Elethiomel builds a Phoenix H5C1

[Elethiomel]

Well, my H5C1 arrived and I have to say, this thing is a beauty. It's so clean and professional looking! It seems a shame to muck it up with my amateur soldering .

I grabbed my donor board, a C103175 and hit my first snag immediately. I knew the H5C1 could take an SMD GLUE and MMU, but didn't realise there were at least two SMT footprints for those chips. The IMP chips are larger, square parts compared to the smaller rectangular chips on one of my other project boards. Unfortunately, it seems the H5C1 supports only these smaller chips and the socketed PLCC versions. Is there enough of these chips around that adding support to them for a future revision would be a good idea or is there perhaps another technical reason they're not already supported?

20220409_173747.jpg (414 KiB) Viewed 5070 times

20220409_173754.jpg (365.19 KiB) Viewed 5070 times

Oh well, off to ebay I go. I scored a C070789 board for £39 "for parts of spares" that shows a scrambled screen from an ST that had been broken up for parts.

s-l1600.jpg (326.22 KiB) Viewed 5070 times

Could be the RAM. Could be a socketed chip knocked loose like the shifter? Either way, even if one of the chips I need from it is borked, I still get a lot of good parts from it for cheap, and I get a decent model of board for possible future projects. Seller is away, so it will be at least another week before I get it :/

Snag 2 was hit pretty quickly too. It looks like the H5 needs a PLCC MFP. All of my ST systems have DIP MFPs. Is a PLCC MFP on an ST a common thing and I just happen to have boards with DIPs on them or is there another reason why only the PLCC is supported? I've ordered the appropriate part (as well as some other goodies) from the store

Next step : HEADER PINS!

[Elethiomel]

Upon closer inspection, it looks like the MMU and GLUE are actually just a PLCC with the J legs soldered directly to the board? I thought they were SOIC type devices with the legs splayed out. I should be able to desolder these, clean them up and pop them in sockets from what I can see. If that's the case, I have a board I don't really need on the way, but it will be fun to debug and fix

[Steve]

You can buy all these chips from Exxos store as far as I know. If he's out of stock Atarifreakz online or eBay has them I think.

[Elethiomel]

Cheers Steve, wanted to get something useful from this duff board so I've desoldered the MMU and GLUE. I just need to clean them up a bit and they should pop right in to the sockets.

All header pins in place and all jumpers placed according to the Exxos thread. I'll try do the audio fix tonight.

[Elethiomel]

I didn't end up fitting the audio fix in the end. The resistors I have are just too big and I couldn't get one to sit right and solder properly.

So, I stripped down my donor board, soldered in the reset switch, the RAM socket, the monitor socket and a 64pin DIL for the CPU and set about populating the board.
Apologies for the terrible lighting. My workshop's lighting is a nightmare for taking photos.

populated.jpg (427.13 KiB) Viewed 4983 times

I then plugged in my trusty exxos PSU, the keyboard and fired up the system. Power light on keyboard, nothing on screen. It remains blank. Oh well. Time for some debugging.

There's two things wrong with the populated board above that I know about that I've since fixed.

1) The crystal oscillator is oriented the wrong way.
2) The ROM is also oriented the wrong way.

There's most definitely more! If anything jumps out please holler.

OK, we now have two things fixed but I'm still getting a blank screen. In my experience I should at least be getting a white screen if the shifter is running, and there's VSYNC and HSYNC from the GLUE, even if the CPU is locked up or the ROM or RAM is bad. OK time to put what I've learned from watching countless hours of Youtube channels like "Adrian's Digital Basement" to good use. Let's see if the chips are getting power. Yup, CPU and shifter look good at least. 5.12 volts. ground pins seem good.

How about the clocks?

NOTHING on pin 15 on the 68K.
Odd.
Pin 2 on the shifter is reading at 32MHz

20220413_005641.jpg (298.78 KiB) Viewed 4983 times

OK, that's good. Now, the shifter should be outputting 16MHz on pin 39. This is then fed to the MMU and the MMU creates an 8MHz clock which goes on to the rest of the system. Various other clocks like a 4MHz and 2MHz one etc. are created from the MMU and GLUE, but they're not doing anything without that 16MHz from the shifter!

OK, maybe I borked the shifter? Swap it for a known good one. Nope. Same deal.
Try switching the MMU and GLUE just in case? Nope.

