officer960's H5 C1 build
-
exxos
- Site Admin

- Posts: 28344
- Joined: 16 Aug 2017 23:19
- Location: UK
Re: officer960's H5 C1 build
Don't those other boosters have caches? They likely give double results for whatever test sysinfo does. Turn caches off and I'd assume you would see much different results.
-
officer960
- Posts: 67
- Joined: 22 Jul 2018 21:42
Re: officer960's H5 C1 build
They do, and that makes sense. Thanks for taking the time to educate me. The caches also seem to wreak havoc on compatibility which is why I wanted this accelerator in this particular computer. Tiny boost for farting around in GEM but can disable and be 100% stock for things that get crashy.exxos wrote: 08 Aug 2025 01:16 Don't those other boosters have caches? They likely give double results for whatever test sysinfo does. Turn caches off and I'd assume you would see much different results.
H5C1, H5C5B, 1040 ST, Mega ST, STe, Mega STe, Falcon, TT030, Amiga 2k V2+, Amiga 2k Video Toaster GVP030, Amiga 500, Amiga 500+ V2, Amiga 500+ Firebird V4, Amiga 1200 PiStorm, FPGA: V4SA, MiST, MiSTer, UnAmiga, U64 Elite
-
officer960
- Posts: 67
- Joined: 22 Jul 2018 21:42
Re: officer960's H5 C1 build
@exxos after my initial faux pau with the 16MHz booster I wanted to do some experiments before coming back with questions.
The TOS decoder works perfectly, but the booster only "kinda" works at 8 or 16Mhz. When running Gem software it seems to behave normally at either 8 or 16Mhz.
When I run games, demos, or basically anything with sound or graphics (including most paint programs Spectrum 512 for example) - all bets are off at either speed. At 16MHz I get a lot of graphical glitches which is somewhat to be expected, but definitely not normal behavior. When I connect the 8MHz ON wire to ground (via a switch) the 8MHz behavior is beyond not normal. The computer gives the same or worse graphical glitches and seems to go into fluctuating slow motion, as if it's unstable - like its stuttering.
As an example with PP's hard drive adapted games at 16MHz the "splash screen" is practically always garbled on any of my accelerated ST's. If I run at 8Mhz the splash screen is stable. With this booster they are garbled at 16Mhz AND at 8MHz. At 8MHz the colors in the games are drastically off, the sound is completely wrong. If the game/demo doesn't outright crash or hang there are random pixels and portions of the screen are in the wrong place - like it clips a quarter of the screen and puts it at the other end of the screen.
I have tried the following:
drastically shortened the 16MHz input wire connected between the booster and pin #39 of the Shifter with no change in behavior. I tried two other 68000 CPU's on the booster and it behaves identically with all three of the CPU's.
tried running the offending software from two different Ultrasatan SD cards one with PP's driver the other with HD Driver 12.50 and ran games from floppy (Gotek) and get the same issues.
removed the booster, put a Dip 68000 in the same socket the booster/decoder had occupied, or a PLCC 68000 in its socket the machine runs all the games/demos/software perfectly.
I currently have the decoder installed (without CPU), connected the solder jumper and disabled the booster with a 68000 in the PLCC socket and it's very stable under TOS 1.04 or 2.06 with any games/demos etc.
I should have grabbed some video of what its doing but hopefully the description will suffice for ideas of what may be going on?
The TOS decoder works perfectly, but the booster only "kinda" works at 8 or 16Mhz. When running Gem software it seems to behave normally at either 8 or 16Mhz.
When I run games, demos, or basically anything with sound or graphics (including most paint programs Spectrum 512 for example) - all bets are off at either speed. At 16MHz I get a lot of graphical glitches which is somewhat to be expected, but definitely not normal behavior. When I connect the 8MHz ON wire to ground (via a switch) the 8MHz behavior is beyond not normal. The computer gives the same or worse graphical glitches and seems to go into fluctuating slow motion, as if it's unstable - like its stuttering.
As an example with PP's hard drive adapted games at 16MHz the "splash screen" is practically always garbled on any of my accelerated ST's. If I run at 8Mhz the splash screen is stable. With this booster they are garbled at 16Mhz AND at 8MHz. At 8MHz the colors in the games are drastically off, the sound is completely wrong. If the game/demo doesn't outright crash or hang there are random pixels and portions of the screen are in the wrong place - like it clips a quarter of the screen and puts it at the other end of the screen.
