Turbo Everdrive V2 does not work in my US Turbo

[bozo55]

Hi all,

My Turbo Everdrive works in my PCE Duo but gives all sorts of errors in my TG16.  It crashes just browsing the files and if you can manage to get a game to load it crashes after a few seconds.  I tried cleaning and reflowing the card connector and that didn't help.

I brought my Everdrive to my friend's TG16 and it worked perfectly on his so I know it's not the Everdrive itself.

I noticed that the board uses Taicon capacitors.  Should I replace these?  The system is not region modded and works fine with original HuCard games.

I do know that this system has sat in storage for YEARS in its box in a basement.  Will that increase the odds of the caps drying out?

Thanks for any help and sorry for creating yet another help thread.

[Keith Courage]

What kind of Ac adapter are you using? Have you tried swapping with your friends? Just wondering if it is  supplying enough power for the everdrive to work correctly.

[mickcris]

I had broke out my duojindance Duo-r, that I had bought a while ago, to test some scart cables I had made as it was the easiest console to get to.  Its odd as it works fine with my Everdrive v1 but just gives a white screen on the Everdrive v2.  I tried every combination of switch on both console and everdrive with no luck.

My v2 has worked on every other console that I have tried it on.  Seems to still be some odd behavior with the v2 on some consoles.

[bozo55]

What kind of Ac adapter are you using? Have you tried swapping with your friends? Just wondering if it is not supplying quite enough power for the everdrive to work correctly.

I have a genuine NEC supply 10.5v 730mA.  It's putting out 11 volts at the + Cap where the AC adapter plugs in.  I tried a genuine Model 1 Genesis adapter 9v 1.2A and that didn't help.  I noticed some voltage droop on pin 38 with the Turbo everdrive inserted.  It's a solid 5v  but then can fluctuate to a lesser amount, like 3.8 but only for a split second.  One of the 1uf50v caps was only showing 1v, but consistently.  Others were showing 2v, others 3.4ish. One of the Taicons was showing 4.6v, others 5v, others 11v.

I never thought to try my system on his power supply.  Damn.

[bozo55]

I had broke out my duojindance Duo-r, that I had bought a while ago, to test some scart cables I had made as it was the easiest console to get to.  Its odd as it works fine with my Everdrive v1 but just gives a white screen on the Everdrive v2.  I tried every combination of switch on both console and everdrive with no luck.

My v2 has worked on every other console that I have tried it on.  Seems to still be some odd behavior with the v2 on some consoles.

Your Doujindance is region modded?  I know these flash carts are super picky with region modded units.

[mickcris]

Yeah but the v1 works fine. The v2 is supposed to use less power so it should be more compatable.

[bozo55]

Yeah but the v1 works fine. The v2 is supposed to use less power so it should be more compatable.

Krikzz claimed that power is not the reason why region mods screw around with Everdrive functionality.  Rather, it distorts the signaling.  If the V2 is more strict with the signaling requirements, it doesn't matter if it uses less power.

[bozo55]

I tried jumpering pin 38 of the HuCard slot directly to the 7805 output with a 22 gauge wire and that did not help at all (I heard doing this could solve the problem).

Seems to me I have no choice but to replace the caps.  I already have 10 of the 16 I need from Nippon-Chemicon and Nichicon.  I'm gonna hunt down the rest from my electronics store and see if a recap helps. 

Unfortunately the power supply is sealed so I won't be recapping that any time soon.  I somehow doubt it's power supply related though, since even the Genesis power supply doesn't help, and that powers a Genesis and Mega Everdrive no problem.

[bozo55]

I hunted down some caps from a local shop but had trouble finding exact values in quality brands.

For the 10 and 100uf caps I ended up huge caps rated at 250v and 160v respectively.  Besides the frustration of fitting the caps, are there issues I should be aware of (i.e. heat, power draw, reliability)?  Would it be wrong to use such overkill type caps?

[wilykat]

No heat issue, they are just huge and rated to handle a lot more volts than you need.

[Keith Courage]

I just wanted to chime in and say that within the past week I have had two people contact me saying they have the exact same issue with the everdrive version 2 on their US turbo grafx systems. Works fine with Duos though

I hope this isn't the sign of a wide spread problem with these things. Please post if the caps fix this problem for you.

