Chesster Challenger Strange Behaviour

This forum is for general discussions and questions, including Collectors Corner and anything to do with Computer chess.

Moderators: Harvey Williamson, Steve B, Watchman

Forum rules
This textbox is used to restore diagrams posted with the fen tag before the upgrade.
Post Reply
afos99
Member
Posts: 50
Joined: Sun Jul 19, 2015 9:49 am
Location: Brighton UK

Chesster Challenger Strange Behaviour

Post by afos99 »

Hi all
I have just received a Chesster that sounds like Mickey Mouse on helium. The tone of the voice is very high and faster than normal. Also the ready LED is flashing faster than my other model with a loud hum when lit. I opened it up suspecting that it was being overclocked but it has the standard 5Mhz crystal.

It does however have an extra capacitor obviously soldered in parallel to one connected to the crystal. I cut one of the legs of the capacitor but this didn't change the speed of the voice.

Is it possible that the crystal is faulty in some way? should I replace it? I hope some of you guys might have some experience or ideas.

Thanks

Dave
Larry
Senior Member
Posts: 2269
Joined: Wed Aug 01, 2007 8:42 am
Location: Gosford, NSW Australia

Post by Larry »

If it was me I'd replace the crystal, they are dirt cheap. However, my
understanding is that crystals rarely give trouble. I believe the
Chester is a Par Excellence, so it's mate solving times should
match the Par. If after changing the crystal there is no change, then perhaps
we can set up a mate problem with a Par, and the same problem with
the Chester. We time them both and see if they match. I'm assuming
here that the Chester is in otherwise good working condition. Regarding
the second capacitor, snip it off completely.
L
afos99
Member
Posts: 50
Joined: Sun Jul 19, 2015 9:49 am
Location: Brighton UK

Post by afos99 »

Larry

I have some data from past tests I have done on the Chesster family - Big Chesster, Little, RS 1900L, Sharper Image and Saitek ATS.

I use the so called Final Position which was from a game between Milano and Mach III. What I like is the long list of times for other machine to get the g7 move.

Chesster has limited levels so I had to use the average time A1 to A8, 5 seconds to 3.5 minutes. Chesster tended to take twice the average time for each level to get a move.

So on Lv2 (10s) it would get the g7 move after 21s. It didn't get it on Lv1 (5s).

Now the mutant Chesster - it stops on each level in about half the time of my original machine. however it doesn't find g7 until level 3.

My guess is the real time clock is running twice as fast but the chess calculating process is the same.

Will changing the crystal change this? I thought programmers got the RT clock by software and the frequency of the crystal.

Dave
yoyo_chessboard
Member
Posts: 126
Joined: Mon Oct 17, 2016 5:44 pm
Location: South of France - cote d'azur

Post by yoyo_chessboard »

hi

could you make a photo of the main board, the component side ?
the higher the resolution, the better it will be.

i think i know well the fidelity boards as i accelerated many families : SC9, excellence, designer, designer display and chesster and kishon (the german chesster) and even the reversi challenger

the problem you meet is very strange, and i don't think it can be linked to an acceleration, and you said thta it has the 5 Mhz crystal which is the standard.

with the photo i might see something obvious , compared to my chesster, which is running at 6 Mhz, after changing some parts.
User avatar
Stray Black Cat
Member
Posts: 14
Joined: Sat May 10, 2014 8:19 pm
Location: Florida

Post by Stray Black Cat »

I was hoping that this thread would have continued. I was to trying to identify which board on my chesster phantom is running the chess program. The reduced number of levels available from the listed earlier 6100 levels make me think that the chesster piggyback board is running the chess program and is parasitically controlling the 6100 for movement only. How it may relate here is though the 6100 board has a 6 MHz max processor and a 5 MHz crystal, The piggyback chesster has a 4 MHz max processor and a 4 MHz crystal on my machine. The voice defiantly comes from this board. Is it possible that the early chessters may be limited to 4 MHz?
https://drive.google.com/file/d/0ByQNdG ... sp=sharing
This is the CPU and Crystal on the Piggyback chesster.
https://drive.google.com/file/d/0ByQNdG ... sp=sharing
This, I'm sure is the sound effects chip. From scattered web data, it's most likely a custom sound chip with its own rom. I keep getting directs to MAME multiple arcade machine emulator trees and these actually refer to Fidelty and Chesster Challenger and other Fidelity taking chess computers. No actual program data though. No specific data on the chip as well. Fidelity does make mention of the voice contributing company on the last page of the 6126 manual. Chessters speech was produced by Electronic Speech Systems, Hayward, California and is covered by 1980 Mozer U.S. Patent, 4,214,125 and others
yoyo_chessboard
Member
Posts: 126
Joined: Mon Oct 17, 2016 5:44 pm
Location: South of France - cote d'azur