OK Let's make sure those chips are still good. Back in a known working ST they go and up it boots. OK. The Shifter, MMU and GLUE are good.

At this point I pulled every single chip on the board except for the shifter. If there's a bad chip doing something odd to say, the data bus, then maybe it's causing an issue.

Still no 16MHz clock

And that's where I'm going to leave it tonight. I did a visual inspection of the sockets I've soldering on, and nothing seems to be bridged. Curious.

[Elethiomel]

Because I can't sit still and leave things alone, I took another visual pass at the board and noticed that there seems to be solder splashes in quite a few places. Maybe there's a short and I just need to give the board a good clean? Here's two random examples

20220413_014336.jpg (235.73 KiB) Viewed 4982 times

20220413_014401.jpg (340.27 KiB) Viewed 4982 times

These are nowhere near where I've been soldering, so I'm quite confident they're artefacts from the manufacturing process. I'll get scrubby with a brush and some IPA either later on or tomorrow.

[Elethiomel]

I just found this thread : https://www.exxosforum.co.uk/forum/viewt ... =78&t=2362
I'll go through it and post results

[exxos]

I just found this thread : https://www.exxosforum.co.uk/forum/viewt ... =78&t=2362
I'll go through it and post results

If you not got 16mhz then at this time there's not much point. Check the 16mhz line isn't shorted to GND or 5v, or tracks next to it first.

[Elethiomel]

Pin 39 is shorted to neither ground nor 5V.

I took readings from all pins without the shifter installed. These readings are a bare H5 board with only the RAM socket, the monitor socket, a single 64pin DIL socket installed and the 32MHz oscillator. No ICs at all. Power is supplied via an Exxos PSU.

Pin - Reading
1 - 0V
2 - 3.5V
3 - 1.23V
4 - 1.23V
5 - 1.23V
6 - 1.23V
7 - 1.23V
8 - 1.23V
9 - 1.23V
10 - 1.23V
11 - a few milliVolts
12 - 1.32V
13 - 1.24V
14 - 1.27V
15 - 1.27V
16 - 1.24V
17 - 1.24V
18 - 1.24V
19 - 1.24V
20 - 0V (pin is correctly connected to ground - sounded out with multimeter)
21 - 1.68V
22 - 1.8V
23 - 1.8V
24 - 1.8V
25 - 1.87V
26 - 1.8V
27 - 1.87V
28 - 1.8V
29 - 1.7V
30 - around 150mV
31 - 5.14V - This pin is NOT shorted to the 5V rail
32 - 5.14V - This pin is NOT shorted to the 5V rail
33 - 5.14V - This pin is NOT shorted to the 5V rail
34 - 5.14V - This pin is NOT shorted to the 5V rail
35 - 5.14V - This pin is NOT shorted to the 5V rail
36 - 5.14V - This pin is NOT shorted to the 5V rail
37 - a few milliVolts
38 - a few milliVolts
39 - a few milliVolts
40 - 5.14V (pin is correctly connected to 5V. Sounded out with multimeter

Only pin 20 sounds as being connected to ground
Only pin 40 sounds as being connected to 5V


After pulling the shifter and swapping it so many times, the socket was starting to look bad and I ended up with no readings on pin 2 where I should have gotten 32MHz. I decided to unsolder the socket and replace it with round socket connector strips.

This was a massive PITA. I desoldered all the chips I needed from the donor ST with my desoldering gun in less than 30 minutes. Getting this socket out took 2.5 hours (not an exaggeration!). I couldn't get all the solder off the pins with the gun, or with braid, or a sucker, or after adding solder again, and ended up breaking the socket up in to small pieces and snipping off the legs. More braid, more desoldering gun and holding the legs with a tweezers and I finally got the last leg out. Some cleanup, some flux and 5 minutes later I had the strips in place.

No change Tried another shifter again. No change.



Am I definitely right that all this shifter needs to produce 16MHz on pin 39 is the following? :

Pin 40 - 5V
Pin 20 - GND
Pin 2 - 32MHZ input.
Pin 1 - 10k to ground (may not even be needed?)

Maybe I can try run it outside the board with some jumpers.

[exxos]

I'm assuming you have 32Mhz on pin 2 as you have 3.5V. But you really need to scope to diagnose switching signals.

AFAIK, 32Mhz into the shifter, plus 0v,power, you should be getting 16mhz output.

You can try carefully bending pin39 shifter up enough to miss the socket, then try 16mhz then..

Have you removed the MMU ?