I'm leaning more towards it being an everdrive issue not the systems since this was never reported using a version 1 everdrive and a US Turbo grafx before.

[bozo55]

I just wanted to chime in and say that within the past week I have had two people contact me saying they have the exact same issue with the everdrive version 2 on their US turbo grafx systems. Works fine with Duos though

I hope this isn't the sign of a wide spread problem with these things. Please post if the caps fix this problem for you.

I'm leaning more towards it being an everdrive issue not the systems since this was never reported using a version 1 everdrive and a US Turbo grafx before.

I'll post results as soon as my cap kit from console5 arrives. I think the fast loading of the RAM on the v2 puts more stress on the system. My friend's working TG16 is a newer serial number than mine I believe but as far as I've heard all TG16s are identical internally. Maybe some system's caps aged worse than others.

[SephirothTNH]

Looks like I'm now in this same boat.  On my TG16 games glitch or freeze after a few min of play.  Some immediately.  However, on my PCE Duo everything works great.  bozo55 let us know if the recap fixes your problem.

Oddly I know that my duo needs re capped.  I have no to very low audio on it.  Yet the ted v2 works perfectly in it.

[bozo55]

Looks like I'm now in this same boat.  On my TG16 games glitch or freeze after a few min of play.  Some immediately.  However, on my PCE Duo everything works great.  bozo55 let us know if the recap fixes your problem.

Oddly I know that my duo needs re capped.  I have no to very low audio on it.  Yet the ted v2 works perfectly in it.

Interesting indeed.  Perhaps the US Turbo was an odd beast due to that board revision not existing in Japan.  Maybe the signaling on it is not as reliable?  All I can say is that I've proven that it can work on a US Turbo.  My friend's unit works perfectly with it.  If recapping it can't fix it I may just let him have it and wait for a new revision (2.3 or whatever) before I buy another Everdrive.

Kevin Horton, a well respected electronics engineer says that Krikzz's designs could be done better and that he's not surprised at how finicky they are (at least with regards to the version 1, see the GametechUS youtube channel for those vids). 

I even experienced an odd glitch today in Air Zonk at the end of Stage 2.  The backgrounds started flickering oddly.  It only lasted for the last section of the stage and then it was gone by stage 3.  Wonder if it's my Duo or just the fact that this cart is wonky.

I am anxiously awaiting the cap kit from Console5. It hasn't crossed the border yet though.  I probably won't have it until next week.  But I won't be working for most of next week so as soon as those caps are in my hands I'll get on it and report back.

[bozo55]

The caps arrived today.  I'm going to get started on this and report back hopefully later today.

[bozo55]

Well guys, I spent the better part of the last few hours recapping the Turbo.  I replaced all 16 caps in the system with the Console5 kit.  The caps were all Nichicon and looked to be perfect replacements.

The results...

...

...

No change whatsoever.  The only cap that was not replaced was the one inside the RF modulator, and Console5 does not provide that cap anyway.

My observations/hypotheses...

1.  The system did not perform any differently with the new caps.  I think anyone looking to recap their Turbo will be wasting their time.

2.  Perhaps the system might work better with the Turbo ED if some of the caps were replaced with ones that were of higher capacitance.  Maybe the Turbo pros here could suggest such a mod.

3.  The Turbo ED v2 is a finicky bastard.  Obviously Krikzz did not test it with the US Turbo very thoroughly because I think this is going to be a common problem.

4.  Why some Turbos work fine with it and others don't is beyond me.  Maybe there was a change in the manufacturing process of the US Turbo down the line.  We should try to isolate working units from non-working with serial numbers.  My unit is Serial # 0107255HB

[DeshDildo]

Hmmmm... I have tested my V2 ED in my Duo and it was fine.  I'll get my Turbo out of the box and let you know.  I like the idea of documenting the serial numbers to see if it's a random occurrence or if there may have been some nominal hardware change at some point.

[bozo55]

Hmmmm... I have tested my V2 ED in my Duo and it was fine.  I'll get my Turbo out of the box and let you know.  I like the idea of documenting the serial numbers to see if it's a random occurrence or if there may have been some nominal hardware change at some point.

Cool.  Thanks!

[Sjlewis78]

I have the same issue on both of my TG16 consoles. I was waiting for bozo on krikz forums to see if a cap replacement would help and no dice. Not sure what to try next or do.