I have tried the following:
drastically shortened the 16MHz input wire connected between the booster and pin #39 of the Shifter with no change in behavior. I tried two other 68000 CPU's on the booster and it behaves identically with all three of the CPU's.
tried running the offending software from two different Ultrasatan SD cards one with PP's driver the other with HD Driver 12.50 and ran games from floppy (Gotek) and get the same issues.
removed the booster, put a Dip 68000 in the same socket the booster/decoder had occupied, or a PLCC 68000 in its socket the machine runs all the games/demos/software perfectly.
I currently have the decoder installed (without CPU), connected the solder jumper and disabled the booster with a 68000 in the PLCC socket and it's very stable under TOS 1.04 or 2.06 with any games/demos etc.
I should have grabbed some video of what its doing but hopefully the description will suffice for ideas of what may be going on?
H5C1, H5C5B, 1040 ST, Mega ST, STe, Mega STe, Falcon, TT030, Amiga 2k V2+, Amiga 2k Video Toaster GVP030, Amiga 500, Amiga 500+ V2, Amiga 500+ Firebird V4, Amiga 1200 PiStorm, FPGA: V4SA, MiST, MiSTer, UnAmiga, U64 Elite
-
exxos
- Site Admin

- Posts: 28344
- Joined: 16 Aug 2017 23:19
- Location: UK
Re: officer960's H5 C1 build
Things like demos, Spectrum 512, overscan stuff etc won't work at 16mhz as they are timing dependent..
I assume you have removed the plcc CPU if running on the H5? And you got the DIP 68K from the store?
8mhz should work like a stock cpu. I can only suggest you have a bad connection somewhere. Maybe run yaartt ram test to make sure your not hitting errors.
EDIT
I half recall a issue with the clock with some cpus.. But it's almost 2am, so it will have to wait now.. I suggest scope the CPU clock on the booster to see I'd it looks good or not compared to the MB clock 8mhz.
I assume you have removed the plcc CPU if running on the H5? And you got the DIP 68K from the store?
8mhz should work like a stock cpu. I can only suggest you have a bad connection somewhere. Maybe run yaartt ram test to make sure your not hitting errors.
EDIT
I half recall a issue with the clock with some cpus.. But it's almost 2am, so it will have to wait now.. I suggest scope the CPU clock on the booster to see I'd it looks good or not compared to the MB clock 8mhz.
-
officer960
- Posts: 67
- Joined: 22 Jul 2018 21:42
Re: officer960's H5 C1 build
Spectrum 512 and overscan stuff etc is not working at 8MHz with the Dip 68000 in the booster set at 8MHz.exxos wrote: 12 Aug 2025 00:58 Things like demos, Spectrum 512, overscan stuff etc won't work at 16mhz as they are timing dependent..
I assume you have removed the plcc CPU if running on the H5? And you got the DIP 68K from the store?
8mhz should work like a stock cpu. I can only suggest you have a bad connection somewhere. Maybe run yaartt ram test to make sure your not hitting errors.
EDIT
I half recall a issue with the clock with some cpus.. But it's almost 2am, so it will have to wait now.. I suggest scope the CPU clock on the booster to see I'd it looks good or not compared to the MB clock 8mhz.
Yes, I was only running one CPU in the system (H5C1) at a time. Yes, I bought the Dip 68000 from your store, but I tried two other Dip 68000's from my collection (they all give same behavior).
I've run YAART and Nembench, Gembench, RAMtest you pretty much name it - they all run fine - its when I try to use anything like a game or demo or software that has sound and its not acting right at 8MHz.
Get some sleep!
H5C1, H5C5B, 1040 ST, Mega ST, STe, Mega STe, Falcon, TT030, Amiga 2k V2+, Amiga 2k Video Toaster GVP030, Amiga 500, Amiga 500+ V2, Amiga 500+ Firebird V4, Amiga 1200 PiStorm, FPGA: V4SA, MiST, MiSTer, UnAmiga, U64 Elite
-
exxos
- Site Admin

- Posts: 28344
- Joined: 16 Aug 2017 23:19
- Location: UK
Re: officer960's H5 C1 build
To hot to sleep :(
Sound is a known problem, its a "bug" in the GLUE IC, hence SND jumper on the H5. Though I've never known 8mhz to cause issues.
Try a 1k resistor from a 5v 68k socket between pins 10 & 14. See if it makes any difference. You could also try a different GLUE IC.
Sound is a known problem, its a "bug" in the GLUE IC, hence SND jumper on the H5. Though I've never known 8mhz to cause issues.
Try a 1k resistor from a 5v 68k socket between pins 10 & 14. See if it makes any difference. You could also try a different GLUE IC.