Post by yoyo_chessboard »

Hi stray black cat

i don't understand exactly what you search so i will reply to what i have understood.

i don't personnaly own a phantom but i have many high res photos of the boards inside.

the daughter board is running the chess program , the brain.
it is very similar to the chesster challenger i have and analysed and dumped in MAME.
there is a 65C02 CPU (4 Mhz-able) and a 3.579 Mhz clock.
the 101-1091A02 is an EPROM containing the voice data.
it is the CPU that uses the data in that ROM to produce the vocabulary.
the words are spelled under IRQ (interruptions).
so there is no special sound chip, all the work is done by the CPU.

the motherboard now
if we agree that the brain is in the daughterboard, the arms are in the main board.
so the mainboard is just moving the figures

but what is interesting is the inter-CPU communication.
the two CPU exchange data through a ROM footprint (U4 on the PCB).

and i would be very interested to study that inter-CPU dialog.

would it be possible for you to DUMP the 27C256 which is on the mainboard and on the daughterboard.
they are both on sockets
i already have the dump of the 101-1091A02 ROM as it is the same as in chesster challenger.

regards.
yoyo_chessboard
Member
Posts: 126
Joined: Mon Oct 17, 2016 5:44 pm
Location: South of France - cote d'azur

Post by yoyo_chessboard »

the photos i own are said to be from a 6124.
so it might differ a little from your 6120 in the clock frequency as what you report is different from what i see.

but i suspect we have the same photos :
on mine you seee the board on a red carpet or something like that.

so may be there are yours or you collected them from the same source.

regards
User avatar
Stray Black Cat
Member
Posts: 14
Joined: Sat May 10, 2014 8:19 pm
Location: Florida

Post by Stray Black Cat »

Hello yoyo_chessboard,
I do not have a 6120. I know that the 6120 tech was combined with the 6100 to get the 6124 and 6126. I just found it odd that the play strength on both was supposed to be the same but the daughterboard on the 6124 is slower than the 6120 boards. I don't have a 6126 to see its speed. Thanks for the ID on the mystery chip. I could find no datasheets on it. Do you have any links. As for dumping the 27C256, That is a little beyond my skill at the moment. I will do it when I learn how to safely do it. For the communication between the two boards, there are also short connections to the cmos static ram of the 6100 board. These tap pins 26 (E2-chip enable input) and 27 (W-write enable input). These may tell the 6100 board what to move.
https://drive.google.com/open?id=0ByQNd ... W9pbGRxT2M
If this is the red background with small squares you referred to then those were pics I had on google drive. The red mat is an ESD mat from Radioshack. I use static protection with any circuits containing CMOS chips. I have a lot more close up pics if needed. Unfortunately the power circuits on the Fidelity 6100 Phantom series computers are horrible. Unlike the Milton Bradleys textbook fullwave rectified circuit from a centertapped 12vac transformer. Fidelity rangled the same voltages from a 9vac transformer through two out of phase half wave rectified circuits. The 4rth tier capacitors are taking a serious beating and I don't see them lasting long, stored on a shelf or active. Sure enough my 6124 would not power up when I pulled it out a couple of weeks ago after three years storage. I will replace the 5 Korean SURGE capacitors with high end Vishays rated at 8000 Hrs @ 125°C shortly. :)
afos99
Member
Posts: 50
Joined: Sun Jul 19, 2015 9:49 am
Location: Brighton UK

Post by afos99 »

Sorry for the long delay.

Now have picture of Chesster board with fast voice and clock.
As mentioned before there was a second capacitor added near the crystal which I removed. Also I noticed R18 has been soldered after manufacture but I think it is the right value.

https://drive.google.com/open?id=0B3SZ8 ... 2tPaTlqUW8

Any ideas for the behaviour.

Thanks

Dave
afos99
Member
Posts: 50
Joined: Sun Jul 19, 2015 9:49 am
Location: Brighton UK

Post by afos99 »

Hi all

Going to change the crystal and see what that does, if anything. However, they do have different capacitances this may be why there was an extra capacitor added to try to rectify the problem.

Dave
afos99
Member
Posts: 50
Joined: Sun Jul 19, 2015 9:49 am
Location: Brighton UK

Post by afos99 »

Have swapped the crystal but get the same result. Therefore something else is causing the program clock speed to be fast. I haven’t put a stopwatch on the levels but the voice is not right.

I do remember when I had two VCCs that the speed of the voices was slightly different. Perhaps clock speed is affected by component quality or variation!?

Hopefully someone may have some idea because I am stumped.

Dave.
Post Reply