[NightWolve]

Kinda doubt a cap replacement would've helped but that should've been done already if you're familiar with the history of NEC consoles. All your caps are likely leaking right now. Do it regardless.

Well, it's too bad about this new design. :/ Seems to work fine in my US TurboExpress (which is practically the same damn thing as the TG-16). I played a few levels of Blazing Lasers with no problems and did quick random loads of other ROMs, nothing thorough though. I guess I should do some more tests before feeling safe.

I only own the TurboExpress and Duo, so hopefully it'll be fine. The TG-16 I won in 1990 died on me at some point, so I threw it away. Anyway, I do like the quick load times, but the main reason I quickly sold my V1 was in the hope this new V2 would support the Arcade Card as krik mulled the idea over but didn't in the end... There's hope Bonknuts and/or elmer might hack a way some day, but that's a longshot I guess.

[elmer]

3.  The Turbo ED v2 is a finicky bastard.  Obviously Krikzz did not test it with the US Turbo very thoroughly because I think this is going to be a common problem.

I think that you've hit the nail on the head there.

The TED2 is krikzz's first design to use a completely different technology based on modern 3.3V chips instead of older 5V chips.

Because of that, the board contains it's own 3.3V power-supply regulator chip ... and so I'd be a little surprised if the problem that you're seeing is because of power issues.

But ... it is probably much more sensitive to signal delay and signal quality issues on the TG16's data and address buses than the old TED1 ... especially since both buses now need to be put through a new 5V-to-3.3V conversion chip which adds it's own delay.

TailChao has already commented on signal issues with the TG16 in his "CD Stupid Card" thread ...

As an aside, I've finally *removed* the region mod in my TG16 since this card now takes care of the only reason I installed it in the first place. However, during that process I did a few quick checks as to how the CD Stupid Card and a real System Card 3.0 react to different propagation delays (either due to wire length or the MC14551s). They both start to fail around the same time (especially with the MC14551s driven with a 5V supply), so I'm not too worried.  But I'll still order the 6ns CPLDs and see if things improve (we're using 15ns right now).

The TG16 has verrry long dataline traces on its mainboard and can be a real jerk in regards to signal integrity. The unmodded white PCE I have seems to run nearly anything. I imagine you Duo owners never have to think about any of this as everything is condensed on one board.

Alright, I just found something seriously messed up:

The reset timing on the PCE and TG16 is different, in that the TG16 will occasionally have the CPU released from reset long before the signal rises on the HuCard slot.

[wilykat]

So NEC's redesigned the already tiny and efficient PCE  to be made larger to appeal to "Bigger is Better" Americans is going to bite Krikzz in his ass.  Still got to wonder what NEC was thinking, spreading out those parts meant longer delay on lines. Could this be why the planned SCSI adapter was never going to work on TG-16 (even with extra RAM added), too long data lines?  I read about it being for Duo only.

Makes me wonder if there's a small spike in TG-16 sale on eBay as people are looking for one that works with their new ED and not realize TG-16 design is the problem.  And spike in PCE and Duo sale for people who are replacing the TG-16 so they can use ED 2.0

[bozo55]

So NEC's redesigned the already tiny and efficient PCE  to be made larger to appeal to "Bigger is Better" Americans is going to bite Krikzz in his ass.  Still got to wonder what NEC was thinking, spreading out those parts meant longer delay on lines. Could this be why the planned SCSI adapter was never going to work on TG-16 (even with extra RAM added), too long data lines?  I read about it being for Duo only.

Makes me wonder if there's a small spike in TG-16 sale on eBay as people are looking for one that works with their new ED and not realize TG-16 design is the problem.  And spike in PCE and Duo sale for people who are replacing the TG-16 so they can use ED 2.0

Nah, I'm hoping Krikzz steps up and addresses this.  Either by allowing people to exchange for another revision or hopefully being able to correct this in firmware somehow.

[Sjlewis78]

Is there a way to test if the signal delay is the root cause? Is there a way to get krikzz to look into this or exchange?

[wilykat]

It may take a good osciloscope, a dual trace model, with one probe on CPU end and another on cart connector end to see what the timing offset looks like.  If it shows significant delay between that 2 points, then the circuit is too long and ED is too "fast" trying to communicate with CPU.  Making it run slower might help.

