15th May 2020, 9:22 am | #121 |
Pentode
Join Date: Jul 2017
Location: Toulon, France
Posts: 239
|
Re: Recreating the Bywood Scrumpi
Hello
Bravo I have already anticipated the II quartz pads for the GND I noticed that it was doing the complete round of the PCB it does not rebound to me it seems but it is possible at R11 level to make a way when thought you? by reading the doc NS we notice that 2V gives 1 in data for the II can be as the think slothie that is enough |
15th May 2020, 6:18 pm | #122 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,364
|
Re: Recreating the Bywood Scrumpi
Thank you. I assume you also have provision for the two resistors and the Capacitor that is needed with the Crystal?
It would not hurt to complete the Ground if the route is easy - it will reduce reflections and noise. Reading it with a DMM it is actually only 1.7v which is a shame... Maybe it will be different with 74C04 rather than the 74LS04 which is all I had - even I got fed up of reading the lights inverted all the time using my 74C07s! Anyway it was yet another long day on this with many problems and I have not really solved any existing (mainly how to write a 1 into memory without a pullup) - only created new ones: I wired up the remaining address/flag LED's and arranged some headers and connectors for the Flags and SIN/SOUT. I also burned out my 5V regulator on my Breadboard supply - lucky I noticed 8.7V on an address line and the SC/MP survived So time to fit a decent power feed direct to the board from my MK14 kit - an LM340 and a CAP - more are on the way anyway. Then I noticed some real problems and intermittent working - removing the CPU showed failed socket connectors - attempts to fix them resulted in bent CPU pins and a lot of swearing in the end I removed the socket and replaced... - a turned pin one could not be persuaded to fit across my hacked together stripboards so a cheap one had to go back in - and guess who had lost part of the track on CPU Pin20 (GND) and didn't notice - anyway fixed and it after going out for a walk in disgust and it once again cycles addresses... I am assuming it is running 00H HALT as I get a Pulse on D7 (on all actually but, I think that is just the tri-state flickers). Of course only in HALT on D7 mode does D7 look even ON where it has R8 to pull it up I suspect the missing resistor on the schematic (R8) is used to provide a pull up for the RESET so it just grounds and goes back HIGH maybe? I will wire up that way and try otherwise the RESET will be very messy even with the two NOT gates as they will float - we have it direct to +5v at the moment. Anyway for your viewing pleasure and yes I know it is an SC/MP chip known as SCAMP not SCRUMPI - I am tired now... https://youtu.be/H-t2mDUwXHA ...and a piccy of the RAM adapters thanks to Slothie |
15th May 2020, 6:33 pm | #123 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Recreating the Bywood Scrumpi
Nice video! When the bus goes tristate the data bus seems to float high - is that how were supposed to write 1s into memory?
|
15th May 2020, 6:42 pm | #124 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Recreating the Bywood Scrumpi
Also you seem to have the OE pin connected to NWDS (R/W) rather than NRDS which would make the memory write only! You'll need a flying lead to NWDS somewhere.
|
15th May 2020, 7:09 pm | #125 | |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,364
|
Re: Recreating the Bywood Scrumpi
Quote:
That may mean that the 1.7v is enough (the data sheet says 2.2v though) and I am writing memory when I use load - I just get nothing back! I must have misunderstood the wiring needed on the adapter - lucky I posted the picture. |
|
15th May 2020, 7:20 pm | #126 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Recreating the Bywood Scrumpi
It might be worth trying connecting the CE and OE together, because with the 2111 if you assert OE and R/W together the write takes precedence.
|
15th May 2020, 8:00 pm | #127 |
Pentode
Join Date: Jul 2017
Location: Toulon, France
Posts: 239
|
Re: Recreating the Bywood Scrumpi
bravo for the reset I have you precise my idea of a pulup with r11 my pcb in construction the fact for gnd ok I will plug it the solution for the data is to reach a draw on the led to have the 2u bus
|
15th May 2020, 8:44 pm | #128 |
Pentode
Join Date: Jul 2017
Location: Toulon, France
Posts: 239
|
Re: Recreating the Bywood Scrumpi
made me your remarks to finalized the pcb
|
15th May 2020, 9:07 pm | #129 | |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,364
|
Re: Recreating the Bywood Scrumpi
Quote:
I think your idea of LED's with resistance in is a good one (Shame they are only 3mm ones). I did some experiments with modern LED's and some I had from the 70's probably from my Dad's stuff - the older ones seemed to draw even more current in the 64mA region (down to 12.8 with a 330ohm resistor) modern ones were around 55mA down to about 7.1 with a resistor. I do wonder if the extra current makes the input different raising the value on the data tri state pullup - maybe he just took the risk of running the chips at double the current... In reality I cannot find the datasheet for the 74C04 only the HC version which may have at 25mA a lower capacity... I think then as you say we will be close to a PCB. I wonder if including space for a LM380-5V, Capacitor and input power socket might be an idea to simplify things for actual use with an SC/MP II? I also wonder if the SC/MP I floats the tri-state slightly higher - maybe we need to study the Scrumpi II board and see if it needed pullups on the data lines. Anyway I will make my RAM readable as well as see how it goes. |
|
15th May 2020, 9:17 pm | #130 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,364
|
Re: Recreating the Bywood Scrumpi
This is looking good but, your resistors are marked wrong.
