Turbo Everdrive V2.4 Region mod fix. For mmmonkey/doujindance region mods

[Keith Courage]

Fix to get a newer V2.4 everdrive to work with this region mod http://www.mmmonkey.co.uk/multi-region-pc-engine-turbo-grafx-16/

This is the region mod I use for most of the systems I work on. Because of this I was able to test on 4 different systems. A pc engine DUO, DUO-R, core grafx, and Turbo Express were all tested doing this fix and all worked. The original symptom was that the everdrive would not even boot to the menu screen. 

It's actually very simple. Instead of using 5V from the voltage regulator to power the region mod, one must simply tap the AC adapter voltage from somewhere on the board instead. So that would be 9V for DUO-R/RX, 10V for regular DUO, 9V pc engine or coregrafx, Express or GT 7V etc....

I checked the chips data sheet and the chips are supposed to be able to handle up to 18V. So there hopefully won't be any long term damage to the region mod by doing this.

Original thread was brought to my attention by ovalfastlx and can be found here. http://krikzz.com/forum/index.php?topic=3733.0 I probably wouldn't tap the power directly from the fuse holder like this guy did but whatever.

I can also confirm that a 8PDT switch is compatible with the newer V2.4 everdrive as well. Tested this with a US TG16, DUO-R, and regular pc engine DUO.

[wilykat]

How about adding another 5v regulator just for the mods?  Region mod, RGB, my BRAM hack, etc all would add up and it would make sense to have another 5v regulator dedicated to mods.  A basic no frill 5v DC converter can be had for a few dollars and should be able to handle something like 300mA to 500mA.

[StarDust4Ever]

How about adding another 5v regulator just for the mods?  Region mod, RGB, my BRAM hack, etc all would add up and it would make sense to have another 5v regulator dedicated to mods.  A basic no frill 5v DC converter can be had for a few dollars and should be able to handle something like 300mA to 500mA.

Good idea. Forgive me for interjecting, but it seems running the modchip logic would be a bad idea. To quote myself from the Everdrive thread:

Even if the logic chip can handle 10V DC at VCC, I feel it would be unwise to tap this source because the logic chip will output 9-10VDC for "high" logic level outputs as well. This means the inputs of chips connected to the logic bus, either on the Everdrive/cartridge or console side, may be getting too much juice. We don't know whether all 5V logic chips (TTL, CMOS, whatever) will be tolerant of this excess voltage signal. It could get pulled down to 5.6V by internal clamping diodes, and this could generate excess current resulting in heating of the logic chips. Operation in such a manner would require inline resistors to limit current flow as is often used when interfacing modern 3.3V logic chips with vintage 5V bus.

So while such a hack may techinically "work," I am concerned that operating the modchip off of 9-10V would place undue stress on the hardware. As to why the voltage mod may work, perhaps the chips are not switching quickly enough at 5V, as increasing the voltage will decrease the response time proportionally. A better solution might be to use high speed chips, ie HC series instead of LS series, but mixing logic types can cause it's own issues which is beyond the scope of this discussion.

[NightWolve]

How about adding another 5v regulator just for the mods?  Region mod, RGB, my BRAM hack, etc all would add up and it would make sense to have another 5v regulator dedicated to mods.  A basic no frill 5v DC converter can be had for a few dollars and should be able to handle something like 300mA to 500mA.

Yeah, that sounds like a better idea (or maybe at 6 VDC), especially considering that an original Turbo Duo power supply is a cheap, linear, basic dry-transformer model (coil, 4 diodes, 1 smoothing capacitor, 1 fuse, that's it) that outputs ~12 VDC (seen it go 13 before) from my tests given input house ~120-128 VAC.

If you upgraded your Duo power supply to a modern 9 VDC 1.5 Amp regulated model (per Steve's expert recommendations), then Keith's idea would feel more comfortable, but still, that 5-6 VDC regulator is 50 cents, so I'd rather have another more replaceable component take the heat/abuse than rely on whatever the custom PCB mod/device has on it which does whatever.

http://www.pcenginefx.com/forums/index.php?topic=14453.msg374635#msg374635

[L-Digital]

I came up with this fix originally on the Everdrive forum after buying a doijindance modded unit. Glad to see it being spread and evolved now that the other chap wrote some decent instructions.

I used a 7.5v Psone psu and had no issues tapping that directly via the fuse. When I tried the regulated 5v Line it didn't provide enough juice for the Everdrive to work.

I no longer have my duo as I have changed over to a classic briefcase unit

[thesteve]

a few notes on the chips
because these dont buffer the signals, the outputs will still be a bit below the inputs regardless of the power source
only danger/issue is the ESD protection afforded by the clamp diodes in the switch chips is disabled by the higher rail, making card/system damage more likely

[Keith Courage]

i'm not saying this doesn't have any risk. Just saying that it works.

How about adding another 5v regulator just for the mods?  Region mod, RGB, my BRAM hack, etc all would add up and it would make sense to have another 5v regulator dedicated to mods.  A basic no frill 5v DC converter can be had for a few dollars and should be able to handle something like 300mA to 500mA.

Unfortunately this will not fix the issue. Tried adding an extra voltage regulator dedicated for just the region mod and 5 volts still doesn't cut it for the newer 2.4 everdrive. At least with this region mod.

[thesteve]

on the bright side the 3383 based mods work on 5V on all but a few systems

[StarDust4Ever]

Unfortunately this will not fix the issue. Tried adding an extra voltage regulator dedicated for just the region mod and 5 volts still doesn't cut it for the newer 2.4 everdrive. At least with this region mod.

It may have something to do with the response time of the chips. Does the region mod use high speed (HC) or low speed (LS) chips? It may be that the turbo Everdrive has additional circuitry creating a slight propigation delay sending requests for ROM address from the ED RAM to the console. If the Everdrive is just barely sending the signal on time, then even a very slight propigation delay by the logic chip could cause a miss, crashing the console CPU.

All digital logic gates have a slew rate, or the time a logic signal takes to transition from low to high or high to low. Increasing the VCC voltage to the chip will improve the slew rate but also consume more current during a transiton. An overvolted logic chip will change states more quickly, a principal often employed with over- or underclocking CPU in modern microprocessors, to change performance vs efficiency.

A modest increase from 5V to say 6V may improve the slew rate of the chip enough so that the console works, but anything over 5.6V and you run the risk of dumping current through the ESD protection diodes within the chips. This will cause unnecessary heating of the components possibly shortening useful service life. So if you're going to overvolt the regionmod logic chip, series resistors need to be added to limit current. 4.7k-10k might be a suitable range. Also, the region mod chip should never be run off the unregulated DC input. A voltage setting slightly over the lowest threshold that yields satisfactory results should be used in this instance, which may warrant a tunable regulator circuit.

It is unfortunate that the logic in the region mods is not adequate to get the job done. A simple 8P2T switch can be installed on the case (or quad ganged DPDT if a rare 8 pole switch cannot be found) as mechanical switches are bidirectional and have no propigation delay, but then you run the risk of longer leads to mount the switch in a convenient location. Too long leads can cause the traces to act as mini antennas, picking up or transmiting interference, have cross talk issues, or can otherwise attenuate high frequency signals.

[thesteve]

ok, propagation delay shouldnt be the issue as these are not gates in the conventional sense
the issue is as you approach VCC the pass-through resistance increases
also because its not a true gate its output voltage is dependent on the signal input not power input
LS is low-power Schottky and has a 9ms delay on switch
HC is high speed cmos having the same pinout and delay as LS
on these chips the propagation delay is only in play during control functions (/CE and A/B) so it is out of play well before reset goes high