If the O-scope shows little variation between PCE and TG-16, the issue is not with long trace but something else.  Is there something about TG-16 that is different from PCE and Duo beside the flipped data line and RF modulator with different frequency?

[elmer]

So NEC's redesigned the already tiny and efficient PCE  to be made larger to appeal to "Bigger is Better" Americans is going to bite Krikzz in his ass.  Still got to wonder what NEC was thinking, spreading out those parts meant longer delay on lines.

To keep the analogy appropriate for the 1980's ... it was like having the "Christie Brinkley" of consoles, and choosing to release "Rosanne Barr" instead. 

[LostFlunky]

i received mine today.  (TED2)

I have 3 U.S. TG's to try it on.  Will do so tonight and report the results with serials.

[Sjlewis78]

The serials on my two that don't work are 0162022HB and 9867404HB

[LostFlunky]

The serials on my two that don't work are 0162022HB and 9867404HB

Yikes!  Those pretty much look like bookends...

[Sjlewis78]

Ha. Yeah maybe the ones in the middle work?

[bozo55]

My friend's working unit starts with 99 i believe 

[SephirothTNH]

The serial number on mine is 9997016HB. No region mod.

It's likely random.  Some sort of marginal error like Kevtris was talking about.  I know from pokeing around in systems for years the 7805s can output anywhere from like 4.8 to 5 volts.  If 40 millivolts can make the deference in region modded systems working with TED a similar thing could be going on here. And it explains why it would work on one tg16 but not another. 

Our serials seem to encompass a wide run.  Maybe the whole run?  Seems like there's not much to benefit from recording serials especially If your friends that worked really starts with 99.

[esteban]

So NEC's redesigned the already tiny and efficient PCE  to be made larger to appeal to "Bigger is Better" Americans is going to bite Krikzz in his ass.  Still got to wonder what NEC was thinking, spreading out those parts meant longer delay on lines.

To keep the analogy appropriate for the 1980's ... it was like having the "Christie Brinkley" of consoles, and choosing to release "Rosanne Barr" instead. 

Roseanne Barr + John Goodman are sexy as hell, dammit.

[NightWolve]

So NEC's redesigned the already tiny and efficient PCE  to be made larger to appeal to "Bigger is Better" Americans is going to bite Krikzz in his ass.  Still got to wonder what NEC was thinking, spreading out those parts meant longer delay on lines.

To keep the analogy appropriate for the 1980's ... it was like having the "Christie Brinkley" of consoles, and choosing to release "Rosanne Barr" instead. 

Hahaha!

[LostFlunky]

So far no good- I've tested it on 2 of my Tg16's and one will not even load the OS (just a random solid coloured screen), while the other loads and will play a game for about 2 minutes before freezing...

I'll post serials later - I may have 2 more of these around...

[Sjlewis78]

Sounds more and more like an issue with the Everdrive on US TG16s

[wilykat]

What if we replaced the stock 7805 with something a little better? Something like Recom R78xx? This one is a drop in replacement that is 5v 1A and I've used it in Atari Lynx 5v mod (the original 5v regulator is a traisnstor and zener diode in emitter-follower design with no protection if one of that fails, the whole system gets 9v and gets cooked)

[SephirothTNH]

I guess it's worth a shot wilykat.  Kevs comments aren't necessarily applicable here though.  I was just using it as an example of how spotty these can be.  His comment was about using the TED v1.x on a region modded PCE.  He found that the TED would fail to write if the system was running on 4.97v or higher but would write fine if it was at 4.95v or less.  I don't think he tested 4.96 I don't remember.

Anyway This is a different situation.  TED v2 and non modded TG16.  It does seem to be write failures though.  At least for some of us.  My TED functions fine in menu but the games seem like they are getting corrupt during the writing process.

If something similar is happening here then yeah we could replace the 7805 but it would still be a crap shoot as to whether or not the system would be supplied with the perfect voltage for the TED to function correctly.  Still I guess it's worth a shot.  It would be better to test if this is a similar situation with a bench power supply though.  That could tell us if this line of thinking is way off or not.

[LostFlunky]

Here is something I don't think has been mentioned - my TG which will normally run a TED2 game for a few minutes before freezing, won't even initialize the TED2 when using the Turbo Booster...

My 3rd Turbo was also a bust, but I used the Turbo Booster the first time so I am going to do some more testing.

