>>>>> the block
>>>>> diagram shows the reset circuit (Q4, Q5) goes to the cpu IC5, the
>>>>> address
>>>>> decoder IC 16, and Buffer IC7.
>>>>>
>>>>> I think its
>>>>> in your reset circuit, attached page.
>>>>>
>>>>> Any thoughts on the reset pins for teh other ICs?  Or why they
>>>>> reset, unless its to sync them all when the voltage is correct.
>>>>>
>>>>>
>>>>>
>>>>> Lorne
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> From: Paul Anderson [mailto:wackyvorlon@me.com]
>>>>>
>>>>> Sent: March-06-14 12:18 AM
>>>>>
>>>>> To: luke aldred
>>>>>
>>>>> Cc: Lorne Hammond; analogue@hyperreal.org
>>>>>
>>>>> Subject: Re: [AH] TR 707 repair
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> It looks like IC11 is a de multiplexer. Here is the data
>>>>> sheet:
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> http://pdf.datasheetarchive.com/datasheetsmain/Datasheets-110/DSAP002952.pdf
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Look at pins 4,5, and 6. If 6 is low, or 4 or 5 are high,
>>>>> then all outputs are pushed high. Including pin 13 which feeds IC16
>>>>> via IC12.
>>>>> The gate on IC12 is configured as an inverter. A high from IC11 pin 13
>>>>> would be
>>>>> low at IC16, holding the output low without regard for the inputs to
>>>>> IC11. I'd
>>>>> say something is making one of those three pins on IC11 sit at an
>>>>> inappropriate
>>>>> level.
>>>>>
>>>>>
>>>>>
>>>>> --------
>>>>>
>>>>>
>>>>>
>>>>> Paul Anderson
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Facilities Manager
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Hackforge
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> On Mar 6, 2014, at 3:12 AM, luke aldred <laproductions@live.com.au>
>>>>> wrote:
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Hey Lorne,
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Thanks again. I can't tell on which pin the reset signal should
>>>>> be on IC7, but in all other locations it is good. IC7 is a buffer
>>>>> for the
>>>>> Cartridge storage, so would it even make a difference? I have the
>>>>> cartridge
>>>>> ribbon unplugged anyway.
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> As for Kenneth Bayes notes, I was noticing small pulses on
>>>>> ground connection, though they are less than 100mv so I wasn't
>>>>> worried. Would
>>>>> this cause dramas?
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Paul, thank you for your input. Pin 10 is connected directly to
>>>>> pin 4, this is the reset signal. Pin 11 is connected to Pin 3, and pin
>>>>> 12.
>>>>> These are fed by an output from IC11. Address data of some sort?
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Luke
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> From: lhammond@uvic.ca
>>>>>
>>>>> To: laproductions@live.com.au; analogue@hyperreal.org
>>>>>
>>>>> Subject: RE: [AH] TR 707 repair
>>>>>
>>>>> Date: Wed, 5 Mar 2014 22:40:32 -0800
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Don't know the machine but I do know the 909 has a reset that
>>>>> stays low and off until there is 5V stable top prevent scrambled
>>>>> memory.
>>>>> So you can have the right voltage at teh power supply but if you reset
>>>>> circuit
>>>>> has a broken trace or dead transistor (on the 909 that is on the PS
>>>>> pcb) it
>>>>> will not allow the cpu to start up.
>>>>>
>>>>>
>>>>>
>>>>> lorne
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> om: luke aldred [mailto:laproductions@live.com.au]
>>>>>
>>>>> Sent: March-05-14 10:00 PM
>>>>>
>>>>> To: analogue@hyperreal.org
>>>>>
>>>>> Subject: [AH] TR 707 repair
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Hi all,
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> I have a
>>>>> TR-707 here which I'm repairing. It does not power boot up. Power
>>>>> supply
>>>>> voltages etc are all good, and scoping the CPU I can see data on the
>>>>> address
>>>>> and data buses.
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> When scoping
>>>>> around the RAM chips, I found that the Chip Select pin (pin 18) for
>>>>> RAM 1,
>>>>> IC14, is sitting on between 4.5-5V. When I look at the same pin on
>>>>> IC13 I can
>>>>> see the chip select data flowing. Now I assumed that the chip address
>>>>> IC 16 may
>>>>> be faulty, so I swapped it out to no avail. Strangely I then noticed
>>>>> when I
>>>>> bridged pins 8 and 9 on IC16 (TC40H010) the unit fired up for the
>>>>> first time.
>>>>> It was mostly functioning, I could hear sounds, swap patterns etc
>>>>> though it
>>>>> seems I could not write any patterns and it was still a bit glitchy in
>>>>> general.
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Anyone
>>>>> got any ideas what else this may be? I have tried new RAM in both
>>>>> positions to
>>>>> no effect, have also tried a new IC11. I wonder if it could be the
>>>>> program ROM
>>>>> or CPU error?
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Thanks
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Luke
>

-- 
=================================
=======   Beatkamp Inc.  ========
=================================
Unit 3, 232 Wilbrod Street
Ottawa, Ontario, K1N-6L6, Canada
Ottawa HQ: 613 518 7402
Vancouver: 604 343 5881
Toronto:   647 479 6231
Palo Alto: 650 353 9193
http://www.beatkamp.com
=================================
BEATKAMP® is a registered
trademark of Beatkamp Inc.
=================================
This e-mail and any attachments
is confidential. If you have
received it in error, please
delete it from your system.
=================================