-
officer960
- Posts: 67
- Joined: 22 Jul 2018 21:42
Re: officer960's H5 C1 build
[
Used a relocator board for illustration purposes only since the machine is put away for the evening: Confirming this is what you’re saying before I try it in the morning. Run a 1k resistor between pin 10 and pin 14 of the 68000 like in above pic?exxos wrote: 12 Aug 2025 02:58 To hot to sleep :(
Sound is a known problem, its a "bug" in the GLUE IC, hence SND jumper on the H5. Though I've never known 8mhz to cause issues.
Try a 1k resistor from a 5v 68k socket between pins 10 & 14. See if it makes any difference. You could also try a different GLUE IC.
You do not have the required permissions to view the files attached to this post.
H5C1, H5C5B, 1040 ST, Mega ST, STe, Mega STe, Falcon, TT030, Amiga 2k V2+, Amiga 2k Video Toaster GVP030, Amiga 500, Amiga 500+ V2, Amiga 500+ Firebird V4, Amiga 1200 PiStorm, FPGA: V4SA, MiST, MiSTer, UnAmiga, U64 Elite
-
exxos
- Site Admin

- Posts: 28344
- Joined: 16 Aug 2017 23:19
- Location: UK
-
officer960
- Posts: 67
- Joined: 22 Jul 2018 21:42
Re: officer960's H5 C1 build
@exxos I installed a 1k resistor between pin 10 and pin 14 of the 68000, removed the solder blob on the disable jumper, reconnected Shifter Pin #39 to the 16MHz and connected the 8MHz ON cable to ground. Confirmed the computer was running at 8MHz with GEMBench. Exactly the same issues as before. I took videos (terrible quality but maybe you can glean something from them) of one game and two demos that work perfectly without the booster attached.
Enchanted Land - Booster Set at 8MHz
DHS Demo #1 - Booster Set at 8MHz
DHS Demo #2 - Booster Set at 8MHz
I disconnected the 16MHz input wire from the Shifter, left the solder jumper open, removed the Dip 68000 (and 1k resistor) and installed the PLCC 68000 and re-ran the game and two demos. They work fine.
Enchanted Land - Booster Disabled
DHS Demo #1 - Booster Disabled
DHS Demo #2 - Booster Disabled
Enchanted Land - Booster Set at 8MHz
DHS Demo #1 - Booster Set at 8MHz
DHS Demo #2 - Booster Set at 8MHz
I disconnected the 16MHz input wire from the Shifter, left the solder jumper open, removed the Dip 68000 (and 1k resistor) and installed the PLCC 68000 and re-ran the game and two demos. They work fine.
Enchanted Land - Booster Disabled
DHS Demo #1 - Booster Disabled
DHS Demo #2 - Booster Disabled
H5C1, H5C5B, 1040 ST, Mega ST, STe, Mega STe, Falcon, TT030, Amiga 2k V2+, Amiga 2k Video Toaster GVP030, Amiga 500, Amiga 500+ V2, Amiga 500+ Firebird V4, Amiga 1200 PiStorm, FPGA: V4SA, MiST, MiSTer, UnAmiga, U64 Elite
-
dml
- Posts: 842
- Joined: 15 Nov 2017 22:11
Re: officer960's H5 C1 build
It is just the syncscroll and related sync effects which are breaking. That looks like a misaligned CPU clock.
I.e. it 'works' from the CPU perspective and the speed is correct - but anything involving writes to the shifter/glue is either too early or too late and the timecritical sync effects break.
Should be fixable by adding some buffers or other delays on the clock arriving from the shifter, to get it realigned.
I.e. it 'works' from the CPU perspective and the speed is correct - but anything involving writes to the shifter/glue is either too early or too late and the timecritical sync effects break.
Should be fixable by adding some buffers or other delays on the clock arriving from the shifter, to get it realigned.
d:m:l
BadMooD d/l: https://www.leonik.net/dml/sec_bm.py
SVO30 d/l: https://www.leonik.net/dml/sec_svo30.py
Q2 engine d/l: https://www.leonik.net/dml/sec_q2.py
AGT project: https://www.leonik.net/dml/sec_agt.py
Atari page: http://www.leonik.net/dml/sec_atari.py
YT: https://www.youtube.com/@dmlTPT
BadMooD d/l: https://www.leonik.net/dml/sec_bm.py
SVO30 d/l: https://www.leonik.net/dml/sec_svo30.py
Q2 engine d/l: https://www.leonik.net/dml/sec_q2.py
AGT project: https://www.leonik.net/dml/sec_agt.py
Atari page: http://www.leonik.net/dml/sec_atari.py
YT: https://www.youtube.com/@dmlTPT
Who is online
Users browsing this forum: ClaudeBot and 1 guest