R3 should be R6 (marked so at the bottom on the photo) R4 Should be R7 (also visible.) R6 is R5 (also visible) R2 is R4 (as R2 is visible on the photo below the IO chips) I cant read the one in the corner on your post but, that should be R3 R5 is R8 |
15th May 2020, 9:32 pm | #131 |
Pentode
Join Date: Jul 2017
Location: Toulon, France
Posts: 239
|
Re: Recreating the Bywood Scrumpi
for scrumpi II you will notice the resistances of pulup for the names of the resistances nothing serious I correct in final the pcb is realized at once so the schemat takes form at the same time or inversely in order to make appear the via gradually c is long us Away from scrumpi I
|
15th May 2020, 9:34 pm | #132 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Recreating the Bywood Scrumpi
I just looked at a photo of a SCRUMPI 2 and there appear to be vertically mounted components (resistors?) By the data switches that could be pullups. I need a better picture to be sure though.
|
15th May 2020, 9:36 pm | #133 |
Pentode
Join Date: Jul 2017
Location: Toulon, France
Posts: 239
|
Re: Recreating the Bywood Scrumpi
see
|
15th May 2020, 9:38 pm | #134 |
Pentode
Join Date: Jul 2017
Location: Toulon, France
Posts: 239
|
Re: Recreating the Bywood Scrumpi
it is indeed resistances
|
15th May 2020, 9:40 pm | #135 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,364
|
Re: Recreating the Bywood Scrumpi
I was studying the same - the vertically mounted ones are added by Mr Austin I think in the RAM expansion space (the bottom chip as 5,6,7,8 are the expansion space and 9/10 are the supplied 256. The seven on the bus though look like pullups...
390K seems quite high but, maybe just enough... |
15th May 2020, 9:42 pm | #136 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,364
|
Re: Recreating the Bywood Scrumpi
There are only seven as I expect the old R8 pullup associated with RUN/HALT is still lurking elsewhere on the board.
|
15th May 2020, 9:57 pm | #137 |
Pentode
Join Date: Jul 2017
Location: Toulon, France
Posts: 239
|
Re: Recreating the Bywood Scrumpi
see
|
16th May 2020, 8:08 am | #138 |
Pentode
Join Date: Jul 2017
Location: Toulon, France
Posts: 239
|
Re: Recreating the Bywood Scrumpi
after proofreading it seems that the system used for the scrumpi I does not work with the 8060 the technology and different mos channel N for the 8a/500d and P for the 8060 the difference is sensitive at the level of the power supply with vgg -7v and vss +5
see the data sheet SLOTHIE had noticed it and I think that the 74c07 driver draw is enough for the 8060 it is noticed that SCRUMPI II uses visible pulup behind the switches the model must therefore be reviewed with an 8a/500d more in the article PE 08/77 there is mention of an improvement for load with a flip flop flip flop when thought you |
16th May 2020, 10:10 am | #139 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Recreating the Bywood Scrumpi
Yes, if you look at the driver and receiver equivalent circuits on page 6 of each pdf the isp8a-500 (SC/MP 1) has a pull up on the databus but the isp8a-600 (SC/MP 2) does not. That is consistent with the appearance of pullups on the SCRUMPI 2 which used the later chip.
So it looks like we need pullups if your going to use the SC/MP 2 - the MK14 has 1.2k pullups on the data bus butt I would probably use something larger 4k-10k. |
16th May 2020, 1:16 pm | #140 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,364
|
Re: Recreating the Bywood Scrumpi
Indeed looking again there is actually a space for an R18 on the Jim Austin one just below R19-R25 so at least we know that for the SCRUMPI 2 project... I really hope we can get some more closeup photos of that board and the underneath though - I will have to write again to the contacts I have as I have heard nothing back.
So I just happened to have a bag of the same resistors so added them as pullups - even though they are very high resistance 390K rather than the 10K suggested by Slothie - after fixing a dry joint on D2 all the lines have a nice clean high on them in the 1 position. Unfortunately when I plug back in the RAM almost all the lines other than D4 and D5 end up pulled low no matter what I do. Maybe I still have an error in the 9111 adapter - I have now made Pin 10 a flying lead (from both RAMs) to NRDS - I assumed they no longer needed to be wired to anything else as well - other than Pin 10 to Pin 8 GND. No shorts are evident on the Data Lines to ground or between each other without the chips there is no contact to Zero unless the Switches are Down. I will take another look at the chip adapter - or maybe the RAM is just fault now - I will test on the MK14 perhaps. Hopefully my new book that arrived this morning will help - his list of tests before installing the CPU are the ones I carried out on this board so that is good news.... |