I still have one more to test after that...

[bozo55]

Here is something I don't think has been mentioned - my TG which will normally run a TED2 game for a few minutes before freezing, won't even initialize the TED2 when using the Turbo Booster...

My 3rd Turbo was also a bust, but I used the Turbo Booster the first time so I am going to do some more testing.

I still have one more to test after that...

This sucks.  Hopefully Krikzz will address this once he's back from vacation.

[SephirothTNH]

Interesting.  No booster on my end.  I did buy this unit with a broken RF port so I removed the RF modulator and s-video modded it.  Surely the s-vid mod doesn't pull more power then the RF modulator did?

Here's another thought.  What power supply is everyone using?  I'm using a Genesis model 1.

[bozo55]

Interesting.  No booster on my end.  I did buy this unit with a broken RF port so I removed the RF modulator and s-video modded it.  Surely the s-vid mod doesn't pull more power then the RF modulator did?

Here's another thought.  What power supply is everyone using?  I'm using a Genesis model 1.

Mine is a mint stock system that sat in storage most of it's life.  Using stock AC adapter and RF cable, zero mods.  I've tried a Genesis AC adapter and that did not help.

[LostFlunky]

All right - full play by play of testing 4 TG16's.

1. Serial #9841711HB

Mods:
-A/V out (composite)
-LEDs under grille

Connected via:  TGCD  Result: TED V2 no start...
Connected via:  TurboBooster+  Result: TED V2 no start
Connected via:  RF Result: TED V2 appears to work fine - ran it for half an hour with Air Zonk
Connected via:  A/V mod  Result: TED V2 starts, but is glitchy/freezing after a short while...

2. Serial #unknown - painted over

Mods:
-permanently wired for JP region
-LEDs under TG16 nameplate

Connected via:  TGCD  Result: TED V2 no start...
Connected via:  TurboBooster+  Result: TED V2 started - with card read error - would not go past this.
Connected via:  RF Result: TED V2 no start

3. Serial #9919190HB - this unit is pretty much shit canned because it will only play TTI era releases... except Bonk III.

Mods:
-RF unit removed
-S-video mod

Connected via:  TGCD  Result: TED V2 no start...
Connected via:  TurboBooster+  Result: TED V2 no start


4. Serial #022339HB

Mods:
-none - virgin


Connected via:  TGCD  Result: TED V2 no start...
Connected via:  TurboBooster+  Result: TED V2 no start
Connected via:  RF Result: TED V2 seems to work fine - ran for over an hour with no freezing.


Summary - I have two units that work via RF and that is it...

With unit #1, I ran it for half an hour with Air Zonk and then switched to the A/V output by plugging in cables - the picture went back and the music kept playing.  I switched back to the RF signal and the screen had indeed turned black.  "run+select" reset just made everything black and the music stopped too...

[bozo55]

All right - full play by play of testing 4 TG16's.

1. Serial #9841711HB

Mods:
-A/V out (composite)
-LEDs under grille

Connected via:  TGCD  Result: TED V2 no start...
Connected via:  TurboBooster+  Result: TED V2 no start
Connected via:  RF Result: TED V2 appears to work fine - ran it for half an hour with Air Zonk
Connected via:  A/V mod  Result: TED V2 starts, but is glitchy/freezing after a short while...

2. Serial #unknown - painted over

Mods:
-permanently wired for JP region
-LEDs under TG16 nameplate

Connected via:  TGCD  Result: TED V2 no start...
Connected via:  TurboBooster+  Result: TED V2 started - with card read error - would not go past this.
Connected via:  RF Result: TED V2 no start

3. Serial #9919190HB - this unit is pretty much shit canned because it will only play TTI era releases... except Bonk III.

Mods:
-RF unit removed
-S-video mod

Connected via:  TGCD  Result: TED V2 no start...
Connected via:  TurboBooster+  Result: TED V2 no start


4. Serial #022339HB

Mods:
-none - virgin


Connected via:  TGCD  Result: TED V2 no start...
Connected via:  TurboBooster+  Result: TED V2 no start
Connected via:  RF Result: TED V2 seems to work fine - ran for over an hour with no freezing.


Summary - I have two units that work via RF and that is it...

With unit #1, I ran it for half an hour with Air Zonk and then switched to the A/V output by plugging in cables - the picture went back and the music kept playing.  I switched back to the RF signal and the screen had indeed turned black.  "run+select" reset just made everything black and the music stopped too...

Strange how using the composite could affect things...  Is it a A/V mod box that attaches to the ext port?

[SephirothTNH]

I don't know if you would call them a mod box.  They do plug into the ext port though.  Official products that gave composite out and added save ram. 

That is odd that it works with RF but not anything else whether mod or official.  It is a little different for me then.  Even though I'm using s-video the ted always boots.  And always lets me load a game. 

Edit:  Lost Monkey I must have glossed over your "AV Mod" result on the first system.  That sounds like what mine is doing.  the game starts up but will eventually glitch up or freeze.

[LostFlunky]

1. Serial #9841711HB

Mods:
-A/V out (composite)
-LEDs under grille

Connected via:  TGCD  Result: TED V2 no start...
Connected via:  TurboBooster+  Result: TED V2 no start
Connected via:  RF Result: TED V2 appears to work fine - ran it for half an hour with Air Zonk
Connected via:  A/V mod  Result: TED V2 starts, but is glitchy/freezing after a short while...

Strange how using the composite could affect things...  Is it a A/V mod box that attaches to the ext port?

No box - just a direct wiring of the A/V outs from the bus to RCA jacks on the side of the TG16...

[LostFlunky]

Here's another thought.  What power supply is everyone using?  I'm using a Genesis model 1.

I forgot to mention this before, but I tried with both an original power supply and a Genesis supply... same results.

[elmer]

Summary - I have two units that work via RF and that is it...

With unit #1, I ran it for half an hour with Air Zonk and then switched to the A/V output by plugging in cables - the picture went back and the music kept playing.  I switched back to the RF signal and the screen had indeed turned black.  "run+select" reset just made everything black and the music stopped too...

There are also 2 guys reporting failures on Core II systems (but only when they attach their Tennokoe2).

We need someone that actually does electronic design work to comment on this ... but from my very limited rememberence of university, this sounds like CPU bus-loading issues.

Everything is OK on your 2 unmodified base systems ... but once you add the extra circuitry or trace-length on to the CPU bus with either the TGCD, the TurboBooster+ (the memory backup), or the region-mod ... then that's when it all falls apart.

It's entirely possible that the TED2's two ALVC16424S 3.3V-5V transceivers just don't have the strength to drive good signals against that load.

It could still be power issues, though.

It really needs someone with the right skills/equipment to look at it.

Strange how using the composite could affect things...  Is it a A/V mod box that attaches to the ext port?

No box - just a direct wiring of the A/V outs from the bus to RCA jacks on the side of the TG16...

This is the result that confuses me.

[NightWolve]

(the original 5v regulator is a traisnstor and zener diode in emitter-follower design with no protection if one of that fails, the whole system gets 9v and gets cooked)

I wish I understood that way back... My SNES kept popping fuses and I ran out of 1 Amp rated, so I went higher cause I thought I solved it, but same thing. Steve told me the regulator burned out so I did a 5 Volt mod to the system, took out the regulator and heat sink, used a new regulated power supply that outputs 5.3 VDC exact and got the system running again. But, I fried one of the SRAM chips with the fuse burnouts and the 12 Volt original power supply, the cheap unregulated style that it was (says 10, goes 12 with my home's 126 VAC supply).

So as wilykat describes, the broken-down regulator in my case, having gotten at the end of its lifespan, sent the full input 12 Volts to the motherboard from its output leg and caused damage!! Only for a second as the new fuse popped again! So yeah, until I swap that SRAM chip out, I have graphical glitches on bitmaps with certain games... :/

Anyway, it would be interesting if this can be solved with a better 5 Volt regulator for the US TG-16!

[DeshDildo]

I finally have news to report.  My untouched (other than cleaning the card slot) Turbografx will sort of run TED V2.  I am running it with the original power supply and original RF box.  Serial number 9Y88229HB. 

I am pretty sure my rom files are good (same ones I ran on my old TED that worked just fine).  Many games will load but crash during a boss fight or other random times during the game.  Other games just won't load at all.  These same games do load fine while running in my Turbo Duo.  Perhaps I will contact Bernie and download his rom file and try some of the same games that won't load. 

Very interesting indeed.